Vous êtes sur la page 1sur 422

Essays on Medieval Computational Astronomy

Time, Astronomy, and Calendars


texts and studies

Editors

Charles Burnett
Sacha Stern

Editorial Board

Dibh Crinn Benno van Dalen Gad Freudenthal Tony Grafton


Leofranc Holford-Strevens Bernard R. Goldstein Alexander Jones
Daryn Lehoux Jrg Rpke Julio Sams Shlomo Sela John Steele

volume 5

The titles published in this series are listed at brill.com/tac


Essays on Medieval
Computational Astronomy
By

Jos Chabs
Bernard R. Goldstein

leiden | boston
Cover illustration: Detail of the star chart in the Alfonsine Libro de las estrellas de la ochava espera in
Madrid, Real Academia de la Historia, ms 9/5707, f. 103. Courtesy Real Academia de la Historia .

Library of Congress Cataloging-in-Publication Data

Chabs, Jos, 1948-


Essays on Medieval computational astronomy / by Jos Chabs, Bernard R. Goldstein.
pages cm. (Time, astronomy, and calendars, ISSN 2211-632X ; volume 5)
Includes bibliographical references and index.
ISBN 978-90-04-28174-5 (hardback : alk. paper) ISBN 978-90-04-28175-2 (e-books) 1. Astronomy,
Medieval. 2. AstronomyTables. I. Goldstein, Bernard R. II. Title. III. Title: Medieval computational
astronomy.

QB26.C43 2015
521.01'51dc23
2014041332

This publication has been typeset in the multilingual Brill typeface. With over 5,100 characters covering
Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more
information, please see www.brill.com/brill-typeface.

issn 2211-632X
isbn 978-90-04-28174-5 (hardback)
isbn 978-90-04-28175-2 (e-book)

Copyright 2015 by Koninklijke Brill nv, Leiden, The Netherlands.


Koninklijke Brill nv incorporates the imprints Brill, Brill Nijhoff and Hotei Publishing.
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive,
Suite 910, Danvers, ma 01923, usa. Fees are subject to change.

This book is printed on acid-free paper.


Contents

List of Figures vii

Introduction 1

part 1
Conjunctions and Oppositions

1 Nicholaus de Heybech and His Table for Finding True Syzygy 9

2 Computational Astronomy: Five Centuries of Finding True


Syzygy 40

3 Transmission of Computational Methods within the Alfonsine


Corpus: The Case of the Tables of Nicholaus de Heybech 57

part 2
Planetary Motions

4 Ptolemy, Bianchini, and Copernicus: Tables for Planetary


Latitudes 73

5 Displaced Tables in Latin: The Tables for the Seven Planets for
1340 99

6 Computing Planetary Positions: User-Friendliness and the Alfonsine


Corpus 150

part 3
Sets of Tables

7 Andalusian Astronomy: al-Zj al-Muqtabis of Ibn al-Kammd 179

8 Early Alfonsine Astronomy in Paris: The Tables of John Vimond


(1320) 227
vi contents

9 John of Murss Tables of 1321 308

10 Isaac Ibn al-adib and Flavius Mithridates: The Diffusion of an


Iberian Astronomical Tradition in the Late Middle Ages 338

part 4
Other Tables

11 Ibn al-Kammds Star List 373

12 Astronomical Activity in Portugal in the Fourteenth Century 389

Index 407
List of Figures

1.1 The mean conjunction ( s = m) at time t takes place after the true
conjunction at time t 11
1.2 The corresponding true conjunction (s = m) at time t takes place
before the mean conjunction at time t 12
1.3 Graphical representations of the entries in the five columns of Nicholaus
de Heybechs Table at intervals of 6 18
3.1 Facsimile of Nicholaus de Heybechs table (excerpt): Vienna,
Nationalbibliothek, ms 2440, f. 74v 59
3.2 Facsimile of tv 7 (excerpt): Madrid, Biblioteca Nacional, ms 3385,
f. 106r 62
3.3 Facsimile of tv 8 (excerpt): Madrid, Biblioteca Nacional, ms 3385,
f. 107r 63
4.1 The functions y =| cos x| and y = cos2 x in the interval (0, 180), where
the cosine function represents c5(x) in formulas (1) and (2) 91
5.a The solar equation displaced vertically 106
5.b Ptolemys second lunar model 110
5.c The model for Mars 120
5.d Mars, equation of center and equatio porcionis 123
5.e Mars, minutes of proportion 125
5.f First station of Mars 130
5.g Equation of anomaly near greatest distance for Mars 133
5.h Venus, equation of anomaly as a function of the true argument of
anomaly 140
6.1 Ptolemys model for Mars 152
6.2 The equation of center of Jupiter 170
7.1 The geometrical model underlying Ibn al-Kammds table for trepidation,
as reconstructed 209
8.1 A geometric interpretation of Vimonds tables for the mean motion for
Venus 253
8.2 Vimonds equation of center, col. 5, for Venus, Mars, Jupiter, and
Saturn 262
8.3 Vimonds equation of center, col. 5, for Mercury 262
8.4 Ptolemys model for the three superior planets and Venus 269
9.1 Correction for Saturn as a function of its mean argument of center 315
9.2 Correction for the Moon for a given age (20 days) 319
9.3 Correction for the Moon for a given value of mean lunar argument of
anomaly (4,0) 319
viii list of figures

12.1 Table of samt in Burgos (from Lib 1 to Ari 1) 393


12.2 Calendaric matters (Madrid, ms 3349, f. 3v) 394
12.3 Domiciles according to al-Brn 401
Introduction

During the Middle Ages and early modern times tables were a most success-
ful and economical way to present mathematical procedures and astronomical
models, facilitating computations based on them. One reason for depending
on these techniques was the absence of modern mathematical notation to
represent the algorithms that astronomers were using. Indeed, some major
sets of tables were the direct result of the development of new astronomical
approaches. In the second century ad Ptolemy composed the Almagest, a com-
prehensive handbook of astronomy and related mathematical procedures: in
it he presented a set of observations on the basis of which he determined the
parameters of his geometrical models for planetary motion. He then compiled
tables which were also included in this treatise. At a later date he revised these
tables, making them easier to use, in a work called the Handy Tables where, for
example, Ptolemy displayed entries at 1-intervals, rather than at 3- and 6-
intervals as he had done previously. Nicolaus Copernicus (14731543) provides
a somewhat different example of this pattern. In 1543 his magnum opus, De
revolutionibus, was published, in which he described a set of planetary mod-
els to replace those based on the Almagest, and included tables for computing
planetary positions with these new models. In 1551 Erasmus Reinhold (1511
1553) compiled the Prutenic Tables, based on the models of Copernicuss De
revolutionibus, presented in ways that facilitate computation. Most of the time,
however, the compilation of complete sets of tablesor of individual tables
just reflected partial changes in the parameters and the models underlying a
particular theory, or new methods to compute the positions of the celestial
bodies without changing the underlying models. This is particularly true in the
Middle Ages, when the Ptolemaic models were rarely challenged.
Astronomical tables are a basic component of astronomy, although they are
frequently neglected (or considered an unintelligible sequence of numbers)
because, more often than not, it is no simple matter to establish the structure
of such tables, and sometimes it is even difficult to identify the problem which
a particular table addresses. Yet, to understand astronomical tables and to
fully tackle the procedures used to compile them, various skills, both linguistic
and mathematical, are required at the same time. The papers included in this
volume give many examples where the meaning and purpose of such tables has
been determined by careful analysis.
As we have shown in our recent monograph, A Survey of European Astronom-
ical Tables in the Late Middle Ages (Brill, 2012), astronomical tables are a primary
source of historical information. Through their analysis it is possible to insert

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_002


2 introduction

them and their compilers in an astronomical tradition, thus displaying unex-


pected links between authors. These analyses also allowed us to discover com-
putational techniques, interpolation methods, and approximation procedures,
as well as to identify changes in the standard parameters and the geometrical
models. It should be stressed that an analyst of a medieval table should only
appeal to modern mathematics very selectively and judiciously, given the risk
of falling into anachronisms that may lead to incorrect results. In particular,
modern mathematical functions are generally to be avoided although medieval
mathematical procedures expressed verbally may legitimately be translated
into algebraic notation. In recomputing a table consistency requires making
use only of computational techniques, concepts, and strategies that were avail-
able at the time. For example, in computing entries in a table for finding the
length of daylight on a given day, one needs to use trigonometrical tables that
can be found in the same set of tables or in a previous set; indeed, this is
how it was done by medieval astronomers. In other words, we have to evaluate
medieval scholars according to criteria consistent with their own time, which
may be different from those appropriate for other periods.
The pioneering work on this subject goes back to the beginning of the twen-
tieth century. In the years between 1899 and 1907 C.A. Nallino published an
edition, translation, and commentary on the zij (i.e., a set of astronomical
tables) of al-Battn (ca. 900). This work was extended in 1956 by E.S. Kennedy
in his survey of Islamic astronomical tables. Moreover, in 1962 O. Neugebauer
provided an extensive commentary on the zij of al-Khwrizm, originally com-
posed in Baghdad in the ninth century that is only extant in a Latin version
by Adelard of Bath (10751160), which was published by H. Suter in 1914. While
the publication and analysis of Islamic astronomical tables have continued to
this day, considerable attention has also been devoted to astronomical tables
in Latin, Byzantine Greek, Hebrew, and the vernacular languages in Europe,
including our own contributions to this field which have been built on those of
our distinguished predecessors. As a result of these scholarly efforts a critical
mass of studies is available which provides a good understanding of the overall
framework for the transmission of astronomical ideas and procedures through
tables in the period from Ptolemy to the early sixteenth century. The study of
many sets of tables in a variety of languages still to be examined in European
libraries (and others in different parts of the world) will surely lead to signif-
icant modifications of this framework in addition to an expansion of its con-
tents. In particular, there are at least three major topics concerning medieval
astronomy in the Latin West, where the analysis of tables is likely to play a crit-
ical role in enhancing our understanding of developments in this domain: the
early stages of astronomy in Latin Europe, based on Arabic materials in the
introduction 3

Iberian Peninsula and the Maghreb; the importance of several sets of tables in
Hebrew compiled in the Iberian Peninsula in the fourteenth and fifteenth cen-
turies; and the shaping of the Parisian Alfonsine Tables in the central decades
of the fourteenth century in Paris.
The twelve articles presented in this collection have been jointly written by
the two of us in a period covering about two decades. They are not the only
ones, for our collaboration has extended to many other papers and mono-
graphs, for the most part devoted to astronomical tables. This four-handed
collaboration, in the sense that all words and sentences as well as all numbers
and computations have been written and checked by both of us, has proved to
be fruitful and stimulating, an accomplishment that was achieved despite an
oceana few thousand kilometers wideseparating us.
The focus of attention in history of science is generally on scientists who
produced new theories, such as Kepler and Newton. In the Middle Ages, how-
ever, the innovations were made in the context of a long tradition. The essays
selected here address the major issues in medieval astronomy, and offer exam-
ples of several of the cleverest solutions in tabular form we have ever found.
The first group of essays concerns syzygies; specifically, the determination
of the time interval from mean to true syzygy, where mean syzygy refers to
a conjunction or opposition of the mean Sun with the mean Moon, and true
syzygy refers to a conjunction or opposition of the true Sun with the true Moon.
The concept of mean Sun and mean Moon (or the mean positions of the Sun
and the Moon) was used by Ptolemy in his Almagest, where the mean position
of the Sun, the Moon, or a planet is the place it would reach on the ecliptic if
its motion were uniform. This mean position is then correctedusing a table
derived from a geometrical modelto yield its true position at a given time.
Finding the time interval from mean to true syzygy is indeed a fundamental
problem in computational astronomy, for it is the first step in computing
the time and circumstances of a solar or lunar eclipse. Ptolemy had given
an approximate solution by means of an iterative procedure without a table,
and medieval astronomers devoted much effort to provide innovative, more
precise, and more user-friendly solutions. One impressive approach to this
problem is due to Nicholaus de Heybech (ca. 1400), who split the time from
mean to true syzygy into two terms, one for the Sun and the other for the Moon,
and managed to present his solution in the form of a single table in five columns
that makes computation easy to do in a few steps, for it requires only addition,
subtraction, and simple interpolation, rather than an iterative procedure with
many steps.
The second topic addressed here concerns the motion of the planets. The
history of tables for computing planetary latitude (that is, the angular distance
4 introduction

from a planet to the ecliptic) and using planetary equations (that is, the differ-
ence between the mean and true positions of a planet, where the Sun and the
Moon are included among the planets) is surveyed in two papers, and many
examples of their evolution are given. The other paper deals with the Tables for
the Seven Planets for 1340, compiled by an anonymous author working within
the framework of the Alfonsine Tables, who clearly had a profound understand-
ing of planetary astronomy. He succeeded in simplifying the astronomers task
by eliminating subtractions in the course of computing planetary positions.
This was very helpful at a time when negative numbers were not yet available,
for many complicated arithmetic rules were needed to express what can now
be represented algebraically using negative numbers. To accomplish this, the
compiler of these tables introduced what we call horizontal and vertical dis-
placements (following the terminology of E.S. Kennedy), such that the results
of the computations agree with those that depend on the standard Alfonsine
Tables, although the intermediate steps are all different.
In the third group we present different sets of tables by several authors,
beginning with the zij by Ibn al-Kammd (Crdoba, ca. 1100). In the Iberian
Peninsula astronomers were heirs to two traditions: a Greek tradition largely
based on the works of Ptolemy, and an Indian tradition introduced into the
Islamic world in the eighth and ninth centuries. The first is represented by the
zij of al-Battn, and the second by the zij of al-Khwrizm. Ibn al-Kammd
relied on both traditions, and we analyze each of his tables in terms of its
structure and sources. We then turn to the sets of tables by John Vimond
(ca. 1320) and John of Murs (ca. 1321), both of whom were active in Paris
and engaged in recasting the tradition of Alfonsine astronomy, specifically,
astronomical tables produced in Toledo in the thirteenth century under the
patronage of Alfonso x of Castile. The Parisian version of the Alfonsine Tables
is extant in many copies and it was the most influential set of tables in Latin
Europe during more than two centuries. Finally, we review the tables compiled
by Isaac Ibn al-adib (d. ca. 1426), who left Spain to live in Sicily in about 1396,
and whose tables, which concentrate on the motions of the two luminaries,
give innovative solutions. It is worth noting that the sets of tables by these four
authors were all available at the time in Latin manuscripts, even though the
work of Ibn al-Kammd was originally written in Arabic and that of Isaac Ibn
al-adib in Hebrew.
Finally, of the two papers listed under Other Tables, one is entirely devoted
to a list of 30 stars that depends on Ptolemys star catalogue. It was compiled
by Ibn al-Kammd, and it was surprisingly successful, for it is preserved in
texts in Arabic, Latin, and Hebrew. The other paper describes the contents of
a manuscript containing a mixture of tables of diverse origins and intended
introduction 5

for different purposes, some of which are strictly astrological. The tables, none
of them for planetary motion, reflect astronomical activity in Portugal during
the fourteenth century. Among the tables in this manuscript is one for the
daily lunar progress in an astrological month and another is for astrological
months, where 13 such months constitute a tropical year; this astrological
Moon (a fictitious celestial body) advances 390 in each astrological month.
We have found that these astrological tables appear in many medieval and
early modern sets of astronomical tables. For an explanation and historical
account of these astrological tables, see our article, Planetary velocities and
the astrological month, Journal for the History of Astronomy, 44 (2013), 465478.
Despite their great variety, underlying the tables reviewed in this collection
of essays are models for the Sun, the Moon, and the planets, which follow
essentially the Ptolemaic tradition in an ingenious and original way. Some of
them also provide elegant and intelligent solutions to standard astronomical
problems, within a general tendency to offer the practitioner user-friendly
tables that facilitate his task of computing planetary positions and celestial
events.

We thank the editors and publishers of the journals in which the chapters
of this book first appeared as independent essays for graciously granting us
permission to reprint them: bibliographic data for the original publication are
given in the first footnote of each chapter. Since it was decided to reset type for
these essays, we have taken the opportunity to make some minor changes in
them.
part 1
Conjunctions and Oppositions


chapter 1

Nicholaus de Heybech and


His Table for Finding True Syzygy*

Introduction

The calculation of eclipses was a major task for medieval astronomers, and the
first step in this procedure was the determination of the time from mean to
true syzygy (where syzygy means either the conjunction or the opposition of
the Sun with the Moon). The basic approach given in Ptolemys Almagest was
refined by subsequent astronomers in various ways. In the Latin West of the
late Middle Ages the corpus of Alfonsine astronomy held a dominant position.
Of special interest in this tradition is the solution to this problem in Chap. 22
of John of Saxonys canons for the Alfonsine Tables (ca. 1327)1 as well as the
solution attributed to Nicholaus de Heybech of Erfurt (ca. 1400).2 The text
and table of Nicholaus de Heybech (see Appendix 1) have not been discussed
previously.3 lt will be shown that the method of Nicholaus de Heybech is much

* Historia Mathematica 19 (1992), 265289.


1 Editio princeps, Ratdolt [1483]; cf. Poulle [1984, 17].
2 In a short note, Thorndike [1948] collected what little is known about Nicholaus de Heybech
of Erfurt. His principal text is the canon and table to find the time of true conjunction and
opposition of the Sun and the Moon. The date of composition of this text is uncertain but
some of the manuscripts of it were copied in the 1440s. Thorndike adds that most of the
manuscripts are of the fifteenth century but one manuscript, dated 1394, was in the Library of
Grenville Kane, Tuxedo Park, New York (the present location of that manuscript is in Prince-
ton). In addition to listing manuscripts of the text which will be discussed below, Thorndike
reports that ms Kln W* 178 (fol. 29v) contains a table of mean conjunctions for the years from
1384 to 1504 computed for the meridian of Paris and ascribed to a certain Nicholaus, sug-
gesting that it is the same author. In the manuscript we find the authors name in the phrase
given in the heading as composita per Nycolaum de Er* where the asterisk represents here
a balloon-shaped symbol. However, it is by no means certain that this author is to be iden-
tified with our Nicholaus. Moreover, Hartmann [1919, 1213] refers to a certain Nicholaus de
Heybech who was registered as a student at Erfurt University in 1421, and he suggests that
this might be our Nicholaus. According to Thorndike, in 1392 Nicholaus de Heybech com-
pleted a copy of Gerard of Cremonas Theorica Planetarum, preserved in ms Cues 213. From
this meager information we conclude that Nicholaus de Heybech was active circa 1400.
3 A translation of the text of Nicholaus de Heybech appears in Appendix ii.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_003


10 chapter 1

simpler than that of John of Saxony. In the same period another way to solve
this problem by means of tables, introduced by Levi ben Gerson (d. 1344)
and followed by Jacob ben David Bonjorn (fl. ca. 1361), has recently received
extensive treatment.4 Levi ben Gersons method requires the use of a set of four
tables and calls for many more computations than the method of Nicholaus de
Heybech.
The Alfonsine tradition is identified with Alfonso x, king of Castile (Spain),
who reigned from 1252 to 1284. He sponsored much scientific activity, and a
number of astronomical works were written in, or translated into, Castilian
at that time. The Alfonsine Tables are a special case because they are only
extant in a Latin version produced in Paris in the 1320s, and derivatives from
it. lt has even been argued that there never was a Spanish version prior to the
Parisian one [see Poulle 1988]. The manuscript tradition of these tables is very
complex and has not been adequately examined. For this reason, it has been
customary to identify the Alfonsine Tables with those published by E. Ratdolt
in 1483 despite the difficulty in deciding which tables were intended to be so
designated by the Parisian group in the 1320s, let alone by the astronomers
serving under Alfonso x. One of the characteristic features of the Parisian
version (and the edition of 1483) is the division of the circle into physical
signs of 60 rather than natural signs of 30 that were prevalent in ancient
and medieval astronomy [Poulle 1988, 100]. We refer to the entire manuscript
tradition associated with these tables as the corpus of Alfonsine astronomy.
We shall first discuss Ptolemys method and then the method presented by
John of Saxony, neither of which is reduced to specific tables for this purpose.
Then we turn to the method of Nicholaus de Heybech, who successfully pre-
sented his solution to this problem in a single table. We begin by introducing
some definitions and notation.
Let s, s, m, and m be the true and mean longitudes of the Sun and the
Moon, respectively, at a mean syzygy that occurs at a given time t. The solar
correction (cs) and the lunar correction (cm) are defined as

cs() = s s

and

cm() = m m

4 On Levi ben Gersons method, see Goldstein [1974, 136144, 229241]; on Jacob ben David
Bonjorn see Chabs [1989, 2639, and 1991].
nicholaus de heybech and his table for finding true syzygy 11

figure 1.1 The mean conjunction ( s = m) at time t takes place


after the true conjunction at time t (see Figure 1.2)

where is the mean solar anomaly and is the mean lunar anomaly. Note that
for 0 180, cs 0; and that in the simple lunar model, for 0 180,
cm 0. However, in Ptolemys complete lunar model cm depends on the mean
elongation = m s, as well as on . At mean syzygy 2 = 0, whereas at true
syzygy 2 = 0, where = m s.5
We add a prime symbol to all variables related to the true syzygy taking place
at a time t, corresponding to a mean syzygy that takes place at time t. Let t =
t t be the interval between the two events. When the true syzygy comes after
mean syzygy, t is positive and m < s. Figure 1.1 illustrates a mean conjunction
where m > s, and t is negative; Figure 1.2 illustrates the corresponding true
conjunction.
From the definition of t it follows that

(1)

5 For a discussion of Ptolemys solar and lunar models see, for example, Pedersen [1974, 122
202] and Neugebauer [1969, 191198].
12 chapter 1

figure 1.2 The corresponding true conjunction ( s = m) at time


t takes place before the mean conjunction at time t
(shown in Figure 1.1)

where the functions vm(t) and vs(t) are the time dependent velocities in longi-
tude of the Moon and the Sun. Since m = s at mean conjunction, Eq. (1) can
be written as

(2)

The difficulties in assigning proper values to these functions in the calculation


of t gave rise to a variety of approaches in the determination of the times of
true syzygies.
The functions vm(t) and vs(t) are not tabulated in Ptolemys Almagest. For the
calculation of t Ptolemy depended on the approximation vs(t)/vm(t) 1/13.6
Then Eq. (1) becomes

(3)

6 Almagest vi.4; Toomer [1984, 281]; Neugebauer [1975, 122]; Pedersen [1974, 221226].
nicholaus de heybech and his table for finding true syzygy 13

According to Ptolemy the next step is to compute vm(t) as a function of the


lunar anomaly by means of a relation which, expressed in modern notation, is

(4) vm(t) = 0;32,56 + 0;32,40 c,

where 0;32,56 /h is the hourly mean lunar velocity in longitude, 0;32,40 /h


is the hourly mean lunar velocity in anomaly, and c = cm() cm( 1) is
the difference in the lunar correction corresponding to one degree of mean
anomaly at the time of mean syzygy. However, since Ptolemys table for the
lunar correction (Almagest iv.10) is arranged for intervals in the argument of
6 in the upper part and 3 in the lower part, the previous expression is stated
as c = (cm() cm( d))/d, where d is 6 or 3, as appropriate. At time t
+ t the true positions of the Sun and the Moon are to be computed anew.
If there is still a sensible difference between them, this procedure is to be
iterated.

The Alfonsine Method according to the Canons of John of Saxony7

If the time t at which a mean syzygy takes place is known, the values for all
the required quantities may be found and the time difference t from mean to
true syzygy may be derived by means of a successive approximation procedure
involving two principal steps.

1. First, it is necessary to compute a time interval which yields the time of true
syzygy, correct to the nearest hour. To do so, John of Saxony appeals to Ptolemys
complete lunar model in which Ptolemy introduces an equation of center q, a
function of . However, since q( ) = 0 at mean syzygy, the canons introduce an
equated lunar anomaly in order to determine a consistent value for vm. The
relation proposed is

(5)

where the mean lunar anomaly and the elongation are to be replaced by
their values at mean syzygy. In the text, the elongation = m s, is called
longitudo lune when > 0 and longitudo solis when < 0. The text offers
no justification for Eq. (5) but clearly its purpose is to account for the change

7 Poulle [1984, 8087].


14 chapter 1

in lunar velocity in the time interval from mean syzygy to true syzygy. The
lunar velocity will now be treated as a function of rather than of , where
corresponds to the midpoint of the time interval from mean syzygy to true
syzygy.
To derive Eq. (5), let t be half the time interval from mean syzygy to true
syzygy and = v t, where v, the hourly mean lunar velocity in anomaly,
is 0;32,40 /h. In Eq. (5), the coefficient 13/24 (= 0;32,30) seems to represent an
approximation of v. We further assume that the velocity in elongation from
mean syzygy to true syzygy can be approximated by its mean value, v = 0;30,29
/h. It follows that the time, 2t, from mean syzygy to true syzygy equals /,
or 2; hence t = . Note that in Eq. (5) is therefore measured in hours. The
value of vm corresponding to is intended to represent the velocity of the
Moon at the midpoint of the time interval from mean syzygy to true syzygy.
The text assumes that the velocity of the Sun remains the same in this time
interval.
According to the canons, the time interval is then computed by means of
the relation

(6)

where = 0;0,1 < (|Int()| 1) and where Int() is the integer part of the
elongation at mean syzygy (in degrees). When = 0, Eq. (6) has the same
structure as Eq. (1); that is, the difference in lunar and solar longitudes is divided
by the difference in their velocities. Note, however, that in Eq. (6) lunar velocity
is a function of the equated anomaly defined in Eq. (5) rather than the mean
anomaly. In the text the difference between the lunar and solar velocities is
called superatio and the denominator of Eq. (6) is called superatio equata.
The solar velocity in the time from mean conjunction to true conjunction
is essentially constant; therefore, the term is intended to modify the lunar
velocity which changes noticeably in that same time interval. This term serves
the same function in the Toledan Tables and Azarquiels almanach, and John
of Saxony apparently followed the same tradition.8 This adjustment to lunar

8 Toomer [1968, 85] and Mills [19431950, 233] display a table of corrections to the hourly
lunar velocity corresponding to ; the argument varies from 1 to 7 and the entries from
0to 6. In alBattns zj a similar table is found [Nallino 1907, 88], but here the argument
varies from 1 to 7 and the corresponding entries from 1to 7. The tabulated function is thus
0;0,1 |Int()|. It would seem that for al-Battn this is a correction to the hourly velocity of
the Moon to account for the change in lunar velocity in the time interval from mean to true
nicholaus de heybech and his table for finding true syzygy 15

velocity is no longer needed because Eq. (5) satisfies the same purpose. Indeed,
has no sensible effect on the value of because its maximum of 0;0,6 is very
small compared with the least difference in lunar and solar velocities (0;27,45
according to the Toledan Tables, or 0;27,47 according to the edition of 1483 of
the Alfonsine Tables).9
The text presents rules for the algebraic sign of , but they are incomplete.
The text just considers the case where the longitude of the Moon is greater
than that of the Sun, i.e., > 0; this implies that true conjunction precedes
mean conjunction (t < 0), and that the lunar anomaly in that interval is always
smaller than its value at mean conjunction. Under this circumstance, the text
gives the correct rule for the algebraic sign of : if 180, then 0, and
vice versa. Unfortunately, the text uses the expression longitudo10 instead of
longitudo lune ( > 0) and omits any discussion of the case where < 0
(longitudo solis).

2. The second step requires the computation of a time interval * to yield the
time, correct to a minute of an hour, for the true syzygy: * + = t. In order to
do so, the canons propose a differential method that involves calculating and
comparing all the quantities at two different times 24 min apart (i.e., 1/60 of a
day): t + and t + + 0;1 d. For a given time dependent variable x, let

dx = x(t + + 0;1) x(t + )

be its variation in a minute (i.e., a sixtieth) of a day. According to the canons,


the time interval * is determined as

(7)

syzygy. Al-Battns maximum argument of 7 reflects the maximum value of which for
him is 7;0,10. The maximum entry in this table of 7is an average (where the range is from
0to 11) for the effect on lunar velocity in degrees per hour corresponding to a change in
lunar anomaly of about 3 (corresponding to half the time for the Moon to reach the
point of true conjunction).
9 Toomer [1968, 82]; Ratdolt [1483, fol. g6rg7r]. In his canons, John of Saxony mentions a
table by John of Lignres for solar and lunar velocity, but it is not clear which set of entries
is intended [cf. Poulle 1984, 82].
10 Ratdolt [1483, fol. b1r, line 5]; cf. Poulle [1984, 82, line 57].
16 chapter 1

where * is the elongation at time t + . The denominator of Eq. (7) can also be
written as

dm ds = m(t + + 0;1) s(t + + 0;1) m(t + ) + s(t + )


= (t + + 0;1) (t + )
= d.

Since the mean elongation is no longer 0, the lunar longitude has to be


computed according to Ptolemys complete lunar model rather than the simple
lunar model, both of which are represented in the Alfonsine Tables.11
The variation in elongation in a minute of a day, d, is very nearly the
instantaneous velocity in elongation at the time of true syzygy. Analogously,
dm represents the instantaneous velocity in lunar longitude in a minute of a
day.
To illustrate the method described in the canons to the Alfonsine Tables
by John of Saxony, we show in Table i some basic magnitudes (in degrees)
involved in the computation of the true conjunction corresponding to the
mean conjunction of 20 July 1327, 3;58,10 h after noon.12 True conjunction is
estimated to take place at time t + + * = 3;58,10 h 8;43,30 h + 0;9,49 h =
19;24,29 h (19 July 1327) in Toledo.

A Nicholaus among All the Johns

The Latin tradition of the Alfonsine Tables was mainly developed by the group
of astronomers who may be called the Johns (i.e., John of Murs, John of
Lignres, and John of Saxony), and also by such lesser known figures as John
of Genoa and John of Montfort who computed tables for lunar velocities.13

11 Ptolemys table for his complete lunar model appears in Almagest v.8 [Toomer 1984, 238].
The corresponding table (with modified parameters) appears in the Alfonsine Tables
[Ratdolt 1483, fol. e4re6v; cf. Poulle 1984, 148153].
12 Poulle [1984, 214218] presents a worked example according to John of Saxonys canons for
this conjunction. We have recomputed all the magnitudes involved (using the tables, as
he did, in the 1483 edition of the Alfonsine Tables) and our results differ somewhat from
his. For example, we compute the equation for the centrum lune according to John of
Saxony and the appropriate tables in [Poulle 1984] at time t + as 1;18,46 whereas he
found it to be 1;13,47. Despite the discrepancies, the resulting time of true conjunction
is essentially the same in both computations.
13 Nothing is known about John of Montfort except that, ca. 1332, he produced a table for
nicholaus de heybech and his table for finding true syzygy 17

table i A worked example for the mean conjunction of 20 July 1327


according to the instructions by John of Saxony

t = 3;58,10 h t+ t + + *
(mean conj.) ( = 8;43,30 h) t + + 0;1 d (* = +0;9,49 h)

s 2, 4;53,30 2, 4;32, 0 2, 4;32,59 2, 4;32,24,13


s 2, 3;40,35 2, 3;19,45 2, 3;20,42 2, 3;20, 8, 2
35;25, 4 35; 3,34 35; 4,34 35; 3,57
m 2, 4;53,30 2, 0; 6, 5 2, 0;19,16 2, 0;11,28,50
m 2, 8;26,14 2, 3;13,15 2, 3;28,37 2, 3;20, 7,57
3,42;26, 7 3,37;41, 8 3,37;54,12 3,37;46,28,42
3,42;26, 7 3,36;22,22 3,36;39, 6 3,36;29,12, 2
2 0 5,51; 8,10 5,51;32,33 5,51;18, 8,55
4;45,39 0; 5,29 0; 7,55 0; 0, 0, 5

Despite the differences of their tables in detail, all the Johns required two
kinds of tables for the determination of the time of true syzygy: tables of
correction and tables of velocity for each of the luminaries. However, some
manuscripts within the Alfonsine corpus contain a single table to solve this
problem where all entries are given in units of time. The author of that table,
according to its heading, is Nicholaus de Heybech of Erfurt, and this table is
usually accompanied by a short canon explaining its use. This canon adds that,
after the time of true syzygy is determined, a correction is required to take
into account the equation of time.14 Both the canon and the table survive in

solar and lunar velocities. John of Genoa was the author of a set of canons and tables for
eclipses (also dated 1332), and a computation for the solar eclipse of March 1337 according
to the Alfonsine Tables [cf. Thorndike and Kibre 1963, Cols. 51, 61, and 1690]. The tables for
lunar velocity by John of Montfort and by John of Genoa are analyzed in Goldstein [1992].
14 Nicholaus de Heybech refers to a table for the equation of time in the canon (see Appen-
dix i [18]). His allusion to a table for the equation of time is simply copied from John of
Saxonys canon (Chap. 22). This table appears in the 1483 edition of the Alfonsine Tables
(fol. k1rk2r) and has the heading Tabula elevationum signorum in circulo directo. It is
arranged for signs of 30, and each entry for the equation of time lies under each sign, as
it is described both by John of Saxony and Nicholaus de Heybech in Appendix 1: [18]. Note
that this table was taken from al-Battn [cf. Nallino 1907, 6164] and already appeared in
the Toledan Tables [cf. Toomer 1968, 3435], and that Poulle [1984, 6, 222] excluded it from
the Alfonsine Tables.
18 chapter 1

figure 1.3 Graphical representations of the entries in the five columns of Nicholaus de
Heybechs Table at intervals of 6

several Latin manuscripts, often together with the Alfonsine Tables and Jacob
ben David Bonjorns tables for syzygies.
Nicholaus de Heybechs canon and table are transcribed in Appendix i,
based on the following manuscripts:

B Basel, Universittsbibliothek f.ii.7, fol. 36r37v (table and canon);


D Dijon, bm 447, fol. 62rv (canon);
P Paris, bn Lat. 7287, fol. 72r73r (table), fol. 86va87ra (canon);
P Paris, bn Lat. 7290a, fol. 103r104r (table).

Thorndike and Kibre [1963, 1390, 1478, 1562] listed the first three as well as
three other manuscripts: Bern 454, Cues 211, and Vat. Pal. 1376. Most of the
information seems to have been taken from E. Zinner [1925]. This table seems
to have had a wide diffusion, for it is extant in quite a number of manuscripts
in addition to those already mentioned: Cracow, Jagiellonian Library, mss 609
(table and canon), 610 (table and canon), 613 (canon), 1852, and 1865; Princeton,
Library of Grenville Kane, ms 51; Vienna, Nationalbibliothek 2440 (table and
canon) and 5227 (table and canon).
The argument of the table is arranged at 1 intervals from 0 to 180 using
signs of 60. Column i has the heading equatio solis and the entries are
displayed in hours and minutes; column ii gives the diversitas equationis solis
in hours and minutes; column iii lists the minuta proportionalia; column iv
displays the equatio lune in hours and minutes; and column v gives the
diversitas equationis lune in minutes of an hour. Figure 1.3 illustrates the five
columns at 6 intervals.
nicholaus de heybech and his table for finding true syzygy 19

The structure of the table suggests that the time interval between mean and
true syzygy given by Eq. (2) is here distributed into two terms, each of which
accounts for the role of one luminary:

(8)

The calculation of the first, or solar term (ts), involves columns i, ii, and
iii, and that of the second, or lunar term (tm), involves columns iii, iv, and
v. The strategy consists in treating each term separately, and within each
term computing a set of minimum and maximum values and then to use an
interpolation scheme for intermediate values.
The solar term can be approximated as

(9)

where vs is the mean solar velocity (/h) and c3() is a coefficient of inter-
polation ranging from 0 to 1, depending on the mean lunar anomaly, , and
tabulated in col. iii in minutes. When = 0, the Moon reaches its minimum
velocity, min(vm), and c3() = 0. Hence Eq. (9) reduces to

(10)

The values computed by means of Eq. (10) correspond to those tabulated in


Col. i. Analogously, when = 180, the Moon reaches its maximum velocity,
max(vm), and c3() = 1. In this case, Eq. (9) reduces to

(11)

The values computed by means of the term in brackets in Eq. (9) correspond
to those tabulated in Col. ii. Let c1 represent an entry in Col. i and c2 an entry
in Col. ii of Nicholaus de Heybechs table: c1 and c2 are both functions of the
mean solar anomaly, . Then

(12) ts = c1() c2() c3().

Similarly, the lunar term in Eq. (8) can be approximated as

(13)
20 chapter 1

where c3() is a coefficient of interpolation ranging from 0 to 1, depending on


the mean solar anomaly, . Column iii serves as the interpolation scheme for
both the Sun and the Moon. When = 0, the Sun is at its minimum velocity,
min(vs), and c3() = 0. Hence Eq. (13) reduces to

(14)

The values computed by means of Eq. (14) correspond to those tabulated in


Col. iv. Analogously, when = 180, the Sun is at its maximum velocity, max(vs),
and = 1. In this case, Eq. (13) reduces to

(15)

The values computed by means of the term in brackets in Eq. (13) correspond
to those tabulated in Col. v. Let c4 represent an entry in Col. iv and c5 an entry
in Col. v of Nicholaus de Heybechs table: c4 and c5 are both functions of the
mean lunar anomaly . Then

(16) tm = c4() c5() c3().

Finally, from Eqs. (12) and (16) we find that the time from mean to true syzygy
is

(17)

In this equation the algebraic signs of c1() and c2() are the same as the
corresponding sign of c3(), and those of c4() and c5() are the same as that of
cm().
In the canon by Nicholaus de Heybech we are first told (sentences [5][8])
to compute the six quantities in Eq. (17); note that the sign convention for c4()
and c5() differs from the one we have used. The rule for computing the sec-
ond term is found in sentence [9], and sentence [10] tells us to compute the
difference between the first and the second terms, that is, the solar term, ts. In
sentence [11] we are told to compute the fourth term, and sentence [12] tells us
to compute the lunar term, that is, tm (according to our convention). In sen-
tences [13] and [14] we are given the rules for adding the solar and lunar terms.
For the example given above (the mean conjunction of 20 July 1327), we find

c1 = 2;41 h c2 = 0;35 h c3() = 0;06,


c4 = +6;24 h c5 = +0;02 h c3() = 0;53,
nicholaus de heybech and his table for finding true syzygy 21

using as arguments = 35;25,4 and = 3,42;26,7. The resulting time interval t


is 8;34 h, whereas the value according to John of Saxonys method is 8;33,41
h.
It is noteworthy that with the table of Nicholaus de Heybech essentially
the same result is obtained with much less work; we also note that his entire
procedure is in units of time restricted to two sexagesimal places. In contrast,
John of Saxonys procedure involves the computation of many intermediate
quantities, most of which are not in units of time, and he does not specify the
degree of precision of these quantities needed to assure the minutes of time in
the ultimate result.
Equation (17) incorporates the instructions in the canon of Nicholaus de
Heybechs table, and we have seen that, at least for the example presented
above, the entries are reasonable. As was frequently the custom among medie-
val astronomers, Nicholaus de Heybech does not describe his method for
computing the entries in his table. We now present our reconstruction of his
method for computing the table.
To recompute the entries in column i, we need to fix a set of values in Eq. (10).
For the solar correction we use the values in the 1483 edition of the Alfonsine
Tables; we consider min(vm) = 0;29,37,13 /h and vs = 0;2,27,51 /h.
The first parameter is the minimum lunar velocity found in John of Genoas
table, arranged at intervals of 6 and given to 3 sexagesimal places.15 The value
for the mean solar velocity vs is derived from the table for mean solar motion
in the editio princeps of the Alfonsine Tables.
To recompute the entries in Col. ii, we also have to fix the value for the
maximum lunar velocity in Eq. (11): max(vm) = 0;36,58,54 /h, which is found in
the same table by John de Genoa mentioned above.16 The values in Col. ii are
computed as the difference between those in Col. i and those found by means
of Eq. (11) with the same argument. These parameters yield the entries in Col. i
and Col. ii quite closely (see Table ii). It should be stressed that, restricting our
attention to attested parameters, we obtain best agreement with the entries
in column ii using John of Genoas velocity table which contains an isolated

15 The Alfonsine corpus has many variant tables for lunar velocity (see Goldstein [1992]) and
there is no reason to assume that Nicholaus de Heybech would have used the version in
the editio princeps of the Alfonsine Tables in which the minimum value is 0;30,21/h.
16 We use this value for the lunar velocity in John of Genoas table which appears in many
manuscripts, and hence it is textually secure. Nevertheless, it may have been miscomputed
according to an analysis of the entire table: see Goldstein [1992]. Other values for the max-
imum lunar velocity in the Alfonsine corpus are: 0;36,4/h [cf. Goldstein 1980], 0;36,25/h
[Ratdolt 1483, fol. g7r] and 0;36,53/h, ascribed to John of Lignres [cf. Goldstein 1992].
22 chapter 1

error for the maximum entry, 0;36,58,54 /h (instead of 0;36,53,20 /h which


is consistent with the rest of the table). The fact that this value is so much
greater than any other strongly supports the claim that Nicholaus de Heybech
indeed depended on this table by John of Genoa. As we shall see, the entries
in column iv in Nicholaus de Heybechs table also agree with recomputation
based on this table by John of Genoa.
The recomputation of columns i and ii of Nicholaus de Heybechs table is
only displayed here for values of the argument in multiples of 6, for it is likely
that he derived the other entries by distributing them uniformly between the
values calculated at 6 intervals. For instance, in Col. i after finding 28 min for
= 6, one way to proceed would be to distribute 28 into six parts according
to the sequence 555544 (leading to the values 5, 10, 15, 20, 24, and 28,
that actually appear in Nicholaus de Heybechs table). After finding 57 min for
= 12, one can distribute 57 28 = 29 into 6 parts according to the sequence
555554 (leading to the values in Nicholaus de Heybechs table: 33, 38, 43,
48, 53, and 57). These sequences generate smooth curves within each interval
of 6, but they are no longer smooth when considered as a whole.
The results are shown in Table ii, where c is the recomputed value and tc
is its difference from the value given in the text.
We now argue that the coefficient c3, tabulated in column iii (minuta
proportionalia), has been recomputed according to the formula

(18)

where, in a simple eccentric model, d = 60 + e (the distance of the luminary


at apogee), d = 60 e (its distance at perigee), d() is its distance for a true
anomaly, , and e is the eccentricity.17 In a simple eccentric model, this inter-
polation coefficient is a function of and e, but, as is shown in Table iii, c3 does
not depend strongly on the eccentricity, for similar results are obtained using
2;16 (the solar eccentricity corresponding to the maximum solar correction in
the Alfonsine Tables: 2;10) or 5;10 (the lunar eccentricity corresponding to the
maximum lunar correction in these tables: 4;56). This justifies the use of the
same interpolation coefficient in both the solar and lunar terms of t (Eqs. 9
and 13). Note also that while the entries in Col. iii have been computed with the
true anomaly as argument, the instructions by Nicholaus de Heybech indicate

17 The argument for the distance is here the true anomaly in an eccentric model, whereas
in the Almagest vi.8 [cf. Toomer 1984, 308] Ptolemy presents a similar interpolation table
based on the mean anomaly in an eccentric model [cf. Toomer 1984, 654].
nicholaus de heybech and his table for finding true syzygy 23

table ii A recomputation of columns i and ii of


Nicholaus de Heybechs table

Arg. i c tc ii c tc
h h min h h min

6 0;28 0;28,28 0 0; 6 0; 6, 4 0
12 0;57 0;56,54 0 0;12 0;12, 9 0
18 1;25 1;25, 4 0 0;18 0;18, 8 0
24 1;53 1;52,50 0 0;24 0;24, 4 0
30 2;19 2;18,58 0 0;30 0;29,38 0
36 2;43 2;43,25 0 0;35 0;34,50 0
42 3; 6 3; 5,58 0 0;40 0;39,40 0
48 3;27 3;27,39 1 0;44 0;44,17 0
54 3;47 3;47,21 0 0;48 0;48,29 0
60 4; 4 4; 4,40 1 0;51 0;52,11 1
66 4;18 4;18,55 1 0;54 0;55,13 1
72 4;29 4;29,38 1 0;57 0;57,30 1
78 4;38 4;37,57 0 0;59 0;59,17 0
84 4;44 4;43,48 0 1; 1 1; 0,31 0
90 4;47 4;47, 7 0 1; 1 1; 1,14 0
96 4;47 4;46,54 0 1; 1 1; 1,11 0
102 4;43 4;43, 2 0 1; 0 1; 0,22 0
108 4;37 4;36,51 0 0;59 0;59, 3 0
114 4;27 4;27, 1 0 0;57 0;56,57 0
120 4;14 4;13,58 0 0;54 0;54, 9 0
126 3;58 3;57,53 0 0;50 0;50,44 1
132 3;39 3;38,37 0 0;46 0;46,37 1
138 3;18 3;18, 0 0 0;42 0;42,13 0
144 2;55 2;54,13 +1 0;38 0;37, 9 +1
150 2;29 2;28,17 +1 0;32 0;31,37 0
156 2; 1 2; 0,14 +1 0;26 0;25,39 0
162 1;32 1;31,52 0 0;20 0;19,35 0
168 1; 3 1; 2,34 0 0;14 0;13,21 +1
174 0;32 0;31,38 0 0; 7 0; 6,45 0
180 0; 0 0 0 0; 0 0 0
24 chapter 1

table iii A recomputation of column iii of


Nicholaus de Heybechs table

Arg. iii tc for e = 2;16 tc for e = 5;10


min min min

10 1 0 0
20 2 0 0
30 4 0 0
40 7 0 1
50 11 0 1
60 15 0 1
70 20 0 1
80 25 0 1
90 31 0 0
100 36 0 1
110 42 1 +1
120 47 2 +1
130 51 1 +1
140 54 1 0
150 56 0 0
160 58 0 0
170 60 0 0
180 60 0 0

that the argument for Col. iii is the mean anomaly. Indeed, these instructions
do not suggest that the true anomalies need be computed. In any event, the
differences between entries computed with true, rather that mean, anomaly
are small (less than 3 min in the worst case), and so there is little practical effect
on the final result.
For our recomputation of Cols. iv and v, we have taken the values of the lunar
correction cm() from the 1483 edition of the Alfonsine Tables, vm() from John
of Genoas lunar velocity table, and from his solar velocity table the following
values for the maximum and minimum velocities: max(vs) = 0;2,33,40 /h and
min(vs) = 0;2,22,30 /h.18

18 See, for example, Paris BnF Lat. 7282, fol. 129rv [cf. Poulle 1984, 210].
nicholaus de heybech and his table for finding true syzygy 25

These parameters are used to evaluate the time given by Eq. (14) which
corresponds to Col. iv and the difference between it and that resulting from
Eq. (15) with the same argument. It would seem that Nicholaus de Heybech
increased the argument of lunar anomaly by 6 in calculating the lunar velocity
in between mean syzygy and true syzygy. Hence Eq. (14) becomes

(19)

and Eq. (15) is modified analogously:

(20)

Again, the recomputation is only displayed for the values of the argument
in multiples of 6. The results are shown in Table iv. The recomputed value
according to Eq. (19) is c, and t c is its difference from the value given in the
text. The shift of 6 significantly diminishes the discrepancy between the text
and the recomputed values, with the result that a value in the text is always
within 1 min of time of the recomputed value. We note again that John of
Genoas table for lunar velocity is arranged at intervals of 6 so that a shift of 6
is a shift of a line in the table.
Column v is recomputed as the difference between the time found in col-
umn iv and the time found by means of Eq. (20) with the same argument.
Table v displays Col. v together with our recomputed values (the differences
between the times found by means of Eq. (19) and Eq. (20)). Since the entries
in Col. v are small and have a very limited range, the rounding procedure plays
a more significant role than in other columns. For this reason we have not dis-
played the differences between text and recomputation.

Conclusion

Our principal goal in studying the canon and the table of Nicholaus de Heybech
of Erfurt was to understand their use in finding the time from mean syzygy to
true syzygy, as well as to reconstruct his table. It turned out that it works very
well for its intended purpose and is quite easy to use.
When we review some of the different approaches prior to that of Heybech,
we see that Ptolemy had already solved the problem by means of an iterative
procedure without recourse to any tables. On the other hand, John of Saxonys
method took advantage of solar and lunar velocity tables, while also intro-
ducing what we have called a differential method for computing an instan-
26 chapter 1

table iv A recomputation of column iv of Nicholaus de


Heybechs table

cm() +6 vm(+6) iv c tc
/h h h min

6 0;28,28 12 0;29,42, 6 1; 3 1; 2,56 0


12 0;56,41 18 0;29,46,17 2; 5 2; 4,59 0
18 1;24,27 24 0;29,51,51 3; 5 3; 5,35 1
24 1;51,27 30 0;29,59,31 4; 3 4; 3,47 1
30 2;17,29 36 0;30, 8,34 4;59 4;59, 5 0
36 2;42,21 42 0;30,19,43 5;50 5;50,48 1
42 3; 5,46 48 0;30,32,15 6;38 6;38,24 0
48 3;27,30 54 0;30,46,53 7;21 7;21,10 0
54 3;47,20 60 0;31, 2,55 7;59 7;58,48 0
60 4; 5, 4 66 0;31,18,55 8;31 8;31,22 0
66 4;20,27 72 0;31,37,45 8;58 8;57,36 0
72 4;33,18 78 0;31,57,15 9;18 9;17,53 0
78 4;43,28 84 0;32,16,45 9;32 9;32,19 0
84 4;50,41 90 0;32,39,45 9;40 9;39,25 +1
90 4;54,54 96 0;33, 8,31 9;40 9;38,36 +1
96 4;55,56 102 0;33,30,36 9;34 9;33,37 0
102 4;53,38 108 0;33,49,21 9;23 9;23,34 1
108 4;48,10 114 0;34,10,19 9; 7 9; 6,58 0
114 4;39,15 120 0;34,35,24 8;44 8;43, 7 +1
120 4;27, 0 126 0;34,58,23 8;15 8;14,16 +1
126 4;11,23 132 0;35,18,25 7;41 7;40,39 0
132 3;52,47 138 0;35,38,47 7; 3 7; 2, 9 +1
138 3;31, 3 144 0;35,57,36 6;20 6;19, 9 +1
144 3; 6,35 150 0;36,13,37 5;33 5;32,32 0
150 2;39,35 156 0;36,28,15 4;42 4;42,22 0
156 2;10,26 162 0;36,38, 0 3;49 3;49,41 1
162 1;39,27 168 0;36,46,22 2;54 2;54,25 0
168 1; 7, 6 174 0;36,53,15 1;57 1;57,17 0
174 0;33,47 180 0;36,58,54 0;59 0;58,53 0
180 0 0; 0 0 0
nicholaus de heybech and his table for finding true syzygy 27

table v A recomputation of column v


of Nicholaus de Heybechs
table

v c Eq.(19) Eq.(20)
min min h h

6 0 0;26 = 1; 2,56 1; 2,30


12 0 0;50 = 2; 4,59 2; 4, 9
18 1 1;15 = 3; 5,35 3; 4,20
24 2 1;39 = 4; 3,47 4; 2, 8
30 2 2; 1 = 4;59, 5 4;57, 4
36 3 2;20 = 5;50,48 5;48,28
42 3 2;38 = 6;38,24 6;35,46
48 3 2;53 = 7;21,10 7;18,17
54 3 3; 6 = 7;58,48 7;55,42
60 3 3;17 = 8;31,22 8;28, 5
66 4 3;25 = 8;57,36 8;54,11
72 4 3;29 = 9;17,53 9;14,24
78 4 3;34 = 9;32,19 9;28,45
84 4 3;34 = 9;39,25 9;35,51
90 4 3;30 = 9;38,36 9;35, 6
96 4 3;20 = 9;33,37 9;30,17
102 4 3;20 = 9;23,34 9;20,14
108 4 3;12 = 9; 6,58 9; 3,46
114 3 3; 1 = 8;43, 7 8;40, 6
120 3 2;50 = 8;14,16 8;11,26
126 3 2;35 = 7;40,39 7;38, 1
132 2 2;22 = 7; 2, 9 6;59,47
138 2 2; 6 = 6;19, 9 6;17, 3
144 2 1;50 = 5;32,32 5;30,42
150 2 1;33 = 4;42,22 4;40,49
156 2 1;15 = 3;49,41 3;48,26
162 1 0;57 = 2;54,25 2;53,28
168 1 0;38 = 1;57,17 1;56,39
174 0 0;19 = 0;58,53 0;58,34
180 0 0 =0 0
28 chapter 1

taneous velocity in elongation at a time very close to true syzygy. To be sure, a


modern definition of a differential would require the evaluation of a function at
two moments separated by an infinitesimal time interval. A value for the veloc-
ity very close to that obtained by the modern definition can be computed by
taking a sufficiently small time interval, and John of Saxony set that interval as
0;1 days.
Less than a century after John of Saxony presented his method, Nicholaus de
Heybech offered a much simpler alternative without sacrificing the accuracy of
the previous methods. First, instead of requiring many steps of computation,
some of which called for tables while other did not, Heybech presented his
solution in the form of a single table whose entries are given in time and whose
arguments are the mean positions of the Sun and the Moon at mean syzygy,
the initial values for all methods of computation. Second, the use of the table
involves very simple arithmetic operations in contrast to the complexity of John
of Saxonys method. Nicholaus de Heybech has produced a user-friendly table
for which he deserves much credit. In this respect he is following Ptolemy who,
for example, established tables for finding planetary positions that replaced
complex trigonometric computations.
Both John of Saxony and Nicholaus de Heybech showed real insight into
the ways Ptolemys models work and, on this basis, they were able to facilitate
the computation of certain astronomical phenomena. This example shows that
astronomers in the Middle Ages could make significant contributions without
introducing any modification of the models for the motions of the celestial
bodies.
Finally, we wish to emphasize that it is inappropriate to define the corpus of
Alfonsine Tables by the tables as they appear in the editio princeps, or by the
canons of John of Saxony, or even to assign either of them any privileged status.
Indeed, it has long been known that there is much more in the vast number
of manuscripts within this corpus, but little attention has been paid to their
contents. In this case study we have indicated the significance of Nicholaus de
Heybechs table and its dependence on a lunar velocity table by John of Genoa,
neither of which is included in the restricted definition of the Alfonsine Tables.
It is reasonable to expect that similar studies of other parts of this unexplored
corpus would reveal hitherto unsuspected riches.
nicholaus de heybech and his table for finding true syzygy 29

Appendix i

a Text in Latin
The text presented here is a transcription of Dijon bm 447, fol. 62rv (D). Variant
readings from other manuscripts are noted: Basel f.11.7, fol. 37v (B) and Paris
BnF Lat. 7287, fol. 86va87ra (P). Sentence numbers and punctuation have been
added by the editors.

Canon super tabulas magistri nicholay de heybech

[1] Tempus uere coniunctionis et oppositionis solis et lune per tabulas a mag-
istro nicholao de heybech de erfordia compositas inuenire. [2] Scias tem-
pus medie coniunctionis si uolueris coniunctionem ueram, uel tempus medie
oppositionis si uolueris oppositionem ueram. [3] Scias eciam argumentum
solis et argumentum lune ad idem tempus. [4] Quibus scitis et habitis intra
primo cum argumento solis in tabulam equationis temporis uere coniunctio-
nis et oppositionis luminarium. [5] Et accipe equationem solis et scribe super
eam m si argumentum solis sit minus tribus signis phisicis, uel scribe a si sit
plus tribus signis. [6] Accipe eciam ibidem diuersitatem equationis solis atque
minuta proportionalia et scribe super ea solis. [7] Quibus habitis et seruatis
intra postea cum argumento lune in easdem tabulas in lineis numeri et accipe
equationem lune in directo existentem et scribe super am a si argumentum
sit minus tribus signis, uel m si sit plus. [8] Accipe eciam ibidem diuersi-
tatem equationis lune et minuta proportionalia et scribe super ea lune. [9]
Istis habitis et seruatis accipe partem proportionalem de diuersitate equatio-
nis solis secundum proportionem minutorum lune ad 60. [10] Quam partem
proportionalem minue ab equatione solis prius seruata et habebis eam bene
equatam. [11] Accipe similiter partem proportionalem de diuersitate equatio-
nis lune secundum proportionem minutorum proportionalium solis ad 60. [12]
Quam partem proportionalem adde super equationem lune prius inuentam
et habebis eam bene equatam. [13] Postea uide si super ambas equationes,
scilicet solis et lune, scriptum sita; tunc adde eas simul cum tempore medie
coniunctionis prius inuento et seruato; sed si super utramque equationem
scriptum fuerit m, tunc minue eas similiter ab eodem tempore. [14] Si uero
super unam scriptum fuerit a et super aliam m, tunc minue minorem equa-
tionem a maiori et residuum adde cum tempore predicto si super maiorem
scriptum fuerit a, uel minue si super maiorem scriptum fuerit m. [15] Et sic
habebis ueram coniunctionem solis et lune si operatus es ad coniunctionem,
et similiter ueram oppositionem si operatus es ad eam reperiendam, et hoc
diebus non equatis, ad meridianum loci ad quem predicta inquirebas. [16] Et
30 chapter 1

ad istud tempus debent queri loca planetarum. [17] Si gradum ascendentem


scire uolueris, oportet equare dies. [18] Intra ergo cum uero loco solis in tab-
ulam equationis dierum cum suis noctibus et inuenies in directo graduum
equationem dierum, scilicet sub signo illo in quo est sol. [19] Et si inueneris
ibi gradus et minuta accipe pro quolibet gradu quatuor minuta hore et pro
quolibet minuto quatuor 2a hore, que adde cum tempore predicto quodlibet
ad suum genus et proueniet tempus uere coniunctionis uel oppositionis diebus
equatis. [20] Et cum illo debet queri ascendens et reliqua que pertinent ad fig-
uram.

Title: canon heybech] om. B; post tabulas add. P coniunctionis solis et lune.
[1] per tabulas compositas] om. B.
[3] eciam] igitur P; idem] iddem B.
[5] sit] om. P.
[7] in easdem] easdem B D P.
[9] post minutorum add. B proportionalium. [13] inuento et] om. B; similiter] simul B.
[14] equationem] om. P; maiorem 1 ipsam B et add. maiorem s. l.; maiorem 2] ipsam B.
[17] si] si uero B.
[18] ergo] igitur B; graduum] graduum ascendentem P; equationem] equationum B.
[20] post figuram add. D et sic est finis secundum uasseun, add. P Et sic finis scripti Maruanti per
me Mertinj anno domini 1447 12 die Julij.

b The Table of Nicholaus de Heybech of Erfurt for Finding the Time of


True Syzygy
The table presented here is a transcription of Basel f.ll.7, fol. 36r37r (B). Variant
readings from two other manuscripts are noted: Paris BnF Lat. 7287, fol. 72r73r
(P) and BnF Lat. 7290a, fol. 103r104r (P). Column iii is found before Col. ii in
P and P.
mss P and P share 28 errors (23 entries of Col. iv are shifted upward one
place); P exhibits 5 additional errors and P just one. Therefore, we suggest that
P was copied from P.

Title

B lncipit tabula equationum temporis uere coniunctionis et oppositionis


solis et lune ordinata per Nycholaum dictum de Heybech de Erfordia
(f. 36r)
P Tabula equationum uere coniunctionis et oppositionis solis et lune (signs
are used for the last four words) secundum magistrum Nycholaum de
Heybech de Erfordia (f. 72v)
nicholaus de heybech and his table for finding true syzygy 31

P Tabula equationum uere coniunctionis et oppositionis solis et lune se-


cundum magistrum Nycholaum de Heybech de Erfordia (f. 103r)

Column Headings

Argument Linee numeri


i Equatio solis
ii Diversitas equationis solis
iii Minuta proportionalia
iv Equatio lune
v Diversitas equationis lune

Argument i ii iii iv v
s /s h min h min min h min min

0 1/5 59 0 5 0 1 0 0 11 0
0 2/5 58 0 10 0 2 0 0 22 0
0 3/5 57 0 15 0 3 0 0 33 0
0 4/5 56 0 20 0 4 0 0 43 0
0 5/5 55 0 24 0 5 0 0 53 0
0 6/5 54 0 28 0 6 0 1 3 0
0 7/5 53 0 33 0 7 0 1 14 0
0 8/5 52 0 38 0 8 0 1 24 0
0 9/5 51 0 43 0 9 1 1 34 0
0 10/5 50 0 48 0 10 1 1 45 0
0 11/5 49 0 53 0 11 1 1 55 0
0 12/5 48 0 57 0 12 1 2 5 0
0 13/5 47 1 2 0 13 1 2 15 0
0 14/5 46 1 7 0 14 1 2 25 1
0 15/5 45 1 12 0 15 2 2 35 1
0 16/5 44 1 17 0 16 2 2 45 1
0 17/5 43 1 21 0 17 2 2 55 1
0 18/5 42 1 25 0 18 2 3 5 1
0 19/5 41 1 30 0 19 2 3 14 1
0 20/5 40 1 35 0 20 2 3 24 1
0 21/5 39 1 40 0 21 3 3 34 1
0 22/5 38 1 44 0 22 3 3 43 2 P, P: 1
0 23/5 37 1 49 0 23 3 3 53 2
0 24/5 36 1 53 0 24 3 4 3 2
32 chapter 1

(cont.)

Argument i ii iii iv v
s /s h min h min min h min min

0 25/5 35 1 58 0 25 3 4 12 2
0 26/5 34 2 3 0 26 4 4 22 2
0 27/5 33 2 7 0 27 4 4 31 2
0 28/5 32 2 11 0 28 4 4 41 2
0 29/5 31 2 15 0 29 4 4 50 2
0 30/5 30 2 19 0 30 4 4 59 2

0 31/5 29 2 23 0 31 5 5 8 2
0 32/5 28 2 27 0 32 5 5 17 2
0 33/5 27 2 31 0 33 5 5 26 2
0 34/5 26 2 35 0 34 5 5 34 3
0 35/5 25 2 39 0 35 6 5 42 3
0 36/5 24 2 43 0 35 6 5 50 3
0 37/5 23 2 47 0 36 6 5 58 3
0 38/5 22 2 51 0 37 7 6 6 3
0 39/5 21 2 55 0 38 7 6 14 3
0 40/5 20 2 59 0 39 7 6 22 3
0 41/5 19 3 3 0 40 8 6 30 3
0 42/5 18 3 6 0 40 8 6 38 3
0 43/5 17 3 10 0 41 8 6 46 3
0 44/5 16 3 14 0 42 9 6 53 3
0 45/5 15 3 17 0 42 9 7 0 3
0 46/5 14 3 20 0 43 9 7 7 3
0 47/5 13 3 24 0 44 10 7 14 3
0 48/5 12 3 27 0 44 10 7 21 3
0 49/5 11 3 31 0 45 10 7 28 3
0 50/5 10 3 35 0 46 11 7 35 3
0 51/5 9 3 38 0 46 11 7 41 3
0 52/5 8 3 41 0 47 11 7 47 3
0 53/5 7 3 44 0 48 12 7 53 3
0 54/5 6 3 47 0 48 12 7 59 3
0 55/5 5 3 50 0 49 12 8 5 3
0 56/5 4 3 53 0 49 13 8 11 3
0 57/5 3 3 56 0 50 13 8 16 3
0 58/5 2 3 59 0 50 14 8 21 3
nicholaus de heybech and his table for finding true syzygy 33

Argument i ii iii iv v
s /s h min h min min h min min

0 59/5 1 4 2 0 51 14 8 26 3
1 0/5 0 4 4 0 51 15 8 31 3

1 1/4 59 4 6 0 52 15 8 36 3
1 2/4 58 4 9 0 52 16 8 41 3
1 3/4 57 4 11 0 52 16 8 46 4
1 4/4 56 4 14 0 53 17 8 50 4
1 5/4 55 4 16 0 53 17 8 54 4
1 6/4 54 4 18 0 54 18 8 58 4
1 7/4 53 4 20 0 54 18 9 2 4
1 8/4 52 4 22 0 55 19 9 6 4
1 9/4 51 4 24 0 56 19 9 9 4 P: 55
1 10/4 50 4 26 0 56 20 9 12 4
1 11/4 49 4 28 0 57 20 9 15 4 P: 56
1 12/4 48 4 29 0 57 21 9 18 4
1 13/4 47 4 31 0 57 21 9 21 4
1 14/4 46 4 32 0 58 22 9 24 4
1 15/4 45 4 34 0 58 22 9 26 4
1 16/4 44 4 35 0 58 23 9 28 4
1 17/4 43 4 37 0 59 23 9 30 4 P: 24
1 18/4 42 4 38 0 59 24 9 32 4
1 19/4 41 4 39 0 59 24 9 34 4
1 20/4 40 4 40 0 59 25 9 36 4
1 21/4 39 4 41 1 0 26 9 37 4
1 22/4 38 4 42 1 0 26 9 38 4
1 23/4 37 4 43 1 0 27 9 39 4
1 24/4 36 4 44 1 1 28 9 40 4
1 25/4 35 4 45 1 1 28 9 40 4
1 26/4 34 4 45 1 1 29 9 40 4
1 27/4 33 4 46 1 1 30 9 40 4
1 28/4 32 4 46 1 1 30 9 40 4
1 29/4 31 4 47 1 1 31 9 40 4
1 30/4 30 4 47 1 1 31 9 40 4

1 31/4 29 4 47 1 1 32 9 39 4
1 32/4 28 4 47 1 1 32 9 38 4
34 chapter 1

(cont.)

Argument i ii iii iv v
s /s h min h min min h min min

1 33/4 27 4 47 1 1 33 9 37 4
1 34/4 26 4 47 1 1 33 9 36 4
1 35/4 25 4 47 1 1 34 9 35 4
1 36/4 24 4 47 1 1 34 9 34 4
1 37/4 23 4 46 1 1 35 9 33 4
1 38/4 22 4 46 1 1 35 9 31 4
1 39/4 21 4 45 1 1 36 9 29 4
1 40/4 20 4 45 1 0 36 9 27 4
1 41/4 19 4 44 1 0 37 9 25 4
1 42/4 18 4 43 1 0 37 9 23 4
1 43/4 17 4 42 1 0 38 9 21 4
1 44/4 16 4 41 1 0 38 9 19 4
1 45/4 15 4 40 0 59 39 9 16 4
1 46/4 14 4 39 0 59 39 9 13 4
1 47/4 13 4 38 0 59 40 9 10 4
1 48/4 12 4 37 0 59 40 9 7 4
1 49/4 11 4 36 0 58 41 9 4 4
1 50/4 10 4 34 0 58 42 9 0 4
1 51/4 9 4 32 0 58 42 8 56 4
1 52/4 8 4 30 0 57 43 8 52 3
1 53/4 7 4 29 0 57 43 8 48 3
1 54/4 6 4 27 0 57 44 8 44 3
1 55/4 5 4 25 0 56 44 8 40 3
1 56/4 4 4 23 0 56 45 8 35 3
1 57/4 3 4 21 0 55 45 8 30 3
1 58/4 2 4 19 0 55 46 8 25 3
1 59/4 1 4 17 0 54 46 8 20 3
2 0/4 0 4 14 0 54 47 8 15 3 P, P: 16

2 1/3 59 4 12 0 53 47 8 10 3
2 2/3 58 4 10 0 52 48 8 4 3
2 3/3 57 4 7 0 52 48 7 59 3
2 4/3 56 4 4 0 51 49 7 53 3
2 5/3 55 4 1 0 50 49 7 47 3 P: 4 58, P: 3 58
2 6/3 54 3 58 0 50 49 7 41 3 P, P: 55
nicholaus de heybech and his table for finding true syzygy 35

Argument i ii iii iv v
s /s h min h min min h min min

2 7/3 53 3 55 0 49 50 7 35 3 P, P: 52
2 8/3 52 3 52 0 48 50 7 29 3 P, P: 49
2 9/3 51 3 49 0 48 50 7 23 3 P, P: 46
2 10/3 50 3 46 0 47 51 7 17 2 P, P: 43
2 11/3 49 3 43 0 46 51 7 10 2 P, P: 39
2 12/3 48 3 39 0 46 51 7 3 2 P, P: 36
2 13/3 47 3 36 0 45 52 6 56 2 P, P: 33
2 14/3 46 3 33 0 45 52 6 49 2 P, P: 29
2 15/3 45 3 29 0 44 52 6 42 2 P, P: 25, P: 0 47
2 16/3 44 3 25 0 43 53 6 35 2 P, P: 22
2 17/3 43 3 22 0 43 53 6 28 2 P, P: 18
2 18/3 42 3 18 0 42 53 6 20 2 P, P: 15
2 19/3 41 3 15 0 41 54 6 13 2 P, P: 11, P: 6 10
2 20/3 40 3 11 0 41 54 6 5 2 P, P: 7
2 21/3 39 3 7 0 40 54 5 57 2 P, P: 3
2 22/3 38 3 3 0 39 54 5 49 2 P, P: 2 59
2 23/3 37 2 59 0 39 55 5 41 2 P, P: 55
2 24/3 36 2 55 0 38 55 5 33 2 P, P: 51
2 25/3 35 2 51 0 37 55 5 25 2 P, P: 47
2 26/3 34 2 47 0 36 55 5 17 2 P, P: 43
2 27/3 33 2 43 0 35 56 5 9 2 P, P: 38
2 28/3 32 2 38 0 34 56 5 0 2 P, P: 34
2 29/3 31 2 34 0 33 56 4 51 2
2 30/3 30 2 29 0 32 56 4 42 2

2 31/3 29 2 24 0 31 56 4 34 2
2 32/3 28 2 19 0 30 57 4 25 2
2 33/3 27 2 15 0 29 57 4 16 2
2 34/3 26 2 10 0 28 57 4 7 2
2 35/3 25 2 6 0 27 57 3 58 2
2 36/3 24 2 1 0 26 57 3 49 2
2 37/3 23 1 57 0 25 58 3 40 2
2 38/3 22 1 52 0 24 58 3 31 1 P: 42
2 39/3 21 1 47 0 23 58 3 22 1
2 40/3 20 1 42 0 22 58 3 13 1
2 41/3 19 1 37 0 21 58 3 3 1
36 chapter 1

(cont.)

Argument i ii iii iv v
s /s h min h min min h min min

2 42/3 18 1 32 0 20 58 2 54 1
2 43/3 17 1 28 0 19 59 2 44 1
2 44/3 16 1 23 0 18 59 2 35 1
2 45/3 15 1 18 0 17 59 2 25 1
2 46/3 14 1 13 0 16 59 2 16 1
2 47/3 13 1 8 0 15 59 2 6 1
2 48/3 12 1 3 0 14 59 1 57 1 P, P: 0
2 49/3 11 0 58 0 12 59 1 47 1 P, P: 0
2 50/3 10 0 53 0 11 60 1 38 0
2 51/3 9 0 48 0 10 60 1 28 0
2 52/3 8 0 43 0 9 60 1 19 0
2 53/3 7 0 38 0 8 60 1 9 0
2 54/3 6 0 32 0 7 60 0 59 0
2 55/3 5 0 27 0 5 60 0 50 0
2 56/3 4 0 22 0 4 60 0 40 0
2 57/3 3 0 17 0 3 60 0 30 0
2 58/3 2 0 12 0 2 60 0 20 0
2 59/3 1 0 6 0 1 60 0 10 0
3 0/3 0 0 0 0 0 60 0 0 0

Appendix ii

Translation
The sentence numbers added to the Latin text have been retained in the
translation; words in square brackets have been added by the editors. Sentences
[16][20] are virtually identical with John of Saxonys canon to the Alfonsine
Tables, Chap. 22 [cf. Poulle 1984, 86, lines 127138].

Canon for the Tables of Master Nicholaus de Heybech

[1] To find the time of true conjunction and opposition of the Sun and the
Moon by means of the tables compiled by Master Nicholaus de Heybech of
Erfurt, [2] you must know the time of mean conjunction if you seek [the time
nicholaus de heybech and his table for finding true syzygy 37

of] true conjunction, or the time of mean opposition if you seek [the time
of] true opposition. [3] You must also know the solar argument and the lunar
argument at that time; [4] and when these are known and obtained, first enter
the table for the equation of the time of true conjunction and opposition of
the luminaries with the solar argument. [5] Take the solar correction [in Col. i]
and write above it m if the solar argument is less than 3 physical signs [i.e.,
180], or write a if it is greater than 3 [physical] signs. [6] Take also in the same
place the difference of the solar correction [in Col. ii] as well as the minutes of
proportion [in Col. iii], and write Sun above them; [7] and when these are
obtained and recorded, then enter the same tables in the column [labeled]
argument [Lat.: linee numeri] with the lunar argument, and take the lunar
correction opposite it [in Col. iv], and write above it a if the [lunar] argument
is less than 3 [physical] signs, or m if it is more. [8] Take also in the same place
the difference of the lunar correction [in Col. v] and the minutes of proportion
[in Col. iii], and write Moon above them; [9] and when these are obtained and
recorded, take the proportional part from the difference of the solar correction
according to the ratio of the minutes of the Moon to 60. [10] Subtract this
proportional part from the solar correction which was recorded previously, and
you will have it well corrected. [11] Similarly, take the proportional part from
the difference of the lunar correction according to the ratio of the minutes
of proportion of the Sun to 60. [12] Add this proportional part to the lunar
correction previously found, and you will have it well corrected. [13] Then,
see if there is an a written above both corrections, the solar and the lunar:
then add together with them the time of mean conjunction previously found
and recorded; but if there has been an m written above both of them, then
subtract them from the same time in the same way. [14] However, if there is an
a written above one of them and an m above the other, then subtract the
smaller correction from the larger and add the remainder together with the
predicted time if there has been an a written above the larger, or subtract [it]
if there has been an m written above the larger one. [15] You will thus obtain
for the meridian of the place for which you were seeking the predictions the
true conjunction of the Sun and the Moon, if you were working to find the
[true] conjunction, and similarly, the true opposition, if you were working to
find it, but without the equation of time [Lat.: diebus non equatis]. [16] And
the positions of the planets must be sought for that particular time. [17] If you
wish to know the degree of the ascendant, it is necessary to equate the days. [18]
Therefore, enter the table for the equation of time [lit.: the equation of the days
with their nights] with the true solar position, and you will find the equation
of time opposite [the number] of degrees, that is, beneath the sign in which
the Sun is. [19] If you find degrees and minutes there, take four minutes of an
38 chapter 1

hour for each degree, and four seconds of an hour for each minute; and add
these together with the predicted time, each to its own rank, and there results
the time of true conjunction or opposition with the equation of time. [20] With
that [time] you must seek the ascendant and the other [magnitudes] relating
to the figure [of the sky for the horoscope].

Acknowledgments

We are grateful to Jos Luis Mancha (Sevilla) for establishing the Latin text of
the canon given in Appendix i, and we are indebted to him for many helpful
remarks. Alan C. Bowen (Princeton) read a draft translation of this canon, and
many of his suggestions have been incorporated in Appendix ii. We have also
benefited from comments by Emmanuel Poulle (Paris) on a previous draft of
this paper.

References

Alfonsine Tables. 1483. Tabulae astronomicae Alfontii regis castelle. Venice: E. Ratdolt.
Chabs, J. 1989. Anlisis del contenido astronmico de las tablas de Jacob ben David
Bonjorn. Barcelona: Ph.D. dissertation, University of Barcelona (u. of b. Microfilm
814).
Chabs, J. 1991. The astronomical tables of Jacob ben David Bonjom. Archive for History
of Exact Sciences 42, 279314.
Goldstein, B.R. 1974. The Astronomical Tables of Levi ben Gerson. Hamden, ct: Archon
Books (Transactions of the Connecticut Academy of Arts and Sciences, Vol. 45).
Goldstein, B.R. 1980. Solar and lunar velocities in the Alfonsine Tables. Historia Mathe-
matica 7, 134140.
Goldstein, B.R. 1992. Lunar velocity in the Ptolemaic tradition. In In the Prime of lnven-
tion: Essays in the History of Mathematics and the Exact Sciences, P. Harman &
A.E. Shapiro, Eds. Cambridge: Cambridge Univ. Press, pp. 317.
Hartmann, J. 1919. Die astronomischen Instrumente des Kardinals Nikolaus Cusanus.
Abhandlungen der kniglichen Gessellschaft der Wissenschaften zu Gttingen, Neue
Folge, 10 (6).
Mills, J.M. 19431950. Estudios sobre Azarquiel. Madrid/Granada: Consejo Superior de
Investigaciones Cientficas.
Nallino, C.A. 1907. Al-Battn sive Albatenii Opus Astronomicum, ii. Milan: Reale Osser-
vatorio di Brera in Milano.
Neugebauer, O. 1969. The Exact Sciences in Antiquity. New York: Dover.
nicholaus de heybech and his table for finding true syzygy 39

Neugebauer, O. 1975. A History of Ancient Mathematical Astronomy. Berlin/New York:


Springer Verlag.
Pedersen, O. 1974. A Survey of the Almagest. Odense: Odense University Press.
Poulle, E. 1984. Les tables alphonsines avec les canons de Jean de Saxe. Paris: Centre
national de la recherche scientifique.
Poulle, E. 1988. The Alfonsine Tables and Alfonso x of Castille. Journal for the History of
Astronomy 19, 97113.
Ratdolt, E. 1483. See Alfonsine Tables.
Thorndike, L. 1948. Nicholaus de Heybech of Erfurt. Isis 39, 5960.
Thorndike, L., and Kibre, P. 1963. A Catalogue of Incipits of Medieval Scientific Writings
in Latin. Cambridge, ma: Medieval Academy of America.
Toomer, G.J., 1968. A survey of the Toledan Tables. Osiris 15, 5174.
Toomer, G.J. 1984. Ptolemys Almagest. New York: Springer-Verlag.
Zinner, E. 1925. Verzeichnis der astronomischen Handschriften des deutschen Kulturge-
bietes. Munich: C.H. Beck.
chapter 2

Computational Astronomy:
Five Centuries of Finding True Syzygy*

1 Introduction

Medieval astronomers were aware of the difficulties and enormous effort


needed to follow the rules for determining astronomically significant events
and, from Ptolemy on, it was generally understood that numerical tables could
be constructed that represented the underlying geometrical models ade-
quately. Indeed, tables were already the result of many computations that
relieved the user of much work and thus reduced the possibility of making mis-
takes.1 Moreover, the use of a table could be described more simply than the
corresponding rules for calculating the same quantity. In short, tables were the
most successful and economical way to present complex mathematical proce-
dures in the Middle Ages. Many solutions to the same problem were generated,
usually without justification or reference to previous treatments of it. For this
reason, the task of the historian is to provide the astronomical significance of
the quantities present in the mathematical relationships, as well as to indicate
the dependence of later scholars on the works of their predecessors.
In this paper we explore the specific case of finding the time from mean
to true syzygy (conjunction or opposition of the Sun and the Moon).2 The
procedures introduced for this purpose rely entirely on arithmetic, and do not
involve any observations. Rather, they illustrate the methods and approaches
of medieval practitioners in computational astronomy, and yield insights into

* Journal for the History of Astronomy 28 (1997), 93105.


1 See, for example, B.R. Goldstein, Descriptions of astronomical instruments in Hebrew, in
From deferent to equant: A volume of studies in the history of science in the ancient and medieval
Near East in honor of E.S. Kennedy, ed. by D.A. King and G. Saliba (New York Academy of
Sciences, New York, 1987), 105141 (espec. p. 128, where Ibn al-adibs view is cited).
2 In addition to the authors whose works are discussed here, we know of a few other treatments
of this question. The most important is probably that of John of Murs noted by E. Poulle, John
of Murs, in Dictionary of scientific biography (New York, 19701980), vii, 128133 (espec. p. 130),
and we have also located tables for this purpose in Hebrew astronomical manuscripts: Moses
Farissol Botarel (c. 1481) in Munich, ms heb. 343, ff. 96b97a, and Isaac ben Elia Kohen of
Syracuse (c. 1491) in London, ms British Library Or. 2806, ff. 43a44b.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_004


computational astronomy: five centuries of finding true syzygy 41

the ways that various traditions in the Middle Ages affected early modern
astronomy. As far as we can tell, solutions to this problem that make use of
double argument tables first appeared in Spain, and later in other parts of
Europe. To be sure, medieval examples of double argument tables have been
noted previously, but they pertain to the calculation of lunar and planetary
equations (beginning with Ibn Ynus, d. 1009),3 and for problems concerning
the time of day.4 We will focus our attention on tabular solutions, while also
noting procedures that were not reduced to tables.
Simply stated, the problem is: given the time at which a mean syzygy takes
place (easily solved using available tables), to determine t, the time from mean
syzygy, t, to true syzygy, t, where t > 0 indicates that true syzygy takes place
after mean syzygy. There are four variables to be taken into consideration: s
and m, the true longitudes of the Sun and the Moon at mean syzygy; and vs
and vm, the velocities in longitude of the Sun and the Moon during the time
from mean to true syzygy. The most serious difficulty is that the lunar velocity
cannot properly be considered constant in this time interval, and there was no
simple way to approximate its average value.
In Almagest vi.4 Ptolemy (c. 150) presented an approximate solution (with-
out reducing it to a table) which may be expressed in modern notation as

(1) t = 13/12vm(t),

where is the true elongation at mean syzygy ( = m s for conjunction,


and = m s + 180 for opposition), together with a rule for computing
vm(t), the velocity of the Moon at the time of mean syzygy. This method makes
the simplifying assumptions that the solar and lunar velocities are constant
over the relevant time interval, and that the ratio of lunar to solar velocity
is 13 to 1. Though these approximations are crude, they can be refined by

3 C. Jensen, The lunar theories of al-Baghdadi, Archive for history of exact sciences, viii (1972),
321328; D.A. King, A double argument table for the lunar equation in Ibn Ynus, Centaurus,
xviii (1974), 129146; D.A. King, Some early Islamic tables for determining lunar crescent
visibility, in King and Saliba (eds), op. cit. (ref. 1), 185225; J.D. North, The Alfonsine Tables
in England, in Prismata, ed. by Y. Maeyama and W.G. Saltzer (Wiesbaden, 1977), 269301;
G. Saliba, The double argument lunar tables of Cyriacus, Journal for the history of astronomy,
vii (1976), 4146; M. Tichenor, Late medieval two-argument tables for planetary longitudes,
Journal of Near Eastern studies, xxvi (1967), 126128, reprinted in E.S. Kennedy, Studies in the
Islamic exact sciences (Beirut, 1983), 122124.
4 See, for example, B.R. Goldstein, A medieval table for reckoning time from solar altitude,
Scripta mathematica, xxvii (1964), 6166, reprinted in Kennedy, Studies (ref. 3), 293298.
42 chapter 2

introducing an iteration and checking procedure. After computing a first value


for t, m and s may be recomputed from the solar and lunar correction tables
for t + t. If they are not equal, the procedure in Equation (1) is repeated,
using the new value for . The process converges rapidly, and it is rarely,
if ever, necessary to perform more than two iterations to achieve equality
of solar and lunar longitudes, to the precision of minutes of are. For most
medieval astronomers this approach (with or without refinement) was quite
satisfactory: it was adopted by al-Battn;5 it appears in the canons to the tables
of al-Khwrizm;6 and it is found in the principal astronomical tables composed
in medieval Spain, namely, the Toledan Tables and the Alfonsine Tables (as
attested in Chapter xxx of the Castilian canons).7

2 The Method of Ibn al-Kammd

The earliest solution to the syzygy problem that we have found in the form of
a table appears in a work of Ibn al-Kammd (c. 1116) called al-Muqtabis.8 This
is one of his three known works (and the only one extant), preserved uniquely
in a Latin manuscript: Biblioteca Nacional de Madrid, ms 10023. The text was
translated into Latin in 1260 in Palermo by John of Dumpno. Ibn al-Kammds
solution consists of a double argument table (f. 52r): its entries, t(vm(t) vs(t),
), are given in hours and minutes, and are functions of the difference between
the hourly velocities of the Moon and the Sun (in minutes and seconds per
hour, from 0;27,30/h to 0;33,30/h at intervals of 0;0,30/h) and the elongation,
(in degrees and minutes, from 0;30 to 12;0 at intervals of 0;30). An excerpt
of this table is reproduced in Table 1. The entries were calculated by means of
the formula

5 O. Neugebauer, A history of ancient mathematical astronomy (New York and Berlin, 1975), 123,
points out that Ptolemy does not indicate whether iteration is necessary, whereas in works
of the Byzantine period it is explicitly stated that the procedure is to be iterated until no
elongation remains. See also C.A. Nallino, Al-Battn sive Albatenii opus astronomicum (2 vols,
Milan, 19031907), i, 94.
6 O. Neugebauer, The astronomical tables of al-Khwrizm (Copenhagen, 1962), 63.
7 On zijes (sets of astronomical tables), see E.S. Kennedy, A survey of Islamic astronomical
tables, Transactions of the American Philosophical Society (Philadelphia, 1956), xlvi/2; see
also F.S. Pedersen, Canones Azarchelis: Some versions, and a text, Cahiers de lInstitut du
MoyenAge grec et latin, liv (1987), 129218 (espec. p. 182); M. Rico y Sinobas, Libros del saber
de astronoma del Rey Alfonso x de Castilla (5 vols, Madrid, 1866), iv, 150151.
8 J. Chabs and B.R. Goldstein, Andalusian astronomy: al-Zj al-Muqtabis of Ibn al-Kammd,
Archive for history of exact sciences, xlviii (1994), 141.
computational astronomy: five centuries of finding true syzygy 43

table 1 Ibn al-Kammd (Madrid, ms 10023, f. 52r)

[vm vs]
[] 27;30 28; 0 33; 0 33;30

0;30 1; 6 1; 4 0;54 0;53


1; 0 2;11 2; 8 1;49 1;47
1;30 3;17 4;12 2;43 2;41

6; 0 13; 5 12;51 10;45 * 10;45

11;30 25; 6 24;39 20;36 20;46 **
12; 0 26;11 25;43 21;30 21;50 **

* Read 10;54.
** These values seem to belong to the column
labelled 33;0, and the values there to this column.

(2) t = /[vm(t) vs(t)].

This solution gives rather crude results (see Section 6, below), because the rela-
tive velocity of the luminaries is assumed to be constant during the time inter-
val from mean to true syzygy, but its simple presentation made it attractive,
and led to its adoption (with minor modifications) by a number of subsequent
astronomers.

The astronomical work of Juan Gil de Burgos (c. 1350), as preserved in


Hebrew in London, Jews College, ms heb. 135, contains a table (ff. 92b93b)
with the same range, intervals, and precision for the horizontal argument
(vm vs) as that of Ibn al-Kammd. The vertical argument () has the
identical range and the same accuracy, but the interval is different, for here
it is 0;6 (rather than 0;30). The entries in this table are more accurately
computed from Equation (2) than those in Ibn al-Kammds table.
The Tables of Barcelona, extant in Catalan, Hebrew, and Latin manuscripts,
and probably completed in 1381,9 depend on the same solution as that

9 J.M. Mills, Las tablas astronmicas del Rey Don Pedro el Ceremonioso (Madrid and Barcelona,
1962); J. Chabs, Astronoma Andalus en Catalua: Las Tablas de Barcelona, in From Bagh-
dad to Barcelona: Studies in the Islamic exact sciences in honour of Prof Juan Vernet, ed. by
44 chapter 2

table 2 Yosef Ibn Waqr (Munich, ms heb.


230, f. 78b)

[]
[vm] 0;10 0;20 1; 0 6; 0

12; 0 0;21 0;42 2;11 13; 6


12;10 0;21 0;42 2; 9 12;54
12;20 0;21 0;42 2; 7 12;40

13; 0 0;20 0;40 2; 0 12; 0

14;50 0;17 0;34 1;43 10;12

given by lbn al-Kammd, although the presentation is somewhat different.


The horizontal argument has a minimum value of 0;28,0/h (rather than
0;27,30/h) and the same upper limit; the same interval (0;0,30/h) is also
used here. The range and interval of the vertical argument differ from those
in the tables of Ibn al-Kammd and Juan Gil: the elongation is given from
0;10 to 6;40 at intervals of 0;10.
Yosef lbn Waqr of Seville (mid-fourteenth century) presents a slightly dif-
ferent solution in the form of tables, uniquely preserved in Munich, ms heb.
230, f. 78b, where the headings are written in Arabic in Hebrew characters. He
treats the effect of the velocities of each luminary in Equation (2) separately
in two double argument tables. In the first table (Munich, ms heb. 230, f. 78b
top) only the lunar velocity varies while it is assumed that the solar veloc-
ity is constant. In this table the elongation, (in degrees and minutes, from
0;10 to 1;0, at intervals of 0;10, and from 1 to 6 at intervals of 1), is the
horizontal argument (rather than vertical, as before), and the lunar veloc-
ity is the vertical argument (in degrees and minutes per day, from 12;0/d
to 14;50/d, at intervals of 0; 10/d). Table 2 displays an excerpt of this table.
The entries, t(, vm), can be recomputed by means of a formula analogous
to Equation (2):

(3) t = 24 / (vm k),

J. Casulleras and J. Sams (Barcelona, 1996), 477525; J. Chabs, Lastronomia de Jacob ben
David Bonjorn (Barcelona, 1992), 23.
computational astronomy: five centuries of finding true syzygy 45

table 3 Yosef Ibn Waqr (Munich, ms heb.


230, f. 78b)

[t]
[vs] 0;10 0;20 1; 0 6; 0

0;57 0;24 0;47 2;22 14;12


0;58 0;25 0;49 2;25 14;30
0;59 0;25 0;49 2;27 14;42
1; 0 0;25 0;50 2;30 15; 0
1; 1 0;26 0;51 2;32 15;12
1; 2 0;26 0;52 2;35 15;30

where k = 1/d, and represents the constant solar velocity.

Ibn Waqrs second double argument table is for determining the position of
the Sun at true syzygy (see the excerpt in Table 3): the vertical argument is vs,
the solar velocity (in degrees and minutes per day, from 0;57/d to 1;2/d, at
intervals of 0;1/d). The horizontal argument (not specified in the ms) is t, the
time from mean to true syzygy found in the previous table, and it is given in
hours, from 0;10h to 1h, at intervals of 0;10h, and from 1h to 6h, at intervals of
1h. The entries, (t, vs), represent the arc-distance (in minutes and seconds)
that the Sun travels during that time, and we have recomputed them by means
of the formula

(4) = (60/24)vs t.

In the canons to Ibn Waqrs tables (Munich, ms heb. 230, ff. 5a5b [Hebrew]
and ff. 14a14b [Arabic in Hebrew characters]) we are told that the true positions
of the Sun and Moon should be computed for noon of day 14 or day 29 of
the month (rather than for mean syzygy), and then the table should be used,
thus indicating that the table gives time from noon to true syzygy. Despite this
statement in the canons, the entries in the table display the time from mean to
true syzygy, rather than the time from noon to true syzygy (for an example, see
Section 6, below).

Two centuries later, the method presented by Ibn al-Kammd appears, with-
out ascribing it to any author, in the printed version of the Alfonsine Tables
edited by P. Du Hamel in Paris (1553). In this case we have exactly the same
46 chapter 2

pattern as in Ibn al-Kammd, but for minor details (p. 155): the entries,
t(vm vs, ), are given here to seconds; the relative difference in velocity
is given from 0;22/h to 0;34/h at intervals of 0;1/h; the elongation ranges
from 1 to 8 at intervals of 1. This table also includes columns for the dif-
ferences in the entries between two consecutive values of the elongation, to
assist in the cumbersome task of interpolation. The entries are computed by
means of Equation (2). On pp. 156157 there is another version of the same
table: its entries are sixtieths of the corresponding previous ones.

3 New Approaches to an Old Problem

John of Saxony (c. 1330), one of the Parisian astronomers who adapted the
Alfonsine Tables and composed canons to it, offered a more sophisticated
solution using Ptolemys lunar models.10 He made allowances for the variation
in the lunar velocity in the time interval between mean and true syzygy, and
introduced a method of successive approximations to t, first to the nearest
hour, and then to the nearest minute of an hour.11 This yields an improvement
in accuracy, but involves a lot of computation that most practitioners in the late
Middle Ages were probably not prepared to follow. Unfortunately, this solution
cannot be displayed in tabular form.
Another attempt to give more accurate solutions was successfully devel-
oped by Levi ben Gerson (12881344) who lived in southern France and wrote
in Hebrew. He depended on his own lunar models rather than Ptolemys, but
remained within the framework defined by Ptolemy. He presented his original
solution in the form of four tables, and all of them contain additive coeffi-
cients to avoid calculations with negative terms.12 The equation of time is also
included in the tables, so that all coefficients add up to 24; 17h. Two different

10 E. Poulle, Les tables alphonsines avec les canons de Jean de Saxe (Paris, 1984), 80ff.
11 J. Chabs and B.R. Goldstein, Nicholaus de Heybech and his table for finding true syzygy,
Historia mathematica, xix (1992), 265289.
12 B.R. Goldstein, The astronomical tables of Levi ben Gerson (New Haven, 1974), 136146.
For examples of medieval planetary correction tables where negative terms are elimi-
nated by adding a constant. see H. Salam and E.S. Kennedy, Solar and lunar tables in
early Islamic astronomy, Journal of the American Oriental Society, lxxxvii (1968), 492
497; E.S. Kennedy, The astronomical tables of Ibn al-Aclam, Journal for the history of
Arabic science, i (1977), 1323 (espec. p. 14); and B.R. Goldstein, The survival of Arabic
astronomy in Hebrew, Journal for the history of Arabic science, iii (1979), 3139 (espec.
p. 37).
computational astronomy: five centuries of finding true syzygy 47

sets of Levis tables for this purpose are known: they share the same structure,
but they are based on different parameters.
The time from mean to true syzygy is dependent on the motions of both
luminaries in a way that renders it difficult to treat them separately. Neverthe-
less, Isaac Ibn aladib, an astronomer of Spanish origin who settled in Sicily
at the end of the fourteenth century,13 managed to treat the effect of each lumi-
nary separately. In his astronomical tables, entitled Ora selulah, he mentioned
the works of Bonfils (Tarascon, c. 1365), Bonjorn (Perpignan, c. 1361), Ibn al-
Kammd, and Ibn alRaqqm (Granada, d. 1315),14 and gave two corrections to
be applied to the time of mean syzygy in order to find that of true syzygy: one
due to the Sun and the other to the Moon, each one in separate tables (Paris, ms
heb. 1086, ff. 7a9b). The first (solar) correction has a maximum of 3;54,22h for
a value of the mean argument of the Sun of 91; the second (lunar) correction
reaches a maximum of 9;42,6h when the true argument of the Moon is 96. Ibn
al-adibs tables also give separate corrections for the positions of the luminar-
ies between mean and true syzygy, making use of Ptolemys values: 0;32,56/h
for the mean hourly lunar velocity in longitude, 0;32,40/h for the mean hourly
lunar velocity in anomaly, and 0;2,28/h for the mean hourly velocity of the Sun.
Within the Spanish astronomical tradition, Abraham Zacut (d. after 1515)
presented yet another approach to this problem.15 In his Almanach perpetuum
printed in Leiria (Portugal) in 1496 there is a double argument table for the
equation of syzygies. Its title is indeed surprising (pp. 65v66v): Tabula ad
verificandum horam aspectuum vel coniuntionis. The corresponding title for the
same table in Zacuts Ha-ibbur ha-gadol is Table for correcting the time of
conjunction and opposition and quarters of the month and all aspects of the
Moon with all the planets (Lyon, ms heb. 14, f. 142r). In this table the vertical
argument is the elongation (arcus distantie), , between the Sun and the Moon.
The values taken for the elongation are 0;5, 0;10, 0;20, , 1, and thereafter
for each half degree to 13, for a total of 31 values. The horizontal argument
goes from 10;36 to 16 at intervals of 0;12, and represents the daily increment
of elongation, . This table contains 868 entries, yielding not the time from

13 B.R. Goldstein, Scientific traditions in late medieval Jewish communities, in Les Juifs au
regard de l histoire: Mlanges en l honneur de M. Bernhard Blumenkranz, ed. by G. Dahan
(Paris, 1985), 235247.
14 On Ibn al-Raqqm, see J. Carandell, Risla fi ilm al-ill de Muammad Ibn al-Raqqm
al-Andalus (Barcelona, 1988).
15 On Zacut, see F. Cantera Burgos, El judo salmantino Abraham Zacut, Revista de la
Academia de Ciencias de Madrid, xxvii (1931), 63398; and F. Cantera Burgos, Abraham
Zacut (Madrid, 1935).
48 chapter 2

mean to true syzygy but the time, t, counted from noon, at which the true syzygy
occurs in Salamanca. The 842 non-zero entries in this table can be recomputed
by means of the following equation:

(5) t = 24 / ,

where and are deduced from the values of the true longitudes of the Sun
and the Moon, at Salamanca for noon, found in other tables of the Almanach
perpetuum (Tables 2 and 7 for the daily positions of the Sun and the Moon). The
results given by Zacut differ by more than 1 in only 26 cases, thus indicating
that the original table was satisfactorily calculated.

4 An Accurate and Easy Solution

Nicholaus de Heybech of Erfurt (c. 1400) seems to have been the first astron-
omer to reach a solution that met the criteria of improved accuracy and a
user-friendly presentation in tabular form (see Table 4 for an excerpt).16 To
do this, he introduced two terms for the time from mean to true syzygy, to be
added algebraically, one for the Sun and one for the Moon, and treated them
separately. Each term can be considered as a combination of three functions,
some of which depend only on the mean solar anomaly, , and some only on
the mean lunar anomaly, . In calculating the entries in his table, Nicholaus de
Heybech made use of Ptolemys second lunar model for computing the under-
lying lunar velocities, in contrast to his predecessors who generally depended
on Ptolemys simple lunar model.17 The result is a single table in 5 columns and
180 rows that gives results as good as those of his predecessors (see Section 6,
below), but with much less work. The rule for computing t from the entries in
the table is as follows:

(6) t = ts tm = [c1() c2() c3()] [c4() c5() c3()],

where c1 , c5 represent entries in columns 1, , 5 of Heybechs table.

16 Heybechs entire table is published in Chabs and Goldstein, op. cit. (ref. 11), together with
an explanation of the way it was computed; see Essay 1.
17 For the use of Ptolemys second lunar model in computing lunar velocities, see B.R. Gold-
stein, Lunar velocity in the Ptolemaic tradition, in The investigation of difficult things:
Essays on Newton and the history of the exact sciences, ed. by P.M. Harman and A.E. Shapiro
(Cambridge, 1992), 317.
computational astronomy: five centuries of finding true syzygy 49

table 4 Nicholaus de Heybech. Note that here s represents


signs of 60. The columns are labelled: Argument:
linee numeri; i: equatio solis; ii: diversitas
equationis solis; iii: minuta proportionalia; iv:
equatio lune; and v: diversitas equationis lune.

Argument i ii iii iv v
s /s h min h min min h min min

0 1/5 59 0 5 0 1 0 0 11 0
0 2/5 58 0 10 0 2 0 0 22 0

1 0/5 0 4 4 0 51 15 8 31 3

2 0/4 0 4 14 0 54 47 8 15 3

3 0/3 0 0 0 0 0 60 0 0 0

Nicholaus de Heybechs method for finding the time from mean to true syzygy
requires a few, rather simple, computations. Not long after him, John of Gmun-
den (d. 1442), lecturer on astronomy at the University of Vienna, fully under-
standing Heybechs approach, presented it in an even simpler way by means
of a double argument table, and it is preserved in a holograph manuscript (ms
Vin. 5151). The table is entitled Tabula ostendens distantiam vere coniunctionis
et oppositionis a media (ff. 119v122r): the horizontal argument is , the mean
solar anomaly (in degrees, from 0s 0 to 11s 24, at intervals of 6), and its verti-
cal argument, , is the mean lunar anomaly (in degrees, from 0s 0 to 6s 0, at
intervals of 6). The table is preceded by a short canon (ff. 117v119r), at the end
of which we read: Iste canon editus et scriptus est Wienne per magistrum Johan-
nem de Gmunden die 20 mensis Maius anno domini 1440 currente. The table, an
excerpt of which appears in Table 5, displays 1,800 entries, t(, ), given in
hours and minutes, for 0 180. These entries can also be used for the
other values of because the following symmetry relation holds:

(7) t(, ) = t(360 , 360 ).

The maximum value, t = 14;0h, corresponds to = 264 and 270, and = 84


and, by Equation (7), the minimum value, t = 14;0h, corrresponds to = 90
and 96, and = 276.
50 chapter 2

table 5 John of Gmunden (ms Vin. 5151, ff. 119v122r). We have added a minus sign where
the text reads m(inue) and nothing where it reads a(dde).

[ ]
[] 0s 0 0s 6 3s 0 6s 0 9s 0 11s 18 11s 24

0s 0 0; 0 0;28 4;47 0; 0 4;47 0;57 0;28


0s 6 1; 3 0;35 3;44 1; 3 5;49 1;59 1;31
0s 12 2; 5 1;37 2;41 2; 5 6;51 3; 1 2;33

2s 24 9;40 9;15 5;23 9;44 14; 0 10;31 10; 5


3s 0 9;40 9;15 5;27 9;44 13;56 10;31 10; 5

5s 24 0;59 0;37 2;48 0;59 5;44 1;44 1;21


6s 0 0; 0 0;22 3;46 0; 0 3;46 0;45 0;22

It is easy to derive this table from that of Nicholaus de Heybech, but note that
John of Gmunden uses signs of 30 whereas Nicholaus de Heybech uses signs of
60. For example, the maximum entry t(270, 84) (see Table 5) can be derived
directly from the entries in Nicholaus de Heybechs table: the result is 13;57h,
which differs from the entry in John of Gmundens table by 3 minutes. In other
cases we found even closer agreement.
In a fifteenth-century manuscript in Rome, ms Casanatense 1673, we have
found another copy of John of Gmundens double argument table (ff. 89v92r).
The table is entitled Tabule distantie vere coniunctionis vel oppositionis luminar-
ium a media: Composita Erfordie Duringie, and shares all the characteristics of
Gmundens except for two: signs of 60 rather than 30 are used, and the dif-
ferences between consecutive entries, whether in the same row or column, are
shown. There is no text in the manuscript explaining the use of this table, and
just after it there is an interpolation table (f. 92v) to be used in connection with
our table, with the indication that it was drawn per dominum nicolaum de
Reichenbach. This is the name of the copyist of the canons to the tables of
Iohannes Bianchinus, also found in this miscellaneous manuscript. The dou-
ble argument table bears no date, but in the manuscript it is found between
an almanac for the planets beginning in 1456 and a table for mean syzygies the
radices of which, as explicitly stated, are taken for the year 1452 (completed)
and the meridian of Vienna.
The same presentation as that given by John of Gmunden is also found in
an earlier work on eclipses, known as The book of six wings, that was written
computational astronomy: five centuries of finding true syzygy 51

table 6 Immanuel ben Jacob Bonfils.


Excerpt from the double
argument table for the time
from mean to true syzygy in
Wing 2 after subtracting
24; 16h.

[]
[ ] 0 90 180 270

0 0; 0 9;54 0; 0 9;54
90 4;10 6; 6 3;24 13;45
180 0; 0 9;57 0; 0 9;57
270 4;10 13;45 3;24 6; 6

by Immanuel ben Jacob Bonfils of Tarascon (southern France) in the mid-


fourteenth century. Originally written in Hebrew, this work was translated into
Latin in 1406 and into Greek in 1435.18 The table that he called Wing 2 is for
determining , the longitude from mean to true syzygy, and t, the time from
mean to true syzygy. Actually, the tabulated entries are + 3 and t + 24; 16h
(which will be called t* in what follows). The additive coefficient to t consists
of 0;16h, to account for the maximum equation of time, and 24h, to avoid cal-
culations with negative terms.19 In Bonfilss double argument table for finding
the time to true syzygies, the horizontal argument is , the mean lunar anomaly
(in degrees, from 0s 0 to 11s 24, at intervals of 6), and the vertical argument
is , the mean solar anomaly (in degrees, from 0s 0 to 12s 0, at intervals of 6).
The entries t*(, ) are given in hours and minutes. The table for t* is sim-
ilar to that of John of Gmunden except for interchanging the positions of the
arguments, and the fact that it has twice as many entries, 3,600.
Table 6 displays an excerpt of Bonfilss table, after subtracting 24;16h. Al-
though very close, the entries do not coincide with those given by John of

18 P.C. Solon, The Hexapterygon of Michael Chrysokokkes, Ph.D. dissertation, Brown Uni-
versity, 1968; P.C. Solon, The six wings of Immanuel Bonfils and Michael Chrysokokkes,
Centaurus, xv (1970), 120.
19 This device is reminiscent of a similar one used by Levi ben Gerson for the same purpose
relating to the same problem. Bonfils was aware of the work of his predecessor in southern
France, and probably depended on him here: cf. B.R. Goldstein, The astronomy of Levi ben
Gerson (12881344) (New York and Berlin, 1985), 9; see also ref. 12, above.
52 chapter 2

table 7 Recomputation of some entries


in Bonfilss table using the
values given by al-Battn

[]
[ ] 0 90 180 270

0 0; 0 9;51 0; 0 9;51
90 4;10 6; 4 3;26 13;38
180 0; 0 9;54 0; 0 9;54
270 4;10 13;38 3;26 6; 4

Gmunden or Nicholaus de Heybech. Indeed, those two astronomers follow the


Alfonsine tradition, whereas Bonfilss table can be derived from al-Battns
tables, very nearly. Table 7 shows the recomputed values of some entries in Bon-
filss table using the values given by al-Battn for the solar and lunar equations
and velocities,20 and not taking into account any correction for the equation of
time.
It has not previously been noted in the scholarly literature that John of
Gmundens approach was retained by early modern astronomers. Georg Peur-
bach (14231461), one of the earliest advocates of humanism at the University
of Vienna, is the author of the Tabulae eclypsium, an extensive set of astro-
nomical tables first printed in 1514 in Vienna, edited by Georg Tannstetter, and
bound together with a work of Peurbachs associate and student, Regiomon-
tanus: Tabula primi mobilis Joannis de Monteregio. In Peurbachs work there is a
forty-eight-page table entitled Tabula distantie vere coniunctionis aut oppositio-
nis a media (ff. a3vd3r) and, in fact, it is a variant of John of Gmundens table
for finding the time from mean to true syzygy. In Peurbachs double argument
table, the mean solar anomaly ranges from 0s 0 to 11s 30, and the mean lunar
anomaly from 0s 0 to 6s 0, as was the case for John of Gmunden. But Peur-
bachs table is much more expanded, for the intervals are 2 for and 1 for ,
yielding an impressive total number of entries of 32,400. This table, together
with the rest of Peurbachs Tabulae eclypsium, was later printed in 1553 in Basel
in a volume entitled Luminarium atque planetarum motuum tabulae octoginta
quinque, omnium ex his quae Alphonsum sequuntur quam faciles together with
tables by Bianchini, Prugnerus, and Peurbach.

20 Nallino, op. cit. (ref. 5), ii, 80, 88.


computational astronomy: five centuries of finding true syzygy 53

By comparing the entries common to all three tables, namely, those by


Nicholaus de Heybech, John of Gmunden, and Georg Peurbach, given at inter-
vals of 1, 6, and 1 of mean lunar anomaly, respectively, it is clear that Peur-
bachs table does not derive directly from Heybechs, but from Gmundens, and
that Peurbach used an interpolation scheme to fill in the intermediary values.
This puts Peurbach within a medieval tradition for computing the time from
mean to true syzygy rather than going back to classical sources, as might be
expected of a Renaissance scholar.

5 Calculations in the Sixteenth Century

Copernicus seems to have had a method of his own to derive the time from
mean to true syzygy;21 it is described in De revolutionibus iv.29. His procedure
depends on Ptolemys second lunar model for computing lunar velocities at
syzygy, as did Nicholaus de Heybechs procedure, and considers the increment
in lunar anomaly in the interval from mean to true syzygy (without introduc-
ing any new tables), in a way that is reminiscent of John of Saxony. Although
Copernicus probably knew the solutions given by his immediate predecessors,
it is not clear that he ever used them. According to Swerdlow and Neugebauer,22
Copernicus had copied parts of Peurbachs Tabulae eclypsium and appended
them to a bound volume containing the 1492 edition of the Alfonsine Tables
and the 1490 edition of Regiomontanuss Tabulae directionum. This handwrit-
ten quire of astronomical tables known as the Uppsala Notebook seems to
come from Copernicuss studies at Cracow University (14911495?).23 Moreover,
the Prutenic tables by Erasmus Reinhold (15111553), based on the Copernican
models and first published in Tbingen in 1551, do not contain any table for the
time from mean to true syzygy.
In December 1590, the Danish astronomer, Tycho Brahe, observed a lunar
eclipse, and displayed the times computed for eclipse middle (which should
be very close to true syzygy) according to five procedures without indicat-
ing any of the intermediate values required by each procedure. The five pro-
cedures are labelled: Alfonso, Peurbach, Copernicus, Maestlin, and Brahes

21 N.M. Swerdlow and O. Neugebauer, Mathematical astronomy in Copernicuss De revolution-


ibus (New York and Berlin, 1984), 276.
22 Swerdlow and Neugebauer, op. cit. (ref. 21), 272.
23 P. Czartoryski, The Library of Copernicus, Studia Copernicana, xvi (1978), 355396 (espec.
p. 366).
54 chapter 2

own.24 Brahes times computed according to the Alfonsine Tables and those
of Peurbach differ from our recomputations, and the five values he presents
for the time for eclipse middle vary from each other by as much as 2 hours.
However, as can be seen from Table 8, below, the different methods we have
used yield results that are much closer together. Hence, we suspect that Brahe
made some mistakes in his calculations. Moreover, it has been claimed that
Brahe intended to test computed times against observed times, but the text is
too terse to draw any conclusions.25

6 A Test of the Methods

To compute the time of true syzygy that was necessary for determining in
advance the circumstances of an eclipse, medieval astronomers had to go
through a complicated procedure. They had to use different tables for the
mean motions, equations, velocities, etc., of the two luminaries, and they had
to follow instructions, whenever available, to perform these calculations step
by step. Throughout the Middle Ages these tables and instructions evolved
in different ways, although they remained within the framework established
by Ptolemy. But they were adapted to different models and parameters, and
thus yielded different numerical results for a given problem. We know of no
case prior to the sixteenth century where the various methods for finding
time from mean to true syzygy proposed by the astronomers considered above
were tested against an observation. Nevertheless, we think it useful to compare
these methods, and so we have derived t for the same syzygy according to
the procedures of each author. To do this, we have used the same data in all
cases, rather than depending on the intermediate data that result from each
authors way of deriving them. As our test case, we have taken a syzygy, the
mean conjunction of 20 July 1327, occurring at 3;58,10h after noon in Toledo.26
The basic magnitudes for that event, as calculated according to the instructions
in the canons to the Alfonsine Tables by John of Saxony, are: = 35;25,4, =
222;26,7, and = 4;45,39. Our results, using the different methods and tables,
are displayed in Table 8.

24 Tychonis Brahe Opera omnia, ed. by J.L.E. Dreyer (15 vols, Copenhagen, 19131929), xii,
2025.
25 V.E. Thoren, Tycho Brahes discovery of the variation, Centaurus, xii (1967), 151166 (espec.
p. 158). Thoren also claimed that this observation led Brahe to the discovery of the lunar
variation, but we do not believe there is enough evidence to support his view.
26 Poulle, op. cit. (ref. 10), 214 ff.; Chabs and Goldstein, op. cit. (ref. 11), 271.
computational astronomy: five centuries of finding true syzygy 55

table 8 A comparison of the various methods

t
Table or method Date hours

Ptolemya c. 150 8;40


lbn al-Kammd c. 1116 8;47
Yosef lbn Waqr mid-14th cent. 8;47
John of Saxonyb c. 1330 8;34
Levi ben Gersonc c. 1340 8;33
Immanuel Bonfils c. 1365 8;44
Nicholaus de Heybechd c. 1400 8;34
John of Gmunden 1440 8;33

a. With iteration; without iteration the value obtained is


8;51,30h (see ref. 5).
b. For details, see Goldstein and Chabs, op. cit. (ref. 11), 271.
c. With Levis second set of tables, the value obtained is 8;6h.
d. For details, see Goldstein and Chabs, op. cit. (ref. 11), 274.

7 Conclusion

The most important point is that observations were not the driving force for
innovations in the treatment of the time interval from mean to true syzygy.
Rather, computational methods were devised whose goal was to simplify com-
putations and to be user-friendly without sacrificing accuracy. We have seen
that a great deal of ingenuity went into these changes in procedure, and that
they were most successfully presented in the form of tables. The first step in
transforming the rules for finding the time from mean to true syzygy was a
rather simple table, and the earliest example of it is found in the tables of Ibn al-
Kammd, a Spanish Muslim astronomer. Other revisions of the rulesusually
in the form of tableswere made in Spain, notably by Ibn Waqr, and then
astronomers in France (and later in Germany) added new devices of their own
while ultimately depending on their Spanish predecessors. So far, we have not
found any examples of this type of table among astronomers from the eastern
Islamic countries. Moreover, as is the case for other scientific matters in Spain,
there is evidence to be gleaned from traditions in Arabic, Catalan, Castilian,
Hebrew, and Latin, reflecting the multicultural setting of the Iberian Penin-
sula. We have also seen that these medieval discussions continued to be copied,
56 chapter 2

printed, and cited in the sixteenth century, i.e., they were still a part of the living
astronomical tradition in the early modern period.

Acknowledgement

We thank R. Mercier for allowing us to consult his notes on Bonfilss table Wing
2 from which we have greatly benefited.
chapter 3

Transmission of Computational
Methods within the Alfonsine Corpus:
The Case of the Tables of Nicholaus de Heybech*

By the end of the fourteenth century the Alfonsine tradition, which origi-
nated in Castile a little more than 100 years earlier, had become the main
computational tool for European astronomers. A great variety of astronomi-
cal tables, often accompanied by texts, followed this tradition, using the same
models to describe the motions of the celestial bodies and the same under-
lying parameters, but differing in presentation. This Alfonsine corpus, as we
have recently named it, dominated the scene of Western astronomy for several
centuries.1
Within this corpus are the Parisian Alfonsine Tables of which hundreds of
copies in manuscript are extant as well as two editions that appeared before
1500. Of particular interest are various methods and tables for finding the time
from mean syzygy (i.e., conjunction or opposition of the Sun and the Moon)
to true syzygy, starting with the method described in the canons by John of
Saxony (1327).2 A specific approach to this problem, summarized below, was
introduced by an otherwise almost unknown Nicholaus de Heybech of Erfurt
(c. 1400).3 Heybechs table was modified as it was transmitted from Erfurt to
Poland, then to Salamanca, and finally to Jerusalem, and it is a remarkable
example of the variety within the Alfonsine corpus that did not involve any

* Journal for the History of Astronomy 39 (2008), 345355.


1 See Jos Chabs and Bernard R. Goldstein, The Alfonsine Tables of Toledo (Dordrecht and
Boston, 2003).
2 Jos Chabs and Bernard R. Goldstein, Computational astronomy: Five centuries of finding
true syzygy, Journal for the history of astronomy, xxviii (1997), 93105.
3 Jos Chabs and Bernard R. Goldstein, Nicholaus de Heybech and his table for finding
true syzygy, Historia mathematica, xix (1992), 265289. We have seen a dozen manuscripts
of Nicholaus de Heybechs tables: Basel, Universittsbibliothek, f.ii.7; Dijon, Bibliothque
Municipale, 447; Paris, Bibliothque nationale de France, lat. 7287 and lat. 7290a; Cues, 211;
Vienna, Nationalbibliothek, 2440; Cracow, Biblioteka Jagielloska, 609, 610, 613, 1852, and 1865
(twice); and Princeton, University Library, Grenville Kane Collection 51. Several authors have
mentioned other manuscripts containing the same material: Bern, 454; Vatican, Pal. lat. 1376;
Vienna, Nationalbibliothek, 5245; and Munich, Clm 14111 and 26666.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_005


58 chapter 3

changes in the theory or the parameters. Moreover, in his computations of


the entries in his table, Heybech used a table for lunar velocity that first
appears in Paris in the 1330s; this is part of the Alfonsine corpus and differs
significantly from the tables for lunar velocity in the traditions of al-Battn
and of al-Khwrizm.4 We thus offer an illustration of the transmission of
a computational technique within the Alfonsine corpus that kept evolving
during its long journey through a considerable part of Europe and beyond.
Heybechs method was presented in the form of a single table in 5 columns
and consisted in taking the time interval between mean syzygy and true syzygy
as the difference between two independent terms, one for the Sun and one for
the Moon. Each term is calculated separately, and both require the computa-
tion of a set of minimum and maximum values and the use of an interpola-
tion scheme for intermediate values. In Heybechs table, this scheme is a list
of interpolation coefficients, ranging from 0 to 1 (column iii, headed minuta
proportionalia, where 60 minutes = 1), depending on the mean lunar anomaly,
when computing the solar term, and on the mean solar anomaly, in the case of
the lunar term. Besides column iii, the computation of the solar term requires
columns i (headed equatio solis) and ii (headed diversitas equationis solis), both
given in hours and minutes. As for the lunar term, besides column iii, its com-
putation requires columns iv (headed equatio lune) and v (headed diversitas
equationis lune), the former given in hours and minutes, and the latter in min-
utes of an hour.
The solar term (ts) can be obtained by means of the expression,

ts = c1() c2() c3(),

where is the mean solar anomaly and the mean lunar anomaly, and c1, c2,
and c3 represent entries in columns i, ii, and iii, respectively. The entries in
col. i depend on and assume that = 0; those in col. ii also depend on and
represent the differences between the values for = 0 and = 180, for a given
; and the entries in col. iii, ranging from 0 to 1, are for interpolation for other
values of between 0 and 180.

4 Bernard R. Goldstein, Lunar velocity in the Ptolemaic tradition, in The investigation of


difficult things: Essays on Newton and the history of the exact sciences, ed. by P.M. Harman and
A.E. Shapiro (Cambridge, 1992), 317; idem, Lunar velocity in the Middle Ages: A comparative
study, in From Baghdad to Barcelona: Studies in the Islamic exact sciences in honour of Prof.
Juan Vernet, ed. by J. Casulleras and J. Sams (2 vols, Barcelona, 1996), i, 181194.
transmission of computational methods 59

figure 3.1 Facsimile of Nicholaus de Heybechs table (excerpt): Vienna, Nationalbibliothek,


ms 2440, f. 74v
60 chapter 3

Similarly, the lunar term (tm) can be obtained by means of the expression,

tm = c4 () c5 () c3 (),

where c4 and c5 represent entries in columns iv and v, respectively. The entries


in col. iv depend on and assume that = 0, and the entries in col. v also
depend on and represent the differences between the values for = 0 and
= 180, for a given . Again, the entries in col. iii are for interpolation. Thus,
according to Heybechs table, the time from mean syzygy to true syzygy is given
by

(1) t = ts tm = c1() c2() c3() c4() + c5() c3().

Heybechs method for determining t was appreciated by many medieval


astronomers, for it simplified their computations without sacrificing accu-
racy.5 In the manuscripts that preserve this table there is usually a short canon
explaining its use, but not the method for computing the entries. In all cases
the tables are identical, but for copyists errors.
Madrid, Biblioteca Nacional, ms 3385, containing a set of tables in Latin
which we call the Tabule Verificate for Salamanca (ff. 104r113r), has much the
same material as in Heybechs table, but with a different presentation.6 In
these tables, fully part of the Alfonsine corpus, the epoch is 1 Jan. 1461. The
name of the author of the Tabule Verificate (henceforth tv) is unfortunately
not known, although we have identified Nicholaus Polonius as the most likely
candidate among the few astronomers working in the Castilian city of Sala-
manca. Polonius was a Polish scholar who came to Salamanca no later than
1460 and held the newly established chair in astronomy/ astrology at the uni-
versity there until 1464. It is reasonably clear that he brought the Tabule Res-
olute (a form of the Parisian Alfonsine Tables) with him from Poland.7 The

5 See Chabs and Goldstein, Computational astronomy (ref. 2). The conventions for the
algebraic signs in Eq. 1 are not well described in Heybechs canons, whereas the versions in
tv and Zacut are unambiguous because the headings tell the user when to add and when to
subtract.
6 Jos Chabs and Bernard R. Goldstein, Astronomy in the Iberian Peninsula: Abraham Zacut
and the transition from manuscript to print (Philadelphia, 2000), especially pp. 2336.
7 On this set of tables, see Jerzy Dobrzycki, The Tabulae Resolutae, in De astronomia Alphonsis
Regis, ed. by M. Comes, R. Puig, and J. Sams (Barcelona, 1987), 7177; Jos Chabs, Astronomy
at Salamanca in the mid-fifteenth century: The Tabulae Resolutae, Journal for the history of
astronomy, xxix (1999), 167175.
transmission of computational methods 61

tables we have labelled tv 7 (f. 106rv) and tv 8 (f. 107rv) in a previous pub-
lication represent Heybechs columns for the lunar correction and the solar
correction, respectively, but they are organized in a different way: column 2
in tv 7 displays c4(), that is, column iv in Heybechs table; column 4 in tv 7
is equivalent to 1 c3(), that is, the complement in 1 to column iii in Hey-
bechs table; column 2 in tv 8 represents c1() c2(), that is, the difference
between columns i and ii in Heybechs table. However, there is no longer a col-
umn equivalent to Heybechs column v. The rest of the columns in tv 7 and
tv 8 contain the arguments and line-by-line differences of the entries of other
columns.
With the entries in tv 7 and tv 8,

(2) t = [c1() c2()] c4() + [1 c3()] c2().

This expression is equivalent to

(3) t = c1 () c2 () c3 () c4 (),

and it agrees with the first three terms in Eq. (1). The suppression of the
fourth term, c5 () c3 (), is indeed an acceptable approximation because in
Eq. (1) its contribution is at most 0;04h, a small amount compared with the
maximum values of the first, second, and third terms (4;47h, 1;01h, and 9;40h,
respectively).
It should be noted that the changes introduced in these two tables by the
unknown author of the Tabule Verificate are not mere variations in the positions
of the columns, but imply a different approach from that in Heybechs table,
among other things because attention shifts from lunar apogee ( = 0), which
is assumed for col. i in Heybechs table, to lunar perigee ( = 180), which is
assumed for col. 2 in tv 8.
The next step in the transformation of this table took place in 1513 in Jerusa-
lem where Abraham Zacut (14521515) had recently arrived.8 Zacuts best
known astronomical work was composed in Hebrew in Salamanca (1478), and
entitled ha-ibbur ha-Gadol (The great composition). It was later published
(with a number of modifications) in Latin and Castilian in Leiria, Portugal
(1496), and entitled Almanach perpetuum. This work, in turn, was later trans-

8 For biographical details, see Chabs and Goldstein, Abraham Zacut (ref. 6), 615.
62 chapter 3

figure 3.2 Facsimile of tv 7 (excerpt): Madrid, Biblioteca Nacional, ms 3385, f. 106r


transmission of computational methods 63

figure 3.3 Facsimile of tv 8 (excerpt): Madrid, Biblioteca Nacional, ms 3385, f. 107r


64 chapter 3

lated into Arabic and diffused in the Islamic world.9 Zacuts tables are mainly
based on the Parisian Alfonsine Tables, but some depend on astronomical
traditions in Hebrew that began in Provence in the fourteenth century. For
his method of finding the time from mean to true syzygy in the Almanach
perpetuum Zacut depended on this Hebrew tradition that derived from the
work of Levi ben Gerson (12881344) and was transmitted to him through
Jacob ben David Bonjorns tables (c. 1360), and it was quite distinct from the
Alfonsine tradition. But in his new set of tables of 1513 for Jerusalem Zacut
included tables for finding t that represent a modified version of Heybechs
table. Ironically, in 1478 Zacut used the Christian calendar for mean motions,
whereas in 1513 he used the Jewish calendar for this purpose. Zacuts tables of
1513 in Hebrew for Jerusalem are extant only in fragments:10 we consulted New
York, Jewish Theological Seminary of America [jtsa], ms 2574 (not dated), 15
folios (containing only Zacuts canons and tables).11 This manuscript contains
three tables for computing t by a method that is similar to the one in the
Tabule Verificate.
Table 1 (f. 8b) is for the solar correction when the Moon is at perigee on
its epicycle. The argument is the solar longitude, that is, the solar anomaly
increased by 90 (under the assumption that the solar apogee is at 90, which
is an adequate value at the time), and it is given in degrees at intervals of 1.
The entries are displayed in hours and minutes. They derive from Heybechs
table, and correspond to the difference between the entries in columns i and
ii, c1() c2(), as was the case in the Tabule Verificate for Salamanca (tv 8,
col. 2). Note however that, contrary to the Tabule Verificate, Zacuts table for
the solar correction is presented as a single table, not as a column of a table, as
in tv 8.
Table 2 (f. 9a) is for the lunar correction when the Moon is at perigee on its
epicycle. The argument is the mean lunar anomaly and it is given in degrees
at intervals of 1. The entries are displayed in hours and minutes. The entries

9 For Zacuts tables in the Islamic world, see Julio Sams, Abraham Zacut and Joseph
Vizinhos Almanach perpetuum in Arabic, Centaurus, xlvi (2004), 8297; idem, In pursuit
of Zacuts Almanach perpetuum in the eastern Islamic world, Zeitschrift fr Geschichte der
Arabisch-Islamischen Wissenschaften, xv (20022003), 6793.
10 See Bernard R. Goldstein, The Hebrew astronomical tradition: New sources, Isis, lxxii
(1981), 237251, p. 248.
11 Another fragment containing Zacuts tables of 1513 is extant in NewYork, jtsa, ms 2567.
There is no hint in Zacuts canons that he was aware of Heybech or that he had direct
access to his table.
transmission of computational methods 65

table 1 An excerpt of the table for the first correction (New


York, jtsa, ms 2574, f. 8b). Table for the correction
of the Sun when the Moon is at the perigee of its
epicycle, in hours and minutes.

3s 4s 5s 6s 7s 8s

subtract
1 0; 4h 1;53h 3;16h 3;47h 3;18h 1;54h
2 0; 8 1;57 3;18 3;47 3;16 1;50
3 0;12 2; 0 3;19 3;47 3;14 1;47
4 0;16 2; 4 3;21 3;47 3;12 1;43
5 0;19 2; 6 3;22 3;47 3;10 1;39

10 0;38 2;22 3;30 3;46 2;58 1;20

15 0;56 2;36 3;37 3;42 2;44 1; 1

20 1;15 2;49 3;42 3;37 2;30 0;42

25 1;32 3; 1 3;46 3;31 2;14 0;21

30 1;50 3;14 3;47 3;22 1;57 0; 0


add

in this table also derive from Heybechs table, and correspond to its column iv,
c4(), as was the case in the Tabule Verificate for Salamanca (tv 7, col. 2). Note
again that Zacuts table for the lunar correction is presented as a single table,
not as a column of a table, as in the Tabule Verificate.
Table 3 (f. 9b) is a double argument table where Zacut combined both
solar and lunar components. The rows were computed for values of the lunar
anomaly at intervals of 10, ranging from 0s 0 to 6s 0, whereas the columns
were computed for values of the solar longitude at intervals of 15, beginning
with 3s 0 (as in Table 1). The entries in this double argument table are given in
minutes of time.
We note that for lunar anomaly, the maximum correction takes place for
argument 0, and vanishes for argument 180. This should indeed be so, for
this correction is to be added to, or subtracted from, the value found in Table 1
when the Moon is at its epicyclic perigee; thus, the correction at argument 180
66 chapter 3

table 2 An excerpt of the table for the second correction (New


York, jtsa, ms 2574, f. 9a). Table for the correction of
lunar anomaly, in hours and minutes.

0s 1s 2s 3s 4s 5s

add
0 0; 0h 4;59h 8;32h 9;42h 8;15h 4;43h
1 0;11 5; 8 8;37 9;42 8; 9 4;34
2 0;22 5;17 8;42 9;42 7;58 4;25
3 0;33 5;26 8;42* 9;42 7;55 4;16
4 0;43 5;35 8;51 9;42 7;53 4; 7

10 1;44 6;23 9;12 9;31 7;16 3;13

15 2;36 7; 1 9;27 9;17 6;42 2;25

20 3;25 7;34 9;36 9; 2 6; 4 1;39

25 4;13 8; 4 9;42 8;40 5;24 0;50



30 4;59 8;32 9;42 8;15 4;43 0; 0
subtract

* With jtsa, ms 2567, f. 65b, read: 8;46.

(i.e., lunar epicyclic perigee) is 0. The maximum effect of solar anomaly should
be at 90 from the solar apogee and this is represented by the column for 6s
0. Hence, this third correction is to improve the first correction displayed
in Table 1, where the only variable considered was solar longitude. Table 3
then takes into account the effect of the change in lunar anomaly in the time
interval due to the solar motion (where the heading is the solar longitude),
and represents the term [1 c3()] c2() in Eq. (2). This is certainly the case:
1 c3() appears as the column for a solar longitude of 6s 0, i.e., when the
Sun is 90 ahead of apogee and the correction reaches its minimum; it is the
complement in 1 of Heybechs column iii, as in tv 7, column 4. Moreover,
c2() appears as the row for a lunar anomaly of 0s 0, and it is column ii in
Heybechs table, as in tv 8, column 4. The product of the entries in this row
and this column generates the rest of the table. For example, consider the
entries in the row for 2s 0: 0, 12, 23, 32, , 47, , 24, 12, 0. Each one is found
transmission of computational methods 67

table 3 An excerpt of the table for the third correction (New York, jtsa, ms 2574, f. 9b). Table
for the correction of all values for the lunar anomaly to be added to its value at the
perigee of its epicycle: a double argument table [lua meubberet; lit.: a combined
table]

Solar
long. 3s 0 3s 15 4s 0 4s 15 6s 0 8s 0 8s 15 9s 0

Lunar
anom. subtract
0s 0 0 15 29 41 60 31 16 0
0s 10 0 15 29 40 59 30 15 0
0s 20 0 14 28 40 58 29 15 0
1s 0 0 14 28 39 56 29 15 0
1s 10 0 13 27 37 44* 28 14 0
1s 20 0 13 25 35 51 26 13 0
2s 0 0 12 23 32 47 24 12 0

3s 0 0 8 16 23 31 17 9 0

5s 20 0 0 0 1 1 1 1 0
6s 0 0 0 0 0 0 0 0 0
add

* With jtsa, ms 2567, f. 66a, read: 54.

by multiplying the corresponding entry for a lunar anomaly of 0s 0 by 47 (the


entry corresponding to a solar longitude of 6s 0). So, for 2s 0 of lunar anomaly
and

3s 0 of solar longitude: 0; 0 0;47 = 0; 0 (entry: 0 min);


3s 15 of solar longitude: 0;15 0;47 = 0;11,45 = 0;12 (entry: 12 min);
4s 0 of solar longitude: 0;29 0;47 = 0;22,43 = 0;23 (entry: 23 min);
4s 15 of solar longitude: 0;41 0;47 = 0;32, 7 = 0;32 (entry: 32 min);

6s 0 of solar longitude: 0;60 0;47 = 0;47 (entry: 47 min);

8s 0 of solar longitude: 0;31 0;47 = 0;24,17 = 0;24 (entry: 24 min);


8s 15 of solar longitude: 0;16 0;47 = 0;12,32 = 0;13 (entry: 12 min);
9s 0 of solar longitude: 0; 0 0;47 = 0; 0 (entry: 0 min).
68 chapter 3

Thus, Zacut replaces two columns in two different tables in the Tabule Ver-
ificate by one double argument table where the multiplication is already done,
thus facilitating the task of the computer; all that is left is linear interpolation
in both the horizontal and vertical directions in the table. The use of a dou-
ble argument table is consistent with Zacuts preference for this kind of table,
which appears already in his ibbur, and this distinguishes him from the tra-
dition, always within the Alfonsine corpus, represented by Heybech and the
Tabule Verificate. On the other hand, the entries in these three tables are not
identical with those in the Tabule Verificate, but both sets are internally consis-
tent and differ very slightly.
Chapter 3 of the canons for these tables (jtsa, ms 2574, ff. 12b13a) concerns
the time interval from mean to true syzygy, with instructions on how to find this
interval from the three tables, but nothing is said about the origin of this table
or the way its entries were computed. Zacut adds a worked example (f. 12b) for
finding true conjunction for Tishri 5274am [= 30 Aug. 1513]: mean conjunction
took place 3(d) 18;6,30h after noon.12 According to the text, the Suns position
was then 5s 16;29 and the lunar anomaly was 8s 12;47. The first correction with
solar longitude 5s 16;29 as argument is 3;38,30h to be subtracted. The result is
14;28h [= 18;6,30h 3;38,30h]. With argument 8s 12 for the lunar anomaly, the
second correction is stated to be about 9;10h [ms: 10, written as a word] to be
subtracted. The result is then given in the text as 5;22h, although it should be
5;18h [= 14;28h 9;10h]. The third correction, with the two arguments, is 0;22h to
be subtracted. Thus, the final result, as given in the text, is 5;0h. This result can
be checked using the tables themselves. In Table 1 the entry for 5s 16 is 3;38h;
in Table 2 the entry for 8s 12 is 9;8h and for 8s 13 it is 9;11h. So 9;10h, the value
given in the text for the second correction, agrees with computation using the
table. In Table 3, with 5s 15 (rounded from 5s 16 for the solar longitude) and
8s 10 (rounded from 8s 12 for the lunar anomaly) as arguments, the entry is
0;22h. Hence, the total correction is 13;10,30h (= 3;38,30h 9;10h 0;22h),13
and the result should be: 18;6,30h 13;10,30h = 4;56h (text: 5;0h).
It is most likely that for this refinement of a technique in the Alfonsine cor-
pus Zacut depended on tables in Latin he had seen in Salamanca many years
before he arrived in Jerusalem. Zacut then transmitted his new method for
finding t in Hebrew, thus contributing to the enlargement of the Alfonsine

12 3(d) means weekday 3, i.e., Tuesday. And indeed 30 Aug. 1513 (jdn 2273923) was a Tuesday.
13 The absolute value of this amount for the total correction is close to its maximum: see
Richard L. Kremer, Wenzel Fabers tables for finding true syzygy, Centaurus, xlv (2003),
305329, p. 314 (table 2).
transmission of computational methods 69

corpus, a body of astronomical material with no theoretical changes or modi-


fications in the models or the parameters that depended, directly or indirectly,
on the Alfonsine Tables compiled in Toledo in the 1270s and diffused from Paris
throughout Europe and beyond in the 1320s.

Epilogue

There is one known instance of a later text that depended on Zacuts tables of
1513 for finding t: a Geniza fragment in Hebrew, ms A 697-1, at the John Rylands
University Library, Manchester, England.14 In this brief fragment of an anony-
mous calendrical text for 5557am (= 17961797),15 the goal is to compute the
times of true opposition (full-moon) for each month in the year 5557 am as a
function of true solar longitude and mean lunar anomaly at mean opposition.
On the first line Zacut is credited for the method of determining true opposi-
tions, and the text includes values computed from his first two correction tables
(but there is no evidence of his third table). For example, the solar longitude in
this text for opposition in Tishri 5557 is 6s 23;47 whose correction is given as
3;33h, and the lunar anomaly at that time is 3s 20;26 whose correction is given
as 9;2h. These are exactly the values for these arguments in Zacuts Tables 1 and
2, where the arc minutes of the arguments have been ignored.

14 We are grateful to Y. Tzvi Langermann for bringing this manuscript to our attention.
15 The date given in the text is not easy to read but 5557am is confirmed by recomputing the
astronomical data with Zacuts tables for 1513.
part 2
Planetary Motions


chapter 4

Ptolemy, Bianchini, and Copernicus:


Tables for Planetary Latitudes*

The study of planetary theory in Ptolemaic astronomy has concentrated on


the models and tables for planetary longitudes, and considerably less attention
has been paid to Ptolemys models and tables for planetary latitudes. There are
good grounds for this imbalance both in medieval sources and in the modern
secondary literature, but it is not our goal here to examine the reasons for
this. Rather, we wish to focus on some special features of tables for planetary
latitude, particularly those of Giovanni Bianchini (d. ca. 1469) that are extant
in many manuscript copies,1 in a printed edition of 1526, and in a copy in
the hand of Copernicus.2 In Almagest xiii Ptolemy gives a full treatment of
planetary latitudes and, in the case of the inferior planets, he refers to three
components which we will call inclination (declinatio), slant (reflexio), and
deviation (deviatio).3 However, Ptolemys tables for the latitudes of Venus and
Mercury in Almagest xiii.5 display only the first two of these components,4
whereas Bianchini has columns for all three of them.
Our plan is first to give a brief survey of the history of tables for planetary
latitude, particularly those that include, for Venus and Mercury, columns for
the deviation. Then we will describe Bianchinis tables for planetary latitude
in detail. Finally, we will discuss Copernicuss copy of Bianchinis tables for
planetary latitude.

* Archive for History of Exact Sciences 58 (2004), 453473, communicated by N. Swerdlow.


1 According to the catalogues we consulted, Bianchinis tables survive in a large number of
manuscript copies in many libraries: Naples, Nuremberg, Milan, Paris, Rome, Venice, Vatican,
among others (see, e.g., Boffito 1908; Thorndike 1950 and 1953; Zinner 1990). Of special interest
is Nuremberg, Stadtbibliothek, ms Cent v 57, copied in Vienna in 1460 by Regiomontanus.
2 Uppsala, University Library, ms Copernicana 4, ff. 276v281r. Since it is most likely that
Copernicus saw a manuscript of these tables while he was a student in Cracow between 1491
and 1495, we will refer to the catalogue of the scientific manuscripts there: Rosiska 1984a.
3 In Swerdlow and Neugebauer 1984, p. 523 et passim, deviatio is translated deflection.
4 O. Pedersen 1974, pp. 355386; Neugebauer 1975, pp. 216226; Riddell 1978; Toomer 1984,
pp. 632634; and Swerdlow 2005.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_006


74 chapter 4

1 Tables for Planetary Latitude Prior to Bianchini

In the Handy Tables, composed after the Almagest, Ptolemy made some
changes in the theory of the planetary latitudes and introduced new tables
with a different presentation. The latitude tables in the Handy Tables had lit-
tle influence on subsequent astronomers, although they were the source for
parameters in the corresponding tables in the Mumtaan zij (9th century).5 In
the Iberian peninsula this tradition is represented by Ibn al-Kammd (ca. 1110)
whose zij (extant in a unique Latin manuscript) provides the only previously
known example of a set of astronomical tables where the planetary latitudes
follow the Handy Tables for the inferior planets.6 In the medieval astronom-
ical literature there is yet another tradition, not based on Ptolemaic models,
where the tables for the planetary latitudes differ substantially from those in
the Almagest and the Handy Tables. This tradition whose roots lie in Indian
astronomy appears in the zij of al-Khwrizm (9th century) and, later on, in
tables headed Tabula bipertalis numeri and Tabula quadripertalis numeri, in the
Toledan Tables (available in Latin in the 12th century, but manuscript copies of
it only begin to proliferate in the 13th century).7

5 For the Handy Tables, see Stahlman 1959, pp. 143155, 325334; Neugebauer 1975, pp. 1006
1016; and Swerdlow 2005. For the Mumtaan zij of Yay ibn Ab Manr, see Kennedy 1956,
pp. 145147, 173; and Vernet 1956. Kennedy 1956, p. 146, indicates that these tables for the lat-
itudes of Venus and Mercury have the same parameters as the corresponding tables in the
Handy Tables, but their structure is much more primitive. The extremal values for the lati-
tudes of Venus and Mercury in the Handy Tables are reported by al-Battn, but he does not
identify his source: see Nallino 19031907, 1:116.
6 For the zij of Ibn al-Kammd, see Chabs and Goldstein 1994, pp. 3132; and Madrid, Bib-
lioteca Nacional, ms 10023, f. 45rv. We have identified another copy of this table in the zij of
Juan Gil of Burgos (14th century) that survives in a single Hebrew manuscript: see f. 139a of
what was formerly London, Beth Din, ms 135 [olim, London, Jews College, ms 135; film no. 4796
at the Institute for Microfilmed Hebrew Manuscripts, Hebrew University (Jerusalem)]. The
same table is also found among the Tables of Barcelona (14th century): see Chabs 1996, p. 505.
7 For the zij of al-Khwrizm, see Suter 1914, pp. 138167; Neugebauer 1962, pp. 101103. For the
Toledan Tables, see Toomer 1968, pp. 6970; Richter-Bernburg 1987; and F.S. Pedersen 2002,
pp. 15, 13091321. On the rule for computing planetary latitudes in the zij of al-Khwrizm
see Kennedy and Ukashah 1969, espec. p. 89. A variant of this tradition for treating planetary
latitudes is found in the zij of Ibn Azzz (Fez, 14th century) where the entries in one column
for each planet are the same as in the zij of al-Khwrizm and those in the other are the
reciprocals of the corresponding entries in the earlier zij: see Sams 1999, pp. 114; and Sams
1997, p. 92.
ptolemy, bianchini, and copernicus 75

While these two traditions were represented in some texts, the mainstream
tradition found in most medieval sets of astronomical tables dealing with plan-
etary latitudes certainly derives from the Almagest and it was transmitted to
the West primarily via the zij of al-Battn (ca. 900) and the Toledan Tables.8
As was the case in the Almagest, neither of these tables displays columns for
the deviation for Venus and Mercury. Interest in the third component of lati-
tude is found in Maghribi sources, notably in a text written by an anonymous
author in Tunisia ca. 1280, a reworking of the canons to the zij of Ibn Isq al-
Tnis (ca. 1222), based on Ibn al-Kammd. Chapter 18 of this text (uniquely
extant in Hyderabad, Andra Pradesh State Library, ms 298) presents a worked
example for the latitude of Venus where the deviation is considered. Further
references to this method for reckoning the latitudes of the inferior planets
with three components are found in the zij of Ibn al-Bann (d. 1321) which,
in turn, depends on the zij of Ibn Isq.9 Written at much the same time,
the canons to the Castilian Alfonsine Tables include explanations for using
tables to compute the latitudes of the planets. These canons were composed no
later than 1272 and they are preserved in a unique manuscript now in Madrid;
the original tables, which are not extant, had an epoch of January 1, 1252.10
For the inferior planets, Chap. 22 specifically mentions columns for deviation,
here called third latitude, for Venus (22:12) and for Mercury (22:28), as fol-
lows:

[11] Mas quando quisieres saber la latitud de Venus. entra con su entro en
las tablas de su latitud y toma la que fuere en su derecho de los minutos
proporionales de la declinaion e de los minutos proporionales del decli-
namiento. et escrive cada uno dellos a su parte y escrive sobre cada uno lo
que hallares en somo de la regla donde lo tomas de alto o de baxo.
[12] E toma otrosi lo que fuere en aquel derecho del entro en la regla de la
latitud tercera et guardala otrosi et la parte que le hallares escripta de suso
es siempre septentrional.
()

8 For the zij of al-Battn, see Nallino 19031907, 1:115116, 2:140141. For the Toledan Tables,
see Toomer 1968, pp. 7172; and F.S. Pedersen 2002, pp. 13221326.
9 For the zij of Ibn Isq al-Tnis, see Mestres 1999, pp. 5659. For the zij of Ibn al-Bann,
see Vernet 1952, pp. 96100. For planetary latitude tables in the Islamic East, see van Dalen
1999, espec. p. 323.
10 Madrid, Biblioteca Nacional, ms 3306; see Chabs and Goldstein 2003a.
76 chapter 4

[26] Mas quando quisieres saber la latitud de Mercurio entra con su


entro en las tablas de su ladeza et toma lo que fuere en su derecho de
los menudos proporionales de la su declinaion y de los menudos pro-
porionales del desviamiento.
[27] E escrivelos con sus titulos segun lo diximos en Venus.
[28] E toma otrosi con el entro su ladesa terera e escrivela con su parte
e siempre la hallaras meredional.

[11] But when you wish to know the latitude of Venus, enter the tables of its
latitude with the center and take the minutes of proportion for inclination
and the minutes of proportion for slant which are opposite it; and write
them down separately and note for each of them what you will find at the
top or at the bottom of the corresponding column.
[12] And also take what is in the column for the third latitude which is
in opposite it and keep it, and what is written is always northern.
()
[26] But when you wish to know the latitude of Mercury enter the
tables of its latitude with the center and take the minutes of proportion
for inclination and the minutes of proportion for slant which are opposite
it.
[27] And write them down with their headings, as we said for Venus.
[28] And also take with its center [as argument what is in the column
for] the third latitude and write it down with the rest, and it is always
southern.

The instructions in the text seem to refer to tables in the style of Almagest
xiii.5. However, in contrast to Ptolemys tables, the columns for the minutes
of proportion for the inclination and for the slant are not the same.
The Castilian Alfonsine Tables arrived in Paris in the early 14th century
and they began to spread in a modified form, in Latin, throughout Europe.
Among the Parisian astronomers, John Vimond, seems to be the first to have
constructed tables for the latitudes of each planet including, for Venus and
Mercury, a column for the deviation. His latitude tables are also in the style of
the Almagest with arguments at 12-intervals from 0s 12 to 12s 0.11 The inclusion
of columns for the deviation is exceedingly rare in the West and, to the best of
our knowledge, Vimonds tables are the earliest to display them. As a matter

11 Paris, Bibliothque nationale de France, ms lat. 7286c, ff. 1r8v; see Chabs and Goldstein
2003b, and Chabs and Goldstein 2004.
ptolemy, bianchini, and copernicus 77

of fact, the third component for Venus and Mercury is not tabulated in the
manuscripts of the Parisian Alfonsine Tables we have seen that date from the
14th century.
The Parisian Alfonsine Tables gave rise to a variety of tables that preserve the
basic parameters of Alfonsine astronomy but differ in presentation. In particu-
lar, the latitude tables were sometimes recast in the form of double argument
tables such that the entries display the latitude of the planet as a function of
both its anomaly and its center. Double argument tables for planetary latitudes
are found in several sets of tables: in the 14th century the Oxford Tables of 1348
ascribed to William Batecombe; and in the 15th century the tables of John of
Gmunden (Vienna), ha-ibbur ha-gadol by Abraham Zacut (Salamanca), and
the Almanach Perpetuum (based on Zacuts tables and printed in 1496 in Leiria,
Portugal).12 We have spot checked corresponding entries in both the Oxford
Tables and in the Almanach Perpetuum and it is clear that these entries all
derive from Ptolemys tables (or a minor variant of them). We can also say
that they were computed independently, based on small divergences between
corresponding entries. In particular, the entries for Venus and Mercury take
into account the deviation: when the anomaly and the center are 0, the incli-
nation and the slant are both 0 and the corresponding entries in the table
only account for the deviation. In the Almanach Perpetuum and in the Oxford
Tables, these entries are not 0, but +0;10 (Venus) and 0;45 (Mercury), which
are the standard extremal values for the deviation (see below). The deviation is
also embedded in the rest of the entries but some recomputations are required
to reveal it. Double argument tables for latitude were certainly an advance over
single argument tables such as those in Almagest xiii.5. The increase in size
offered more possibilities to the computer (Ptolemys table has 675 entries for
all 5 planets, whereas the Oxford Tables have 8,220 and those by Zacut 8,680,
of which 5,611 are for the inferior planets). But, above all, the latitude tables
with double arguments gained in user-friendliness, for with them the com-
puter could often find the latitude he sought (or, at least, a first approximation
to it) simply by inspecting the table. In this way he could avoid the tedious and
risky computations using the table where the entries in each column are a func-
tion of a single argument, looking up many entries (7 in the case of an inferior

12 For the Oxford Tables of 1348, see North 1977; for the tables of John of Gmunden, see
Porres 2003; for ha-ibbur ha-gadol and the Almanach Perpetuum, as well as a survey of
such double argument tables, see Chabs and Goldstein 2000, pp. 137143. According to
Tichenor 1967, p. 128, al-Ksh (d. 1429) also had double argument tables for planetary
latitudes. On al-Ksh, see now Kennedy 1998.
78 chapter 4

planet), each with its proper sign, and combining them (paying attention to the
algebraic signs) according to complicated rules.
John of Murs, who was also active in Paris in the early 14th century, compiled
various sets of tables, and among them are tables for the latitudes of Venus and
Mercury. For both planets we are given double argument tables for the first two
components of latitude, whereas the deviation is presented in a separate col-
umn headed, 3a latitudo.13 A contemporary of John de Lignres, also working
in Paris, John of Saxony, wrote canons to the Parisian Alfonsine Tables in 1327 in
which planetary latitudes are not even mentioned. His canons were frequently
copied in the 14th and 15th centuries, and published together with the edi-
tio princeps of the Alfonsine Tables (Ratdolt 1483). Despite their absence from
these canons, this edition of the Alfonsine Tables has tables for the planetary
latitudes but the deviation for Venus and Mercury is not taken into account.
The second edition of the Alfonsine Tables, edited by J.L. Santritter in 1492 with
a new set of canons, also has tables for the planetary latitudes which are the
same as those in the editio princeps (i.e., in both cases, the arguments are given
at intervals of 6, and the entries are the same), differing only in presentation.14
In his canons, chapters 25 and 26 (for Venus and Mercury, respectively), Santrit-
ter discussed planetary latitudes with instructions for computing the deviation
(que proveniet ex deviatione deferentis ab ecliptica) for the inferior planets. It
turns out that these instructions for computing the deviation were copied from
the canons by John of Lignres (1322), Priores astrologi motus corporum, chap-
ters 22 and 23, almost verbatim.15

13 Lisbon, Biblioteca de Ajuda, ms 52-xii-35, ff. 63r64r. The entries for the deviation for
Venus, given to minutes, are based on a maximum of +0;10, as explained in the text of
the Almagest. But those for Mercury have an extremal value of 0;23, rather than the
value in the Almagest of 0;45 (see below).
14 See Ratdolt (ed.) 1483, f. h1v; and Santritter (ed.) 1492, ff. e5r, f1r, f5r, g1r, and g5r. Rather than
keeping the components of latitude for all planets in columns in a single table, Santritter
collected columns for different phenomena (such as latitude, unequal daily motion, and
retrograde motion) in a separate table for each planet.
15 See Saby 1987, pp. 209, 211. For Venus, John of Lignres says: Postea accipe de minutis
proportionalibus in altero locorum servatis 6am partem que erit latitudo Veneris tertia
examinata que provenit ex deviatione deferentis ab ecliptica; et est semper hec tertia
latitudo septentrionalis, while Santritter (f. c2r) has: Postea accipe de minutis propor-
tionalibus in altero supra loco servatis sextam partem, que erit latitudo Veneris tertio
examinata que proveniet ex deviatione deferentis ab ecliptica, est quem [read: que] sem-
per ista latitudo septentrionalis.
ptolemy, bianchini, and copernicus 79

2 Planetary Latitude in the Tables of Bianchini

Bianchini was a prominent astronomer, active in Ferrara, whose precise dates


are not known.16 One of his earliest works, written in 1442, concerns the con-
struction and use of a surveying instrument, and it includes an explanation of
the use of the decimal point.17 In the 1460s he corresponded with Regiomon-
tanus on problems in astronomy in which Regiomontanus indicated difficulties
in Ptolemys treatment of various phenomena.18 Bianchinis tables were pub-
lished twice in Venice (ed. S. Beuilaqua 1495; and ed. L. Gaurico 1526),19 and
both editions include planetary latitudes along with planetary longitudes in
large double argument tables where the anomaly is given in days and the cen-
ter is given in degrees. The number of entries for the latitudes of all 5 planets
is 10,584, far greater than in the other double argument tables described above.
As we found in the Oxford Tables and the Almanach Perpetuum, Bianchinis
entries for the planetary latitudes are based on Ptolemys tables in Almagest
xiii.5 (with slightly different parameters), and take into account the deviation
for the inferior planets.
We now turn our attention to Bianchinis set of auxiliary tables for comput-
ing planetary latitudes in the ed. of 1526 that was not included in the editio
princeps of 1495, despite the fact that these tables appear in manuscript copies.
This set consists of two tables, one for the superior planets and one for the
inferior planets. These two tables have not been studied previously which is
somewhat surprising since they appear in a manuscript copied by Copernicus
that was printed by Curtze in 1875 and by Prowe in 1884.20 However, this copy
was not identified with the tables of Bianchini until 1984 by Rosiska.21 A brief
description of Bianchinis tables, with notes on the manuscripts in Cracow, was
also published by Rosiska, but her goal was to catalogue the tables in these
manuscripts, not to analyze their mathematical structure.22
Bianchinis first table is for the superior planets and has 10 columns (see
Table 1). The description that follows is based on Cracow, Biblioteka Jagiel-

16 Federici Vescovini 1968; Zinner 1990, p. 37.


17 Rosiska 1996, p. 57.
18 Swerdlow 1990.
19 There was a third edition, edited by N. Pruckner (Basel, 1553), but we have not seen it.
20 Curtze 1875, pp. 230238; Prowe 18831884, 2:231240. The transcription of Copernicuss
canon for the latitudes of Venus and Mercury appears in Curtze on p. 238, and in Prowe
on pp. 239240. Cf. Swerdlow and Neugebauer 1984, p. 524, n. 44.
21 Rosiska 1984b.
22 Rosiska 1984a, pp. 485486.
80 chapter 4

loska, ms 555 (ff. 237240), a manuscript dating from about 1453 and copied
in Perugia by a scholar from Cracow:23

col. 1: argument at 1-intervals, from 1 to 360;


col. 2: minutes of proportion for Saturn in minutes and seconds;
col. 3: northern latitude for Saturn in degrees and minutes, from 2;2
(for argument 360, or 0) to 3;4 (for argument 180);
col. 4: southern latitude for Saturn in degrees and minutes, from 2;1
(for argument 0) to 3;5 (for argument 180);
col. 5: minutes of proportion for Jupiter in minutes and seconds;
col. 6: northern latitude for Jupiter in degrees and minutes, from 1;6
(for argument 0) to 2;5 (for argument 180);
col. 7: southern latitude for Jupiter in degrees and minutes, from 1;4
(for argument 0) to 2;8 (for argument 180);
col. 8: minutes of proportion for Mars in minutes and seconds;
col. 9: northern latitude for Mars in degrees and minutes, from 0;5
(for argument 0) to 4;21 (for argument 180);
col. 10: southern latitude for Mars in degrees and minutes, from 0;2
(for argument 0) to 7;7 (for argument 180).

Note that the northern or southern latitudes are functions of the true anomaly,
whereas the minutes of proportion are functions of the true center. Table 1
displays an excerpt at 10-intervals from 0 to 180 of Bianchinis table for the
superior planets.
According to Rosiska, this table is found in ten manuscripts in Cracow.24 In
the printed edition of 1526, Table 1 is separated into three tables: the first is for
Saturn (324r326v) and displays, in this order, columns 1, 2, 3, and 4; the second
is for Jupiter (327r329v) and displays, in this order, columns 1, 5, 6, and 7; the
third is for Mars (330r332v) and displays, in this order, columns 1, 8, 9, and
10. The entries in columns 3, 4, 6, 7, 9, and 10 (for the northern and southern
latitudes of the three planets) follow very closely the corresponding values in
Almagest xiii.5. The three columns for the minutes of proportion (columns 2,
5, and 8) in fact are the same (but for shifts: see below), and they are based on
the column in Almagest xiii.5 for the minutes of proportion that applies to the
latitude for all planets.25 This particular column in Almagest xiii.5 is usually

23 Rosiska 1984b, p. 644. Henceforth bj, ms 555.


24 Rosiska 1984a, p. 485.
25 See Toomer 1984, pp. 632634.
ptolemy, bianchini, and copernicus 81

table 1 Bianchinis table for the planetary latitudes of the superior planets

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

0* 38;24 2; 2 2; 1 56;12 1; 6 1; 4 60; 0 0; 5 0; 2


10 30 0 2 4 2 3 58 56 1 8 1 6 58 56 0 9 0 4
20 20 24 2 6 2 4 60 0 1 8 1 6 56 12 0 12 0 6
30 10 24 2 8 2 6 58 54 1 10 1 8 52 0 0 14 0 8
40 0 0 2 11 2 8 56 12 1 12 1 10 45 44 0 17 0 11
50 10 24 2 13 2 11 52 0 1 13 1 12 38 24 0 22 0 16
60 20 24 2 16 2 15 45 44 1 16 1 16 30 0 0 28 0 22
70 30 0 2 20 2 20 38 24 1 20 1 20 20 24 0 35 0 28
80 38 24 2 25 2 25 30 0 1 25 1 25 10 24 0 42 0 38
90 45 44 2 30 2 30 20 24 1 30 1 30 0 0 0 52 0 49
100 52 0 2 35 2 35 10 24 1 35 1 35 10 24 1 3 1 2
110 56 12 2 40 2 40 0 0 1 40 1 40 20 24 1 17 1 16
120 58 56 2 45 2 45 10 24 1 45 1 45 30 0 1 34 1 37
130 60 0 2 49 2 50 20 24 1 49 1 50 38 24 1 56 2 3
140 58 56 2 54 2 54 30 0 1 53 1 55 45 44 2 22 2 41
150 56 12 2 57 2 58 38 24 1 58 2 0 52 0 2 55 3 29
160 52 0 3 0 3 1 45 44 2 1 2 4 56 12 3 30 4 39
170 45 44 3 1 3 3 52 0 2 3 2 6 58 56 4 6 6 0
180 38 24 3 4 3 5 56 12 2 5 2 8 60 0 4 21 7 7

* The entries for 0 appear in the table as entries for 360.

referred to as c5(), for it is the fifth column in that table. Note that is the
argument of latitude counted from the northern limit on the deferent (i.e., 90
from the nodes where the deferent crosses the ecliptic) and that no algebraic
signs are associated with the coefficients c5(). In this paper c5() will be called
p(x), and Bianchini uses it for various purposes of interpolation. We note that
p(x) |cos x|.26
In the instructions to compute the planetary latitudes for the superior plan-
ets (Almagest xiii.6), Ptolemy indicates that the northern limits on the def-
erent differ in each case from the apogees of the superior planets by +50
(Saturn), 20 (Jupiter), and 0 (Mars). For example, according to Ptolemy, the

26 Neugebauer 1975, p. 219. The entries in Ptolemys col. 5 are very nearly those of a cosine
function; the divergence from the values of the cosine reaches a maximum of 0;0,11 (where
cos 0 is taken to be 1) at about 42.
82 chapter 4

ascending node for Saturn is at longitude 90, i.e., the northern limit (n) is at
longitude 180, and the apogee (a) of Saturns deferent is 230 (rounded from
233).27 Hence

a n = 50.

Bianchinis table takes these differences into account: the entries in column
c2 for Saturn are shifted 50 with respect to the corresponding entries in the
Almagest; column c5 for Jupiter +20; and column c8 for Mars has no shift. Thus,
if p(x) is an entry in the table in the Almagest for argument x, and ci(x) an entry
in col. i in Bianchinis table, then

p(x) = c2(x 50) = c5(x + 20) = c8(x).

Bianchinis inclusion of these shifts makes his table more user-friendly than
Ptolemys since there is one less step for the user who can enter the column for
interpolation with the true center in all cases. In other words, there is no longer
any need to compute an argument of latitude.
The second table is for the inferior planets and, as given in Cracow, Bib-
lioteka Jagielloska, ms 555 (ff. 241244), it has 11 columns (see Table 2):

col. 1: argument at 1-intervals, from 1 to 360;


col. 2: inclination (declinatio) for Venus in degrees and minutes with extremal
values 1;3 (for argument 360, or 0) and 7;22 (for argument 180);
col. 3: slant (reflexio) for Venus in degrees and minutes with extremal value
2;30 (for arguments 131139 and 221229);
col. 4: minutes of proportion for the inclination of Venus in minutes and
seconds;
col. 5: minutes of proportion for the slant of Venus in minutes and seconds;
col. 6: deviation for Venus in minutes and seconds, ranging from +0;10,0 (for
arguments 0 and 180) to 0;0 (for arguments 90 and 270);
col. 7: inclination (declinatio) for Mercury in degrees and minutes with ex-
tremal values 1;46 (for argument 0) and 4;4 (for argument 180);
col. 8: slant (reflexio) for Mercury in degrees and minutes with extremal value
2;30 (for arguments 112117 and 243247);
col. 9: minutes of proportion for the inclination of Mercury in minutes and
seconds;

27 Neugebauer 1975, p. 208.


ptolemy, bianchini, and copernicus 83

col. 10: minutes of proportion for the slant of Mercury in minutes and seconds;
col. 11: deviation for Mercury in minutes and seconds, ranging from 0;45,0
(for arguments 0 and 180) to 0;0 (for arguments 90 and 270).

The inclination and slant are functions of the true anomaly, whereas the min-
utes of proportion and the deviation are functions of the true center. Table 2
displays an excerpt at 10-intervals from 0 to 180 of Bianchinis table for the
inferior planets.
According to Rosiska, this table is found in eight manuscripts in Cracow.28
In the printed edition of 1526, Table 2 is separated into two tables. The first is
for Venus (333r338v) and displays, in this order, columns 1, 2, 3, 4, 5, and 6; the
second is for Mercury (339r344v) and displays, in this order, columns 1, 7, 8, 9,
10, and 11.
We note that there is a column for the minutes of proportion (col. 10),
not found in Almagest xiii.5, but the three other columns for the minutes of
proportion (columns 4, 5, and 9) are the same (but for shifts) as that for the
minutes of proportion in the planetary latitude tables of Almagest xiii.5, as
was the case for the superior planets. Nevertheless, entries in columns c4 and
c9 are shifted +90 with respect to the corresponding entries in the Almagest.
Thus, if p(x) is an entry in the table in the Almagest, then

p(x) = c4(x + 90) = c5(x) = c9(x + 90).

Therefore, column 5 in Table 2 is identical to column 8 in Table 1, and both


are the same as the column for the minutes of proportion in Almagest xiii.5.
Again, the shifts are introduced to facilitate computation. In fact, this is also
the reason for repeating the same column (but for shifts) for each planet.
The entries in column 10, the minutes of proportion for the slant of Mercury,
decrease from 54;0 (at 0) to 0;0 (at 90), increase to 66;0 (at 180), decrease to
0;0 (at 270), and increase back to 54;0 (at 360). There is no indication of the
algebraic sign, and they may all be taken to be positive. The extremal values
for this correction are indeed 110 more and 110 less than 60 minutes, which
corresponds to the instructions given by Ptolemy in Almagest xiii.6 (but there
is no corresponding table in the Almagest). It is readily seen that the entries in
col. 10 can be computed from p(x), as follows:

9
c10(x) = 10 c5(x) for 0 x 90, 270 x 360

28 Rosiska 1984a, p. 486.


84 chapter 4

table 2 Bianchinis table for the planetary latitudes of the inferior planets

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)

0* 1; 3 0; 0 0; 0 60; 0 10; 0 1;46 0; 0 0; 0 54; 0 45; 0


10 1 1 0 14 10 24 58 56 9 49 1 44 0 18 10 24 53 2 44 12
20 1 0 0 27 20 24 56 12 9 22 1 42 0 36 20 24 50 36 42 9
30 0 57 0 41 30 0 52 0 8 40 1 36 0 55 30 0 46 48 39 0
40 0 52 0 55 38 24 45 40 7 38 1 26 1 13 38 24 41 10 34 18
50 0 45 1 8 45 44 38 24 6 24 1 13 1 29 45 44 34 33 28 48
60 0 36 1 20 52 0 30 0 5 0 0 59 1 44 52 0 27 0 22 30
70 0 25 1 33 56 12 20 24 3 24 0 42 1 57 56 12 18 22 15 18
80 0 14 1 45 58 56 10 24 1 44 0 22 2 9 58 56 9 22 7 48
90 0 0 1 57 60 0 0 0 0 0 0 0 2 20 60 0 0 0 0 0
100 0 16 2 7 58 56 10 24 1 44 0 25 2 27 58 56 11 26 7 48
110 0 36 2 17 56 12 20 24 3 24 0 54 2 29 56 12 22 26 15 18
120 0 59 2 25 52 0 30 0 5 0 1 25 2 29 52 0 33 0 22 30
130 1 25 2 29 45 44 38 24 6 24 2 0 2 23 45 44 42 15 28 48
140 2 7 2 29 38 24 45 44 7 38 2 33 2 8 38 24 50 18 34 18
150 3 3 2 22 30 0 52 0 8 40 3 7 1 45 30 0 57 12 39 0
160 4 12 2 1 20 24 56 12 9 22 3 38 1 17 20 24 61 50 42 9
170 5 32 1 15 10 24 58 56 9 50 3 58 0 40 10 24 64 50 44 13
180 7 22 0 0 0 0 59 0 9 0 4 4 0 0 0 0 66 0 45 0

* The entries for 0 appear in the table as entries for 360.


c5(180). bj, ms 555, reads 59;0 instead of 60;0 (as in ed. 1526).
c6(180). bj, ms 555, reads 9;0 instead of 10;0 (as in ed. 1526).
c5(40). bj, ms 555, reads 45;40 instead of 45;44 as in c4(50), c4(130), c5(140), c9(50),
and c9(130). This is just one example of miscopying in the table as it is found in this
manuscript; we have not indicated the others.
The number of errors in these two tables is low despite the fact that they contain
more than 7,500 entries (360 10 + 360 11), most of them with two sexagesimal
digits.

and

11
c10(x) = 10 c5(x) for 90 x 270.

As for the inclination and slant of the two inferior planets, the entries in
columns 2, 3, 7, and 8 follow very closely the corresponding values in Almagest
ptolemy, bianchini, and copernicus 85

xiii.5. Now, in Table 2, columns 6 and 11 give the deviation for Venus and
Mercury, respectively: the entries in col. 6 show an extremal value of +0;10,0
and those in col. 11 an extremal value of 0;45,0.29 It is readily seen that both
columns 6 and 11 can be computed from p(x), the fifth column in Almagest
xiii.5 or, equivalently, from c5(x) in Table 2, where the following holds:

c6(x) = +0;10 c5(x)

and

c11(x) = 0; 45 c5(x).

In sum, all the entries in Bianchinis Tables 1 and 2 ultimately derive from
Almagest xiii.5, but Bianchini presented them differently. In particular, he
introduced one column for Venus (Table 2, col. 6) and two columns for Mercury
(Table 2, cols. 10 and 11) which are not found in Almagest xiii.5.
It is almost impossible to determine the specific source used by Bianchini
in compiling his tables for planetary latitudes because, globally, all such tables
in the tradition of the Almagest share the same entries (with minor variants),
except in some cases for the extremal values and entries near them. Table 3
displays the significant extremal entries in various sets of tables, all of them in
the style of the Almagest (e.g. Oxford, Bodleian Library, ms Can. Misc. 27, con-
taining tables by John of Lignres), except for Lisbon, ms Ajuda (mentioned
above), and the Oxford Tables of 1348,30 which are in the form of double argu-
ment tables. A glance at this table suggests that Bianchini may have depended
on a copy of the Toledan Tables.
The variation in extremal values from one set of tables to another suggests
more variation in the other entries than is the case. For example, in the column
for the inclination of Venus in different sets of tables, most of the entries are the
same. In Table 4 a sample of such entries is presented (Ptolemy tabulated the
inclination of Venus at 6-intervals from 6 to 90, and at 3 intervals from 93
to 180; the Alfonsine Tables [ed. 1483 and ed. 1492] at 6-intervals from 6 to
180; and Bianchini at 1-intervals from 1 to 360).
Both editions of Bianchinis tables are preceded by canons explaining their
use. Chapter 34 includes short comments on planetary latitudes and worked
examples for Jupiter, Venus, and Mercury. In the case of the inferior planets,

29 For an explanation of these parameters, see Neugebauer 1975, pp. 222224.


30 For these tables, we have consulted Vienna, Nationalbibliothek, ms Vin. 2440.
86 chapter 4

table 3 Extremal values in tables for planetary latitudes

Toledan Can. Ajuda Oxford Alf. t.


Almagest Tables Misc. 27 52-xii-35 Tables Bianchini 1483

Sat. n. 3; 2 3; 4 3; 2 3; 2 3; 2 3; 4 3; 3 c
s. 3; 5 3; 5 3; 5 3; 5 3; 4 3; 5 3; 5
Jup. n. 2; 4 2; 5 a 2; 5 2; 5 2; 5 2; 5 2; 8
s. 2; 8 2; 8 2; 8 2; 8 2; 8 2; 8 2; 8
Mars n. 4;21 4;21 4;21 4;21 4;31 4;21 4;21
s. 7; 7 7; 7 7;30 6;30 7; 7 7; 7 7;30
Ven. Incl. 6;22 6;22 a 7;22 7;22 7;22 7;22 b 7;12
Slant 2;30 2;30 2;30 2;25 2;30 2;30
Mer. Incl. 4; 5 4; 4 4; 5 4; 5 4; 5 4; 4 4; 5
Slant 2;30 2;30 2;30 2;44 2;30 2;30

a. Other mss read 2;4 for Jupiter and 7;22 for Venus (Pedersen 2002, p. 1326).
b. bj, ms 555: 7;22; ed. 1526: 7;12.
c. Santritters ed. (1492) reads 3;2.

Bianchini used the terms declinatio, reflexio, and deuiatio for the three com-
ponents of latitude, but we note that the edition of 1495 renders the first term
as declaratio, a mistake that was corrected in the edition of 1526. We present
Bianchinis worked examples for Venus and Mercury, without following them
word-for-word.
In the example for Venus the true anomaly = 258 and the true center
= 121. For this , Table 2 gives c2() = 0;20 (inclination) and c3() = 2;9
(slant). For this , Table 2 gives c4() = 51;24 (minutes of proportion for the
inclination), c5() = 30;52 (minutes of proportion for the slant), and c6() =
0;5,9 = 3 (deviation). The three components of latitude are 1 = c2() c4() =
0;17,8, 2 = c3() c5() = 1;6,22, and 3 = 0;5,9. The latitude of Venus is thus:
= 1 + 2 + 3 = 0;54,23, in agreement with the edition of 1495 (although the
edition of 1526 gives the result as 0;54,22).31
In the example for Mercury = 224 and = 189. Table 2 gives c7() = 2;20
(inclination), c8() = 2;15 (slant), c9() = 9;24 (minutes of proportion for the

31 If Bianchini had followed the instructions in the Almagest (see below), he would have
multiplied the entry in col. 6 by c5() and found 3 = 0;5,9 0;30,52 = 0;2,39; the final
result would have been = 0;50,53.
ptolemy, bianchini, and copernicus 87

table 4 Inclination of Venus

Argument Almagest xiii.5 Bianchini Alf. t. 1483

6 1; 2 1; 2 1; 2
12 1; 1 1; 1 1; 1
18 1; 0 1; 0 1; 0
24 0;59 0;59 0;59
30 0;57 0;57 0;57
36 0;55 0;55 0;55
42 0;51 0;51 0;51
48 0;46 0;46 0;46

150 3; 3 3; 3 3; 3
153 3;23 3;24
156 3;44 3;44 3;43
159 4; 5 4; 5
162 4;26 4;26 4;26
165 4;49 4;49
168 5;13 5;13 5;24
171 5;36 5;42
174 5;52 6;12 6;24
177 6; 7 6;46
180 6;22 7;22 a 7;12

a. bj, ms 555: 7;22; ed. 1526: 7;12.

inclination), c10() = 65;1 (minutes of proportion for the slant), and c11() =
0;44,21 (deviation). The three components of latitude are 1 = c7() c9()
= 0;22, 2 = c8() c10() = 2;26, and 3 = 0;44. The latitude of Mercury is
thus: = 1 + 2 + 3 = 3;32, and this is the value computed by Bianchini.32
Bianchini (as well as Vimond, John of Lignres, and Batecombe before him,
Zacut at about the same time, and Santritter afterwards) assumed that the
formula for the deviation is

32 Analogously, if Bianchini had followed the instructions in the Almagest (see below), he
would have computed 3 = c5() c6() and found 3 = 0;44; in this case his final result
would have been the same.
88 chapter 4

(1) 3 = c5(x) d

where d = 0;10 for Venus and 0;45 for Mercury. However, Ptolemys model
requires the formula

(2) 3 = c5(x) c5(x) d

and the instructions in Almagest xiii.6 indicate this (although Ptolemy ex-
pressed himself in a rather convoluted way):

Then we take these same sixtieths which were found by the second entry
with the longitude, calculate the amount which is the same fraction of
them as they are of 60, and, for Venus, take 16 of this and set it out
too, always with a northerly direction; but for Mercury we take of the
amount and set it out, always in a southerly direction.33

The critical phrase is calculate the amount which is the same fraction of them
as they are of 60. Ptolemy seems to indicate taking a fraction of a fraction which
is the product of the fractions; in this case, the fractions being the same, the
result is the square of the fraction used to compute 2 for Venus, i.e., the square
of the sixtieths (c5 c5). In the medieval Latin translation of the Almagest by
Gerard of Cremona (d. 1187), based on an Arabic version (rather than the Greek
original), this passage appears as follows:

Deinde post illud tendemus ad hec minuta eadem etiam que invenimus
mittendo longitudinem secundo, et accipiemus ex eis secundum quanti-
tatem partis qua ipsa sunt ex sexaginta partibus, et eius quod provenerit
in Venere accipiemus semper sextam et firmabimus in septentrione; et in
Mercurio semper accipiemus medietatem et quartam et firmabimus in meri-
die.34

After this we shall attend to the same minutes which we found by entering
with the longitude for the second time, and we shall take [a part] of them
according to the size of the part which they themselves are of 60 parts,
and of that which results in [the case of] Venus, we shall always take a

33 Toomer 1984, p. 636; cf. Swerdlow and Neugebauer 1984, p. 523.


34 Liechtenstein (ed.) 1515, f. 149v. On Gerard of Cremonas translation, see Kunitzsch 1974,
pp. 83112.
ptolemy, bianchini, and copernicus 89

sixth and write it down35 as northern; and in [the case of] Mercury we
shall always take one half plus one quarter and put it down as southern.

Since this translation is reasonably faithful to the original Greek, it cannot be


the source for Bianchinis computation of the deviation according to formula
(1).
In contrast to the western tradition, eastern Islamic astronomers under-
stood Ptolemys instructions to require computations of the deviation for Venus
and Mercury according to formula (2).36
The western tradition for computing 3, accepted by Bianchini, seems to
have its origin in the zij of al-Battn.37 In the canons to the Toledan Tables,
which depend on al-Battn in this respect, the instruction is:

After this, if your computation is for Venus, take one sixth of the minutes
you wrote in the other place, and this [sixth] is always northern; and if it
is for Mercury, take one quarter plus a half [of them], and this is always
southern.38

This provides additional evidence in support of the suggestion that Bianchini


depended on the Toledan Tables and their canons although he replaced the
rule given in words with a table.
For those using formula (1), the coefficient c5 was always taken to be positive
and, as we noted above, c5(x) |cos x|. But this function has a singularity at

35 Presumably the underlying Arabic word is a derived form of athbata which means to write
down although the root means to be firm; see Nallino 19031907, 2:325, and n. 38, below.
36 We have checked the zij al-Sanjar (ca. 1120) by al-Khzin (London, British Library, ms Or.
6669, ff. 148v, 152r; and Vatican, ms Arab. 761, ff. 179v, 186r), and the zij al-Khqn (ca. 1420)
by al-Ksh (Istanbul, Aya Sofia ms 2692, ff. 98v, 99v; and London, India Office, ms 430,
ff. 139v, 140v). In both zijes the deviation, called first latitude, was computed according to
formula (2).
37 Nallino 19031907, 1:116.
38 F.S. Pedersen 2002, pp. 268269: Post haec accipe ex ipsis minutis quae scripsisti in alio
loco, si fuerit numerus tuus Veneris, sextam partem eorum, et est semper septentrionalis;
et si fuerit Mercurii, quartam partem [eius] et dimidiam, et est semper meridiana; cf.
p. 515 [ 05]. Ibn al-Bann gives the same instructions: Take a sixth of the minutes of
proportion that you wrote down (athbata) in the second place, and it is the third latitude,
and it is always to the north [for Venus], and keep it (Vernet 1952, Arabic text, p. 38);
and these instructions are also found, almost verbatim, in the anonymous zij preserved
in Hyderabad, Andra Pradesh State Library, ms 298 (see Mestres 1999, Arabic text, p. 91:
chapter 18:12).
90 chapter 4

90 which has no physical meaning. On the other hand, the function y = cos2
x does not exhibit such a singularity. Figure 4.1 shows the behavior of the
functions represented by formulas (1) and (2).
Indeed, in De revolutionibus vi.8, Copernicus comments on this very point,
arguing that if c5 is the coefficient for the deviation (rather than the square of
c5), then for Venus and Mercury the angle between the deferent and the eclip-
tic would be constant (contrary to Ptolemys model for the inferior planets) and
this [component of] latitude () would suddenly leap back from the intersec-
tion [i.e., the nodal line] into the same latitude that it previously left.39 Clearly,
at the time when this passage was written Copernicus understood the behav-
ior of the deviation properly, but his approach had been different in previous
discussions of the latitude for the inferior planets.40

3 Tables for Planetary Latitude in Copernicus

Uppsala, University Library, ms Copernicana 4, contains a set of ten tables for


planetary latitudes in the hand of Copernicus: ff. 276v277v [superior planets];
278r279r, 280r [inferior planets]; 279v [all planets]; and 280v281r [canons
for the inferior planets only]. These tables are included in a quire bound
together with the 1492 edition of the Alfonsine Tables (edited by Santritter)
and Regiomontanuss Tabulae directionum (Augsburg, 1490), and it is almost
certain that they were copied at the time when Copernicus was a student in
Cracow.41 Comparison with the tables of Bianchini described above shows that
Copernicus copied and rearranged these tables that had been compiled by his
Italian predecessor.42 A brief description of the 10 tables follows, indicating the
corresponding table of Bianchini:

c1. Saturn ( f. 276v)


There are two sub-tables with a column for the argument at 1-intervals, from
1to 30, and six columns, one for each of the zodiacal signs. One sub-table
displays the entries in Bianchinis Table 1, col. 3 (northern latitude), and the

39 See Swerdlow and Neugebauer 1984, p. 523; Copernicus 1543, ff. 191v192r.
40 Cf. Swerdlow and Neugebauer 1984, pp. 508, 536.
41 For a description of this manuscript see Czartoryski 1978, p. 366.
42 Note that Copernicus has arguments at 1-intervals in agreement with Bianchini, in con-
trast to the 6-intervals in the corresponding tables edited by Ratdolt (1483) and by
Santritter (1492).
ptolemy, bianchini, and copernicus 91

figure 4.1 The functions y =| cos x| and y = cos2 x in the interval (0, 180),
where the cosine function represents c5(x) in formulas (1) and
(2). The maximum difference between these functions is 0;15 at
60 and 120

other, those in Table 1, col. 4 (southern latitude). The last few entries for
northern latitudes are not exactly the same as those in Bianchinis table: the
maximum value given by Copernicus is 3;3, and not 3;4, but this is probably
due to miscopying (or a faulty archetype).

c2. Jupiter ( f. 277r)


There are two sub-tables presented in the same way as the corresponding
table for Saturn. One sub-table displays the entries in Bianchinis Table 1, col. 6
(northern latitude), and the other, those in Table 1, col. 7 (southern latitude).

c3. Mars ( f. 277v)


There are two sub-tables presented in the same way as the corresponding tables
for Saturn and Jupiter. One sub-table displays the entries in Bianchinis Table 1,
col. 9 (northern latitude), and the other, those in Table 1, col. 10 (southern
latitude). In the latter, the entries for the minutes corresponding to arguments
151180 are all mistakenly shifted downwards 1, so that, at the end of the
column, Copernicus had to write the last number outside the frame of the table.

c4. Venus ( f. 278r)


This table has a column for the argument at 1-intervals, from 1 to 30, and
six double columns, one for each of the first six zodiacal signs. In each double
92 chapter 4

column we are given the entries in Bianchinis Table 2, columns 2 (inclination)


and 3 (slant). The heading for the left side of the table is Prima pars tabule, and
that for the right side, Altera pars.

c5. Mercury ( f. 278v)


This table is headed Tabella latitudinis Mercurii. It has a column for the argu-
ment at 1-intervals, from 1 to 30, and six double columns, one for each of the
first six zodiacal signs. In each double column we are given the entries in Bian-
chinis Table 2, columns 7 (inclination) and 8 (slant). The heading for the left
side of the table is Superior pars circuli, and that for the right side, Inferior pars
circuli.

c6. Venus ( f. 279r)


This table is headed northern deviation (deviacio borealis). It has a column for
the argument at 1-intervals, from 0 to 30, and three columns, one for each
of the first three zodiacal signs. The entries are taken from Bianchinis Table 2,
column 6 (deviation).

c7. Mercury ( f. 279r)


This table is headed southern deviation (deviacio australis). It has a column for
the argument at 1-intervals, from 0 to 30, and three columns, one for each
of the first three zodiacal signs. The entries are taken from Bianchinis Table 2,
column 11 (deviation).

c8. Venus and Mercury ( f. 279r)


This table is headed Minuta ad declinacionem. It has a column for the argument
at 1-intervals, from 0 to 30, and three columns, one for each of the first three
zodiacal signs. The entries are taken from Bianchinis Table 2, column 4, which
is identical to column 9.

c9. The Five Planets ( f. 279v)


This table is headed Tabella M(inutorum) proporcionabi(lium) 5 planeta(rum).
It displays the minutes of proportion, and it has a column for the argument at
1-intervals, from 0 to 30, and three columns, one for each of the first three
zodiacal signs. The entries are taken from Bianchinis Table 1, column 8, which
is identical to Table 2, column 5. Part of the column for 0s/6s and 5s/11s has
been crossed out, and the first 18 entries have been accurately copied in the left
margin. Note that Copernicus correctly gives the entry for 40 (45;44) rather
than 45;40 (see Table 2).
ptolemy, bianchini, and copernicus 93

c10. Mercury ( f. 280r)


This table is headed M(inuta) prop. ad reflectionem, and it has a column for the
argument at 1-intervals, from 0 to 30, and six columns, one for each of the
first six zodiacal signs. The entries are taken from Bianchinis Table 2, column 10.
The heading for the left side of the table is Prima pars ta(bule), and that for the
right side, Altera pars.

In his set of 10 tables, Copernicus gathered all information contained in Bian-


chinis Tables 1 and 2, avoiding all unnecessary repetitions by taking advantage
of the numerous symmetries they contain. This is not to say that he presented
his tables in a more compact way, for Copernicus split the two tables given by
Bianchini into ten, and treated the shifts differently. Moreover, the extremal
values in Copernicuss tables are the same as those of Bianchini in Table 1 (the
one exception for Saturn has already been noted above).
Copernicuss canons for Venus and Mercury are not very different from
those that we have discussed, beginning with the canons to the Toledan Tables.
In particular, he accepts formula (1) for computing 3. Hence, it is not easy
to decide whether he depended on Santritters canons in the edition of the
Alfonsine Tables (1492) with which these tables are bound,43 or on Bianchinis
canons, or on some other set of canons. Copernicus expresses himself in his
own way and so linguistic criteria are not sufficient to establish filiation. Santrit-
ters canons for 3 do not refer to a table; rather, they give instructions for taking
fractional parts of p(x) for Venus and for Mercury (with the directions north
and south, respectively), as in the canons to the Toledan Tables.44 Copernicus,
however, refers to tables of the kind he displays and does not give rules with
the fractions, a sixth or three-quarters. This suggests that Copernicus is closer
to Bianchini than to Santritter, although it is not impossible that he relied on a
different source.

43 Santritter (ed.) 1492, f. c2rv.


44 For Mercury, Santritters instructions (again taken, almost verbatim, from John of Lign-
res) suggest computing with a quarter plus a half of a quarter (i.e., 38) rather than a half
and quarter (34). In this case, it would have been difficult for Copernicus (or any other
reader of Santritters canons) to compute the third latitude of Mercury correctly without
checking some other relevant text. See Santritter (ed.) 1492, f. c2v: Et accipe de postea
minutis proportionalibus in altero locorum servatis quartam partem et dimidium quarte,
que pars cum suo dimidio est latitudo Mercurii tertio examinata, que est semper merid-
ionalis. Where Santritter has que pars cum suo dimidio, John of Lignres has que 4a pars
cum suo dimidio (Saby 1987, p. 211). Note that John of Lignress instruction for using the
coefficient 38 (= 0;22,30) is consistent with the extremal value in his table for the deviation
of Mercury: 0;23 (see note 13, above).
94 chapter 4

In contrast to these student notes, Copernicus introduced a new theory for


planetary latitudes in his Commentariolus (ca. 1512), and modified it in his De
revolutionibus (1543). His aim was to transform Ptolemys geocentric models
into heliocentric models, but his success was far from complete.45 There are
no tables in the early work, but among the tables for planetary latitude in De
revolutionibus is a column that displays p(x)2 for the deviation.46
Despite the misunderstanding of the deviation for Venus and Mercury by
medieval and early modern astronomers, Georg Peurbach (d. 1461) correctly
described the models in the Almagest for the latitudes of both superior and
inferior planets, including the deviation: see his Theoricae novae planetarum
(first printed by Regiomontanus in Nuremberg in 1472). But this work does not
contain any tables.47

4 Conclusion

In the West the deviation for Venus and Mercury was generally not included
in tables for planetary latitudes and it was probably ignored by many users of
those tables as well. Bianchini was one of a very small number of astronomers
working in the Latin tradition of the Almagest who called attention to this third
component of latitude for Venus and Mercury and produced tables for it. Both
the zij of al-Battn and the canons to the Toledan Tables give instructions for
computing the deviation, but no tables accompany them. In the 13th century
the Castilian canons to the Alfonsine Tables mention the deviation, but the
tables associated with these canons are not extant. In the 14th century we
have identified John Vimond as having compiled tables for the deviation. John
of Lignres also compiled tables for the deviation but, as we noted, with an
idiosyncratic extremal value for Mercury; he also included instructions for
computing the deviation in his Priores astrologi but these instructions do not
refer to a table. In the 15th century, Peurbach described Ptolemys model for
the latitudes of Venus and Mercury correctly although he did not produce any
tables for this purpose. Moreover, Santritter gave instructions for computing
the deviation for Venus and Mercury, but did not display any table for it.
We have seen that prior to 1500 Ptolemys instructions for computing the

45 See Swerdlow 1973, pp. 482489, 494499, 509510; and Swerdlow and Neugebauer 1984,
pp. 483537, espec. pp. 535537.
46 Copernicus 1543, ff. 194v195r. See Swerdlow and Neugebauer 1984, pp. 530535.
47 Peurbach 1472, reprinted in Schmeidler (ed.) 1972, pp. 783787; Aiton 1987, pp. 3435.
ptolemy, bianchini, and copernicus 95

deviation were usually misunderstood in the West, and computations based,


e.g., on Bianchinis worked examples would yield slightly incorrect results.
In his student days Copernicus accepted a table based on formula (1) for
the deviation, but in his magnum opus of 1543 he argued that the previous
interpretation of Ptolemys instructions (which he once shared) by earlier
astronomers was faulty, and that one needed to invoke formula (2) for the
deviation of Venus and Mercury.
Given the medieval tradition to which he adhered, Bianchini compiled a set
of user-friendly tables that simplified the computations required for using
the Alfonsine Tables. Many astronomers in the late Middle Ages (e.g., John
of Lignres, William Batecombe, John of Gmunden, and Abraham Zacut) also
had this as a goal (each interpreting it in his own way), and Bianchini fits
nicely in this group. Finally, there can be little doubt that early in his career
Copernicus depended on Bianchinis tables for planetary latitude which, in
turn, are based on Ptolemys models in the Almagest. Hence, Bianchinis tables
can be considered a source for Copernicuss knowledge of astronomy.

Acknowledgments

We are grateful to Montserrat Lamarca and Neus Verger (Biblioteca de la Uni-


versitat de Barcelona, Spain) for providing us with a copy of the Tables of Bian-
chini (ed. 1526); to the Biblioteka Jagielloska (Cracow) for a partial copy of
ms 555; to Noel M. Swerdlow for a partial copy of Uppsala, University Library,
ms Copernicana 4; and to Benno van Dalen for partial copies of the Sanjar zij
by al-Khzin and the Khqn zij by al-Ksh. We also thank Fritz S. Pedersen
for supplying us with the text and a preliminary translation of the passage in
the Latin version of the Almagest (ed. 1515), Julio Sams for sending us a partial
copy of Ibn Azzzs zij, and Alan C. Bowen for comments on the instructions
for computing the deviation by John of Lignres and J.L. Santritter.

References

Aiton, E.J. 1987. Peurbachs Theoricae novae planetarum: A Translation with Commen-
tary, Osiris, 3:544.
Beuilaqua, S. (ed.) 1495. Tabularum Joannis Blanchini Venice: Augustinus Morauus.
Bianchini. See Beuilaqua 1495; Gaurico 1526; and Pruckner 1553.
Boffito, G. 1908. Le Tavole astronomiche di Giovanni Bianchini, La Bibliofilia, 9:378
388, 446460.
96 chapter 4

Casulleras, J. and J. Sams (eds.) 1996. From Baghdad to Barcelona: Studies in the Islamic
Exact Sciences in Honour of Prof. Juan Vernet. Barcelona.
Chabs, J. 1996. Astronoma andalus en Catalua: Las Tablas de Barcelona, in J. Ca-
sulleras and J. Sams (eds.) 1996, pp. 477525.
Chabs, J. and B.R. Goldstein 1994. Andalusian Astronomy: al-Zj al-Muqtabis of Ibn al-
Kammd, Archive for History of Exact Sciences, 48:141.
Chabs, J. and B.R. Goldstein 2000. Astronomy in the Iberian Peninsula: Abraham Zacut
and the Transition from Manuscript to Print. Transactions of the American Philo-
sophical Society, 90.2. Philadelphia.
Chabs, J. and B.R. Goldstein 2003a. The Alfonsine Table of Toledo. Dordrecht.
Chabs, J. and B.R. Goldstein 2003b. John Vimond and the Alfonsine Trepidation
Model, Journal for the History of Astronomy, 34:163170.
Chabs, J. and B.R. Goldstein 2004. Early Alfonsine Astronomy in Paris: The Astronom-
ical Tables of John Vimond (1320), Suhayl, 4:207294. [See essay 8, below]
Copernicus, N. 1543. De revolutionibus. Nuremberg.
Curtze, M. 1875. Reliquiae Copernicanae, Zeitschrift fr Mathematik und Physik, 20:
221248.
Czartoryski, P. 1978. The Library of Copernicus, Studia Copernicana, 16:355396.
Dalen, B. van 1999. Tables of Planetary Latitude in the Huihui li: Part ii, in Yung Sik
Kim and Francesca Bray (eds.), Current Perspectives in the History of Science in East
Asia. Seoul, pp. 316329.
Federici Vescovini, G. 1968. Bianchini, Giovanni, Dizionario Biografico degli Italiani,
10:194 196.
Gaurico, L. (ed.) 1526. Tabule Joa. Blanchini bononiensis Venice: Lucas Antonius
Giunta.
Kennedy, E.S. 1956. A Survey of Islamic Astronomical Tables. Transactions of the Ameri-
can Philosophical Society, 46.2. Philadelphia.
Kennedy, E.S. 1998. On the Contents and Significance of the Khqn Zj by Jamshd
Ghiyth al-Dn al-Ksh. Frankfurt.
Kennedy E.S. and W. Ukashah 1969. Al-Khwrizms Planetary Latitude Tables, Cen-
taurus, 14: 8696. Reprinted in King and Kennedy (eds.) 1983, pp. 125135.
King, D.A. and M.H. Kennedy (eds.) 1983. Studies in the Islamic Exact Sciences by E.S.
Kennedy, his colleagues, and former students. Beirut.
Kunitzsch, P. 1974. Der Almagest: Die Syntaxis Mathematica des Claudius Ptolemus in
arabisch-lateinischer berlieferung. Wiesbaden.
Liechtenstein, P. (ed.) 1515. Almagestu[m] Cl. Ptolemei Pheludiensis Alexandrini astrono-
mo[rum] principis. Venice.
Mestres, A. 1999. Materials Andalusins en el Zj dIbn Isq al-Tnis. University of
Barcelona (Ph.D. dissertation).
Nallino, C.A. 19031907. Al-Battn sive Albatenii Opus Astronomicum, 2 vols. Milan.
ptolemy, bianchini, and copernicus 97

Nauta, L. and A. Vanderjagt (eds.) 1999. Between Demonstration and Imagination: Essays
in the History of Science and Philosophy Presented to John D. North. Leiden.
Neugebauer, O. 1962. The Astronomical Tables of al-Khwrizm. Copenhagen.
Neugebauer, O. 1975. A History of Ancient Mathematical Astronomy. Berlin.
North, J.D. 1977. The Alfonsine Tables in England, in Y. Maeyama and W.G. Salzer,
(eds.), Prismata: Festschrift fr Willy Hartner. Wiesbaden, pp. 269301. Reprinted in
North 1989, pp. 327359.
North, J.D. 1989. Stars, minds and fate: Essays in ancient and medieval cosmology. London
and Ronceverte.
Pedersen, F.S. 2002. The Toledan Tables: A review of the manuscripts and the textual
versions with an edition. Copenhagen.
Pedersen, O. 1974. A Survey of the Almagest. Odense.
Peurbach, G. 1472. Theoricae novae planetarum. Nuremberg. Reprinted in Schmeidler
(ed.) 1972, pp. 755793.
Porres, B. 2003. Les tables astronomiques de Jean de Gmunden: dition et tude compara-
tive. Paris: cole pratique des hautes tudes (Ph.D. dissertation).
Prowe, L. 18831884. Nicolaus Coppernicus, 2 vols. (v. 1 is in two parts). Berlin. Reprinted
Osnabrck 1967.
Pruckner, N. (ed.) 1553. Luminarium atque planetarum motuum tabulae octoginta quin-
que Basel.
Ratdolt, E. (ed.). 1483. Tabulae astronomice illustrissimi Alfontij regis castelle. Venice.
Richter-Bernburg, L. 1987. id, the Toledan Tables, and Andalus Science, Annals of
the New York Academy of Sciences, 500:373401.
Riddell, R.C. 1978. The Latitudes of Venus and Mercury in the Almagest, Archive for
History of Exact Sciences, 19:95111.
Rosiska, G. 1984a. Scientific Writings and Astronomical Tables in Cracow: A Census of
Manuscript Sources (xivthxvith Centuries). Wrocaw.
Rosiska, G. 1984b. Identification of Copernican Tables of the Latitudes of the Planets,
Qwartalnik Historii Nauki i Techniki, 29:637644 [in Polish with an English sum-
mary].
Rosiska, G. 1996. The Fifteenth-Century Roots of Modern Mathematics, Qwartalnik
Historii Nauki i Techniki, 41:5370.
Saby, M.-M. 1987. Les canons de Jean de Lignres sur les tables astronomiques de 1321.
Unpublished thesis: cole Nationale des Chartes, Paris. A summary appeared as:
Les canons de Jean de Lignres sur les tables astronomiques de 1321, cole Nationale
des Chartes: Positions des thses, pp. 183190.
Sams, J. 1997. Andalusian Astronomy in 14th Century Fez: al-Zj al-Muwfiq of Ibn
Azzz al-Qusann, Zeitschrift fr Geschichte der Arabisch-Islamischen Wissen-
schaften, 11:73110.
Sams, J. 1999. Horoscopes and History: Ibn Azzz and His Retrospective Horoscopes
98 chapter 4

Related to the Battle of El Salado (1340), in L. Nauta and A. Vanderjagt (eds.) 1999,
pp. 101124.
Santritter, J.L. (ed.) 1492. Tabule Astronomice Alfonsi Regis. Venice.
Schmeidler, F. (ed.) 1972. Joannis Regiomontani Opera Collectanea, facsimiles of nine
works by Regiomontanus and one by Peurbach that was printed by Regiomontanus.
Osnabrck.
Stahlman, W.D. 1959. The Astronomical Tables of Codex Vaticanus Graecus 1291. Provi-
dence: Brown University (Ph.D. dissertation). University Microfilms, No. 625761.
Suter, H. 1914. Die astronomischen Tafeln des Muammad ibn Ms al-Khwrizm.
Copenhagen.
Swerdlow, N.M. 1973. The Derivation and First Draft of Copernicuss Planetary Theory:
A Translation of the Commentariolus with Commentary, Proceedings of the Ameri-
can Philosophical Society, 117:423512.
Swerdlow, N.M. 1990, Regiomontanus on the critical problems of astronomy, in T.H.
Levere and W.R. Shea (eds.), Essays on Galileo and the History of Science in Honour of
Stillman Drake. Dordrecht.
Swerdlow, N.M. 2005. Ptolemys Theories of the Latitude of the Planets in the Almagest,
Handy Tables, and Planetary Hypotheses, in J. Buchwald and A. Franklin (eds.),
Wrong for the Right Reasons. Dordrecht, pp. 4171.
Swerdlow, N.M. and O. Neugebauer 1984. Mathematical Astronomy in Copernicuss De
Revolutionibus. New York.
Thorndike, L. 1950. Giovanni Bianchini in Paris Manuscripts, Scripta Mathematica
16:512, 169180.
Thorndike, L. 1953. Giovanni Bianchini in Italian Manuscripts, Scripta Mathematica
19:517.
Tichenor, M.J. 1967. Late Medieval Two-Argument Tables for Planetary Longitudes,
Journal of Near Eastern Studies, 26:126128. Reprinted in King and Kennedy (eds.)
1983, pp. 122124.
Toomer, G.J. 1968. A Survey of the Toledan Tables, Osiris, 15:5174.
Toomer, G.J. 1984. Ptolemys Almagest. New York.
Vernet, J. 1952. Contribucin al estudio de la labor astronmica de Ibn al-Bann. Tetun.
Vernet, J. 1956. Las Tabulae Probatae, in Homenaje a Mills-Vallicrosa. Barcelona,
pp. 501522. Reprinted in Vernet 1979, pp. 191212.
Vernet 1979. Estudios sobre historia de la ciencia medieval. BarcelonaBellaterra.
Zinner, E. 1990. Regiomontanus, his life and work. Translated by E. Brown. Amsterdam.
chapter 5

Displaced Tables in Latin:


The Tables for the Seven Planets for 1340*


The displaced tables are typical of a pervasive tendency in Islamic science
to provide extensive and elegant numerical tables for the convenience of
practitioners. The underlying astronomical theory is neither questioned
nor affected.1
edward s. kennedy


Introduction

In 2012 we published A Survey of European Astronomical Tables in the Late Mid-


dle Ages in which we discussed a wide range of tables, many of which were
previously known. In this monograph we mentioned an unusual set of tables,
whose significance had not been appreciated hitherto, that depends on a prin-
ciple of displacement to eliminate subtractions in the course of computing
planetary positions. This principle was employed in some zijes in the Islamic
world, but until now there was nothing comparable in the Latin West. As we
will see, this set of tables is a most ingenious reworking of the Parisian Alfon-
sine Tables, rather than a translation of an Arabic zij (where the term zij is
used in Arabic to refer to a set of astronomical tables with instructions for their
use). What makes this set of tables so unusual is that, for example, the mean
planetary motions are defined differently from those in the standard Alfon-
sine tables, and some of the functions for computing true planetary longitudes
from their mean motions also differ noticeably from those in the standard

* Archive for History of Exact Sciences 67 (2013), 142, communicated by George Saliba.
1 Kennedy 1977, p. 16.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_007


100 chapter 5

Alfonsine tables; nevertheless, computation with these displaced tables yield


the same results as computation with the standard Alfonsine tables. More-
over, the tables for first station are presented in a completely different way
from those we have seen in the Ptolemaic tradition, although the underlying
parameters are unchanged. Given the absence of instructions, it was not an
easy task to unravel the cleverness of the construction of these tables. We had
some helpful guidance from a paper by Kennedy (1977), but it was not suffi-
cient for uncovering various subtleties in these tables for which there are no
counterparts in the tables that Kennedy analyzed.
Paris, Bibliothque nationale de France, ms lat. 10262, is a 15th-manuscript
containing two sets of astronomical tables: an anonymous set for 1340 (ff. 2r
46v); and another set, called Tabule frequentine (ff. 47v71r), composed by Mel-
chion de Friquento of Naples in 1438 (f. 46v: 1437 completus).2 Neither set has
previously been examined. In this paper we focus on the first set, entirely com-
posed of tables with no accompanying text. At the bottom of the last table
(f. 46v), we read: Expliciunt tabule de septem planetis et de veris locis eius (Here
end the tables for the seven planets and their true positions). Therefore, we
shall refer to this set as the Tables for the Seven Planets (where the Sun and the
Moon are considered planets). The yearly tables begin in March, and we are
told that the tables are valid for 1340 completus, or 1340 (complete), meaning
that the epoch is noon of February 28, 1341, the last day of the year, counting
from March 1, 1340.3
The name of the author of these tables is not given in the manuscript, and no
locality is mentioned. However, on f. 9r, at the bottom of a table for the mean
motion of the Sun, we are told that the radix for the Sun, anno domini 1340
completo, is 8s 16;41,49 (= 256;41,49). In contrast to the use of physical signs
of 60 in the Parisian Alfonsine Tables, we note the use here of zodiacal signs
of 30. If we compute the mean motion of the Sun using the Parisian Alfonsine
Tables for Toledo, we find its mean longitude to be 346;20,53 and its mean
argument of center, 256;43,54.4 Therefore, the given radix refers to the mean

2 For a description of this manuscript, see Thorndike 1957, especially pp. 144145. On the Tabule
frequentine, see Kremer (forthcoming).
3 The last day of February belongs to 1340 according to the convention of this text (in fact, it is
the last day of that year), but to 1341 according to our modern convention (it is the 59th day
of this year). The same convention applies to all subsequent years in this text. Note also that
the common astronomical practice in the Middle Ages was to take noon as the beginning of
the day.
4 According to this computation, 89;36,59 is the longitude of the solar apogee, a, which is the
difference between the solar longitude and the solar mean argument of center at a given time.
displaced tables in latin 101

argument of center of the Sun. The difference in the mean argument of center,
0;2,5, is the amount traveled by the Sun in about 0;50h, which corresponds
to a difference in geographical longitude of about 12;40 east of Toledo. This
figure agrees, although not exactly, with the longitude of the meridian of Paris,
taken as 12;0 east of Toledo in many astronomical tables of the 14th and 15th
centuries.5
The Tables for the Seven Planets give multiple examples of displaced tables.
A table is said to be displaced with respect to another when its entries are the
same as those in the standard table after adding a constant to its argument
(horizontal displacement), or derive from the entries in the standard table
by adding a constant (vertical displacement). This can also be expressed in
algebraic terms: if y = f(x) is the function underlying a given table, then the
function embedded in the displaced table is y = f(x + kh), for a displacement
on the x-axis, or y = f(x) + kv, for a displacement on the y-axis. Of course, both
displacements can occur at the same time, leading to an equation such as

(1) y = f(x + kh) + kv.

The purpose of displaced tables is to avoid subtractions, that is, the use of
complicated rules for handling negative numbers before they were available
to astronomers who computed planetary positions by means of astronomical
tables.6
This problem was already felt by astronomers in the 9th century, who called
the standard table al and the displaced table wa.7 Displaced tables were
clever computational devices, with implications for the method of computa-
tion of astronomical quantities, but did not challenge either the parameters
or the models on which the original tables were based. Kennedy (1977) pre-
sented an explanation of displaced planetary tables in the medieval Islamic
world, and demonstrated the equivalence of computations using one such set
with computations using Ptolemys tables. Kennedy focused on the tables of

5 See, e.g., Kremer and Dobrzycki 1998.


6 It is important to distinguish between displaced tables and shifted tables. A shifted table
contains the same information as the standard table (the same columns and rows, and the
same entries) but the first row is not that for argument 0 or 1 but for some other convenient
number, to stress the fact that the entry reaches a specific value such as a maximum or
a minimum for that particular value of the argument. Shifts occur, for instance, for the
planetary latitudes of certain planets in John of Murss Tables of 1321 (Chabs and Goldstein
2009, see esp. pp. 308309).
7 Salam and Kennedy 1967, p. 497. See also Debarnot 1987, p. 43 (table 9); and Jensen 1971.
102 chapter 5

Ibn al-Aclam (d. 985) and called attention to the fundamental relation of three
displacements for each planet (see eq. 26, below).8 Byzantine astronomers
apparently depended on their Muslim predecessors for displaced tables: see
Tihon 19771981, espec. 68:76 and 110. As far as we can determine, none of the
displaced tables compiled by astronomers in the Islamic World and Byzantium
were known in the West. Two Jewish astronomers in Provence, Levi ben Gerson
(d. 1344) and Immanuel ben Jacob Bonfils of Tarascon (fl. 1350), composed zijes
in Hebrew in which they used the principle of displaced tables for the times and
longitudes of syzygies (but not for the motion of the planets); however, there is
no evidence to suggest that they depended on Islamic or Byzantine sources.9
In astronomy written in Latin the tables described here are the first to use dis-
placed tables systematically, although we know of an earlier use of this type
of table by John Vimond around 1320, limited to his table for trepidation (or
access and recess).10
As mentioned above, the Tables for the Seven Planets have no accompanying
text; therefore, the following comments are based entirely on the information
provided by the tables themselves. As will be seen, the terminology in these
tables often differs from that commonly used at the time, and many of the
tables have a presentation that diverges from other tables with the same param-
eters that are based on the same model.
There follows a table of contents, arranged by section number.

1. Multiplication table
2. Mean motion of the Sun
3. Solar equation
4. Length of daylight, diurnal seasonal hours, and the equation of time
5. Mean motions of the Moon
6. Lunar latitude
7. Lunar equations

8 Displaced tables in Islamic astronomy are also discussed in Jensen 1971, Saliba 1976, Saliba
1977, Mercier 1989, and Van Brummelen 1998. For displaced tables in the Maghrib, see
Sams and Mills 1998, and Sams 2003.
9 For Levi ben Gerson, see Goldstein 1974, pp. 136146, 229241; for Bonfils, see Solon 1970,
pp. 34, 11, and Kremer (forthcoming). It is noteworthy that the Tabule frequentine in this
very same manuscript contain an adaptation of Bonfilss tables, where displaced tables
occur.
10 On Vimond, see Chabs and Goldstein 2003, pp. 275277; and Chabs and Goldstein 2004,
pp. 265267.
displaced tables in latin 103

8. Lunar node
9. Precession/trepidation
10. Planetary mean motions
11. Planetary equations and stations
11.1. Equation of center and equatio porcionis
11.2. Equation of anomaly near greatest distance
11.3. Equation of anomaly near least distance
12. Latitudes of the superior planets
13. Planetary visibility
14. Possibility of an eclipse
15. Eclipsed fraction of the solar and lunar disks
16. Latitudes of Venus and Mercury.

Displacements are discussed in 3, 5, 7, 9, 10, and 11.

1 Multiplication Table

The first table in this manuscript (ff. 2r7v) is a multiplication table for base
60 arithmetic, with entries for each integer from 1 to 60. This table, presented
as a 60 60 square matrix, is quite common in manuscripts containing sets
of astronomical tables (see, e. g., Chabs and Goldstein 2012, p. 227). The fact
that f. 8rv is blank makes it uncertain if this multiplication table belongs to the
set examined here, for it could very well have been inserted by the copyist to
facilitate computation.

2 Mean Motion of the Sun

Folio 9r contains five tables for the mean motion of the Sun. The first lists the
Radices ad 32 annos post annum 1340, i.e., the values of the mean motion of the
Sun at the beginning of each year for a period of 32 consecutive years. The entry
for year 1 is 8s 16;26,51, and it corresponds to the mean argument of center
of the Sun at noon of February 28, 1342; the entry for year 32 is 8s 16;36,12. It
should be emphasized that a period of 32 years for the motion of the Sun is
very unusual; in any case, we have not found it in the previous literature, where
periods of 20, 24, and 28 abound (see Chabs and Goldstein 2012, p. 54). We also
note that the usual phrase anni expansi (expanded years, that is, those within
a cycle, in contrast to anni collecti for the years at the interval of a cycle) does
not appear here or in the headings of the other tables.
104 chapter 5

The other four tables for the Sun have a general title, beginning Tabula
porcionis solis ad annos radicum , where porcio has to be understood here as
argument of center.11 The second table on this folio gives the mean motion
of the Sun for accumulated months; its first entry corresponds to March (1s
0;33,18) and the last one to February (11s 29;45,39). These entries agree exactly
with those given by John of Lignres in his Tabule magne for a year beginning
in January (see Chabs and Goldstein 2012, Table 5.1b, pp. 5556), who also
used zodiacal signs of 30. The third table is headed ad annos perpetuacionis,
from 1372 to 1852, at 32-year intervals, and again we see the use of non-standard
terminology, for the phrase anni collecti is totally absent from this set of tables.
The fourth and fifth tables are, respectively, for the mean motion of the Sun for
days from 1 to 31, and for hours and fractions of an hour from 1 to 60. The entry
for argument 1d is 0s 0;59,8.
As mentioned above, at the bottom of the folio we are told that the Radix
porcionis solis ad anno domini 1340 completo is 8s 16;41,49. From this value and
the entry for year 32 (8s 16;36,12) we derive a mean motion in the argument
of center of 0;59,8,13,32,49/d. As indicated previously, the entries in these
tables represent the mean argument of center of the Sun, , and this is another
peculiar characteristic of the Tables for the Seven Planets, because in other sets
the mean solar motion, that is, , is usually tabulated. The two quantities are
related by means of a, the longitude of the solar apogee, for = + a. A small
table, found later in the text (f. 17r), gives the motion of the solar apogee. It
is entitled Motus augium equatus and the entries are yearly values, probably
corrected for precession. The entry for 32 year is 0;19,42,35, which implies a
daily motion of 0;0,0,6,4/d. By adding this value to the one derived for the
mean motion in the argument of center of the Sun, the result is 0;59,8,19,37/d,
in good agreement both with the value used by Vimond (0;59,8,19,37,4/d) and
the value in the Parisian Alfonsine Tables (0;59,8,19,37,19,13,56/d) for the mean

11 We know of no prior usage of porcio for the argument of center of the Sun. There are,
however, examples where it means anomaly in the case of the planets. On porcio (or
portio) in the sense of anomaly, see Goldstein, Chabs, and Mancha 1994, pp. 6364
(see also Nallino 19031907, 2:328), where a reference is given to John of Lignress use
of this word in the canons to his tables, whose incipit is Priores astrologi motus corporum
celestium (1322). The Arabic for anomaly is ia, which means portion and one of the
Hebrew terms for anomaly, manah, also means portion. Both the Latin and Hebrew terms
apparently come from Arabic. The fact that in our text the argument of center of the Sun
is considered the anomaly indicates that the author had in mind an epicyclic model for
the Sun, which is equivalent to the eccentric model according to Apolloniuss theorem
(see Almagest iii.3; Toomer 1984, pp. 141153).
displaced tables in latin 105

solar motion in longitude.12 It would thus seem that the Tables for the Seven
Planets belong to this tradition. We will show that the rest of the tables provide
additional evidence of this relationship.

3 Solar Equation

The solar equation is given on ff. 9v10r. The title of the corresponding table
(Tabula equacionum solis cum auge eius equata) indicates that the value for
the solar apogee has already been taken into consideration. The argument,
which we call k , in parallel with the standard usage of ,13 begins at 0, not
at 1 as was the common practice at the time. The entry for 0 is 2s 29;37,9
(= 89;37,9), and corresponds to the longitude of the solar apogee at epoch,
89;36,59 (see n. 4). The minimum 2s 27;27,9 is reached at 9294, and the
maximum 3s 1;47,9 at 267269. Subtracting algebraically either of these two
values from the entry for 0, one gets 2;10,0, which is the standard Alfonsine
parameter for the maximum solar equation.14 Therefore, all entries for the solar
equation, c(k ), are displaced upwards by 89;37,9, compared to the entries in
the standard tables in the Alfonsine corpus, c(), to make all entries positive,
thus allowing the user to avoid dealing with a set of complicated rules for
adding and subtracting various terms:

(2) c(k ) = c() + 89;37,9.

This is the first example of a displaced table in this set, where the argument is
the mean argument of center, k , and the entries represent c(k ), which is the
sum of the Alfonsine solar equation (whether positive or negative), c(), and
the longitude of the solar apogee, a. In this case the displacement is vertical
(see Figure 5.a).

12 For Vimond, see Chabs and Goldstein 2004, p. 221; for the Parisian Alfonsine Tables, see
Ratdolt 1483, f. d5r. For a modern edition of these tables, based on Ratdolt 1483, see Poulle
1984.
13 Throughout this paper capital letters are used to represent variables and functions in the
Tables of the Seven Planets.
14 As in the tables of John Vimond (Chabs and Goldstein 2004, p. 223), the solar equation
with the same maximum of 2;10is not explicitly given. For the maximum solar equa-
tion, 2;10, in the Parisian Alfonsine Tables, see Ratdolt 1483, f. e3r.
106 chapter 5

figure 5.a The solar equation displaced vertically

The canons to this table, which are not extant, should have had an instruc-
tion indicating something like to find the true position of the Sun, enter the
table with the mean argument and add what you find to it or, in algebraic
terms,

(3) l = k + c(k ).

And, indeed, in Ptolemys model for the equation of the Sun, , the true longi-
tude of the Sun is obtained adding its mean longitude, , to the solar equation:
= + c(). Now the solar equation is also the difference between the true argu-
ment of center of the Sun, , and its mean argument of center, : it is negative
if 0 180, and positive if 180 360. The Tables of the Seven Planets
yield the same result as that deduced from Ptolemy, for

= + c()
= ( + a) + c()
= + (a + c())
= k + c(k ) = l,

provided that k = , which is the case here. The two terms in L are positive and
both are explicitly tabulated on ff. 9v10r.
displaced tables in latin 107

4 Length of Daylight, Diurnal Seasonal Hours, and the Equation of


Time

On ff. 10v11v there is a table with a column for the argument from 1 to
30, and three other columns for each of the twelve zodiacal signs, begin-
ning with Aries. The headings are Hora equalis, tempus horarum, and equatio
dierum. The first entries for Aries 1 are 12;3h, 15;4, and 8;28 min, respec-
tively. The entries in the first column, under Hora equalis, display the length
of daylight (i.e., the time interval from sunrise to sunset) for a given locality;
it reaches a maximum, 15;10h, at Gem 28 Cnc 2, and a minimum, 8;50h, at
Sgr 28 Cap 2. If we consider the obliquity of the ecliptic, , to be 23;33,
the resulting geographical latitude, , for which the table is valid is = 42;44,
which agrees fairly well with the parallel through Toulouse (rather than with
that of Paris, where the longest daylight is 16;0h). The geographical longitude
(see above) and latitude that we derive from the tables do not yield good
agreement with any place where astronomy was practiced in the 14th cen-
tury, but a locality approximately fitting both computed coordinates is Perpi-
gnan.
The entries in the second column, under tempus horarum, display the length
of a diurnal seasonal hour (i.e., a twelfth of the length of daylight); it reaches a
maximum, 18;57h, at Gem 27 Cnc 3, and a minimum, 11;3h, at Sgr 27 Cap
3. These two columns are mutually consistent.
The third column displays the equation of time (i.e., the difference between
apparent and mean time where apparent time is counted from true noon,
that is, the moment that the true Sun crosses the meridian, and mean time
is counted from mean noon). The argument is the solar longitude expressed
in degrees and the entries are given in minutes and seconds of an hour. The
extremal values are 0;21,24h (at Tau 26), 0;11,16h (at Leo 5), 0;31,48h (at Sco 8),
and 0;0,0h (at Aqr 2023). When converting these values into time-degrees,
that is, multiplying each entry by 360/24h, we obtain 5;21, 2;49, 7;57, and
0,0, respectively, which are the characteristic values found in the equation
of time ascribed to Peter of Saint Omer and used by John of Lignres, among
others (see, e. g., Chabs and Goldstein 2012, pp. 3740).

5 Mean Motions of the Moon

The mean motions of the Moon are addressed in five tables on ff. 12r13r. In
addition to columns for the various arguments, each table has three columns
headed centrum lune (here meaning double elongation), porcio lune (argument
108 chapter 5

of lunar anomaly), and medius locus lune (mean longitude of the Moon). Again,
for the first two quantities this is not the standard terminology.
The first table lists the radices at the beginning of each consecutive year
from 1342 for a period of 32 years. The entries for year 1 are 6s 3;36,2 (double
elongation), 3s 17;36,22 (anomaly), and 8s 10;16,38 (longitude). The second
table gives the mean motion of the Moon for the months; the first entries,
for March, are 1s 5;49,36 (double elongation), 1s 15;0,53 (anomaly), and 1s
18;28,6 (longitude). The third and fourth tables are, respectively, from 1372 to
1852, at 32-year intervals, and for hours and parts of an hour, from 1 to 60. The
fifth table is for days from 1 to 31, and the first entries for argument 1d for the
three variables are 0s 24;22,54 (double elongation), 0s 13;3,54 (anomaly), and
0s 13;10,35 (longitude), confirming that the first column is indeed the double
elongation, for 0s 24;22,54 = 2 (0s 13;10,35 0s 0;59,8).
As was the case for the Sun, at the bottom of f. 12v we are given the radices for
the three variables, anno Christi 1340 perfecto: 9s 14;21,15,14 (double elongation),
0s 18;53,8,0 (anomaly), and 4s 0;53,34,57 (longitude). We note that the values
are given here to thirds and that the last digit for the anomaly is 0, indicating
that the author was sensitive to accuracy. From the values of the radices and
those for year 32 in the first table for the mean motions of the Moon, we
derive the following values for the mean motions: 24;22,53,23,16/d (double
elongation),15 13;3,53,57,30/d (anomaly),16 and 13;10,35,1,15/d (longitude).17 All
of them are typical parameters of the Parisian Alfonsine Tables (Ratdolt 1483,
ff. d5vd7r).
We have recomputed the radices using the Parisian Alfonsine Tables for
noon of February 28, 1341 for the meridian of Paris, 12 east of Toledo, that is,
at 0;2d before noon in Toledo. The results are presented in Table 1.
The agreement is good for the double elongation and the mean argument of
center of the Sun, but much better results are obtained when recomputing the

15 The value for year 32 given in the text is 4s 15;17,11, and the difference between this value
and that of the radix is 210;55,56. Thus the amount of the double elongation in 32 years
of 365;15 days is 210;55,56 and 791 complete revolutions, leading to a daily mean motion
of 24;22,53,23,16 /d.
16 The value for year 32 given in the text is 12s 22;28,24 (read 2s 22;28,24), and the difference
between this value and that of the radix is 63;35,6. Thus the amount traveled by the
Moon in the argument of anomaly in 32 years of 365;15 days is 63;35,6 and 424 complete
revolutions, leading to a daily mean motion of 13;3,53,57,30/d.
17 The value for year 32 given in the text is 1s 16;35,39, and the difference between this value
and that of the radix is 285;42,4. Thus the amount traveled by the Moon in longitude in
32 years of 365;15 days is 285;42,4 and 427 complete revolutions, leading to a daily mean
motion of 13;10,35,1,15/d.
displaced tables in latin 109

table 1 Recomputation for the Moon at epoch: noon, Feb. 28, 1341

Text Computation TextComp.

Mean argument of center of the Sun 8s 16;41,49 256;41,46 0; 0, 3


Double elongation 9s 14;21,15,14 284;20,13,41 0; 1, 1,33
Mean argument of anomaly 0s 18;53, 8, 0 32; 1,34,53 13; 8,26,53
Lunar mean longitude 4s 0;53,34,57 128;29, 2, 0 7;35,27, 3

table 2 More detailed recomputation for the Moon at epoch: noon, Feb. 28, 1341

Text Computation TextComp.

Mean argument of center of the Sun 8s 16;41,49 256;41,49 0; 0, 0


Double elongation 9s 14;21,15,14 284;21,14,39 +0; 0, 0,35
Mean argument of anomaly 0s 18;53, 8, 0 32; 2, 7,32 13; 8,59,32
Lunar mean longitude 4s 0;53,34,57 128;29,34,57 7;36, 0, 0

radices for 0;1,57,30d before noon in Toledo, that is, for a locality 11;47,45 east
of that Spanish city. The results are presented in Table 2.
From Table 2 it follows that the mean argument of anomaly of the Moon
tabulated here is diminished by 13;9 from that obtained from the Parisian
Alfonsine Tables for a locality near the meridian of Paris, and that the mean
lunar longitude is diminished by exactly 7;36. As will be seen in 7, the
displacements applied to the two quantities correspond, on the one hand, to
the maximum value of the lunar equation of center in the Parisian Alfonsine
Tables (13;9) and, on the other hand, to the sum (7;36) of two parameters, the
maximum values for the equation of anomaly (4;56) and for the increment
(2;40). So, if we call the mean argument of anomaly in the Tables of the Seven
Planets, and that in the Parisian Alfonsine Tables, we have

(4) = 13;9.

Similarly, if l is the mean lunar longitude in the Tables of the Seven Planets,
and that in the Parisian Alfonsine Tables, we have

(5) l = 7;36.
110 chapter 5

figure 5.b

Ptolemys second lunar model. o is the


observer, d is the center of the deferent circle
whose radius is dc, c is the center of the
epicycle, s is the direction to the mean Sun, l is
the true position of the Moon, v is the direction
to the vernal point, and d is 180 from d on a
circle about o with radius od. The angle is
the elongation of the mean longitude of the
Moon from the mean longitude of the Sun,
angle is the mean argument of anomaly
counted from the mean apogee e , whose
position is fixed by the direction dc, ae is the
true lunar apogee, angles c3 and c are the
corrections to the mean lunar longitude, and
angle is the true longitude of the Moon.

6 Lunar Latitude

The table for the lunar latitude (f. 13r) has a maximum of 5;0,0 at 90; it is
the standard table found in many zijes, including those used by Alfonsine
astronomers (see Chabs and Goldstein 2012, pp. 103104).

7 Lunar Equations

The lunar equations are presented in three separate tables. The author, well
aware of Ptolemys second lunar model (see Figure 5.b), has split the treatment
of the lunar equations, c3 and c, according to the two independent variables
involved: the double elongation, 2, and the true argument of anomaly, . The
true longitude of the Moon is the sum of four positive terms, all of which are
tabulated (see eq. 14, below), whereas in the standard Alfonsine tables the true
longitude of the Moon is the algebraic sum of three terms (see eq. 8, below).
Despite the differences in these procedures, we show that the results are the
same.
The longitude of the Moon, , is found by applying an equation, c, to its mean
longitude, , which is a linear function of time: = + c. Moreover, the mean
argument of anomaly, , is also a linear function of time. In Ptolemys second
lunar model (Almagest v.8), the lunar equation, c, depends on two indepen-
dent variables: the true argument of anomaly, (angle aecl), and the double
elongation, 2 (twice the angular distance between the mean longitude of the
Moon and the mean longitude of the Sun). In this context the lunar equation
displaced tables in latin 111

is a combination of functions of one variable represented by the 4th, 5th, and


6th columns of Ptolemys table in Almagest v.8. The vast majority of later sets
of tables used Ptolemys approach, although some of them have slightly modi-
fied parameters or rearrangements of the order of the columns. In the Parisian
Alfonsine Tables (pat) these three columns correspond respectively to the 6th,
5th, and 4th columns; in the following, we will use the conventions of pat to
identify the columns ci:

(6) c = c6() + c5() c4(2)

where

(7) = + c3(2).

In eqs. 6 and 7, c3(2) is the equation of center, c4(2) represent the minutes of
proportion, c5() is called the increment, and c6() is the equation of anomaly.
Thus, the true longitude of the Moon is given by

(8) = + c6() + c5() c4(2).

The first table (ff. 13v14r) is for the equation of center; it has a column for
the argument, from 0 to 29, and then columns for each sign, from 0s to
11s. For each sign and each degree of the argument, we are given entries for
the equation of center and the minutes of proportion (c3 and c4, respectively,
in pats arrangement, both variables depending on the double elongation).
For the tabulated equation of center, the entry for 0s 0 is 13;9. It reaches
a maximum of 26;18 (twice 13;9) at 114115, and a minimum of 0;0 at
245246. So the equation of center of the Moon is tabulated with a vertical
displacement, for its entries are displaced upwards by 13;9 with respect to the
standard Alfonsine table (see Chabs and Goldstein 2012, Table 6.2a, p. 71). If
we call c3(2) the equation of center in the corresponding Alfonsine table, the
tabulated entries, c3(2), for the equation of center in the Tables for the Seven
Planets are given by

(9) c3(2) = c3(2) + 13;9.

It is worth noting that eq. 7 and the definition of in eq. 4 imply that a = .18

18 a = + c3(2) = ( 13;9) + c3(2) = ( 13;9) + (c3(2) +13;9) = + c3(2) = .


112 chapter 5

For the minutes of proportion the entry for 0s 0 is 0. The entries increase
monotonically to 60 at 173187, and decrease to 0 at 349371. Except
for copyists errors, they agree with those in the Parisian Alfonsine Tables.
Therefore,

(10) c4(2) = c4(2).

Let us now turn to the third table (ff. 14v15r), which is for the equation of
anomaly; it has a column for the argument, from 0 to 29, and then columns
for each zodiacal sign, from 0s to 11s. For each sign and each degree of the
argument, we are given entries for the equation of anomaly and the dyversitas
dyametri proporcionalis19 (corresponding to c6 and c5, respectively, in pats
arrangement, both variables depending on the true argument of anomaly). For
the equation of anomaly, the entry for 0s 0 is 4;56,0. It reaches a minimum
of 0;0,0 at 9199 and a maximum of 9;52,0 (twice 4;56,0) at 264, that is,
the vertical displacement is 4;56. This parameter was systematically used by
Parisian Alfonsine astronomers, but its origin goes back much earlier (Chabs
and Goldstein 2003, pp. 252253). Thus, if we call c6() the equation of anomaly
in the corresponding Parisian Alfonsine table, the tabulated entries, c6(a),
for the equation of anomaly in the Tables for the Seven Planets are given
by

(11) c6(a) = c6() + 4;56.

The first entry in the column for the dyversitas dyametri proporcionalis, usu-
ally called increment is 2;40; it reaches a minimum of 0;0 at 113119 and
a maximum of 5;20 (twice 2;40) at 251257. Except for copyists errors, the
entries agree with those in the Parisian Alfonsine Tables, but for the vertical
displacement of 2;40. Again, if we call c5() the increment in the correspond-
ing Parisian Alfonsine table, the tabulated entries, c5(a), for the increment in
the Tables for the Seven Planets are given by

(12) c5(a) = c5() + 2;40.

In the second table (f. 14v), headed dyversitas dyametri centralis, the argument is
given at intervals of 1, from 0 to 60, and the entries, d(m), decrease monotoni-

19 The word dyversitas, used in the headings of the lunar and the planetary equations, is often
spelled diversitas.
displaced tables in latin 113

table a dyversitas dyametri


centralis (excerpt)

Arg. Dyv. dyam. centralis


() ()

0 2;40, 0

10 2;13,20

20 1;46,40

30 1;20, 0

40 0;53,20

50 0;26,40

60 0; 0, 0

cally from 2;40,0 (at 0) to 0;0,0 (at 60). Moreover, the entries show a constant
decrease of 0;2,40 for each degree of the argument (see Table a); hence,

(13) d(m) = 2;40 (1 c4(2)).

The tables described above yield the same results as those obtained from the
Parisian Alfonsine Tables for the true position of the Moon, and were computed
with a similar expression (see eq. 8). If and l are the true lunar longitudes in
the Parisian Alfonsine Tables and the Tables for the Seven Planets, respectively,
then using eqs. 5, 8, 10, 11, and 12, can be written as

= + c6() + c5() c4(2)


= (l + 7;36) + (c6(a) 4;56) + (c5(a) 2;40) c4(2)
= (l + c6(a) + 2;40) + c5(a) c4(2) 2;40 c4(2)
= (l + c6(a) + c5(a) c4(2)) + 2;40 2;40 c4(2) = l.

The term, 2;40 (1 c4(2)) = d(m), is found in the table for the dyversitas
dyametri centralis, and therefore the expression to compute the true longitude
of the Moon is
114 chapter 5

(14) l = l + c6(a) + c5(a) c4(2) + d(m),

where all the terms are tabulated and none of them is to be subtracted. Note
that there is no counterpart to the function d(m) in the standard Alfonsine
tables.
The canons to this set of tables, which are not extant, should have had
an instruction more or less in the following terms: To determine the true
position of the Moon, find the mean longitude of the Moon, and keep it; find
the double elongation, and enter with it in the table for the equation of center
and the minutes of proportion, and keep them. Find the mean argument of
anomaly, and add to it the equation of center to obtain the true argument of
anomaly, and enter with it in the table for the equation of anomaly and the
increment, and keep them. Enter the table for the dyversitas dyametri centralis
with the minutes of proportion, and keep what you find there. Then add the
mean longitude to the true argument of anomaly, and add the result to what
is obtained from multiplying the increment by the minutes of proportion. To
the value obtained add what you found in the table for the dyversitas dyametri
centralis.
To illustrate eq. 14 consider the Moon at epoch (noon, February 28, 1341),
when the double elongation, 2, was 4,45;9,0 (= 285;9,0). The equation of
center in the standard Alfonsine tables is c3(2) = 10;26 and, because of
eq. 9, c3(2) = 2;43 (= 10;26 + 13;9). The corresponding minutes of pro-
portion are c4(2) = 19 and c4(2) = 19, because of eq. 10. In this case the
term d(m) = 1;49,20. Now, the argument of anomaly, , according to the stan-
dard Alfonsine tables, is 32;27,43, and thus = 22;1,43 (= 32;27,43 10;26).
From eq. 4, = 19;18,43, and thus a = 22;1,43 (= 19;18,43 + 2;43). The tab-
ulated values for the equation of anomaly, c6(a), and the dyversitas dyametri
proporcionalis, c5(a), are 3;13,19 and 1;48, respectively, whereas the values
for the equation of anomaly, c6(), and the increment, c5(), found in the
standard Alfonsine tables are 1;42,41 and 0;52, respectively. We note that
eqs. 11 and 12 hold. According to eq. 6, using the standard Alfonsine tables
the correction to be applied is c = (1;42,41) + (0;52) (19/60) = 1;59,9,
whereas using the Tables of The Seven Planets c = 3;13,19 + 1;48 (19/60) =
3;47,31.
With the standard Alfonsine tables the mean lunar longitude at epoch, ,
is computed to be 128;55,23; thus, with eq. 8, the true lunar longitude of the
Moon at epoch is

= + c6() + c5() c4(2)


= 128;55,23 1;59,9 = 126;56,14.
displaced tables in latin 115

With the Tables for the Seven Planets we obtain l = 121;19,23 (see eq. 5); thus,
with eq. 14, the true longitude of the Moon at epoch is

l = l + c6(a) + c5(a) c4(2) + d(m)


= 121;19,23 + 3;47,31 + 1;49,20 = 126;56,14,

in agreement with .

8 Lunar Node

Folio 16r contains five tables for the mean motion of the lunar node. The first
lists the values of the mean motion of the node at the beginning of each
year, beginning with 1342, for a period of 32 years. The entry for year 1, i.e.,
1342, or more precisely, Feb. 28, 1341, is 3s 22;3;17. The second table gives the
mean motion of the Sun for the months; its first entry corresponds to March
(0s 1;38,30) and the last one to February (0s 19;19,42). The third table lists
values from 1372 to 1852, at 32-year intervals. The fourth and fifth tables are,
respectively, for the mean motion of the lunar node for days from 1 to 31, and
for hours and parts of an hour from 1 to 60. The entry for argument 1d is 0s
0;0,3,11. As was the case for other mean motions, at the bottom of the folio
we are told that the radix at anno Christi 1340 perfecto is 3s 2;43,34,43. From
this value and that for year 32 we derive 0;3,10,38,7/d as the mean motion of
the lunar node,20 in full agreement with the corresponding parameter in the
Parisian Alfonsine Tables.

9 Precession/Trepidation

On folios 16v17r there are various tables for precession/trepidation in the


framework of the Parisian Alfonsine Tables. In general, the approach found
in the Alfonsine corpus is to treat two terms separately: a linear term, usually
called motion of the apogees and the fixed stars, based on a mean motion
of 0;0,0,4,20,41,17,12/d, and presented in a single table; and a periodic term,

20 The value for year 32 given in the text is 11s 21;39,20, and the difference between this
value and that of the radix is 258;55,45,17. Thus the amount traveled by the lunar node in
32 years of 365;15 days is 285;42,4 and one complete revolution, leading to a daily mean
motion of 0;3,10,38,7/d.
116 chapter 5

usually called motion of access and recess of the 8th sphere, requiring the use
of two tables (one for the mean motion, based on a value of 0;0,0,30,24,49,0/d,
and another for its equation, found in a separate table with a maximum of
9;0,0): see Chabs and Goldstein 2012, pp. 4852.
The author of the Tables for the Seven Planets follows the same approach,
uses the same parameters, but gives a different presentation. Folio 16v has four
tables under the general title Tabula motus augium et stellarum fixarum atque
8 spere, displaying the two mean motions of the components of Alfonsine
variable precession. Each table has two columns, in addition to that for the
argument: the mean motion of the apogees and the fixed stars on the one hand,
and the mean motion of the 8th sphere on the other. The motion of trepidation
is to be applied to the positions of the fixed stars as well as to the positions of
the planetary apogees.
The first table gives entries for both quantities at the beginning of each
year for a period of 32 consecutive years. The entries for year 1 are 0s 0;51,12
(apogees) and 2s 8;10;51 (8th sphere), and those for year 32 are 0s 1;4,52
(apogees) and 2s 9;46;31 (8th sphere). The second table gives the mean motion
of both quantities for accumulated months. The third table, headed ad annos
perpetuacionis, has entries for years 1372 to 1852, at 32-year intervals, of both
quantities, whereas the fourth table is for their mean motions for days from
1 to 31. As was the case for other quantities, at the bottom of the folio we are
given the radices for both quantities, anno Christi 1340 perfecto: 0s 9;50,45,59,27
(apogees), and 2s 8;7,45,20,10 (8th sphere).21 From these radices and the en-
tries for year 32 we derive the following mean motions: 0;0,26,28 /y (apogees)
and 0;3,5,11/y (8th sphere), which correspond to about 0;0,0,4,20,51/d (apo-
gees) and 0;0,0,30,25,12 /d (8th sphere), that is, 1 revolution in 49,000 years
and 7,000 years, respectively. These are the standard values found in Alfonsine
astronomy for precession/trepidation. Combining the mean motion of the 8th
sphere and its radix, we find that the argument of the 8th sphere was 0 about
1324 years and 9 months before the epoch of these tables (February 28, 1341),
that is, in May 16ad. This date is explicitly given by Giovanni Bianchini (15th
century) as the epoch of the Alfonsine model; see Chabs and Goldstein 2009,
p. 32, and the Tables for the Seven Planets offer a justification of it.
It remains to compute the periodic term of Alfonsine trepidation, that is,
the equation of the 8th sphere, which is tabulated on f. 17r. Indeed, the entries
given in the table are based on the standard Alfonsine equation of the 8th

21 We note that the radix for the apogees has the wrong number of degrees (probably a copy-
ists error); to be consistent with the 32 values in the table, it should be 0s 0;50,45,59,27.
displaced tables in latin 117

sphere which reaches a maximum of 9;0,0 at 90, a minimum of 9;0,0 at 270,


and vanishes at 0 and 180 (see also Ratdolt 1483, f. d3v). The corresponding
entries on f. 17r are 9;0,0 at 0, 18;0,0 (maximum) at 90, 9;0,0 at 180,
0;0,0 (minimum) at 270, and 0;0,0 at 360; thus, the entries are displaced
upwards by 9 from the standard table, and can be represented by the modern
expression (where y is the entry and x is the argument):

(15) y = 9 + arcsin (sin 9 sin x).

This is another case of a vertically displaced table we find in this manuscript.


Interestingly, the only example in Latin of a displaced table previously known
to us is that of John Vimond (see n. 10), whose table for the equation of
the 8th sphere presents both a vertical displacement (8;17) and a horizon-
tal displacement (113). However, Vimonds purpose was not to avoid nega-
tive numbers in his table (otherwise he would have used a vertical displace-
ment of at least 9;0, as here); rather, he wished to set the origin of the equa-
tion of the 8th sphere at the value (8;17) it had at the time he composed
his tables (1320). Apparently, the author of the Tables for the Seven Planets
did not have the same goal as Vimond, although he used the same princi-
ple.
As mentioned above, on folio 17r there is also a small table entitled Motus
augium equatus ad 32 annos post annum 1340, listing 32 values. The first entry is
0;0,37,11 and the last entry, for 32 years, is 0;19,42,35. The entries show a steady
increase but the line-by-line differences do not reveal a clear pattern. A com-
plement to this table is found on f. 18r among the tables for the mean motions,
where we find one headed Motus augium equatus ad annos perpetuacionis, list-
ing entries from 1372 to 1852 at intervals of 32 years. The first entry is 0;19,42,35
but, as in the previous case, the line-by-line differences do not vary in a smooth
way. With all these cautions, from the entry for 32 years it is possible to derive
a motion of the apogee of 0;0,36,57/y, or 0;0,0,6,4/d, which is the amount to
be added to the mean motion in solar anomaly to obtain the mean motion in
the argument of solar longitude (see 2). The value we deduce from the table is
certainly close to Ptolemys value for precession of 1 in 100 years, that is 0;0,36
in 1 Egyptian year of 365 days.

10 Planetary Mean Motions

Next come tables for the mean motions, equations, and stations of the five
planets, on ff. 17v22r (Saturn), ff. 22r26v (Jupiter), ff. 27r31r (Mars), ff. 31r35v
118 chapter 5

table 3 Radices for year 1340 (complete)

Centrum Porcio

Saturn 0s 4;56,59 2s 8;46,40


Jupiter 0s 1;38,17,48 4s 29;28,59,11
Mars 1s 25; 6,26, 4 2s 24;48,42,[..]
Venus 6s 25;41,48,36 7s 17; 6,11
Mercury 3s 19;27,49 3s 23;27,56

(Venus), and ff. 35r40r (Mercury). In this section we review the mean motions
of the planets, for each of which we are given five tables. As we will see, all the
mean motions are displaced with respect to those in the standard Alfonsine
tables.
The first of these five tables lists the radices of two variables, centrum (argu-
ment of center) and porcio (here meaning the displaced argument of anomaly,
as will be seen below), for the beginning of each year after 1340 for a period
of 32 consecutive years. The entries for both variables for year 32 are, respec-
tively: 1s 6;7,52 and 1s 14;30,11 (Saturn); 8s 12;53,57 and 8s 18;7,43 (Jupiter);
2s 0;5,0 and 2s 19;44,28 (Mars); 6s 25;36,12 and 7s 22;56,20 (Venus); and 3s
19;22,12 and 2s 4;36,10(Mercury). The second of these tables gives the mean
motion of the planets for accumulated months, beginning in March, for the
two variables. The third of these tables is headed ad annos perpetuacionis and
displays entries for the two variables from 1372 to 1852, at 32-year intervals.
The fourth and fifth of these tables are, respectively, for the mean motion of
the planets for days from 1 to 31, and for hours and parts of an hour from 1 to
60.
From the values of the radices and those for year 32 in the first table for the
mean motions of the planets, we derive the following values for the daily mean
motions: 0;2,0,29,13 /d (Saturn, argument of center); 0;4,59,9,23/d (Jupiter,
argument of center); and 0;31,26,32,35/d (Mars, argument of center). These
three parameters differ by 0;0,0,6,4/d from the standard Alfonsine mean
motions, as was the case for the argument of solar anomaly, indicating that the
precession of 1 in 100 years was applied to the planets as well as to the Sun.
We have also derived the daily mean motions of anomaly for the inferior plan-
ets: 0;36,59,27,24/d (Venus) and 3;6,24,7,43/d (Mercury), which agree exactly
with the corresponding parameters in the Parisian Alfonsine Tables.
As was the case for the Sun and the Moon, we are also given the radices anno
christi 1340 completo, which we display in Table 3.
displaced tables in latin 119

table 4 Recomputation for the planets at epoch: noon, Feb. 28, 1341

Argument of Difference Rounded Argument of Difference Rounded


center TextComp. value anomaly TextComp. value

Saturn 18;56,58 13;59,59 14 75;46,38 6;59,58 7


Jupiter 19;18,21 17;40, 3 * 18 155;28,57 5;59,58 6
Mars 116; 6,24 60;59,58 61 96;48,44 12; 0, 2 12
Venus 256;41,56 51; 0, 7 51 230; 6, 9 2;59,58 3
Mercury 137;27,46 27;59,57 28 117;27,52 3;59,56 4

* The difference is 18;0,3 if we take the radix for the argument of center in Table 3 to
be 1;18,17,48, which is the result one would obtain from a set of standard Alfonsine
tables.

We have recomputed the mean positions of the planets using the Parisian
Alfonsine Tables for noon of February 28, 1341 for the meridian of Paris, 12 east
of Toledo. The results are presented in Table 4.
From Table 4 it follows that the arguments of center of the five planets given
in these tables are diminished, but for a few seconds, by 14, 18, 61, 51, and
28 from those obtained from the Parisian Alfonsine Tables for a locality near
the meridian of Paris. Table 4 also indicates that the entries for the argument
of anomaly are diminished, but for a few seconds, by 7, 6, 12, 3, and 4 from
those derived from the Parisian Alfonsine Tables. As will be shown in 11, all
these integer values correspond to the horizontal and vertical displacements
(kh3 and kv3, respectively) used in the tables for the equations of the planets. If
we call and the planetary arguments of center and anomaly in the Parisian
Alfonsine Tables, the corresponding quantities in the Tables for the Seven
Planets are defined by:

(16) k = kh3,

and

(17) = kv3.
120 chapter 5

figure 5.c

The model for Mars, where o is the


observer, d is the center of the deferent
circle acp, a is the apogee, p is the
perigee, e is the equant point, ov is the
direction to the vernal point on the
ecliptic, s is the direction to the mean
Sun, e is the mean apogee on the
epicycle whose center is c, ae is the true
apogee on the epicycle, and are the
mean arguments of center and anomaly,
respectively, m is the mean position of the
planet, and m is its true position (where,
for an outer planet, cm is parallel to os).

11 Planetary Equations and Stations

In this section we review the equations and the stations of the five planets, for
each of which we are also given five tables, in which the equation of center
(one table; see 11.1) and the equation of anomaly (four tables; see 11.2 and
11.3) are treated separately. Figure 5.c displays the model for Mars. As we will
see, the true longitude of a planet is found by adding four positive terms, all of
which are tabulated (see eqs. 43 and 45, below). These terms are all different
from those in the standard Alfonsine tables, where the true longitude is found
as the algebraic sum of four terms. The author of these tables has considered
two cases: (1) the epicyle is near the apogee of the deferent, and (2) the epicycle
is near the perigee of the deferent. In case 1 he has the distance of the center of
the epicycle vary from its maximum to its mean value, whereas in case 2 he has
the distance of the center of the epicycle vary from its mean to its minimum
value.

11.1 Equation of Center and equatio porcionis


In the first table the argument ranges from 0 to 29 for each zodiacal sign, and
for each degree we are given three entries: equatio centri, equatio porcionis, and
minutes of proportion. The first two are given in degrees and minutes, and the
third entry is given in minutes (see Table b).
Table 5 displays significant values of the equation of center for the five
planets. To determine the relation between the equation of center presented
here and that in the Parisian Alfonsine Tables, let us consider the case of
Mars. The entry for 0 is 2;29; the minimum (0;36) is reached at 3135
displaced tables in latin 121

table b equatio centri, equatio porcionis,


and minutes of proportion for
Saturn (excerpt)

Equatio Equatio Minutes


centri porcionis of proportion
() () () ()

0 5;30 8;30 0
1 5;24 8;36 0
2 5;17 8;43 1

60 0;48 13;12 37

79 0;29 13;31 58
80 0;29 13;31 59
81 0;30 13;30 1
82 0;30 13;30 2

180 8;39 5;21 60

254 13;31 0;29 0


255 13;31 0;29 0

264 13;25 0;35 2


265 13;23 0;37 1
266 13;22 0;38 59
267 13;21 0;39 58

359 5;36 8;24 0

and the maximum (23;24) at 203207. The difference between the two
extremal values is 22;48, that is, twice 11;24, the standard parameter used
for Mars in the Parisian Alfonsine Tables. The entries in the table are dis-
placed exactly 12 upwards as compared with the corresponding entries in
the Parisian Alfonsine Tables, and 12 is the minimum integer for a vertical
displacement ensuring that all entries for the equation of center are positive.
But the entries for the equation of center have a second displacement, a hori-
zontal displacement, of 61 with respect to the table in the Parisian Alfonsine
122 chapter 5

table 5 Equation of center for all planets

Minimum Maximum Under- Vertical Horizontal


Entry equation equation lying displace- displace-
for 0 of center of center parameter ment (kv3) ment (kh3)

Saturn 5;30 0;29 13;31 6;31 7 14


(7680) (252256)
Jupiter 4;15 0; 3 11;57 5;57 6 18
(7278) (246252)
Mars 2;29 0;36 23;24 11;24 12 61
(3135) (203207)
Venus 1;21 0;50 5;10 2;10 3 51
(3747) (211221)
Mercury 2;47 0;58 7; 2 3; 2 4 28
(6569) (235239)

Tables. With the notation used above, k = 61. If we call c3() the equa-
tion of center in the corresponding Parisian Alfonsine table,22 the tabulated
entries, c3(k ), for the equation of center in the Tables for the Seven Planets
are given by c3(k ) = c3(k + 61) + 12. This is the first example of a horizon-
tal displacement in these tables. For each planet the corresponding displace-
ment applied to (see Table 5) accounts for the difference between the entries
given in the Tables for the Seven Planets and the Parisian Alfonsine Tables.
Thus, the general rule for the equation of center of the planets can be written
as

(18) c3(k ) = c3(k + kh3) + kv3,

where kh3 and kv3 are the horizontal and vertical displacements associated
with c3(), the equation of center in the Parisian Alfonsine Tables. Once c3(k )
is known, the true argument of center, k, is defined as:

(19) k = k + c3(k ),

22 In the Parisian Alfonsine Tables the equation of center is displayed in the third column in
the tables for planetary equations, hence the notation used here.
displaced tables in latin 123

figure 5.d Mars, equation of center and equatio porcionis

on analogy with the expression for the true argument of center, , used in the
Alfonsine tables,

(20) = + c3(),

where c3() 0 when 0 180, as shown in Figure 5.d.


All the underlying parameters displayed in Table 5 are strictly in the tradition
of the Alfonsine corpus. Interestingly, the function for Venus is presented with
displacements that differ from that for the Sun (see 3, above), although they
share the same maximum equation, 2;10.
As for the column labeled equatio porcionis, note that for a given argument
its entries and those corresponding to them in the equation of center always
add up to an integer number of degrees: 14 (Saturn), 12 (Jupiter), 24 (Mars),
6 (Venus), and 8 (Mercury). This is exactly twice the vertical displacement
applied to each planet, that is, 2 kv3. Thus, if we call the equatio porcionis
e(k ), it follows that the author of the Tables of the Seven Planets defined it
as

(21) e(k ) + c3(k ) = 2 kv3.

This definition was intended to eliminate subtractions, as will be justified


below: there is no counterpart to the function e(k ) in the standard Alfonsine
tables. Figure 5.d shows the equation of center and the equatio porcionis in the
case of Mars. The other planets follow the same pattern (see Table 5).
When e(k ) is known, the true anomaly, a, is defined as:
124 chapter 5

(22) a = + e(k ),

on analogy with the expression for the true argument of anomaly, , used in the
Alfonsine tables,

(23) = c3()

where, again, c3() 0 when 0 180.


In the third column of this table the entries for the minutes of proportion
are also dependent upon those in the Parisian Alfonsine Tables (see Figure
5.e for the case Mars) except for two features: a horizontal displacement, dif-
ferent for each planet, and a vertical displacement of 60, which is limited to
the positions when the planet is near its apogee. Note that in the Parisian
Alfonsine Tables the minutes of proportion for these positions are subtrac-
tive, whereas they are additive for the positions near perigee. Consequently,
in order to avoid subtractions, the author of the Tables for the Seven Plan-
ets only applied a vertical displacement of 60 to half of the column, that
is, where the center of the epicycle is near apogee, and left unchanged the
additive part (for details see Table 6). With this approach, if an entry, c4(),
gives the minutes of proportion in the Parisian Alfonsine table,23 the cor-
responding entry in the Tables for the Seven Planets in the case of Mars is
c4(k) = c4(k + 49) + 60, near apogee, and c4(k) = c4(k + 49), near perigee.
Thus, the general rule for the equation of center of the planets can be written
as

(24) c4(k) = c4(k + kh4) + kv4,

near apogee, and simply

(25) c4(k) = c4(k + kh4),

near perigee, where kh4 and kv4 (= 60) are the horizontal and vertical displace-
ments, respectively, associated with c4(), the minutes of proportion. Figure 5.e

23 In the Parisian Alfonsine Tables the minutes of proportion are displayed in the fourth
column in the tables for planetary equations; hence the notation used here. Note, however,
that in the corresponding table in Ptolemys Almagest they are found in the eighth column
and depend on the mean argument of center, whereas in the Parisian Alfonsine Tables
they are a function of the true argument of center. For this function in the Almagest, see
Neugebauer 1975, pp. 185186.
displaced tables in latin 125

figure 5.e Mars, minutes of proportion

displays the minutes of proportion for Mars; the other planets follow the same
pattern (see Table 6).
As in previous cases, the vertical displacements are intended to avoid cum-
bersome rules for addition and subtraction corresponding to the simple rules
we now give by means of algebraic signs. The horizontal displacements are
intended to counterbalance another displacement in a different column or
table. That this is indeed the case can be seen from the fact that the horizontal
displacements applied to the argument of center (the column on the right side
in Table 5) and to the minutes of proportion (the column on the right side in
Table 6) add up algebraically to the vertical displacement applied to the equa-
tion of center (see Table 5):

(26) kh3 kh4 = kv3.

From eqs. 16, 18, 20, and 26 it follows that

(27) k = kh4.

Therefore kh4 is the displacement to be applied to the Alfonsine true argument


of center to obtain the true argument of center in the Tables of the Seven
Planets.24

24 Indeed, k = k + c3(k ) = kh3 + c3() + kv3 = + c3() kh3+ kv3 = (kh3+ kv3) = kh4.
126 chapter 5

table 6 Displacements for the minutes of proportion

Entry Vertical displacement Horizontal displacement


for 0 (kv4) (kh4)

Saturn 0 60 7
(265266 to 360 and 0 to 8081)
Jupiter 0 60 12
(259260 to 360 and 0 to 7778)
Mars 20 60 49
(223224 to 360 and 0 to 3839)
Venus 19 60 48
(223224 to 360 and 0 to 4041)
Mercury 10 60 24
(271272 to 360 and 0 to 4142)

11.2 Equation of Anomaly near Greatest Distance


For each planet the equation of anomaly is displayed in four tables: two, here
called (i) and (ii), associated with greatest distance (i.e., between greatest and
mean distance) and two, with exactly the same presentation, associated with
least distance (i.e., between mean and least distance: see 11.3). In the Almagest
and sets of tables related to it the equation of anomaly is treated separately
when the center of the epicycle is near apogee in contrast to the case when the
center of the epicycle near perigee. Ptolemy first assumed that the center of
the epicycle is at mean distance and computed the equation of anomaly under
this condition (Ptolemys column 6). He then considered the case where the
center of the epicycle lies between maximum distance (apogee), where the
equation of anomaly is least, and mean distance. Ptolemys column 5 gives
the subtractive difference to be applied to the equation of anomaly at mean
distance. This is because at maximum distance the equation of anomaly is
least. To interpolate between mean distance and maximum distance Ptolemy
applied the minutes of proportion in his column 8 to the subtractive difference;
thus, in this case, the total equation is

(28) c() = c6() c5() c8()

where is near apogee. As Neugebauer (1975, pp. 185186) demonstrated, Ptol-


emy used different formulas for c8() near apogee and near perigee, but put the
values in a single column. Between mean distance and perigee Ptolemy com-
displaced tables in latin 127

puted the entries in his column 7 which is the amount to the added to the value
for mean distance to get the equation of anomaly at least distance (perigee)
where the equation of anomaly is greatest; thus, in this case, the total equation
is

(29) c() = c6() + c7() c8(),

where is near perigee. So in some cases one has to subtract and in other cases
one has to add. The author of the Tables for the Seven Planets wished to avoid
subtractions and he took a different approach without changing the model
or the parameters. Instead of first tabulating the values for mean distance
with a subtractive difference, he tabulated the equation of anomaly at greatest
distance, where the equation is least. Hence, all corrections are positive, and
the total equation is

(30) c(a) = d(a) + c5(a) c4(k),

where d(a) is the equation of anomaly at apogee, c5(a) corresponds to Ptol-


emys c5(), and c4(k) (defined in eq. 24) corresponds to Ptolemys c8(); all
terms are positive. For values of near perigee, the author of the Tables for
the Seven Planets adhered more closely to Ptolemys formula since it did not
involve subtraction. Thus, between mean distance and minimum distance, our
author used

(31) c(a) = c6(a) + c7(a) c4(k),

where c6(a) corresponds to Ptolemys c4(), c7(a) corresponds to Ptolemys


c7(), and c4(k) (defined in eq. 25) corresponds to Ptolemys c8(). We will see
that the entry for mean distance is the same whether one uses the formula near
apogee or the formula near perigee.

(i) The argument is the minutes of proportion, as indicated in the heading, dis-
played here as integers from 0 to 60. There are two other columns, one labeled
diversitas dyametri centralis and another for the first station near greatest dis-
tance (see Table c). The entries corresponding to 0 refer to greatest distance,
and the entries corresponding to 60 refer to mean distance.
This table is unprecedented in the astronomical literature and, in the ab-
sence of canons explaining its use, it is difficult to interpret the meaning of the
expression, diversitas dyametri centralis. Table 7 shows the extremal values for
all planets.
128 chapter 5

table c Diversitas dyametri centralis


and first station near greatest
distance for Jupiter (excerpt)

Minutes Diversitas First


of prop. dyametri centralis station
() (s, ) (s, )

0 5 22;37,40 4 4; 5
1 5 22;37,10 4 4; 7

15 5 22;30,10 4 4;30

30 5 22;22,40 4 4;53

45 5 22;14,40 4 5;17

59 5 22; 8,10 4 5;38


60 5 22; 7,40 4 5;39

A close inspection of the entries for the diversitas dyametri centralis indicates
that they are strongly dependent on the apogees of the planets at epoch (noon,
February 28, 1341, Toledo). A comparison of our computation of the planetary
apogees for the city of Toledo at that time and the entries for 0 for each planet
is displayed in Table 8.
From Table 8 it follows that the entries for 0 tabulated here for the diversitas
dyametri centralis are obtained by adding a constant to the longitude of the
apogee of each planet at epoch, 0. This constant is specific to each planet
and results in turn from the displacements applied to its variables, kh4 max
(c6 c5): see Table 10. Therefore, these entries play a crucial role in computing
the true planetary longitudes but have no direct astronomical meaning. On
the other hand, for each planet the range for the entries is 0;20,42 (= 8s
12;42,18 8s 12;21,36); or 0;21 (if we consider 8s 12;21,18 as the intended entry
for 60) for Saturn, 0;30 for Jupiter, 5;38 for Mars, 1;42 for Venus, and 3;12
for Mercury. It turns out that these values agree with the vertical displacement
applied to the difference at greatest distance, c5, associated with the argument
of anomaly. We have called this quantity kv5 (see Table 11). In the case of Mars,
its value is 5;38, and the diversitas dyametri centralis near greatest distance,
d5, which is a function of the minutes of proportion, m, can be written as
displaced tables in latin 129

table 7 Diversitas dyametri centralis and first station near greatest distance for all planets

Diversitas dyametri centralis First station


Entry for 0 Entry for 60 Entry for 0 Entry for 60

Saturn 8s 12;42,18 8s 12;21,36 * 3s 22;44 3s 24; 7


Jupiter 5s 22;37,40 5s 22; 7,40 4s 4; 5 4s 5;39
Mars 4s 25;38,49 4s 20; 0,49 5s 7;28 5s 13;11
Venus 3s 2;48, 9 3s 1; 6, 9 5s 15;51 5s 17; 9
Mercury 7s 3;50, 9 7s 0;38, 9 4s 27;14 4s 25; 8

* Probably an error for 8s 12;21,18. For justification, see below. One should also note
that the number of seconds in the entries for 0 and 60 in columns 1 and 2 agree for
each planet (but for Saturn here), both in this table and in the equivalent, Table 12.

table 8 Planetary apogees near greatest distance

Computed Entry for 0 comp. Rounded value *

Saturn 251;35,19 1; 6,59 1; 7 (= 7 5;53)


Jupiter 171;31,37 1; 6, 3 1; 6 (= 12 10;54)
Mars 133;23,50 12;14,59 12;15 (= 49 36;45)
Venus 89;37, 0 3;11, 9 3;11 (= 48 44;49)
Mercury 208;51,10 4;58,59 4;59 (= 24 19; 1)

* As will be seen below, the rounded value for each planet should result from subtract-
ing the maximum of c6 c5 (see Table 10) from kv3 (see Table 5). In the case of Jupiter,
this gives 1;26 (= 12 10;34), whereas we obtain 1;6 when using the entry for 0 in the
table, 5s 22;37,40 (see Table 7), and a set of standard Alfonsine tables. If the author
had taken 5s 22;57,40 as the entry for 0, the rounded value would have been fully
consistent with those for the other planets.

d5(m) = d5(0) 5;38 c4(k). Thus, the general rule for the diversitas dyametri
centralis near greatest distance is given by

d5(m) = d5(0) kv5 c4(k),

that is,

(32) d5(m) = 0 + kh4 max (c6 c5) kv5 c4(k).


130 chapter 5

figure 5.f First station of Mars

The third column in this table displays the first station as a function of the
minutes of proportion. The entries for the first station represent the true argu-
ment of anomaly of the first stationary point, and those for 0 agree with the
entries in the Almagest (see, e.g., Toomer 1984, p. 588) and sets of tables related
to it. They correspond to the greatest distance of the center of the epicycle,
when the mean argument of center, , is zero. It should be recalled that the
table in Almagest xii.8 gives the positions of the stationary points on the epicy-
cle as a function of (see Neugebauer 1975, pp. 202205), not as a function of the
minutes of proportion, as is the case here. The entries increase (except for Mer-
cury) monotonically and vary in a range different for each planet: 1;23 (Saturn),
1;34 (Jupiter), 5;43 (Mars), 1;18 (Venus), and 2;6 (Mercury). Figure 5.f dis-
plays the first station of Mars, both near apogee (see Table 7) and near perigee
(Table 12).
If we call s5(m) the tabulated argument of anomaly of the first station near
apogee, the entries are can be computed by means of a linear function of m, the
minutes of proportion:

(33) s5(m) = s5(0) + [s5(60) s5(0)] c4(k),

which, in the case of Mars, turns into

s5(m) = 157;28 + 5;43 c4(k).

(ii) In the second table the argument ranges from 0 to 29 for each zodiacal
sign. There are two columns, one for the equation of anomaly (the heading
has argumentum for anomaly) and another for the diversitas dyametri. The
displaced tables in latin 131

table d Equation of anomaly and diversitas dyametri for Mars (excerpt)

Diversitas Diversitas
Equation of dyametri (directus Equation of dyametri (directus
anomaly or retrogradus) anomaly or retrogradus)
() () () () () ()

0 36;45 5;38 180 36;45 5;38


1 37; 7 5;40
191 21;19 1;34
90 67;40 8; 5 statio
192 20; 3 1;20
125 73;29 9;48
126 73;30 9;52 203 9;15 0; 1
127 73;30 9;55 secunda
128 73;29 9;59 directus
204 8;31 0; 0
154 66;22 11;16 205 7;48 0; 0
155 65;42 11;16 206 7; 8 0; 0
156 64;59 11;16
157 64;15 11;15 232 0; 1 1;17
statio 233 0; 0 1;21
158 63;29 11;14 234 0; 0 1;24
235 0; 0 1;28
169 52;11 9;42
prima 270 5;50 3;11
retro.
170 50;55 9;26 358 36; 0 5;35
359 36;23 5;36

title indicates that the entries are given for greatest distance of the center of
the epicycle (ad longitudinem longiorem), that is, near apogee. The heading of
the entries for the diversitas dyametri is generally diversitas dyametri directus,
and sometimes diversitas dyametri retrogradus, a change which is indicated by
the insertion of the words statio, prima, and secunda, among the entries. See
Table d.
The relevant information for the equation of anomaly is summarized in
Table 9, where the underlying parameter was derived by taking half the
difference between maximum and minimum equation of anomaly.
132 chapter 5

table 9 Equation of anomaly near greatest distance for all planets

Greatest Entry Minimum equation Maximum equation Underlying


distance for 0 of anomaly of anomaly parameter

Saturn 5;53 0;0 11;46 5;53


(261266) (9297)
Jupiter 10;34 0;0 21; 8 10;34
(258263) (97102)
Mars 36;45 0;0 73;30 36;45
(233234) (126127)
Venus 44;49 0;0 89;38 44;49
(225) (135)
Mercury 19; 1 0;0 38; 2 19; 1
(250252) (108)

The underlying parameters in this table are the maximum values of the dif-
ference c6 c5, that is, the maximum difference between the equation of
anomaly at mean distance and the subtractive difference in the Parisian Alfon-
sine Tables.25 All these values are shown in Table 10.
Figure 5.g displays the tabulated values of the equation of anomaly for Mars
near greatest distance, here called d(a) and the difference between c6 and c5 in
the Parisian Alfonsine Tables. It is readily seen that d(a) is displaced upwards
with respect to c6 c5. The other planets follow the same pattern.
In the case of Mars the expression for the equation of anomaly at greatest
distance can be written as

d(a) = c6() c5() + 36;45,

which, in the case general, corresponds to

(34) d(a) = c6() c5() + max (c6 c5).

25 In the Parisian Alfonsine Tables the fifth and sixth columns in the tables for planetary
equations display, respectively, the subtractive difference and the equation of anomaly at
mean distance; hence the notation used here.
displaced tables in latin 133

table 10 Maximum values of c6, c5 and c6 c5 in the Parisian Alfonsine Tables

Maximum Maximum Maximum


of c6 of c5 of c6c5

Saturn 6;13 0;21 5;53


(at 9499) (at 100106) (at 9499)
Jupiter 11; 3 0;30 10;34
(at 99102) (at 107117) (at 97102)
Mars 41;10 5;38 36;45
(at 131132) (at 153156) (at 126127)
Venus 45;59 1;42 44;49
(at 135136) (at 161162) (at 135)
Mercury 22; 2 3;12 19; 1
(at 111112) (at 129131) (at 108)

figure 5.g Equation of anomaly near greatest distance for Mars

The diversitas dyametri tabulated for each planet is the difference near
greatest distance. The entries in this column agree with those for the same
quantity, c5(), in the Parisian Alfonsine Tables, but for a vertical displacement,
which differs from one planet to another (see Table 11).
In the case of Mars this displacement amounts to 5;38. In general, if c5() is
the difference near greatest distance in the corresponding Parisian Alfonsine
table, that in the Tables for the Seven Planets is given by
134 chapter 5

table 11 Difference near greatest distance

Greatest Entry Minimum difference Maximum difference Vertical dis-


distance for 0 at greatest distance at greatest distance placement (kv5)

Saturn 0;21 0;0 0;42 0;21


(253261) (99104)
Jupiter 0;30 0;0 1; 0 0;30
(242252) (108118)
Mars 5;38 0;0 11;16 5;38
(204207) (153156)
Venus 1;42 0;0 3;24 1;42
(198199) (161162)
Mercury 3;12 0;0 6;24 3;12
(229231) (129131)

(35) c5(a) = c5() + kv5.

We note that a = , which is a consequence of equations 17, 18, 21, and 22.26
Therefore, the displacements applied by the author of the Tables for the Seven
Planets to the standard quantities used in Alfonsine planetary astronomy are
such that they keep invariant the argument of anomaly.
As in eq. 28, the total correction when the planet is near apogee in Ptolemys
notation is

c() = c6() c5() c8(),

which, in standard Alfonsine notation, is equivalent to

c() = c6() c5() c4().

In the Tables for the Seven Planets the corresponding expression near apogee
is

(36) c(a) = d(a) + c5(a) c4(k),

26 Indeed, a = + e(k ) = ( kv3) + (2 kv3 c3()) = + kv3 [(c3() +kv3)] = c3() = .


displaced tables in latin 135

table e Diversitas dyametri centralis and first station


near least distance for Venus (excerpt)

Minutes Diversitas First


of prop. dyametri centralis station
() (s, ) (s, )

0 3 1;38, 9 5 17; 9
1 3 1;36,17 5 17;10

15 3 1;10, 9 5 17;25

30 3 0;42, 9 5 17;43

45 3 0;14, 9 5 18; 4

59 2 29;48, 1 5 18;20
60 2 29;46, 9 4 18;21

an expression only involving positive terms and leading to the same true lon-
gitude of the planets as the standard Alfonsine procedure.

11.3 Equation of Anomaly near Least Distance


The following two tables, for least distance (i.e., between mean and least dis-
tance), have the same presentation as those for greatest distance (i.e., between
greatest and mean distance), reviewed in 11.2 (i) and (ii).

(i) As in Table c the argument is the minutes of proportion from 0 to 60,


and there are two other columns, one labeled diversitas dyametri centralis and
another for the first station near least distance (see Table e). Here the entries
for 0 correspond to mean distance, and the entries for 60 correspond to least
distance.
The extremal values of the two columns (diversitas dyametri centralis and
first station) for all planets are listed in Table 12.
As was the case for greatest distance, the entries for 0 tabulated here for the
diversitas dyametri centralis are obtained by adding a constant to the longitude
of the apogee of each planet at epoch, 0. This constant is specific to each planet
and results in turn from the displacements applied to its variables: kh4 kv6
(see Table 13).
136 chapter 5

table 12 Diversitas dyametri centralis and first station near least distance for all planets

Diversitas dyametri centralis First station


Entry for 0 Entry for 60 Entry for 0 Entry for 60

Saturn 8s 12;22,18 8s 11;57,18 3s 24; 8 3s 25;27


Jupiter 5s 22; 8,40 5s 21;35,40 4s 5;40 4s 7;11
Mars 4s 21;13,49 4s 13;10,49 5s 13; 6 5s 19;15
Venus 3s 1;38, 9 2s 29;46, 9 5s 17; 9 5s 18;21
Mercury 7s 0;49, 9 6s 28;48, 9 4s 25; 8 4s 24;28

table 13 Planetary apogees at least distance

Computed Entry for 0 comp. Rounded value *

Saturn 251;35,19 0;46,59 0;47 (= 7 6;13)


Jupiter 171;31,37 0;37, 3 0;37 (= 12 11;23)
Mars 133;23,50 7;49,59 7;50 (= 49 41;10)
Venus 89;37, 0 2; 1, 9 2; 1 (= 48 45;59)
Mercury 208;51,10 1;57,59 1;58 (= 24 22; 2)

* As will be seen below, the rounded value for each planet should result from subtract-
ing kv6 (see Table 14) from kv3 (see Table 5). In the case of Jupiter, this gives 0;57
(= 12 11;3), whereas we obtain 0;37 when using the entry for 0 in the table, 5s
22;8,40 (see Table 12) and a set of standard Alfonsine tables. If the author had taken
5s 22;28,40 as the entry for 0, the rounded value would have been fully consistent with
those for the other planets.27

On the other hand, the range for the entries is: 0;25 for Saturn, 0;33 for Jupiter,
8;3 for Mars, 1;52 for Venus, and 2;1 for Mercury. The entries for least distance
are not exactly a continuation of those for greatest distance (Table 7), because
they were computed by means of different expressions and different ranges

27 This is the same situation we noted in Table 8; thus, it would seem that the author either
miscopied an entry for Jupiter on which he built up his two tables for the diversitas
dyametri centralis, or that he had at his disposal a table for Jupiter generating a difference
of 0;20.
displaced tables in latin 137

were used. In the case of Mars the vertical displacement applied here, kv7,
is 8;3 rather than kv5 = 5;38, as was the case at greatest distance. Then the
diversitas dyametri centralis near least distance, d7, can be written as d7(m)
= d7(0) 8;3 c4(k), where m is a value for the minutes of proportion. Thus,
the general rule for the diversitas dyametri centralis near least distance is given
by

d7(m) = d7(0) kv7 c4(k)

that is,

(37) d7(m) = 0 + kh4 kv6 kv7 c4(k).

The third column in this table displays the argument of anomaly of the first
station near least distance as a function of the minutes of proportion (Table 12).
The entries follow the same pattern as those near greatest distance, but we note
that the entries for 60 at greatest distance (see Table 7) do not coincide with
the entries for 0 near least distance (see Table 12) in three cases: Saturn, Jupiter,
and Mars. The entries for 60 essentially agree with those in the Almagest and
sets of tables related to it, corresponding to the least distance of the center of
the epicycle, when the mean argument of center, , is 180. The entries, graphed
in Figure 5.f, increase (except for Mercury) monotonically and vary in a range
which is different for each planet: 1;19 (Saturn), 1;31 (Jupiter), 6;9 (Mars), 1;12
(Venus), and 0;40 (Mercury).
If we call s7(m) the tabulated argument of anomaly of the first station near
perigee, the entries are can be computed by means of a linear function of m,
the minutes of proportion,

(38) s7(m) = s7(60) + [s7(0) s7(60)] (1 c4(k)),

which, in the case of Mars, turns into

s7(m) = 159;15 6;9 (1 c4(k)).

(ii) In the second table the argument ranges from 0 to 29 for each zodiacal
sign. There are two columns (equation of anomaly and diversitas dyametri). The
title indicates that the entries are given for least distance of the center of the
epicycle (ad longitudinem propiorem), that is, near perigee (see Table f). As was
the case for Table d, the words statio, prima, and secunda are inserted among
various entries.
138 chapter 5

table f Equation of anomaly and diversitas dyametri for Mercury (excerpt)

Diversitas Diversitas
Equation of dyametri (directus Equation of dyametri (directus
anomaly or retrogradus) anomaly or retrogradus)
() () () () () ()

0 22; 2 2; 1 180 22; 2 2; 1


1 22;19 2; 2
213 5;27 0; 8
90 42;35 3;30 secunda
214 5; 7 0; 6
110 44; 3 9;48 215 4;48 0; 4
111 44; 4 9;52 statio
112 44; 4 9;55 directus
113 44; 3 9;59 216 4;30 0; 3

130 42;46 4; 1 223 2;38 0; 1
131 42;37 4; 2 224 2;25 0; 0

136 41;39 4; 2 229 1;27 0; 0
137 41;26 4; 1 230 1;18 0; 1

144 39;34 3;59 247 0; 1 0; 9
prima 248 0; 0 0;10
145 39;16 3;58 249 0; 0 0;11
146 38;57 3;56 250 0; 1 0;12
147 38;37 3;54
statio 270 1;29 0;32
retro.
148 38;16 3;52 358 21;29 1;59
359 21;45 2; 0

The relevant information for the equation of anomaly is summarized in


Table 14, where the underlying parameter was derived by taking half the
difference between maximum and minimum equation of anomaly.
For each planet the underlying parameter agrees with that in the Parisian
Alfonsine Tables for mean distance, and so do the entries for the equation of
anomaly, but for a vertical displacement. If c6() represents the equation of
displaced tables in latin 139

table 14 Equation of anomaly near least distance

Minimum Maximum Vertical


Least Entry equation of equation of Underlying displacement
distance for 0 anomaly anomaly parameter (kv6)

Saturn 6;13 0;0 12;26 6;13 6;13


(261266) (9499)
Jupiter 11; 3 0;0 22; 6 11; 3 11; 3
(258261) (99102)
Mars 41;10 0;0 82;20 41;10 41;10
(229) (131)
Venus 45;59 0;0 91;58 45;59 45;59
(224225) (135136)
Mercury 22; 2 0;0 44; 4 22; 2 22; 2
(248249) (111112)

anomaly in the corresponding Parisian Alfonsine table, that in the Tables for
the Seven Planets is given by

(39) c6(a) = c6() + kv6,

where kv6 = 45;59 in the case of Venus.28 Figure 5.h displays the tabulated
values of the equation of anomaly for Venus near least distance in the Tables
for the Seven Planets and at mean distance in the Parisian Alfonsine Tables.
The terminology in the Tables of the Seven Planets may be confusing, but
the author sought to maintain a parallel structure in his treatment of the
equation of anomaly near apogee and near perigee. The entries near apogee
range from greatest distance to mean distance, and those near perigee range
from mean distance to least distance. Hence, what is here called the equa-
tion of anomaly near least distance actually refers to the equation of anomaly
at mean distance, to be applied to distances between mean and least dis-
tance.
The diversitas dyametri tabulated for each planet is the difference near
least distance to be added to the corresponding equation of anomaly at mean
distance (see the extremal values in Table 15). The entries in this column agree

28 This expression is equivalent to c6(a) = c6(a) + kv6, because a = .


140 chapter 5

figure 5.h Venus, equation of anomaly as a function of the true argument of anomaly

table 15 Difference at least distance

Least Entry Minimum difference Maximum difference Vertical dis-


distance for 0 at least distance at least distance placement (kv7)

Saturn 0;25 0;0 (249258) 0;50 (102111) 0;25


Jupiter 0;33 0;0 (240253) 1; 6 (107120) 0;33
Mars 8; 3 0;0 (201) 16; 6 (159) 8; 3
Venus 1;52 0;0 (197199) 3;44 (161163) 1;52
Mercury 2; 1 0;0 (224229) 4; 2 (131136) 2; 1

with those for the same quantity, c7(), in the Parisian Alfonsine Tables, but
for a displacement, which differs from one planet to another. If c7() is the
difference at least distance in the corresponding Parisian Alfonsine table, the
difference at least distance in the Tables for the Seven Planets is given by

(40) c7(a) = c7() + kv7.

Note that at least distance of the center of the epicycle, i.e., perigee, the equa-
tion of anomaly is greatest.29

29 These tables bear some similarity with the tables of John Vimond, who tabulated c6
displaced tables in latin 141

As indicated previously in eq. 29, in Ptolemys notation the total correction


when the planet is near perigee is

c() = c6() + c7() c8(),

which, in standard Alfonsine notation, is equivalent to

c() = c6() + c7() c4().

In the Tables for the Seven Planets, the corresponding expression near perigee
is

(41) c(a) = c6(a) + c7(a) c4(k).

This is an expression involving only positive terms and leads to the same true
longitude of the planets as the standard Alfonsine procedure.
To find the true longitude, , of the planet at that time according to the
Parisian Alfonsine Tables one has to add the correction c() to its mean lon-
gitude, that is, to the sum of the longitude of the apogee, 0, and the true
argument of center of the planet, :

= 0 + + c(),

where c() = c6() + c7() c4().

Now, the author of the Tables of the Seven Planets introduced an analogous
expression to find the true longitude of the planet:

(42) l = d7(m) + k + c(a),

where c(a) = c6(a) + c7(a) c4(k): see eq. 41. The term d7(m) is tabulated under
diversitas dyametri centralis.
The general expression for the longitude of a planet near perigee is therefore
given as

(43) l = d7(m) + k + c6(a) + c7(a) c4(k),

c5, but he did not apply any displacements to the planets. Vimond also tabulated c5 + c7,
which has no counterpart in these tables (see Chabs and Goldstein 2004, pp. 248256).
142 chapter 5

which can be obtained from

(44) = 0 + + c6() + c7() c4().

To be sure,

= 0 + + c6() + c7() c4()


= 0 + (k + kh4) + (c6() kv6) + (c7() kv7) c4(k)
= 0 + kh4 kv6 kv7 c4(k) + k + c6() + c7() c4(k) = l,

provided that d7(m) = 0 + kh4 kv6 kv7 c4(k), which is the case: see eq. 37. All
terms appearing in eq. 43 are positive and are found directly in the tables. Note
that d5(m) and d7(m) have no counterparts in the standard Alfonsine tables.
To illustrate eq. 43 consider the position of Mars at epoch (noon, February 28,
1341), when the planet is near perigee. The mean argument of center, k , and the
mean argument of anomaly, , are given as 55;6,26 and 84;48,42, respectively,
rounded to the seconds (see Table 3). From eqs. 16 and 17 it follows that
and in the standard Alfonsine Tables are 116;6,26 and 96;48,42, respectively
(where kh3 = 61 and kv3 = 12: see Table 5). The corresponding values of the
equation of center in both sets of tables are c3(k ) = 1;23 and c3() = 10;37, and
eq. 18 holds. Thus, the true arguments of center are k =56;29 and = 105;29,
and eq. 27 holds, where kh4 = 49 (see Table 6). The tabulated equatio porcionis
for k = 55;6,26 is e(k ) = 22;37, and thus a = 84;48,42 + 22;37 = 107;26. The
argument of anomaly, , according to the Parisian Alfonsine Tables, is 107;26
(= 96;48,42 + 10;37), and we note that a = . The minutes of proportion
corresponding to k = 56;29 and = 105;29 are c4(k) = 17/60 and c4() = 17/60,
and we note that c4(k) = c4(), as indicated in eq. 25, valid near perigee. The
tabulated values for the equation of anomaly, c6(a), and the diversitas, c7(a), are
79;11 and 11;48, respectively, whereas the values for the equation of anomaly,
c6(), and the additive difference, c7(), found in the Parisian Alfonsine Tables
are 38;1 and 3;45, respectively. We note that eqs. 39 and 40 hold, where kv6 =
41;10 and kv7 = 8;3 (see Tables 14 and 15). From eq. 41, c(a) = 79;11 + 11;48
(17/60) = 82;32, whereas the Parisian Alfonsine Tables yield c() = 38;1 + 3;45
(17/60) = 39;5.
Now, the true longitude derived with the standard Alfonsine Tables is =
133;23,50 + 105;29 + 39;5 = 277;57,50 (see Table 8). On the other hand, from
eq. 37, d7(17) = 141;13,49 8;3 17/60 = 138;56,58 (see Table 12). Thus, l =
138;56,58 + 56;29 + 82;32 = 277;57,58, in agreement with the previous result.
Similarly, the general expression for the longitude of a planet near apogee is
given as
displaced tables in latin 143

(45) l = d5(m) + k + d(a) + c5(a) c4(k),

and it can be obtained from the standard Alfonsine procedure

(46) = 0 + + c6() c5() c4().

To be sure,

= 0 + + c6() c5() c4()


= 0 + (k + kh4) + c6() c5() + c5() + c5() c4()
= 0 + (k + kh4) + d(a) max (c6 c5) + c5() + c5() c4()
= 0 + kh4 max (c6 c5) + k + d(a) + c5() (1 + c4())
= 0 + kh4 max (c6 c5) + k + d(a) + (c5() kv5) c4(k)
= 0 + kh4 max (c6 c5) + k + d(a) kv5 c4(k) + c5(a) c4(k)
= 0 + kh4 max (c6 c5) kv5 c4(k) + k + d(a) + c5(a) c4(k) = l,

provided that d5(m) = 0 + kh4 max (c6 c5) kv5 c4(k), which is the case:
see Table 7 and eq. 32. All terms appearing in eq. 44 are positive and are found
in the tables.
To illustrate eq. 44, consider the position of Mars at noon, October 1, 1340,
when = 37;31,6, and thus the planet is near apogee. From eq. 16 it follows that
k = 336;31,6 (where kh3 = 61; see Table 5). In the Parisian Alfonsine Tables, the
corresponding equation of center is c3() = 6;26,40, whereas the tabulated
equatio centri and equatio porcionis for k = 336;31,6 are c3(k ) = 5;33,20 and
e(k ) = 18;26,40, respectively. We note that eq. 18 holds, where kv3 =12 (see
Table 5). We also note that eq. 27 holds, where kh4 = 49 (see Table 6). Then,
= 31;4,26 because of eq. 18, and k = 342;4,26 because of eq. 19. In the Parisian
Alfonsine Tables, the corresponding minutes of proportion is c4() = 51/60,
whereas the minutes of proportion tabulated here is c4(k) = 9/60. Therefore eq.
24 holds, where kv4 = 60. The mean argument of anomaly of Mars for that date
is = 27;35,24, and thus the true argument of anomaly is = 34;2,4, because
of eq. 23, and = 15;35,24, because of eq. 17, where kv3 =12. Therefore, a =
34;2,4, because of eq. 22. We note that a = . In the Parisian Alfonsine Tables,
the corresponding entries for the equation of anomaly and the difference at
greatest distance are c6() = 13;26 and c5() = 0;48, respectively. Thus, the
total correction is c() = 13;26 + 0;48 (51/60) = 12;45,12. On the other hand,
the tabulated values for the diversitas dyametri and the equation of anomaly
are c5(a) = 6;26 (we note that eq. 35 holds, where kv5 = 5;38: see Table 11)
and d(a) = 49;23. From eq. 36 it follows that c(a) = 49;23 + 6;26 9/60 =
50;20,54.
144 chapter 5

table 16 Displacements in the Tables for the Seven Planets

kv kv3 kv5 kv6 kv

Sun 89;37,9 Moon 13; 9 2;40 4;56 Eq. 8th sphere 9

kv3 kh3 kv4 kh4 kv5 kv6 kv7

Saturn 7 14 60 7 0;21 6;13 0;25


Jupiter 6 18 60 12 0;30 11; 3 0;33
Mars 12 61 60 49 5;38 41;10 8; 3
Venus 3 51 60 48 1;42 45;59 1;52
Mercury 4 28 60 24 3;12 22; 2 2; 1

Now, the true longitude derived with the standard Alfonsine Tables is =
133;23,50 + 31;4,26 + 12;45,12 = 177;13,28 (see Table 8). On the other hand,
from eq. 32, d5(9) = 145;38,49 5;38 9/60 = 144;48,7 (see Table 7). Thus, l
= 144;48,7 + 342;4,26 + 50;20,54 = 177;13,27, in agreement with the previous
result.
In Table 16 we present a summary of the values for the displacements, both
vertical and horizontal, applied by the anonymous author of the Tables of the
Seven Planets to the Sun, the Moon, the 8th sphere, and the planets.

12 Latitudes of the Superior Planets

Folio 40r displays a table for the latitudes of the three superior planets (the
inferior planets are addressed in a very different way on ff. 41r46v: see 16,
below). This table is in the Almagest tradition, and is found in many other sets
of tables such as the zij of al-Battn and the Toledan Tables (see Chabs and
Goldstein 2012, Table 9.2b, p. 109). For the zij of al-Battn, see Nallino 19031907;
and for the Toledan Tables, see Toomer 1968 and Pedersen 2002.

13 Planetary Visibility

On f. 40v there is a table entitled Tabula visionis et occultationis for the three
superior planets and the two inferior planets. It is a table for the visibility of
the planets, also called a table of planetary phases, which is already found in
displaced tables in latin 145

table g Lunar latitude at eclipse (excerpt)

Argument of latitude Latitude


(s, ) ()

0 0 6 0 12 0 6 0 0; 0
0 1 5 29 11 29 6 1 5;13

0 12 5 18 11 18 6 12 62;16
0 13 5 17 11 17 6 13 67;23

Almagest xiii.10, as well as in many other sets of astronomical tables, such as


the Handy Tables, the zij of al-Battn, and the Toledan Tables (see Chabs and
Goldstein 2012, Table 11.2, p. 125).

14 Possibility of an Eclipse

On f. 40v there is a small table entitled Tabula latitudinis lune in principio medio
et fine eclipsis, an excerpt of which we reproduce above (see Table g). The
argument, which is the argument of lunar latitude, is presented in four columns
and, as indicated in the title, it is restricted to the values for which an eclipse
is possible, that is, 13 from the lunar nodes. The purpose of the table is to
show the correspondence between the argument of latitude and the latitude
of the Moon, and the entries can be recomputed by means of the modern
formula

= arcsin (sin i sin ),

where is the argument of latitude, is the latitude, and i is the inclination of


the lunar orb to the ecliptic, taken here as 5;0 (the same parameter as in the
table for lunar latitude on f. 13r).

15 Eclipsed Fraction of the Solar and Lunar Disks

A table for the eclipsed fraction of the solar and lunar disks is also found
on f. 40v. The argument ranges from 1 to 12 linear digits (where the diameter
of the eclipsed body is 12 digits), and the entries are the corresponding area
146 chapter 5

table h Latitude of Venus (excerpt)

Anomaly
0s6/11s24 3s0/9s0 5s24/6s6 6s0/6s0
Center 1st 2nd 1st 2nd 1st 2nd 1st 2nd 3rd
(s, ) (s, ) (s, ) (s, ) () () () () () () () () ()

0 6 6 6 5 24 11 24 0; 7 0; 1 0;0 0;12 0;38 0; 5 0;46 0;0 1


0 12 6 12 5 18 11 18 0;13 0; 2 0;0 0;23 1;15 0;10 1;30 0;0 2
0 18 6 18 5 12 11 12 0;19 0; 2 0;0 0;35 1;52 0;14 2;14 0;0 3

1 0 7 0 5 0 11 0 0;31 0; 4 0;0 0;59 3; 6 0;24 3;41 0;0 5

2 0 8 0 4 0 10 0 0;54 0; 7 0;0 1;41 5;26 0;42 6;25 0;0 9

2 18 8 18 3 12 9 12 1; 0 0; 8 0;0 1;53 6; 1 0;47 7;10 0;0 10


2 24 8 24 3 6 9 6 1; 1 0; 8 0;0 1;55 6; 8 0;48 7;16 0;0 10
3 0 9 0 3 0 9 0 1; 2 0; 8 0;0 1;57 6;12 0;48 7;22 0;0 10

digits (where the area of the eclipsed body is 12). It is very common in sets of
astronomical tables, such as the zij of al-Battn, the Toledan Tables, and it is
already found in Almagest vi.8. However, this was not one of the tables included
in the editio princeps of the Alfonsine Tables (see Chabs and Goldstein 2012,
Table 15.4, p. 175).

16 Latitudes of Venus and Mercury

The tables for the latitude of Venus and Mercury are presented as six sub-tables
for each planet, on ff. 41r43v and ff. 44r46v, respectively. Both are double
argument tables. The vertical argument (center) is given in four columns at
6-intervals and altogether there are 30 columns for the horizontal argument
(anomaly), also at 6-intervals. Each of these columns contains entries for the
inclination (latitudo prima) and the slant (latitudo secunda). The deviation
(latitudo tercia) is presented in another column, which remains invariant in
all sub-tables for each planet. Table h displays an excerpt of the table for the
latitude of Venus.
The maximum values for Venus and Mercury are, respectively, 7;22 and
4;5 for the inclination, 2;30 and 2;30 for the slant, and +10 and 13 for the
displaced tables in latin 147

deviation. This table differs, strongly in presentation and slightly in the basic
parameters, from tables by Parisian astronomers who included the third com-
ponent of latitude, John Vimond and John of Murs.30 For the maximum val-
ues of deviation, Vimond had +10and 45 (Ptolemys values in the Almagest),
for Venus and Mercury, respectively, whereas John of Murs had +10and 23,
in contrast to +10and 13 in the tables reviewed here. John of Lignres was
aware of the deviation, for he mentions it in the chapters on the latitude
of Venus and Mercury in the canons of his Priores astrologi motus corporum
celestium,31 but we do not know of any tables by him similar to those pre-
sented here. All in all, the tables for planetary latitudes in the Tables for the
Seven Planets, although certainly in the same tradition, are not simply related
to those by John Vimond and John of Murs, or by any other known table-
maker.

Conclusion

Throughout the time from the reception of the Alfonsine Tables in Paris (no
later than 1320) to the publication of the editio princeps in 1483 in Venice, an
intense effort was made to adapt tables in the Alfonsine framework to the needs
of practitioners. These adaptations focused on presentation rather than on the
parameters underlying the tables of the Alfonsine corpus of tables. The set of
tables which we have called the Tables for the Seven Planets for 1340 is an early
example of this kind of work, with zodiacal signs of 30 (rather than physical
signs of 60), vacillation in the beginning of the year, and cyclical radices of
32 years, with special characteristics such as displacements, different param-
eter for the latitude of Mercury, and different terminology. The use of double
argument tables and the extensive use of displacements, both vertical and hor-
izontal, show a deep insight into planetary astronomy and great skill in pro-
ducing astronomical tables. This set of tables is also the first example that has
been discovered in the Latin world of a systematic use of displaced tables, of
which only a few examples were previously known in the medieval astronomi-
cal literature. Unfortunately, no name is associated with this set, but the author
was an astronomer working around 1340, probably in Southern France (judg-
ing by the geographical coordinates underlying these tables), who deserves the
highest praise for his skill in providing clever and complex solutions to many

30 Chabs and Goldstein 2004, pp. 257258; Chabs and Goldstein 2009, p. 309.
31 An edition of chapters 22 and 23 is found in Saby 1987, pp. 207211.
148 chapter 5

problems, and for constructing a compact and consistent set of tables for the
planets, building upon the work done by the Parisian group of astronomers in
the early 1320s.

References

Chabs, J. and B.R. Goldstein 2003. The Alfonsine Tables of Toledo. Archimedes: New
Studies in the History and Philosophy of Science and Technology, 8. Dordrecht and
Boston.
Chabs, J. and B.R. Goldstein 2004. Early Alfonsine Astronomy in Paris: The Tables of
John Vimond (1320), Suhayl 4:207294.
Chabs, J. and B.R. Goldstein 2009. John of Murss Tables of 1321, Journal for the History
of Astronomy 40:297320.
Chabs, J. and B.R. Goldstein 2012. A Survey of European Astronomical Tables in the Late
Middle Ages. Boston.
Debarnot, M.-T. 1987. The Zj of abash al-sib: A Survey of ms Istanbul Yeni Cami
784/2, in King and Saliba 1987, pp. 3569.
Goldstein, B.R. 1974. The Astronomical Tables of Levi ben Gerson. Transactions of the
Connecticut Academy of Arts and Sciences, 45. New Haven.
Goldstein, B.R., J. Chabs, and J.L. Mancha 1994. Planetary and Lunar Velocities in
the Castilian Alfonsine Tables, Proceedings of the American Philosophical Society
138:6195.
Hogendijk, J. and A.I. Sabra 2003. The Enterprise of Science in Islam. Cambridge, ma.
Jensen, C. 1971. The Lunar Theory of al-Baghdd, Archive for History of Exact Sciences
8:321328.
Kennedy, E.S. 1977. The Astronomical Tables of Ibn al-Aclam, Journal for the History of
Arabic Science 1:1323.
King, D.A. and M.H. Kennedy (eds.) 1983. Studies in the Islamic Exact Sciences by E.S.
Kennedy, Colleagues and Former Students. Beirut.
King, D.A. and G. Saliba (eds.) 1987. From Deferent to Equant: A volume of studies in the of
history science in the ancient and medieval Near East in honor of E.S. Kennedy. Annals
of the New York Academy of Sciences, 500.
Kremer, R.L. (forthcoming). Melchion de Friquentos eclipse tables of 1437, a revised
Latin version of Immanuel Bonfilss Six Wings.
Kremer, R.L. and J. Dobrzycki 1998. Alfonsine meridians: Tradition versus experience
in astronomical practice c. 1500, Journal for the History of Astronomy 29:187199.
Mercier, R. 1989. The parameters of the Zj of Ibn al-Aclam, Archives Internationales
dHistoire des Sciences 39:2250.
Nallino, C.A. 19031907. Al-Battn sive Albatenii Opus Astronomicum, 2 vols. Milan.
displaced tables in latin 149

Neugebauer, O. 1975. A History of Ancient Mathematical Astronomy. Berlin.


Pedersen, F.S. 2002. The Toledan Tables: A review of the manuscripts and textual versions
with an edition. Copenhagen.
Poulle, E. 1984. Les tables alphonsines avec les canons de Jean de Saxe. Paris.
Ratdolt, E. (ed.) 1483. Tabule astronomice illustrissimi Alfontij regis castelle. Venice.
Saby, M.-M. 1987. Les canons de Jean de Lignres sur les tables astronomiques de 1321.
Unpublished thesis: cole Nationale des Chartes, Paris. A summary appeared as:
Les canons de Jean de Lignres sur les tables astronomiques de 1321, cole Nationale
des Chartes: Positions des thses, pp. 183190.
Salam, H. and E.S. Kennedy 1967. Solar and Lunar Tables in Early Islamic Astronomy,
Journal of the American Oriental Society 87:493497. Reprinted in King and Kennedy
1983, pp. 108113.
Saliba, G. 1976. The Double-Argument Lunar Tables of Cyriacus, Journal for the History
of Astronomy 7:4146.
Saliba, G. 1977. Computational Techniques in a Set of Late Medieval Astronomical
Tables, Journal for the History of Arabic Science 1:2432.
Sams, J. 2003. On the Lunar Tables in Sanjaq Drs Zj al-Sharf , in Hogendijk and
Sabra 2003.
Sams, J. and E. Mills 1998. The computation of planetary longitudes in the zj of Ibn
al-Bann, Arabic Sciences and Philosophy 8: 259286.
Solon, P. 1970. The Six Wings of Immanuel Bonfils and Michael Chrysokokkes, Centau-
rus 15:120.
Thorndike, L. 1957. Notes on some Astronomical, Astrological and Mathematical
Manuscripts of the Bibliothque Nationale, Paris, Journal of the Warburg and Cour-
tauld Institutes 20:112172.
Tihon, A. 19771981. Un trait astronomique chypriote du xive sicle, Janus 64:281
308; 66:4981; 68:65127. Reprinted in Tihon 1994, Essay vii.
Tihon, A. 1994. tudes dastronomie byzantine. Aldershot.
Toomer, G.J. 1968. A survey of the Toledan Tables, Osiris 15:5174.
Toomer, G.J. 1984. Ptolemys Almagest. New York.
Van Brummelen, G. 1998. Mathematical Methods in the Tables of Planetary Motion in
Kshyr ibn Labbns Jmic Zj, Historia Mathematica 25:265280.
chapter 6

Computing Planetary Positions:


User-Friendliness and the Alfonsine Corpus*

Astronomical tables are ways to turn the treatment of complex problems into
elementary arithmetic. Since Antiquity astronomers have addressed many
problems by means of tables; among them stands out the treatment of plan-
etary motion as well as that for the motions of the Sun and the Moon. It was
customary to assign to the planets constant mean velocities to compute their
mean longitudes at any given time in the past or the future, and to add to
these mean longitudes corrections, called equations, to determine their true
longitudes. In this paper we restrict our attention to the five planets,1 with an
emphasis on their equations. Part 1 deals with what we call the standard tra-
dition, beginning with Ptolemys Handy Tables, and Part 2 deals with the new
presentations that proliferated in Latin Europe in the fourteenth and fifteenth
centuries, some of which reflect a high level of competence in mathematical
astronomy.2

1 The Standard Tradition

By the middle of the second century ad Ptolemy displayed tables for the
equations of the five planets with specific layouts and based on specific models,
algorithms, and parameters. We argue that this category of tables, as is the case
for many others, provides a clear example of user-friendliness, the driving force
that prevailed in the history of table-making.
In Almagest xi.11 Ptolemy presented tables for the planetary equations, one
for each of the five planets.3 Each table has eight columns, of which the first two

* Journal for the History of Astronomy 44 (2013), 257276, 479480.


1 Unless otherwise specified, by planets we mean the five visible planets of Antiquity, although
we are well aware that at the time the Sun and the Moon were also considered planets.
2 We do not treat tables in Islamic zijes systematically but, occasionally, we refer to some
of them. For a survey of these zijes, see D.A. King and J. Sams, with a contribution by
B.R. Goldstein, Astronomical Handbooks and Tables from the Islamic World (7501900): An
Interim Report, Suhayl, ii (2001), 9105.
3 G.J. Toomer, Ptolemys Almagest (New York and Berlin, 1984), 549553.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_008


computing planetary positions 151

are for the argument (one from 6 to 180 and the other for its complement in
360). The argument is given at intervals of 6, from 6 to 90 (and for 270 to
354), and at intervals of 3, from 90 to 180 (and for 180 to 270). According to
Toomer, Ptolemy computed the entries at 6-intervals, even where the function
is tabulated at 3-intervals.4 Columns 3 and 4 are for the equation in longitude
and the difference in equation, respectively. Column 3 assumes an eccentric
model, which Ptolemy rejected in favor of an equant model. Column 4 displays
the difference between the equation for an equant model and the equation for
an eccentric model. The sum of corresponding entries in these two columns is
the equation of center, which replaced columns 3 and 4 that appear in Almagest
xi.11 (see Table a, col. 3).5 Columns 5 and 7 give the subtractive and additive
differences to be applied to the equation of anomaly (displayed in col. 6), when
the planet is at greatest and least distance, respectively. Column 8 is for the
minutes of proportion, to seconds, used for interpolation purposes. We note
that, in the case of Venus, the entries for the equation in longitude (col. 3) are
exactly the same as those for the solar equation, although Ptolemy does not call
attention to this fact.6 We display Ptolemys model for Mars to illustrate how a
planets position can be computed directly from the model: see Figure 6.1. To
do this, one must solve plane triangles by means of trigonometric procedures
that were already available in Ptolemys time. The solution is as follows. Given
, we wish to compute the correction angle, c3, by solving triangle eco. But,
before we can do this, we have to find the length of ec, where dc, the radius of
the deferent, is 60. So first we must solve triangle edc to find ec, where angle
ced is the supplement to angle and ed is the eccentricity (a given parameter
in the model). With , ec and eo (twice the eccentricity), we can solve triangle
eco, which yields the values for c3 and co. We then have to solve triangle mco
to find c(). In this triangle two sides and an angle are known: angle mco is
equal to 180 ( c3), cm is the radius of the epicycle (a given parameter in
the model), and co has already been determined. Then

= (a) + + c3 + c(),

4 See Toomer, Almagest (ref. 3), 548, n. 54.


5 O. Neugebauer, A History of Ancient Mathematical Astronomy (Berlin, 1975), 183.
6 Although the equation of center for the Sun and the equation in longitude for Venus are
the same in the Almagest (Toomer, Almagest (ref. 3), 167, 552), their apogees differ: the solar
apogee is 65;30 and tropically fixed (Toomer, Almagest (ref. 3), 155), whereas the apogee of
Venus is 55 in Ptolemys time and sidereally fixed, and thus subject to precession (Toomer,
Almagest (ref. 3), 470; cf. Neugebauer, Ancient Astronomy (ref. 5), 58, 154, 182).
152 chapter 6

figure 6.1

Ptolemys model for Mars. The observer is at


o; the center of the deferent, ac, is at d; the
equant is at e such that ed = do; the
direction to the vernal equinox is ov; a is
the apogee, c is the center of the epicycle
such that dc = 60. The radius of the epicycle
is cm; the mean apogee of the epicycle is at
e and the true apogee of the epicycle is at
ae ; the mean longitude of Mars is in the
direction om , and its true position is at m.
The mean argument of center is and the
mean argument of anomaly is ; the
equation of center is c3, and the equation of
anomaly is c(). The true longitude of Mars,
, is angle vom. For an outer planet the
direction cm is always parallel to os, where
s is the direction to the mean Sun.

where (a), the longitude of the apogee, is a given parameter in the model.
Using the planetary equation tables takes trigonometric functions out of the
computational scheme.
In the Handy Tables Ptolemy did not modify the models or the param-
eters for the planetary equations, but he introduced a series of changes to
make the tables more suitable for calculation. Firstly, the arguments are now
given at intervals of 1, rather than at intervals of 3 or 6, as was the case
in the Almagest.7 This certainly simplifies interpolation. Secondly, he merged
columns 3 and 4 in the Almagest into a single column representing the equa-
tion of center, thus reducing the number of operations required for using these
tables. This also reduced the number of columns, from 8 to 7. Thirdly, the col-
umn for the minutes of proportion was also modified by avoiding unnecessary
precision (the entries are given to seconds in the Almagest but only to minutes
in the Handy Tables) and by changing the argument (mean argument of center
in the Almagest and true argument of center in the Handy Tables).8 This new

7 W.D. Stahlman, The Astronomical Tables of Codex Vaticanus Graecus 1291 (unpublished Ph.D.
thesis, Brown University, 1959; University Microfilms no. 625761), 295324. This dissertation
includes an edition of Ptolemys Handy Tables. A. Tihon and R. Mercier are currently editing
the Handy Tables; only two volumes have been published so far, but they do not deal with
planetary equations.
8 See Neugebauer, Ancient Astronomy (ref. 5), 183186 (Almagest) and 10021003 (Handy
Tables).
computing planetary positions 153

presentation (see Table a for Mars) set the standard for most tables dealing
with planetary equations for about 14 centuries.

table a Equations for Mars in the Handy Tables (excerpt)9

(1) (2) (3) (4) (5) (6) (7)


Equation Min. Subtractive Equation Additive
Argument of center prop. difference of anom. difference
() () () () () () ()

1 359 0;11 60 0; 2 0;24 0; 2


2 358 0;22 60 0; 3 0;48 0; 3
3 357 0;33 60 0; 4 1;12 0; 4

86 274 2
87 273 1
88 272 1
89 271 2
90 270 2;28 33;22 2;49

92 268 11;24
93 267 11;25

96 264 11;25
97 263 11;24

130 230 41; 9


131 229 41;10
132 228 41; 9

152 208 5;37


153 207 5;38

156 204 5;38


157 203 5;37
158 202 8; 2
159 201 8; 3

9 Stahlman, Tables (ref. 7), 307312.


154 chapter 6

table a Equations for Mars in the Handy Tables (excerpt) (cont.)

(1) (2) (3) (4) (5) (6) (7)


Equation Min. Subtractive Equation Additive
Argument of center prop. difference of anom. difference
() () () () () () ()

160 200 8; 2

178 182 0;27 60 0;51 3;52 1;35


179 181 0;14 60 0;26 1;57 0;48
180 180 0; 0 60 0; 0 0; 0 0; 0

In Table a, the mean argument of center, , serves as argument (columns 1 and


2) for the equation of center (col. 3), c3; the true argument of center, , can then
be computed, for

= + c3(),

where c3() 0 when 0 180. The mean argument of center serves


also as argument for the minutes of proportion (col. 4), which are necessary
to compute the true position of the planet when not found at maximum or
minimum distance of the epicycle from the observer. Now the true argument
of anomaly, , serves as argument (columns 1 and 2) for the equation of anomaly
(col. 6), and is obtained from the mean argument of anomaly:

= c3().

The tabulated equation of anomaly, c6(), was originally computed assuming


that the center of the epicycle is at mean distance. When the epicycle lies
between maximum distance (apogee) and mean distance, a subtractive differ-
ence (col. 5) must be applied. Similarly, when the epicycle is between minimum
distance (perigee) and mean distance, an additive difference (col. 7) must be
applied. The true argument of anomaly serves as argument for both subtractive
and additive differences. Then, the total equation of anomaly is

c(, ) = c6() c5() c4(),

when ranges from 270 to 90, that is, when the planet is near apogee, and
computing planetary positions 155

c(, ) = c6() + c7() c4(),

when ranges from 90 to 270, that is, when the planet is near perigee. The
combined effect of the equation of center and the total equation of anomaly is
thus c3() + c(, ), and the true position of the planet, , at a given time is:

= + + c3() + c(, ),

where , the mean longitude of that planet at a given time t since epoch, is
defined as:

= 0 + t,

0 being the planets mean longitude at epoch, and the planets mean
motion in longitude.
In the early Islamic world, the Zj al-Sindhind of al-Khwrizm (fl. 830) fol-
lowed the Indo-Iranian tradition, which was not based on Ptolemaic models
and parameters, and made no use of equants.10 This tradition was represented
by the Zj al-Shh, a work composed in Sasanian Persia and translated into Ara-
bic c. 790, where the maximum value for the equation of Venus is set equal to
that of the solar equation (2;13 or 2;14); the identity of these parameters is
also found in the Almagest.11 Accordingly, the tables for the planetary equa-
tions are quite different, both with respect to the entries and the presentation,
from those in the Almagest or the Handy Tables.
The Greek tradition was represented in the eastern Islamic world by the Zj
al-bi of al-Battn (d. 929) which is strongly Ptolemaic; indeed, the tables in
it for the planetary equations followed exactly those in the Handy Tables, but
for the equation of center of Venus.12 Both the Almagest and the Handy Tables
have 2;24 as the maximum value for Venuss equation of center, whereas it
is 1;59 in the zij of al-Battn. This change in the equation of center of Venus
was not due to new observations of Venus; rather, it was the result of a new
value found from observations for the eccentricity of the solar model that

10 O. Neugebauer, The Astronomical Tables of al-Khwrizm (Copenhagen, 1962).


11 In this tradition the apogees of Venus and the Sun are the same and both are sidereally
fixed: see B.R. Goldstein and F.W. Sawyer, Remarks on Ptolemys Equant Model in Islamic
Astronomy, in Prismata: Festschrift fr Willy Hartner, ed. by Y. Maeyama and W.G. Salzer
(Wiesbaden, 1977), 165181, p. 168.
12 C.A. Nallino, Al-Battn sive Albatenii Opus astronomicum (2 vols, Milan, 19031907), ii,
108137.
156 chapter 6

implied a maximum solar equation of 1;59,10. This new solar parameter was
simply applied to the equation of center for Venus, where it was rounded to
1;59. In this modification al-Battn followed other Islamic zijes, such as that
of abash al-sib (fl. 850).13 Toomer pointed out that modifying the entries
for the equation of Venus was inconsistent with leaving unchanged the entries
for the subtractive and additive differences (at greatest and least distances,
respectively), because they also depend on eccentricity.14 In any case, in the
zij of al-Battn only the entries for the equation of center of Venus differ from
those in the Handy Tables whereas all the rest remain unchanged.
The Toledan Tables were compiled in the second half of the eleventh century,
but the original Arabic version is not extant. In the Latin versions of the Toledan
Tables the presentation and the numerical entries agree with those in the zij
of al-Battn, but for (in most cases) an added column for the first station of
each of the planets.15 In Almagest xii.8 Ptolemy displayed the first and second
stations of the five planets in a single table, using the mean center as argument,
with entries at intervals of 6.16 In the Handy Tables, Ptolemy gave more entries,
at 3-intervals, and presented a table for the two stations for each planet. He
also introduced a change in the argument (true argument of center, instead of
mean argument of center), thus making the entries slightly different from those
in the Almagest.17 In his zij al-Battani reproduced in a separate table the entries
for the first and second stations in the Handy Tables, and only displayed them at
6-intervals. The compilers of the Toledan Tables probably realized that it was
unnecessary to give entries for both the first and the second stations (because
corresponding entries add up to 360) and just included a specific column for
the first station of each of the planets. Thus, in the tables for the planetary
equations, ultimately derived from the Handy Tables, the number of entries

13 See Goldstein and Sawyer, Equant (ref. 11), 168. abash identified both the eccentricities
and the apogees of Venus and the Sun, despite the lack of justification based on obser-
vations or based on anything said by Ptolemy in the Almagest (or elsewhere). In modern
terms, this would mean that the solar orb serves as the deferent for Venus; but this claim
was not made by any medieval scholar. Nevertheless, the medieval tradition was to keep
the apogee and eccentricity of Venus equal to those of the Sun, such that whenever the
parameters for the Sun were changed, the same changes were applied to Venus.
14 See G.J. Toomer, A survey of the Toledan Tables, Osiris, xv (1968) 5174, p. 67.
15 Toomer, Toledan Tables (ref. 14), 6068; and F.S. Pedersen, The Toledan Tables: A review
of the manuscripts and the textual versions with an edition (Copenhagen, 2002), 12651306.
16 Toomer, Almagest (ref. 3), 588.
17 See Neugebauer, Ancient Astronomy (ref. 5), 10051006; and J. Chabs and B.R. Goldstein,
A Survey of European Astronomical Tables in the Late Middle Ages (Leiden, 2012), 118.
computing planetary positions 157

increased, for they are given here at intervals of one degree, and gained one
column which was eliminated as a separate table.18 The Toledan Tables were
by far the most popular tables in Latin Europe, and the presentation in them
for tables of planetary equations can be considered standard.
The Maghrib astronomers Ibn Isq al-Tnis (c. 11931222), Ibn al-Bann
of Marrakesh (12651321), and Ibn al-Raqqm (Tunis and Granada, d. 1315) used
new parameters for the equations of center of Saturn, Jupiter, and Venus. In
contrast, the values given to the equations of anomaly agreed precisely with
those in the standard tradition, namely that of the Handy Tables, the zij of
al-Battn, and the Toledan Tables, but for the fact that in this tradition the
columns displayed are combinations of cols. 5, 6, and 7.19 We further note that
the tables for the planetary equations of Ibn Isq and his followers depart
from the standard tradition not only in the three basic parameters already
mentioned, but also in presentation. Indeed, for each of the planets there are
two tables of equations: one for quantities that depend on the argument of
center and one for those that depend on the argument of anomaly.20
The Castilian Alfonsine Tables were produced in Castile by two astronomers
working under the patronage of Alfonso x (reigned: 12521284), Judah ben
Moses ha-Cohen and Isaac ben Sid. We do not know how the tables for the
planetary equations were presented in these tables, because the tables them-
selves are not extant. However, the canons have been preserved in Castilian,
and chapter 18 (De la equaion de los v planetas) describes the way to com-
pute planetary longitudes by means of tables. Although no numerical values are
given, the description agrees perfectly with the layout of tables in the standard
tradition of the Handy Tables, the zij of al-Battn, and the Toledan Tables.21
This tradition was transmitted from the Iberian Peninsula to the rest of
Europe. The earliest astronomer to depend on this Iberian tradition was John
Vimond, who was active in Paris c. 1320. He compiled a set of tables which

18 The only known example of this kind of table where the entries are given at intervals of
half a degree is preserved in a double folio now in the General Archive of Navarre: see
J. Chabs, The Toledan Tables in Castilian: Excerpts of the planetary equations, Suhayl,
xi (2012), 179188.
19 See J. Sams and E. Mills, The computation of planetary longtiudes in the zj of Ibn
al-Bann, Arabic Sciences and Philosophy, viii (1998), 259286; reprinted in J. Sams,
Astronomy and Astrology in al-Andalus and the Maghrib (Aldershot, 2007), Essay viii.
20 A. Mestres, Materials Andalusins en el Zj dIbn Isq al-Tnis (unpublished Ph.D. thesis,
Universitat de Barcelona, 1999), 5051 and 234235.
21 J. Chabs and B.R. Goldstein, The Alfonsine Tables of Toledo (Dordrecht, 2003), 3839 and
157160.
158 chapter 6

appear to be at the interface of the astronomy rooted in al-Andalus and the


Maghrib and developed in Castile in the late thirteenth century on the one
hand, and the activity of the astronomers working in Paris in the 1320s and
the 1330s that resulted in the Parisian Alfonsine Tables on the other.22 In many
ways Vimonds tables follow a tradition unattested in Latin prior to 1320; for
example, his tables for the planetary equations are also split into two tables for
each planet, much as the Maghrib-Andalusian astronomers did.
In addition to changes in structure, which will be examined later, the main
modification in Vimonds tables is found in the entries for the equations of cen-
ter of Jupiter and Venus, with maximum values of 5;57 and 2;10, respectively.
Not much can be said about the value 5;57 other than it does not appear in any
text or table prior to Vimonds tables, and no medieval discussion of its origin
has been found. However, the value 2;10, also used by Vimond as the maximum
solar equation, appears in previous texts: implicitly in a table for the daily solar
positions for 1278 contained in the Libro del astrolabio llano composed by the
astronomers in the service of King Alfonso x of Castile,23 and explicitly in an
account in John of Murss Expositio of two observations of autumnal equinox,
one by Ptolemy in 132 and the other attributed to Alfonso in 1252, where John
explains that he has seen this observational report in what he calls the Tables
of Alfonso.24 We thus think it likely that these two new values for Jupiter and
Venus/Sun were taken from an earlier work, and the most reasonable candidate
is the Alfonsine Tables in the original Castilian version.25
The Parisian Alfonsine Tables, produced in the 1320s by a group of astrono-
mers working in Paris, were built on material coming from the Iberian Penin-
sula. They are best known today from the editio princeps that appeared in
Venice in 1483. While each part of this printed edition has a complicated his-
tory, the planetary equation tables in it are faithful to the Parisian Alfonsine
Tables as they were presented in the 1320s. The layout of the tables for the
planetary equations conforms to the standard tradition, although they have no
additional column for the first station. We will refer to the presentation and
parameters of this version of the Alfonsine Tables as standard. The entries

22 J. Chabs and B.R. Goldstein, Early Alfonsine Astronomy in Paris: The Tables of John
Vimond (1320), Suhayl, iv (2004), 207294, pp. 236256.
23 J. Chabs, Were the Alfonsine Tables of Toledo first used by their authors?, Centaurus, 45
(2003), 142150.
24 E. Poulle, Jean de Murs et les tables alphonsines, Archives dhistoire doctrinale et littraire
du moyen ge, xlvii (1980), 241271, p. 253.
25 Chabs and Goldstein, Toledo (ref. 21), 251254.
computing planetary positions 159

table b Maximum values of the equations of center and anomaly in various sets of tables
(new values are shown in boldface type)

Saturn Jupiter Mars Venus Mercury


Eq. of Eq. of Eq. of Eq. of Eq. of Eq. of Eq. of Eq. of Eq. of Eq. of
center anom. center anom. center anom. center anom. center anom.

Almagest * 6;31 6;13 5;15 11; 3 11;25 41; 9 2;24 45;59 3; 2 22; 2
Handy Tables 6;31 6;13 5;15 11; 3 11;25 41;10 2;24 45;59 3; 2 22; 2
al-Khwrizm 8;36 5;44 5; 6 10;52 11;13 40;31 2;14 47;11 4; 2 21;30
al-Battn 6;31 6;13 5;15 11; 3 11;25 41; 9 1;59 45;59 3; 2 22; 2
Toledan Tab. 6;31 6;13 5;15 11; 3 11;24 41; 9 1;59 45;59 3; 2 22; 2
Maghrib astr. 5;48 6;13 5;41 11; 3 11;25 41; 9 1;51 45;59 3; 2 22; 2
Vimond ** 6;31 6;13 5;57 11; 3 11;24 41; 9 2;10 45;59 3; 2 22; 2
Parisian Alf. 6;31 6;13 5;57 11; 3 11;24 41;10 2;10 45;59 3; 2 22; 2

* The values for the equation of center shown here are found by adding algebraically the equation
in longitude and the difference in equation in Alm. xi.11 (cols. 3 and 4).
** The values for the equation of center shown here result from subtracting the motus completus
(col. 2) from the mean argument of center (col. 1). The values for the equation of anomaly shown
here result from adding the motus completus (col. 2) to the correction for maximum distance (col. 5
in the standard tradition); see below.

are given at intervals of one degree, as was already established in the Handy
Tables.26 Moreover, out of ten basic parameters for the five planets, only two
differ from those defined by Ptolemy, namely, the equations of center of Jupiter
and Venus, and both of these parameters are already found in John Vimonds
tables. It is difficult to find other examples of such great stability in the trans-
mission of astronomical tables for more than 13 centuries. Table b provides a
summary of the main parameters for the equations of center and anomaly used
by different authors.

26 Characteristic of the Parisian Alfonsine Tables is the consistent use of sexagesimal days
and angles. Angles are given in physical signs of 60 (in contrast to zodiacal signs of 30):
an angle a,b means a 60 + b degrees (where a and b are integers such that 0 a 5 and
0 b 59; and 6,0 = 360). In our notation 10s 25 means 10 30 + 25 = 325, that is, s
signifies a zodiacal sign of 30. Sexagesimal fractions of a degree are used in the same way
with both physical signs and zodiacal signs.
160 chapter 6

2 A Proliferation of New Presentations

Prior to the first edition of the Parisian Alfonsine Tables in 1483, a variety of
original approaches for presenting tables for the planetary equations were
undertaken within the Alfonsine corpus. They coexisted with the standard
tradition, which is preserved in a number of manuscripts dating from the
fourteenth and fifteenth centuries.27 The goal of these new approaches that
depart from the standard tradition was, once again, to facilitate computa-
tion.
Let us return to about 1320, the date of John Vimonds tables, in which the two
equations for each planet are displayed in different tables. In those where all
the tabulated functions depend on the mean argument of center (see Table c),
the entries are given at 6-intervals. Vimond displayed the true argument of
center (col. 2: motus completus) and added columns for the increment of the
true argument per degree of the argument (col. 3: motus gradus), planetary
velocity (col. 4: motus diei), minutes of proportion (col. 5: diametri), and first
station (col. 6). Moreover, the equation of center incorporates a displacement
which is the difference between the apogee of each of the planets and that
of the Sun (no displacement is therefore needed in the case of Venus, for its
apogee is assumed to be the same as that for the Sun). Analysis of Vimonds
tables shows that the motion of the solar apogee was included in the motions
of the planetary apogees, thus following a theory for which there was no
previous evidence outside al-Andalus and the Maghrib.28 With this particular
arrangement Vimond intended to present a more user-friendly table than the
standard table for the equation of center.
Now, in the tables where all the tabulated functions depend on the argu-
ment of anomaly (given at 6-intervals for Saturn, Jupiter, and Mercury, and at
3-intervalsand at 2-intervals in the vicinity of 180for the other two plan-
ets), Vimond also added columns for planetary velocities and other corrections,
such as col. 5 (see Table d), which results from adding the correction for max-
imum distance to the correction for minimum distance (cols. 5 and 7, respec-
tively, in the Handy Tables, the zij of al-Battn, and the Toledan Tables). As
was the case for the equation of center, the entries for the equation of anomaly
are not explicitly displayed. Rather, we are given entries for the motus comple-
tus (col. 2), which is the difference between the equation of anomaly and the

27 For example, for a list of manuscripts extant in Spain or of Spanish origin, see Chabs and
Goldstein, Toledo (ref. 21), 292303.
28 Sams and Mills, Ibn al-Bann (ref. 19), 268270.
computing planetary positions 161

table c John Vimonds equation of center and first station of Mars


(excerpt)

(1) (2) (3) (4) (5) (6)


Motus Motus Motus First
Argument completus gradus diei Diametri station
s () s () min min min s ()

0 6 0 12;31 50;50 26 6 5 8;41


0 12 0 17;36 50; 0 26 4 5 8;21

1 12 1 12;22 49; 0 26 0 5 7;29


1 18 1 17;16 49;10 25 0 5 7;31

4 18 4 6;36 60;30 31 32 5 13;46

7 12 7 11;33 73;30 38 60 5 19;14


7 18 7 18;54 73;10 38 59 5 19;13

10 12 10 23;24 58;50 30 31 5 13;36

11 24 12 2;13 51;50 27 10 5 9;31


12 0 12 7;24 51;10 26 8 5 9; 6

correction for maximum distance. When we compute the differences between


cols. 6 and 5 in the standard tradition, we find agreement with Vimonds motus
completus, indicating that he kept all the basic parameters for the equation
of anomaly in this tradition. It is noteworthy that, as indicated by North, this
implies that Ptolemys eccentricities underlie these tables even though, in
the case of Venus and Jupiter, the eccentricities were modified for computing
the equation of center.29 The only text of which we are aware that treats
the equation of anomaly in this way is of Maghrib origin: the Minhj of Ibn
al-Bann, dependent on the zij of Ibn Isq. In the Minhj the tables for the
equations of anomaly of Saturn and Jupiter give entries for al-mufrad (c6 c5
in the standard terminology for columns) and al-bucd (c5 + c7).30 These are

29 J.D. North, Richard of Wallingford (3 vols, Oxford, 1976), iii, 196.


30 Sams and Mills, Ibn al-Bann (ref. 19), 278285.
162 chapter 6

table d John Vimonds equation of anomaly for Mars (excerpt)

(1) (2) (3) (4) (5) (6) (7)


Motus Motus Motus Motus Motus
Argument compl. gradus diei Diametri grad. diei
s () s () () min min min sec sec

0 3 11 27 1; 8 23 11 0; 8 3 1
0 6 11 24 2;16 23 11 0;17 3 1

4 3 7 27 36;40 1 1 8;53 9 4
4 6 7 24 36;44 0 0 9;19 9 4
4 9 7 21 36;43 3 1 9;46 9 4

5 6 6 24 28;15 46 21 13;30 0 0
5 9 6 21 25;56 53 25 13;37 6 2
5 12 6 18 23;17 62 29 13;19 13 6

5 28 6 2 3; 1 90 42 2;29 74 35
6 0 6 0 0; 0 90 42 0; 0 74 35

precisely two of the columns found in Vimonds tables (cols. 2 and 5). This
particular choice of columns was intended to facilitate the computation of the
planetary equations of anomaly.
In addition to the Expositio, already mentioned, John of Murs, a key fig-
ure in the Parisian milieu for the transmission of Alfonsine astronomy, was
responsible for a set of tables, called the Tables of 1321, devoted exclusively
to the planets and the two luminaries. With these tables the computation of
true planetary positions is entirely different from that described in any other
text of which we are aware.31 The most significant feature of the Tables of
1321 is a new organizational principle, which does not require the equations
of the planets to be displayed explicitly. To be sure, the mean motions of
the planets are here presented in tables for the mean conjunctions of each
planet with the Sun (tabula principalis), and the corrections to be applied for
times between consecutive conjunctions are given in double argument tables

31 J. Chabs and B.R. Goldstein, John of Murss Tables of 1321, Journal for the History of
Astronomy, xl (2009), 297320.
computing planetary positions 163

(contratabula). This particular approach meant that astronomers could avoid


the typically cumbersome computations for determining true planetary posi-
tions, compared with using tables previously available in Latin. One unusual
feature of these double argument tables is that the horizontal argument is the
mean argument of center of the planet (at intervals of 12) and the vertical
argument is the age of the planet, that is, the time after a mean conjunc-
tion with the Sun, expressed as a number of days. It is also noteworthy that
for each planet, besides the tabula and contratabula, we are given a table for
its equation of center and first station. The values of the maximum equation
of center agree in all cases with those used by Vimond, and so do the rest of
the entries (given at 6-intervals in both sets of tables, but presented differ-
ently).
The tables of Vimond and those of John of Murs for 1321 certainly made
the computation of the true positions of the planets much easier, but their
approaches do not seem to have been very popular. The main improvement
in that direction came from double argument tables, which greatly simplified
computations and only required linear interpolation.32 John of Lignres (also
active in Paris) was probably the first astronomer in Latin Europe to draw
up a double argument table combining the equations of center and anomaly

32 There were a few double argument tables in Islamic zijes prior to 1320, e.g., Ibn Ynus
(c. 990) had such a table for the lunar equations as did al-Baghdd (c. 1285): see D.A. King,
A Double-Argument Table for the Lunar Equation Attributed to Ibn Ynus, Centaurus,
xviii (1974), 129146 and C. Jensen, The Lunar Theories of al-Baghdd, Archive for History
of Exact Sciences, viii (1972), 321328. Double argument tables for planetary latitudes
attributed to Ibn al-Bayr, who is otherwise unknown, are preserved in Hyderabad, Andra
Pradesh State Library, ms 298, Tables 6677, where the horizontal headings are the true
arguments of center and the vertical headings are the true arguments of anomaly (both at
intervals of 6): see A. Mestres, Maghrib Astronomy in the 13th Century: a Description of
Manuscript Hyderabad Andra Pradesh State Library 298, in From Baghdad to Barcelona:
Studies in the Islamic Exact Sciences in Honour of Prof. Juan Vernet, ed. by J. Casulleras and
J. Sams (Barcelona, 1996), 383443, p. 428. These tables are also cited in B. van Dalen,
Tables of Planetary Latitude in the Huihui li (ii), in Current Perspectives in the History
of Science in East Asia, ed. by Y.S. Kim and F. Bray (Seoul, 1999), 316329, p. 327. This
manuscript contains the zij of Ibn Isq (early thirteenth century), and Ibn al-Bayr is
mentioned in chapter 18 of the canons to this zij. This implies that Ibn al-Bayr was active
no later than the time of Ibn Isq. A summary of chapter 18 appears in Mestres 1996,
pp. 396397. We are most grateful to van Dalen for sharing with us his translation of the
Arabic text of chapter 18, the Arabic text of chapter 18 and of Tables 6677, as well as his
notes on this material.
164 chapter 6

in a single table for each planet in his Tabule magne (c. 1325).33 The vertical
argument is the mean argument of anomaly (at 6-intervals in all planets, and
also at 3-intervals from 150 to 180 in the case of Mars and Venus),34 and the
horizontal argument is the mean argument of center (at 6-intervals). In Table
e we reproduce an excerpt from the table for the combined equation of Venus
in John of Lignress Tabule magne, as found in Lisbon, ms Ajuda 52-xii-35,
ff. 83r87v, with the title Tabula equationum ultimarum veneris. We note the
use of physical signs of 60, and the inclusion of columns for the differences,
to minutes, of successive entries for a fixed value of the argument of anomaly
(not displayed here).35
For each planet there is a total of at least 1860 entries (2160 in the case of Mars
and Venus) presented as 31 60 or 36 60 matrices, not taking into account the
columns and rows that display the successive differences. None of the entries
explicitly corresponds to the maximum values of the equations of center or
anomaly, which could lead to the identification of the tradition to which it

33 E. Poulle, John of Lignres, Dictionary of Scientific Biography (16 vols, New York, 1970
1980), vii (1973), 122128, pp. 123124; J.D. North The Alfonsine Tables in England, in Pris-
mata: Festschrift fr Willy Hartner, ed. by Y. Maeyama and W.G. Salzer (Wiesbaden, 1977),
269301, pp. 273274, 278; reprinted in J.D. North, The universal frame: historical essays in
astronomy, natural philosophy, and scientific method (London, 1989), 327359.
34 The reason is that in the range 150210 the entries for these two planets vary quite
rapidly, and thus the accuracy of interpolation is increased by doubling the number of
entries. Vimond was already aware of this rapid variation and, in his tables for the equation
of anomaly for Mars and Venus given at 3-intervals, he displayed entries in the range
168192 at 2-intervals.
35 Only three manuscripts containing theses tables are known. The other two use zodiacal
signs of 30 (Erfurt, Biblioteca Amploniana, ms ca 2 388, and London, British Library, ms
add. 24070). All three have a column at the far right for the mean argument of anomaly
from 180 to 360, but the numbers in this column, and only in this one, are inverted in
different ways in two manuscripts; for example, 234 is written as 54 3 (meaning 3,54) in
ms Lisbon; 24 7s in ms Erfurt; and 7s 24 in ms London. It is possible that this reflects
an archetype in Arabic or Hebrew, but see C. Burnett, Why we read Arabic numerals
backwards, in Ancient and Medieval Traditions in the Exact Sciences, ed. by P. Suppes et
al. (Stanford, 2000), reprinted in C. Burnett, Numerals and arithmetic in the Middle Ages
(Aldershot, 2010), Essay vii, for some examples of this inversion in medieval Latin texts.
Moreover, the three manuscripts differ in another respect: ms Lisbon has a column for
successive differences, ms Erfurt has no such column, and ms London has columns and
rows for successive differences. This is definitely a good example of the intervention of
copyists when transmitting the very same table, without altering its presentation or any
of its essential features.
computing planetary positions 165

table e John of Lignress double argument table for the combined equation of Venus
(excerpt)


0,6 0,12 3,54 4,0 4,6 5,54 0,0
() () () () () () () ()

m* m a a a a a
0, 0 0; 8 0;16 1; 0 ** 1; 4 1; 8 0; 8 0; 0
a a
0, 6 2;22 2;14 3;35 3;39 3;42 2;38 2;30
0,12 4;50 4;43 6; 6 6;10 6;13 5; 7 4;58

2, 6 44; 2 43;51 47;23 47;24 47;22 44;26 44;14


2,12 44;32 44;20 48;10 48;10 48; 8 44;57 44;44
2,18 44;29 40;18 48;33 48;33 48;30 44;59 44;43
2,24 43;41 41;44 48;13 48;16 48;14 44;13 43;57

2,54 13; 9 12;26 22;21 16;21 12;21 14;32 13;52


2,57 6;23 5;38 27;14 19;14 15;14 7;44 7; 7
m m
3, 0 0;44 1;28 31; 6 24; 7 19; 7 0;39 0; 0

* m stands for minue (to be subtracted) and a for adde (to be added).
** ms Erfurt: 1;1.

belongs (see Table b), but a few entries are easy to track. Let us consider the
case when = 0 or 180. Then c3() = 0 and = , and the entries for = 90
reduce to c6(90) c5(90) and c6(90) + c7(90), respectively, in the usual termi-
nology for columns. We find agreement in all cases, except for the equation
for Mercury at greatest distance (the entry reads 21;32, whereas computation
with the tables in the standard tradition give 22;2).36 In all other cases there
is good agreement, but it is not always perfect because columns 5, 6, and 7,
which depend exclusively on the argument of anomaly sometimes vary in the

36 In the case of Mars, for instance, c5(90) = 2;28, c6(90) = 33;22, and c7(90) = 2;49 in the
standard tradition beginning in the Handy Tables (see Table a). Thus, c6(90) c5(90) =
30;54 and c6(90) + c7(90) = 36;11. The entry in John of Lignress table of Mars for = 0
and = 90 is 30;54 and that for = 180 and = 90 is 36;11.
166 chapter 6

minutes. To show that the entries in John of Lignress table are specifically
based on the values used by John Vimond and, in particular, on those maximum
values for the equation of center found for the first time in Vimonds tables,
we have recomputed a few critical entries.37 The maximum entries in John of
Lignress tables could not have been computed with values as low as those
in the tradition represented by the Toledan Tables, and we conclude that they
were calculated with Vimonds tables, or that both astronomers had a common
source.
It should be noted that in his tables of 1322 John of Lignres had used the
parameters 1;59 (Venus) and 5;15 (Jupiter) that are found in the Toledan
Tables for the maximum equations of center, replacing them with 2;10 (Venus)
and 5;57 (Jupiter) in his double argument tables for the planetary equations
in 1325. As a matter of fact, John of Lignress tables for planetary equations for
1322, as presented in Bibliothque nationale de France, ms 7286c (ff. 33r47v),38
share the same entries and layout, including a column for first station, with
the Toledan Tables. This change was much the same as John of Murs had done
a few years previously, given that in his earliest astronomical work of 1317,
beginning Auctores calendarii , he had praised the Tables of Toulouse and
seemed unaware of Alfonsine material.39

37 For Venus, the maximum entry in John of Lignress table of 1325 is 48;33 (at = 3,54 and
4,0; and = 2,18) as displayed in Table e. If = 4,0 = 240, then c3(240) = 1;55, using
an equation of center with a maximum of 2;10 (Vimonds value), and c4(240) = 31. Thus,
= 138 1;55 = 136;5. Therefore, c5(136;5) = 1;11, c6(136;5) = 45;59, and c7(136;5) = 1;16.
Finally, the combined equation is 1;55 + 45;59 + (1;16 31/60) = 48;32, in agreement with
the entry. However, when performing the same calculation using an equation of center
with a maximum of 1;59 (as in the Toledan Tables), one finds a combined equation of
48;19. As we shall see in the computation that follows, the results are also unambiguous
in the case of Jupiter. The maximum entry in John of Lignress table is 17;1 (at = 4,24
and = 1,48). If = 4,24 = 264, then c3(264) = 5;57, which is the maximum equation
of center in Vimonds tables, and c4(264) = 7. Thus, = 108 5;57 = 102;3. Therefore,
c5(102;3) = 0;29, c6(102;3) = 11;3, and c7(102;3) = 0;32. Finally, the combined equation is
5;57 + 11;3 + (0;32 7/60) = 17;4, very close to the entry in John of Lignress Tabule magne.
However, if we use a table with a maximum equation of center of 5;15 (as in the Toledan
Tables), one finds a combined equation of 16;21.
38 For a description of this manuscript, see M.-M. Saby, Les canons de Jean de Lignres sur
les tables astronomiques de 1321, (Unpublished thesis: cole Nationale des Chartes, Paris,
1987), 516520. A summary appeared as Les canons de Jean de Lignres sur les table
astronomiques de 1321, cole Nationale des Chartes: Positions des thses (1987), 183190.
39 Chabs and Goldstein, Toledo (ref. 21), 278. The Tables of Toulouse are an adaptation of
the Toledan Tables for the Christian calendar (instead of the Muslim calendar): see Poulle,
computing planetary positions 167

Double argument tables undoubtedly facilitated computation, because they


displayed in a compact and clever way intermediate calculations needed to
obtain a final numerical result.40 This kind of presentation was not an inven-
tion of the Parisian astronomers, for it is already found in Arabic sources, e.g.,
it was used by Ibn al-Kammd (Crdoba, c. 1100) in his tables for the time from
mean to true syzygy as a function of the difference between the hourly veloc-
ities of the Moon and the Sun and the elongation.41 Double argument tables
proliferated in fourteenth-century Europe and were not restricted to the plan-
etary equations: they were also used to display true planetary positions (the
Tabule anglicane, also called the Oxford Tables of 1348, associated with William
Batecombe); planetary conjunctions (John of Murss Tables of 1321); planetary
latitudes (John of Murss Tables of 1321, and the Oxford Tables); syzygies (John
of Murs and Firmin of Beauval in their Tabulae permanentes, Immanuel ben
Jacob Bonfils of Tarascon, Levi ben Gerson, Juan Gil of Burgos, Joseph Ibn
Waqr of Seville, and the Tables of Barcelona); lunar motion (Levi ben Gerson);
and lunar and planetary velocities (Judah ben Asher ii of Burgos).42

Un tmoin de l astronomie latine du xiiie sicle: les Tables de Toulouse, in Comprendre et


matriser la nature au moyen ge: Mlanges dhistoire des sciences offerts Guy Beaujouan
(Geneva and Paris, 1994), 5581.
40 For interpolation in double argument tables, see M. Husson, Ways to read a table: reading
and interpolation techniques of early fourteenth-century double-argument tables, Jour-
nal for the History of Astronomy, xliii (2012), 299319.
41 These tables are only extant in Latin and Hebrew versions; the Latin version was com-
posed by John of Dumpno in 1260 in Palermo and survives uniquely in Madrid, Biblioteca
Nacional de Espaa, ms 10023. See J. Chabs and B.R. Goldstein, Andalusian Astronomy:
al-Zj al-Muqtabis of Ibn al-Kammd, Archive for History of Exact Sciences, xlviii (1994),
141; and B.R. Goldstein, Solomon Franco on the Zero Point for Trepidation, Suhayl, x
(2011), 7783.
42 For the Tabule anglicane see J.D. North, England (ref. 33); for the Tables of 1321, see
J. Chabs and B.R. Goldstein, John of Murs (ref. 31); for the Tabulae permanentes, see
B. Porres and J. Chabs, John of Murss Tabulae permanentes for finding true syzygies,
Journal for the History of Astronomy, xxxii (2001), 6372; for Immanuel ben Jacob Bonfils
of Tarascon, see P. Solon, The Hexapterygon of Michael Chrysokokkes (Brown University,
Ph.D. thesis (unpublished), 1968; Proquest, umi, aat 6910019), and P. Solon, The Six Wings
of Immanuel Bonfils and Michael Chrysokokkes, Centaurus, xv (1970), 120; for Levi ben
Gerson, see B.R. Goldstein, The Astronomical Tables of Levi ben Gerson (Hamden, ct, 1974);
for Juan Gil of Burgos and Joseph Ibn Waqr of Seville, see J. Chabs and B.R. Goldstein,
Computational Astronomy: Five Centuries of Finding True Syzygy, Journal for the History
of Astronomy, xxviii (1997), 93105, pp. 9496; for the Tables of Barcelona, see J.M. Mills,
Las Tablas Astronmicas del Rey Don Pedro el Ceremonioso (Madrid and Barcelona, 1962),
and J. Chabs, Astronomia andalus en Catalua: Las Tablas de Barcelona, in From
168 chapter 6

Another set of tables in the Alfonsine corpus that adheres strictly to its
parameters and models is the set we call the Tables for the Seven Planets for
1340; they are a most ingenious reworking of the Parisian Alfonsine Tables and
include several displaced tables. The purpose of displaced tables is to elimi-
nate all subtractions in the derivation of planetary positions, thus facilitating
computations.43 This anonymous set of tables, most likely of French origin, is
uniquely preserved in Paris, Bibliothque nationale de France, ms 10262 (ff. 2r
46v). The two planetary equations are given in separate tables for each planet,
and are not explicitly displayed. Rather, for the equation of center we are given
entries which are displaced both vertically and horizontally with respect to
those in the standard Parisian Alfonsine Tables, whereas the entries for the
equation of anomaly are only displaced vertically. In modern algebraic terms,
the vertical and horizontal displacements of a function underlying a displaced
table are such that y = f(x + kh) + kv, where y = f(x) is the original function to
which the displaced table is compared, kh is the displacement on the x-axis,
and kv is the displacement on the y-axis. Tables f and g display excerpts of
the equation of center of Jupiter in the Tables for the Seven Planets and in the
Parisian Alfonsine Tables, respectively (where kh = 18 and kv = 6).
Figure 6.2 illustrates the situation for Jupiter. The graph labeled ms 10262
displays the entries in the Tables for the Seven Planets for 1340, and that labeled
pat corresponds to those in the Parisian Alfonsine Tables.
In general the vertical displacements are intended to avoid complicated
rules for addition and subtraction corresponding to the simple rules we now
give by means of algebraic signs. The horizontal displacements are intended to
counterbalance other displacements, such as those applied to the minutes of
proportion. It is easy to recognize that the vertical displacements of the entries
for the equation of anomaly agree with the parameters found in the Parisian
Alfonsine Tables: see Table b. The Tables for the Seven Planets use a total of 40
different displacements for the planets (including the Sun and the Moon): see
Table h.

Baghdad to Barcelona: Studies in the Islamic Exact Sciences in Honour of Prof. Juan Vernet,
ed. by J. Casulleras and J. Sams (Barcelona, 1996), 477525; for Judah ben Asher ii
of Burgos, see B.R. Goldstein, Abraham Zacut and the Medieval Hebrew Astronomical
Tradition, Journal for the History of Astronomy, xxix (1998), 177186, pp. 179181.
43 For details, see J. Chabs and B.R. Goldstein, Displaced Tables in Latin: The Tables for the
Seven Planets for 1340, Archive for History of Exact Sciences, lxvii (2013), 142. Note that the
term displaced applied to tables was coined by E.S. Kennedy in 1977 as a translation of
the Arabic wa (see his The Astronomical Tables of Ibn al-Aclam, Journal for the History
of Arabic Science, i (1977), 1323, espec. p. 16).
computing planetary positions 169

table f Equation of center of table g Equation of center of Jupiter


Jupiter in the Tables for the in the Parisian Alfonsine
Seven Planets44 Tables45

Equatio Minutes Equation Minutes


centri of proportion of center of proportion
() () () () () ()

0 4;15 0 1 0; 6 60
1 4; 9 1 2 0;12 60
2 4; 3 1 3 0;12 60

71 0; 4 53 88 5;56 1
72 0; 3 54 89 5;56 1
90 5;57 2
77 0; 3 59
78 0; 3 2 96 5;57 7
79 0; 4 3 97 5;56 8

168 6;39 60 180 0; 0 60

245 11;56 14 263 5;56 8
246 11;57 13 264 5;57 7

252 11;57 7 270 5;57 2
253 11;56 6 271 5;56 1
272 5;56 1
259 11;52 1 273 5;55 2
260 11;51 58
358 0;12 60
359 4;20 0 359 0; 6 60

44 Note that the equatio centri is always positive; it reaches a minimum of 0;3 at 7278,
and a maximum of 11;57 at 246252. For the minutes of proportion there are two
discontinuities (from 60 to 0 between 77 and 78, and from 0 to 60 between 259 and
260), to keep them positive in all cases.
45 Note that the equation of center is negative between 0 and 180, and positive between
180 and 360; it reaches a minimum of 5;57 at 9096, and a maximum of 5;57 at
264270. The minutes of proportion are positive between 89 and 271, and negative for
the rest of the arguments.
170 chapter 6

figure 6.2 The equation of center of Jupiter. The graph labeled ms 10262 is displaced
vertically by 6 and horizontally by 18 with respect to the graph labeled
pat. The maximum of the upper graph is 11;57 and takes place at
arguments 246252; the maximum of the lower graph is 5;57 and takes
place at arguments 264270.

In any case, computation with this compact and consistent set of tables gives
the same results as those obtained with the Parisian Alfonsine Tables, while
avoiding subtractions at any stage in the computation.
In the fifteenth century the Paduan astronomer, Prosdocimo de Beldomandi
(d. 1428), compiled a new set of tables that belong to the Alfonsine corpus.46 His
tables for the planetary equations follow the Parisian Alfonsine Tables, includ-
ing the 2;10 and 5;57 used by Vimond for Venus and Jupiter, in agreement with
those that were printed in 1483 in the editio princeps.
Giovanni Bianchini (d. after 1469) spent most of his life in Ferrara where he
served as administrator for the estate of the prominent dEste family. About
1442 he compiled an extensive set of astronomical tables which depend on the
Alfonsine Tables, but have a completely different presentation.47 Bianchinis
tables offer a whole new approach for computing the true positions of the

46 J. Chabs, From Toledo to Venice: The Alfonsine Tables of Prosdocimo de Beldomandi,


Journal for the History of Astronomy, xxxviii (2007), 269281.
47 J. Chabs and B.R. Goldstein, The Astronomical Tables of Giovanni Bianchini (Leiden, 2009).
computing planetary positions 171

table h Displacements of the planetary equations in the Tables


for the Seven Planets for 1340

Eq. of center Eq. of anomaly


Vert. displac. Horiz. displac. Vert. displac.

Saturn 7 14 6;13
Jupiter 6 18 11; 3
Mars 12 61 41;10
Venus 3 51 45;59
Mercury 4 28 22; 2

planets. Although tables for the planetary equations are not explicitly given, the
true positions of the planets are computed by means of double argument tables
where the vertical argument is the mean anomaly, represented here by the time
within an anomalistic period for each planet, and the horizontal argument is
the mean center. These tables were first published in 1495 in Venice under the
title Tabulae astronomiae, and again in 1526 and 1553.48
The Tabulae resolutae were compiled in central Europe, and circulated
widely in manuscripts during the fifteenth century and in print during the
sixteenth century.49 One of their characteristics is that the mean motions are
arranged according a system of cyclical radices at intervals of 20 year. The Tabu-
lae resolutae are also strictly based in the Parisian Alfonsine Tables; in fact, they
are a particular form of presenting them. The tables for the planetary equations
display the same parameters as the Parisian Alfonsine Tables but, unlike them,
add a column for first station, thus following the layout of most versions of the
Toledan Tables.
In Vienna John of Gmunden (c. 13801442) collected a great variety of tables
within the framework of the Parisian Alfonsine Tables. He displayed them in

48 For a list of manuscripts that contain Bianchinis tables, see Chabs and Goldstein, Bian-
chini (ref. 47), 14. The owners of manuscript copies of these tables include Johannes
Regiomontanus (Nuremberg, Stadtbibliothek, Cent v 57) and Johannes Virdung (Vatican,
Biblioteca Apostolica, ms Pal. lat. 1375).
49 J. Dobrzycki, The Tabulae Resolutae, in De Astronomia Alphonsi Regis, ed. by M. Comes,
R. Puig, and J. Sams (Barcelona, 1987), 7177; J. Chabs, Astronomy in Salamanca in
the Mid-Fifteenth Century: The Tabulae Resolutae, Journal for the History of Astronomy,
xxix (1998),167175; and J. Chabs, The Diffusion of the Alfonsine Tables: The Case of the
Tabulae resolutae, Perspectives on Science, x (2002), 168178.
172 chapter 6

various sets, called First Version, Tabulae maiores, and Tabulae breviores.50 He
presented his tables for the equations of the planets in three different ways:
at 1-intervals following the standard tradition; at 3-intervals in an abridged
form of the latter; and as double argument tables, reproducing those by John of
Lignres. Therefore, with respect to the planetary equations, John of Gmunden
was not an innovator; rather, he offered table users several possibilities that
were already known in Latin Europe.
In the early sixteenth century Johannes Angelus, a follower of Peurbach and
Regiomontanus, claimed that these two authors had compiled a new table of
planetary equations giving better results than the standard Alfonsine Tables,
but this new table has not been found in any manuscript or printed edition.51
As already mentioned, the Alfonsine Tables were first printed in 1483 by
Erhard Ratdolt in Venice. A few years later (1492) and in the same town, a sec-
ond edition appeared, edited by Johannes Lucilius Santritter. The entries for
the planetary equations are the same in both sets of tables but, in the second
edition, the planets were inexplicably presented in the order Venus, Mercury,
Mars, Jupiter, and Saturn (rather than in the usual order where Mercury pre-
cedes Venus). Also in the 1492 edition, the second column for the argument,
displaying the complement in 360, was eliminated; this left enough space on
the page to include five extra columns for the differences between successive
entries in the remaining columns.
In 1503 Petrus Liechtenstein printed another set of tables in Venice, the Tab-
ule Astronomice Elisabeth Regine. It was much less popular than the standard
version of the Parisian Alfonsine Tables, but it is historically significant because
in his Commentariolus Copernicus cited its author, Alfonso de Crdoba, who
was in the service of Pope Alexander vi in Rome.52 The tables for the planetary

50 B. Porres, Les tables astronomiques de Jean de Gmunden: dition et tude comparative


(unpublished Ph.D. thesis, cole pratique des hautes tudes, Section iv, Paris, 2003).
51 J. Dobrzycki and R.L. Kremer, Peurbach and Margha Astronomy? The Ephemerides of
Johannes Angelus and Their Implications, Journal for the History of Astronomy, xxvii
(1996), 187237, pp. 187188.
52 Alfonso de Crdoba, known as Hispalensis, came from Seville (Latin: Hispalis) and his
origin is well attested in various printed texts of the early sixteenth century. His place of
origin is explicitly given as patria hispalensis and he is cited as Alfonso hispalensi de Cor-
duba. Hence there is no reason to emend the text to hispaniensis, that is, from Spain, as
has been suggested: see N.M. Swerdlow, The Derivation and First Draft of the Copernicuss
Planetary Theory: A Translation of the Commentariolus with Commentary, Proceedings
of the American Philosophical Society, cxvii (1973), 423512, pp. 451452. Copernicuss Com-
mentariolus is undated, but it is usually taken to be from about 1514. Alfonso dedicated his
computing planetary positions 173

table i Alfonso de Crdobas


equation of center of
Mars (excerpt)

Mars
Longitude () min

Leo 15 Leo 15 0; 0 60
Leo 20 Leo 10 0;55 60
Leo 25 Leo 5 1;49 59

Sco 15 Tau 15 11;23 3
Sco 20 Tau 10 11;24 3
Sco 25 Tau 5 11;21 8

Aqr 5 Aqr 25 2;13 58
Aqr 10 Aqr 20 1; 7 59
Aqr 15 Aqr 15 0; 0 60

equations, as well as all the others in this set, depend on the Parisian Alfonsine
Tables, both for models and parameters. However, this is not true for the
presentation. First, for each planet the equation of center is given in a different
table from that for the equation of anomaly. As we have seen, John Vimond
had also used this two-fold presentation, which was most uncommon in Latin
astronomy,53 but the two sets differ in several important aspects (see Table i).
Second, the argument in the tables for the equation of center is given at
5-intervals (as is the case in the tables for the equation of anomaly), in contrast
to the tables in the standard tradition (1-intervals). But most important of all
is the fact that the argument in the table for the equation of center represents

work to Queen Isabella of Castile and Aragon (14511504), whose name in Latin was Elis-
abeth. On this set of tables, see J. Chabs, Astronomy for the Court in the Early Sixteenth
Century: Alfonso de Crdoba and his Tabule Astronomice Elisabeth Regine, Archive for
History of Exact Sciences, lviii (2004), 183217.
53 This separation was intended to distinguish clearly between the columns that depend
on one variable from those that depend on the other. We know of another example of
this two-fold presentation in Erfurt, Biblioteca Amploniana, ms q 362 (ff. 28r36r), also in
the Alfonsine corpus, whose layout seems unrelated to those by either John Vimond or
Alfonso de Crdoba.
174 chapter 6

the mean longitude of the planet, , that is, the mean argument of center plus
the longitude of the planets apogee. Thus, the argument is shifted by a quantity
that, in each case, corresponds to the longitude of the apogee (Leo 15 in the
case of Mars). Again, the purpose is to facilitate calculation. In turn, the tables
for the equation of anomaly display the usual columns of the tables in the
standard tradition (cols. 1, 2, 5, 6, and 7).54
The tabular innovations developed to facilitate computation of the true
longitude of the planets paved the way to a substantial increase in the number
of almanacs in the fourteenth and fifteenth centuries. Although there are some
earlier examples of this genre, the various new presentations of the tables for
planetary equations (such as double argument tables) made almanacs much
easier to compile. In turn, since almanacs and ephemerides display directly
the true positions of the planets at successive times, the user did not have the
difficult task of computing planetary equations; hence, they were very popular,
for they could be used even by those who had not mastered all the subtleties of
astronomy.55
Perhaps the most elaborate and influential almanac in the late Middle Ages
was the Almanach perpetuum.56 Its tables, together with a short explanatory
text, were first printed in two editions (one in Latin and the other in Castilian)
in Leiria, Portugal, in 1496. The tables were derived from a set of astronomical
tables in Hebrew called ha-ibbur ha-gadol (The Great Composition) compiled
by Abraham Zacut of Salamanca (14521515).57 As regards the positions of the
planets, Zacuts work was compiled in the framework of the Parisian Alfonsine
Tables with 1473 as epoch. For each planet it gives the true longitude, the true
argument of center, and the true argument of anomaly for several days in each

54 The fact that the argument in the table for the equation of center was chosen to be the
mean longitude of the planet makes the table less useful in the long term, for it does not
take into account the slow motion of the apogee due to precession.
55 An almanac lists successive true positions of each planet at intervals of one day or a few
days, that is, each planet is listed separately; hence, the information for a given day is
scattered among various tables. An ephemeris displays the true positions of all planets
in a single row at intervals of one day or a few days for a certain number of years. That is,
the difference between an almanac and an ephemeris is only one of presentation.
56 J. Chabs and B.R. Goldstein, Astronomy in the Iberian Peninsula: Abraham Zacut and the
Transition from Manuscript to Print (Transactions of the American Philosophical Society,
90.2, Philadelphia, 2000).
57 On the various editions of the Almanach perpetuum and its impact on the Jewish com-
munity as well as on Christian and Muslim scholars, see Chabs and Goldstein, Zacut (ref.
56), 161171.
computing planetary positions 175

month (sometimes daily) for periods as long as 125 years in the case of the
longitude of Mercury. For these three quantities there is a total of more than
42,100 entries and, in each case, the sign, the degrees, and the minutes are
specified. We have certainly come a long way from the tables for the planetary
equations in the Handy Tables!

Conclusion

As regards planetary equations, the standard tradition, going back to Ptolemys


Handy Tables, survived at least until the first printed editions of the Parisian
Alfonsine Tables. Ptolemys underlying models and most of the parameters
involved were rarely challenged from about the middle of the second century
to the end of the fifteenth century. Only two parameters appearing in the
tables were changed in that period, and John Vimond seems to have been the
first astronomer to have used them in Latin Europe.58 Vimond depended on
material from the Iberian Peninsula, most likely of Arabic origin.
Astronomers in Latin Europe in the fourteenth and fifteenth centuries were
actively engaged with this well-defined tradition, but they did not simply repro-
duce the tables and texts of their predecessors, and many of them devel-
oped innovative approaches to facilitate computational tasks, such as double
argument tables, displaced tables, separated tables, or shifted variables. User-
friendliness, rather than improvement of the models or enhancement of preci-
sion, was the driving force for most of the efforts developed by table-makers
in the computation of planetary positions. Nevertheless, the editio princeps
of the Parisian Alfonsine Tables did not incorporate any of the various new
presentations. These innovations in presentation have only been recognized
in recent years and, taken together, they indicate that astronomers in Latin
Europe reached a high level of mathematical competence in the late Middle
Ages.

58 To be sure, the apogee of Venus was also changed, but this does not modify the table for
the equations for this planet.
part 3
Sets of Tables


chapter 7

Andalusian Astronomy:
al-Zj al-Muqtabis of Ibn al-Kammd*1

Introduction

In 1956 E.S. Kennedy published his A Survey of Islamic Astronomical Tables in


which he described, briefly in most cases, over 100 sets of tables, called zijes
(after the Arabic: zj), from the 8th to the 15th centuries and from all parts of
the medieval Islamic world. At that time only two of them had been published,
and it was clear that our understanding of scientific activity in the Middle Ages
would be greatly enhanced by detailed treatment of the others. Indeed, such
has proven to be the case, as we learn from the many studies that have followed
this pioneering essay.
Astronomers in Islamic Spain, al-Andalus, composed zijes and, beginning
in the 12th century, they were adapted and translated into Hebrew, Latin,
Castilian, and Catalan, the most famous examples being the Toledan Tables
(see Toomer (1968)) and the Alfonsine Tables. In Spain, as elsewhere in the
Islamic world, these zijes were largely based on the work of predecessors going
back to Ptolemy on the one hand, and Hindu astronomers on the other. More
often than not, a table comes with instructions for using it, rather than the
method used to construct it. For this reason, much scholarly energy has been
devoted to describing the methods underlying these tables, as well as their
lines of descent. By such analysis, guided by textual material, one can now
distinguish tables that are based on entirely new models, tables that are merely
copies of tables in previous zijes (at the two extremes), from tables based on
previous models but with new parameters and tables composed by modified
or new mathematical methods.
The work of Ibn al-Kammd, an Andalusian astronomer of the 12th century,
illustrates most of these characteristics. He composed 3 zijes, none of which
survives in the original Arabic, but a Latin manuscript contains a translation
of what appears to be one complete zij with references to the others. Ibn

* Archive for History of Exact Sciences 48 (1994), 141, communicated by J. North.


1 The authors thank Professor J. Sams (Barcelona) for his valuable comments.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_009


180 chapter 7

al-Kammd depended on sources that ultimately go back to Ptolemy and to


Hindu astronomers as they were known in zijes prior to his time, and his
influence was felt by later astronomers writing in Arabic, Latin, and Hebrew. As
is often the case, this work preserves parts of otherwise lost texts: in particular,
we gain valuable information on the solar theory of the Andalusian astronomer
Azarquiel (or Ibn al-Zarqlluh), who lived in the 11th century (see Toomer
(1969)). In this article our primary intention is to describe the astronomical
work of Ibn al-Kammd as it is preserved in Latin in ms Madrid 10023.
Ibn al-Kammd is cited or criticized by a number of his successors: Ibn
al-Him writing in Arabic ca. 1205 (see Sams (1992), pp. 321 ff.); Ab l-asan
cAl al-Marrkush (ca. 1262) (see Mills (1950), p. 347); Juan Gil (ca. 1350) whose
astronomical tables are preserved in a Hebrew version (see Goldstein (1985),
p. 237); Ibn al-adib, a 14th-century astronomer from Spain who went to Sicily
where he wrote astronomical works in Hebrew (see Goldstein (1985), p. 239);
Joseph Ibn Waqr, a Spanish astronomer of the mid-14th century who wrote in
Arabic and Hebrew (see Goldstein (1985), p. 237); and most importantly, as we
shall see, Ibn al-Kammd had a profound influence on the Tables of Barcelona
(see Mills (1962)).
In the discussions that follow new information of particular interest for the
history of astronomy in Spain is presented: see, e.g., the solar equation table
(Section ii); the preservation of material that goes back to the zij al-Mumtaan
of Yay ibn Ab Manr who lived in the 9th century (Section iii, c, f and j);
the table for trepidation (Section iv, b); and the tables for planetary latitude
(Section iv, e).
In sum, Ibn al-Kammd was a major player in medieval Spanish astronomy;
his achievements and the extent of his legacy have not yet been sufficiently
appreciated.

i Ibn al-Kammd: Life and Works

This is the name by which the astronomer Ab Jacfar Amad ben Ysuf Ibn
al-Kammd is known, although a variety of names has been associated with
him (Vernet (1949), pp. 7273). He was probably from Sevilla, but active in
Crdoba in the 12th century. In a remarkable study, Mills (1942), pp. 231247,
called attention to a 14th-century Latin manuscript at the Biblioteca Nacional
de Madrid, ms 10023, containing one of Ibn al-Kammds works; he described it
and gathered all the information then available on this astronomer. Not much
more is known now.
Ibn al-Kammd is the author of 3 zijes:
andalusian astronomy 181

1. al-Kawr al al-dawr (the periodic rotations [?]) (in 60 chapters). This trea-
tise has been partially preserved in Arabic in ms Escorial Ar. 939,4 (not seen
by us). There is also a short text in Castilian in Segovia, Biblioteca de la Cat-
edral, ms 115, ff. 218vb220vb, attributed to Yuaf Benacomed, and entitled
Libro sobre ircunferencia de moto.
2. al-Amad al al-abad (the eternally valid [tables])
3. al-Muqtabis (the compilation [of the two previous works]). ms Madrid 10023
contains the Latin translation of al-Muqtabis. In the explicit it is clearly
stated that the translation was done by John of Dumpno in 1260 in Palermo.
In Arabic only chapter 28 has survived: see ms Alger 1454,2, ff. 6263.

The dates for Ibn al-Kammd are uncertain. Mills (1950), p. 346, considered
the period towards the end of the 12th century, and suggested the year 1195
as that of his death, probably following Ahlwardt (1893), p. 219, where this date
is given with no specification of his source. More recently, Ibn al-Kammd has
been taken to be a direct disciple of Azarquiel. This claim is based on a note in
the margin of f. 30r of ms lat. 7281, a 15th-century manuscript at the Bibliothque
nationale de France, and it has been argued that this claim is supported by
the date ah480 (10871088ad) that appears in ms Madrid 10023, f. 65v: see
Section v, m, below. The marginal note, already transcribed in Mills (1950),
p. 14, reads: Post uenit Alcamet discipulus , referring to Azarquiel. According
to Mills, the same hand, or a similar one, has added: Similiter discipulus
Messalle. It does not seem at all warranted to deduce from this expression
that Ibn al-Kammd was a direct pupil of Azarquiel. Instead, we understand
this to mean only that Ibn al-Kammd was a follower of Azarquiels methods.
On the other hand, the date in ms Madrid 10023, f. 65v, is not the only one
mentioned in the last section of this manuscript (e.g., on f. 66r there is a table
for ah550: see Section v, r, below); as we shall see, the last section of this ms
contains a variety of tabular material not directly related to al-Muqtabis. Sams
(1992), p. 322, noted that Ibn al Him (fl. 1205) criticized Ibn al-Kammd. Since
the available evidence suggests that Ibn al-Kammd lived after Azarquiel and
before Ibn al-Him, we conclude that Ibn al-Kammd lived in the 12th century,
without offering any greater precision.

Texts in ms Madrid 10023


a al-Muqtabis
Text in Latin, presented in two columns.

Introduction (f. 1rava). Transcribed in Mills (1942), pp. 231232.


Index (ff. 1va2rb). Transcribed in Mills (1942), pp. 234235.
182 chapter 7

Canons, 1 to 30, each of them called porta (ff. 2rb18vb). Three chapters
have been published so far: canon 1 is transcribed in Mills (1942), pp. 235
236, canon 28 in Vernet (1949), pp. 7478, and canon 30 in Mills (1942),
pp. 237238. Note that canon 30 explicitly mentions the other two works of
Ibn al-Kammd: al-Kawr al al-dawr and al-Amad al al-abad.
Explicit (f. 18vb). Transcribed in Mills (1942), p. 238.

b Other Texts
From f. 18vb to f. 24rb there is a set of chapters, in some disorder, that are dis-
tinct from those of al-Muqtabis. Some of them are associated with al-Kawr
al al-dawr, and were also translated by John of Dumpno in 1262 in Palermo;
their incipits and explicits were published in Mills (1942), pp. 238242. Mil-
ls also transcribed some of the texts therein, and Toomer (1969), pp. 323324,
transcribed and translated a text concerning the variation of solar eccentric-
ity.

Tables in ms Madrid 10023


Two sets of tables can easily be distinguished in the manuscript:

a al-Zj al-Muqtabis (ff. 27r54v)


The tables are mentioned, or their use is explained, in the text in 30 canons
called al-Muqtabis. The tables are calculated for the meridian of Crdoba.
Folio 54v contains the last table of al-Muqtabis (a geographical table), as we
learn from canon 10 (f. 6va): tabulam longitudinum terrarum positam in
ultimo huius canonis. We will discuss all the tables in al-Muqtabis as follows:
the solar equation in Section ii, eclipse theory in Section iii, and the remaining
tables in Section iv.

b Other Tables (ff. 55r66r)


These tables do not form a homogeneous set. They are not mentioned in the 30
canons of al-Muqtabis, and are presumably distinct from that work. Some of the
tables are related to al-Kawr al al-dawr, some are attributed to astronomers
other than Ibn al-Kammd, and still others are calculated for places other than
Crdoba. These tables will be discussed in Section v.

ii The Solar Equation in Al-Muqtabis

Tabula directionis centri solis et augis eius ad primordium annorum


seductionis (f. 35r)
andalusian astronomy 183

Above the heading, the longitude of the solar apogee, presumably for the
Hijra, is given: Aux 2s 16;45,21. This table displays the solar equation in
degrees, minutes, and seconds as a function of mean solar anomaly. The max-
imum solar equation, which occurs at 92, is 1;52,44, thus differing from the
more common values: 1;59 (abash al-sib, Yay ibn Ab Manr), 1;59,10
(Toledan Tables, al-Battn), 2;14 (al-Khwrizm), 2;23 (Ptolemy).
The use of this table is explained in canon 13 (f. 7vb).
In his Tractatus super totam astrologiam, Bernardus de Virduno (ca. 1300)
attributes an eccentricity of 1;58 to Azarquiel, which yields a maximum equa-
tion of 1;52,42 (see Toomer (1987), pp. 515517 and Sams (1992), p. 213). How-
ever, this is not the only parameter used by Azarquiel for, in the Alfonsine trans-
lation of his treatise on the construction of the equatorium, 1;52,30 is explicitly
called a rounded parameter for the maximum solar equation (see Sams (1987),
p. 468). It is therefore likely that Ibn al-Kammd accepted a parameter from
Azarquiel.
The columns in Table 1, The Solar Equation in al-Muqtabis, are arranged as
follows:

(1) Mean solar anomaly, , for each integer degree from 1 to 180.
(2) Entries in the text (f. 35r): solar equation, c(), in degrees, minutes, and
seconds.
(3) Line-by-line differences in (2): c( + 1) c(). This column allows us to
recognize quite a number of errors in the entries for the solar equation
in the text, for these should increase monotonically from anomaly 0
to reach a maximum at anomaly 92, and then decrease monotonically
to anomaly 180. Consequently, the set of line-by-line differences should
follow a smooth decreasing pattern. In a great number of cases, pairs of
successive line-by-line differences in (3), whether erroneous or not, are
identical, thus strongly suggesting that the table for the solar equation
was originally computed for every other degree, and that interpolation
was used to derive the rest of the entries.
(4) Reconstructed line-by-line differences. Not all the line-by-line differences
which do not fit smoothly in (3) have been reconstructed; we have only
made suggestions in those cases where the tabulated values for the solar
anomaly can be easily derived from the reconstructed values. A question
mark indicates that, although there is a peculiar value in (3), we do not
offer any alternative value.
(5) Reconstructed values for the solar equation. These values result from an
easy explanation of the way the copying, or the calculational error was
made. There follows a list of various kinds of such errors:
184 chapter 7

misreading of a digit: c(22) = 0;40,19 instead of 0;40,59 (19 and 59 are


easily confused in Arabic); c(28) = 0;50,27 instead of 0;51,27;
inversion of the order of the entries: in the text, the seconds for c(175)
and c(176) are apparently inverted;
incorrect interpolation between computed entries: c(30) and c(32)
were computed correctly, but c(31) seems to result from an incorrect
interpolation between them;
displacement of columns; from c(145) to c(150) the entries for the
minutes have been shifted one line downwards.
We have left unchanged those erroneous entries of the solar equation
for which we do not have an easy explanation; some of them may be due
to incorrect computation by the author of the table rather than to copyist
errors.
(6) Recomputed values: for the recomputation of the entries, we used a
simple eccentric model, and the following expression:

[1] tan(c) = e sin()/(60 + e cos())

where the eccentricity used is e = 1;58,2 = 60 sin (1;52,44).


(7) Differences in seconds: T(ext)C(omp.), i.e., col. (2) col. (6); in those
cases where we have confidence in our reconstructed values, this column
displays the differences between the reconstructed values in col. (5), and
computation in col. (6).

table 1 The Solar equation in al-Muqtabis

(1) (2) (3) (4) (5) (6) (7)

1 0; 1,53 0; 1,53 0; 1,54 1


2 3,47 1,54 3,49 2
3 5,40 1,53 5,43 3
4 7,34 1,54 7,37 3
5 9,25 1,51 1,56 9,30 9,31 1
6 11,25 2, 0 1,55 11,24 +1
7 13,18 1,53 13,18 0
8 15,11 1,53 15,12 1
9 17, 4 1,53 17, 5 1
10 0;18,57 0; 1,53 0;18,58 1

11 20,50 1,53 20,50 0


andalusian astronomy 185

(1) (2) (3) (4) (5) (6) (7)

12 22,42 1,52 22,42 0


13 24,39 1,57 ? 24,34 +5
14 26, 9 1,30 ? 26,26 15
15 28,13 2, 4 ? 28,16 3
16 30,17 2, 4 ? 30, 7 30, 7 0
17 32, 1 1,44 ? 31,57 4
18 33,45 1,44 ? 33,46 1
19 35,35 1,50 35,35 0
20 0;37,25 0; 1,50 0;37,24 +1

21 39,12 1,47 39,11 +1


22 40,19 1, 7 1,47 40,59 40,58 +1
23 42,45 2,26 1,46 42,45 0
24 44,30 1,45 44,30 0
25 46, 1,46 46,16 46,15 +1
26 48,16 1,44 48, 0 47,59 1
27 49,44 1,28 1,44 49,43 +1
28 50,27 0,44 1,44 51,27 51,25 +2
29 53, 7 2,40 1,40 53, 7 0
30 0;54,48 0; 1,41 0;54,48 0

31 56,58 2,10 1,40 56,28 56,27 +1


32 58, 7 1, 9 1,39 58, 7 0
33 0;59,45 1,38 0;59,44 +1
34 1; 1,23 1,38 1; 1,21 +2
35 2,38 1,15 1,35 2,58 2,57 +1
36 4,33 1,55 1,35 4,32 +1
37 6, 7 1,34 6, 6 +1
38 7,37 1,30 7,38 1
39 9, 9 1,32 9,10 1
40 1;10,41 0; 1,32 1;10,40 +1

41 12,10 1,29 12, 9 +1


42 13,39 1,29 13,37 +2
43 15, 7 1,28 15, 4 +3
44 16,35 1,28 16,29 +6
45 17,56 1,21 17,53 +3
186 chapter 7

table 1 The Solar equation in al-Muqtabis (cont.)

(1) (2) (3) (4) (5) (6) (7)

46 19,17 1,21 19,15 +2


47 20,33 1,16 1,21 20,38 20,37 +1
48 21,58 1,25 1,20 21,57 +1
49 23,45 1,47 1,17 23,15 23,15 0
50 1;24,32 0; 0,47 0; 1,17 1;24,33 1

51 25,47 1,15 25,48 1


52 27, 2 1,15 27, 3 1
53 28,19 1,17 1,13 28,15 28,15 0
54 29,27 1, 6 1,12 29,26 +1
55 30,37 1,10 30,36 +1
56 31,46 1, 9 31,44 +2
57 32,50 1, 4 32.51 1
58 33,54 1, 4 33,56 2
59 34, 7 0,13 1, 3 34,57 34,59 2
60 1;36, 0 0;1,53 0; 1, 3 1;36, 1 1

61 36,58 0,58 ? 37, 1 3


62 37,59 1, 1 ? 37,59 0
63 38,53 0,54 38,56 3
64 39,49 0,56 39,51 2
65 40,48 0,59 40,44 +4
66 41,37 0,49 41,35 +2
67 42,25 0,48 42,25 0
68 43,13 0,48 43,12 +1
69 43,59 0,47 43,58 +1
70 1;44,43 0; 0,44 1;44,43 0

71 45,25 0,42 45,25 0


72 46, 6 0,41 46, 5 +1
73 46,44 0,38 46,44 0
74 47,21 0,37 47,21 0
75 47,57 0,36 47,55 +2
76 48,33 0,36 48,28 +5
77 49, 1 0,28 48,59 +2
78 49,29 0,28 49,28 +1
andalusian astronomy 187

(1) (2) (3) (4) (5) (6) (7)

79 49,54 0,25 49,55 1


80 1;50,19 0; 0,25 1;50,20 1

81 50,51 0,32 0,22 50,41 50,43 2


82 51, 3 0,12 0,22 51, 4 1
83 51,22 0,19 51,23 1
84 51,41 0,19 51,40 +1
85 51,58 0,17 51,56 +2
86 52,15 0,17 0,12 52,10 52, 9 +1
87 52,14 0, 1 0, 9 52,19 52,20 1
88 52,19 0, 5 0,10 52,29 52,29 0
89 52,29 0,10 0, 6 52,35 52,35 0
90 1;52,40 0; 0,11 0; 0, 5 1;52,40 0

91 52,42 0, 2 52,43 1
92 52,44 0, 2 52,44 0
93 52,42 0, 2 52,43 1
94 52,39 0, 3 52,39 0
95 52,33 0, 6 52,34 1
96 52,27 0, 6 52,26 +1
97 52,16 0,11 52,17 1
98 52, 6 0,10 52, 6 0
99 51,50 0,16 51,52 2
100 1;51,34 0; 0,16 1;51,36 2

101 51,22 0,12 ? 51,18 +4


102 50,59 0,23 ? 50,58 +1
103 50,41 0,18 50,36 +5
104 50,22 0,19 50,12 50,12 0
105 49,48 0,34 0,26 49,46 +2
106 49,18 0,30 49,18 0
107 48,50 0,28 ? 48,48 +2
108 48,16 0,34 48,16 0
109 47,31 0,45 0,35 47,41 47,41 0
110 1;47, 5 0; 0,26 0,36 1;47, 5 0

111 46,55 0,10 0,40 46,25 46,27 2


188 chapter 7

table 1 The Solar equation in al-Muqtabis (cont.)

(1) (2) (3) (4) (5) (6) (7)

112 46,45 0,10 0,40 45,45 45,41 +4


113 45,33 1,12 0,42 45, 3 45, 4 1
114 44,20 1,13 0,43 44,20 0
115 43,12 1, 8 0,48 43,32 43,33 1
116 42,22 0,50 42,42 42,45 3
117 42,22 0, 0 0,50 41,52 41,55 3
118 41, 1 1,21 0,51 3 41, 3 2
119 40,56 0, 5 0,55 40, 6 40, 9 3
120 1;39,52 0; 1, 4 0; 0,54 1;39,12 1;39,13 1

121 38,52 1, 0 38,12 38,15 3


122 37,53 0,59 37,13 37,15 2
123 36,51 1, 2 36,11 36,13 2
124 35,51 1, 0 35,11 35,10 +1
125 33,25 1,26 1, 6 34, 5 34, 5 0
126 32,58 0,27 1, 7 32,57 +1
127 31,50 1, 8 31,48 +2
128 30,39 1,11 30,38 +1
129 29,25 1,14 29,25 0
130 1;28,11 0; 1,14 1;28,11 0

131 27,25 0,46 1,15 26,55 26,55 0


132 26,40 0,45 1,15 25,40 25,37 +3
133 24,50 1,50 1,20 24,20 24,20 0
134 23, 0 1,50 1,20 22,57 +3
135 21,54 1, 6 1,26 21,34 21,35 1
136 20,11 1,43 1,23 20,10 +1
137 18,44 1,27 18,45 1
138 17,18 1,26 17,17 +1
139 15,49 1,29 15,49 0
140 1;14,20 1,29 1;14,18 +2

141 12,47 1,33 12,47 0


142 11,13 1,34 11,13 0
143 9,39 1,34 9,39 0
144 8, 6 1,33 8, 3 +3
andalusian astronomy 189

(1) (2) (3) (4) (5) (6) (7)

145 7,26 0,40 1,40 6,26 6,26 0


146 6,49 0,37 1,37 4,49 4,47 +2
147 5, 8 1,41 3, 8 3, 7 +1
148 3,27 1,41 1; 1,27 1; 1,26 +1
149 1; 1,10 2,17 1,37 0;59,50 0;59,43 +7
150 0;59,19 1,51 0;57,59 0;58, 0 1

151 57,21 1,58 1,38 56,21 56,15 +6


152 54,31 2,50 1,50 54,29 +2
153 52,28 2, 3 ? 52,42 14
154 50,37 1,51 ? 50,54 17
155 48,47 1,50 ? 49, 6 19
156 47,17 1,30 ? 47,15 +2
157 44,55 2,22 1,52 45,25 45,24 +1
158 43,32 1,23 1,53 43,33 +1
159 41,39 1,53 41,40 1
160 0;39,46 1,53 0;39,46 0

161 37,51 1,55 37,52 1


162 35,57 1,54 35,57 0
163 34, 2 1,55 34, 1 +1
164 32, 6 1,56 32, 5 +1
165 30,14 1,52 ? 30, 8 +6
166 28,19 1,55 28,10 +9
167 26,16 2, 3 ? 26,11 +5
168 24,13 2, 3 ? 24,13 0
169 22,14 1,59 22,13 +1
170 0;20,15 1,59 0;20,14 +1

171 18,14 2, 1 18,13 +1


172 16,43 1,31 2, 1 16,13 16,13 0
173 14,12 2,31 2, 1 14,12 0
174 12,11 2, 1 12,11 0
175 10, 8 2, 3 2, 1 10,10 10,10 0
176 8,10 1,58 2, 2 8, 8 8, 8 0
177 6, 6 2, 4 2, 2 6, 6 0
178 4, 4 2, 2 4, 4 0
190 chapter 7

table 1 The Solar equation in al-Muqtabis (cont.)

(1) (2) (3) (4) (5) (6) (7)

179 2, 2 2, 2 2, 2 0
180 0; 0, 0 2, 2 0; 0, 0 0

iii Eclipse Theory in Al-Muqtabis

All the tables related to eclipse theory are mentioned in canons 25 (ff. 13rb14rb)
and 27 (ff. 15ra16va). Some specific terminology in Latin is used in canon 25:
longitudo for elongation, preuentio for opposition, peruenencia solis for
the fraction of the elongation that belongs to the Sun, precessio for the
hourly relative velocity of the luminaries.

a Solar and Lunar Velocities

Tabula diuersi motus solis in una hora quod est respectus (f. 51v)

Tabula diuersi motus lune in una hora quod est respectus (f. 51v)

The use of these two tables is explained in canon 25 (f. 13va). Both tables
coincide, except for copying errors, with those in al-Khwrizm (Suter (1914),
pp. 175180, tables 6166). The extremal values are:

vs(1) = 0;2,22 /h, and vs(180) = 0;2,24 /h (read: 0;2,34 /h).

vm(1) = 0;30,12 /h, and vm(180) = 0;35,40 /h.

The same two tables are also found in the tables of Juan Gil (London, Jews
College, ms Heb. 135, f. 91r), and the Tables of Barcelona (Mills (1962), table
43), as well as in a manuscript containing the Tables of Toulouse (Paris, BnF,
ms Lat. 16658, ff. 90v93r). The Toledan Tables (cf. Toomer (1968), p. 82) and
the Almanac of Azarquiel (Mills (1950), p. 174) have tables for solar and lunar
velocities that agree with those in al-Battn but differ from the present ones.
andalusian astronomy 191

For the solar velocity table, the formula used in recomputing the entries is:

[2] v = v + v .

where v = 0;59,8/24 and = c( + 1) c(); c is the solar equation as a function


of the mean anomaly, , taken from al-Khwrizms tables 2126, col. 2. The
recomputed values (which are not displayed here) are in good agreement with
the entries in the table. Equation [2] has been used by analogy with equation
[3], used for computing the lunar velocity table (cf. Goldstein (1992); Goldstein
et al.):

[3] v() = 0;32,56 + 0;32,40 ,

where = c( + 1) c(); c is the lunar equation as a function of the anomaly,


, taken from al-Khwrizms tables 2126, col. 3. The results are shown in
Table 2, together with a comparison between the entries in the tables of Ibn
al-Kammd and al-Khwrizm. The original table seems to be better preserved
by al-Khwrizm than by Ibn al-Kammd. lt should be noticed that previously
this table had not been recomputed successfully (cf. as-Saleh (1970), p. 162, and
Neugebauer (1962), p. 106).

The columns in table 2, Lunar Velocity, are arranged as follows:

(1) Lunar anomaly, , for each integer degree from 1 to 180.


(2) Entries in the text (f. 51v): lunar velocity in minutes and seconds of are per
hour.
(3) Variant readings in al-Khwrizms lunar velocity table (Suter (1914), tables
6166, col. 3); only the seconds are displayed here, except for 72.
(4) Recomputed values at multiples of 5 of anomaly; only the seconds are
displayed here, except for 135.

table 2 Lunar velocity

(1) (2) (3) (4) (1) (2) (3) (4)

1 30;12 12 91 32;58
2 30;12 92 33; 0
3 30;12 93 33; 2
4 30;13 94 33; 5
5 30;13 13 95 33; 7 15
192 chapter 7

table 2 Lunar velocity (cont.)

(1) (2) (3) (4) (1) (2) (3) (4)

6 30;13 96 33;10
7 30;14 13 97 33;13
8 30;14 98 33;16
9 30;14 99 33;20
10 30;15 14 14 100 33;23 26

11 30;15 101 33;26


12 30;15 102 33;29
13 30;16 103 33;32
14 30;17 104 33;36
15 30;18 17 17 105 33;39 44
16 30,19 18 106 33;42
17 30;20 19 107 33;45
18 30;21 20 108 33;48
19 30;22 21 109 33;51
20 30;23 22 22 110 33;54 54

21 30;24 23 111 33;57


22 30;25 24 112 34; 0
23 30;26 25 113 34; 3
24 30;27 114 34; 5 6
25 30;28 26 115 34; 9 12
26 30;29 116 34;12
27 30;30 117 34;15
28 30;32 118 34;17
29 30;33 119 34;20
30 30;34 34 120 34;22 24

31 30;36 121 34;26 25


32 30;37 122 34;27
33 30;38 123 34;30
34 30;40 124 34;32
35 30;41 39 125 34;36 34 36
36 30;43 126 34;37
37 30;44 127 34;40
38 30;46 128 34;45 42
andalusian astronomy 193

(1) (2) (3) (4) (1) (2) (3) (4)

39 30;47 129 34;45


40 30;49 47 130 34;48 49

41 30;51 131 34;50


42 30;52 132 34;53
43 30;54 133 34;55
44 30;56 134 34;58
45 30;58 56 135 35; 0 34,57
46 31; 0 136 35; 1
47 31; 2 137 35; 3
48 31; 4 138 35; 4
49 31; 6 139 35; 5
50 31; 8 7 140 35; 6 7

51 31;10 141 35; 8


52 31;12 142 35; 9
53 31;14 143 35;10
54 31;16 144 35;12
55 31;21 18 17 145 35;13 17
56 31;23 21 146 35;14
57 31;25 23 147 35;16
58 31;27 25 148 35;17
59 31;29 27 149 35;18
60 31;29 29 150 35;20 22

61 31;32 151 35;21


62 31;35 152 35;22
63 31;37 153 35;24
64 31;40 154 35;25
65 31;42 42 155 35;26 49
66 31;45 156 35;27
67 31;47 157 35;28
68 31;49 158 35;29
69 31;51 52 159 35;30
70 31;54 55 160 35;31 31

71 31;57 161 35;32


194 chapter 7

table 2 Lunar velocity (cont.)

(1) (2) (3) (4) (1) (2) (3) (4)

72 31; 1 32; 1 162 35;33


73 32; 5 163 35;33
74 32; 8 164 35;34
75 32;12 12 165 35;35 36
76 32;17 16 166 35;35
77 32;23 20 167 35;36
78 32;27 24 168 35;37
79 32;27 169 35;37
80 32;31 32 170 35;38 38

81 32;34 171 35;38


82 32;37 172 35;38
83 32;39 173 35;39
84 32;42 174 35;39
85 32;46 45 55 175 35;39 39
86 32;47 176 35;39
87 32;49 177 35;39
88 32;52 178 35;40
89 32;54 179 35;40
90 32;56 57 180 35;40 40

b Time from Mean to True Syzygy

Tabula horarum longitudinis ad dirigendum tempus coniunctionis et


preuentionis (f. 52r)

The use of this table is explained in canon 25 (f. 13va). It is a double argument
table: the vertical argument is the elongation (e), given in degrees and minutes,
from 0;30 to 12;0 at intervals of 0;30. The horizontal argument is the velocity
of the Moon relative to that of the Sun (vm vs), in minutes and seconds of arc
per hour, from 0;27,30 /h to 0;33,30 /h, at intervals of 0;0,30/h.
Each entry (t) can be computed from the following equation:

[4] t = e/(vm vs),


andalusian astronomy 195

where t is the time, given in hours and minutes, that the Moon takes to travel
the longitudinal arc between the Sun and the Moon at mean syzygy, i.e., the
time interval from mean to true syzygy. For a discussion of tables for finding
the time from mean to true syzygy, see Chabs and Goldstein.
The entries in the column for 0;32,0 /h are also copied, erroneously, in the
column for 0;32,30/h. A similar, but not quite identical, table is found in the
Tables of Barcelona (Mills (1962), table 42).

c Solar Eclipses

Tabula rectitudinum ad eclipses solares (f. 52v)

This table in 7 columns (see table 3) is mentioned in canon 27 (f. 15rb), and
gives the declination of midheaven, i.e., the intersection of the ecliptic and the
local meridian, in degrees and minutes, as a function of the longitude of the
ascendant. The manuscript uses septentrionalis for north, and meridionalis
for south, which we have transcribed by assigning positive or negative signs,
respectively, to the tabulated entries.
The declination reaches its maximum of 23;50 at Libra 1 and its minimum
of 23;51 at Aries 0. Note that 23;51 is the value used by Ptolemy in the
Handy Tables for the obliquity of the ecliptic. This table is closely related to
a table found in an Arabic manuscript (ms Escorial Ar. 927, ff. 9v, 12r, 12v: see
Kennedy (1986)), where it is called the Table of samt for determining solar
eclipses ( jadwal al-samt li-cilm kusf al-shams). Kennedy and Faris (pp. 2124)
have describedand presented a graphof this table, based on the copy
in the Escorial where it appears among tables attributed to Yay ibn Ab
Manr (9th century). The entries in this table in ms Madrid 10023, f. 52v, are
displayed in table 3 and, for comparison, those in the Arabic manuscript in the
Escorial are displayed in a separate table (see table 4), where the underlining
of entries indicates that they differ from those in table 3. Both tables are clearly
variants of the same archetype despite many discrepancies that can be ascribed
to copyist errors. The Arabic copy displays 180 additional entries because it
fails to recognize the symmetries (see Kennedy and Faris, p. 24). In addition
to the value adopted for the obliquity of the ecliptic, the table depends on
the latitude of the place for which it is intended. In this case, the latitude
used seems to be 35;55,48 (cf. ms Escorial Ar. 927, f. 8v), even though the
use of seconds for geographical latitude was not meaningful at the time; this
latitude corresponds to Yays native abaristan (cf. Kennedy and Faris, p. 24).
However, it is quite likely that Ibn al-Kammd, who is associated with Crdoba
(latitude = 38;30 as it appears in the geographical table on f. 54v), did not know
196 chapter 7

table 3 Table of samt for determining solar eclipses (ms Madrid 10023, f. 52v)

Degree of the
ascendant Lib Sco Sgr Cap Aqr Psc

1 29 +23;50 +19;25 +6;47 8;35 18;34 22;47


2 28 23;50 19; 5 6;23 8;32 18;50 22;49
3 27 23;48 18;50 6; 0 9;27 19; 5 22;52
4 26 23;47 18;35 5;53 9;50 19;19 22;57
5 25 23;45 18;33 5;50 10;12 19;20 22;59
6 24 23;42 18; 2 4; 1 10;32 19;35 23; 0
7 23 23;39 17;46 3;37 10;53 19;50 23; 5
8 22 23;36 17;29 3;33 11;18 20; 3 23;12
9 21 23;32 17;12 2;50 11;59 20;16 23;17
10 20 23;23 16;55 2; 1 12; 1 20;33 23;18
11 19 23;16 16;36 1;36 12;22 20;35 23;23
12 18 23;12 16;18 0;48 12;48 20;39 23;24
13 17 23; 5 15;41 +0;24 13; 4 20;47 23;26
14 16 22;59 15; 3 0; 0 13;23 20;52 23;27
15 15 22;52 14;24 0;24 13;43 20;55 23;32
16 14 22;44 14; 5 0;48 14; 5 20;58 23;36
17 13 22;37 13;24 1;36 14;24 21; 0 23;37
18 12 22;23 13; 4 1; 1 14;44 21; 4 23;39
19 11 22;18 12;22 2;20 15; 3 21; 7 23;39
20 10 22;11 12;39 3;53 15;23 21;18 23;42
21 9 21;51 11;39 3;37 15;41 21;29 23;42
22 8 21;40 11;33 4; 1 16; 0 21;40 23;45
23 7 21;38 10;36 4;25 16;18 21;41 23;45
24 6 21;18 10;12 5;13 16;36 21;51 23;47
25 5 21; 7 9;52 5;36 16;44 21; 1 23;48
26 4 20;55 9;44 6; 0 16;56 22;10 23;48
27 3 20;42 8;42 6;24 17;12 22;10 23;50
28 2 20;29 8;19 7;10 17;27 22;19 23;50
29 1 20;16 7;23 7;18 17;46 22;28 23;51
30 0 +20; 3 +7;10 7;16 18; 2 22;35 23;51

Vir Leo Cnc Gem Tau Ari


andalusian astronomy 197

table 4 Table of samt for determining solar eclipses (ms Escorial Ar. 927, ff. 9v, 12r, and 12v)*

Degree of the
ascendant Lib Sco Sgr Cap Aqr Psc

1 29 +23;50 +19;25 +6;47 8;39 18;34 22;39


2 28 19; 5 6;23 8;42 18;50 22;45
3 27 23;48 18;50 6; 0 9;27 19; 5 22;51
4 26 23;46 18;35 5;53 9;50 19;15 22;56
5 25 23;45 18;33 5; 7 10;12 19;25 23; 0
6 24 23;42 18; 2 5; 1 10;32 19;35 23; 4
7 23 23;39 17;56 4;37 10;56 19;50 23; 8
8 22 23;36 17;29 4;13 11;18 20; 3 23;11
9 21 23;32 17;12 3;50 11;39 20;18 23;14
10 20 23;23 16;55 3; 1 12; 1 20;33 23;17
11 19 23;16 16;36 2;36 12;22 20;35 23;20
12 18 23;12 16;18 1;48 12;48 20;39 23;23
13 17 23; 5 15;41 +0;24 13; 5 20;42 23;27
14 16 22;59 15; 3 0; 0 13;23 20;47 23;28
15 15 22;52 14;24 0;24 13;43 20;55 23;30
16 14 22;44 14; 5 0;48 14; 5 20;58 23;32
17 13 22;37 13;24 1;36 14;24 21; 0 23;34
18 12 22;27 13; 4 2; 1 14;44 21; 5 23;36
19 11 22;18 12;22 2;25 15; 3 21; 7 23;38
20 10 22;11 12;30 3;53 15;23 21;18 23;40
21 9 21;51 11;39 3;36 15;41 21;29 23;41
22 8 21;40 11;33 4; 1 16; 0 21;40 23;42
23 7 21;30 10;36 4;24 16;18 21;41 23;43
24 6 21;18 10;12 5;13 16;36 21;51 23;44
25 5 21; 7 9;50 5;36 16;44 21; 1 23;45
26 4 20;55 9; 5 6; 0 16;55 22;10 23;46
27 3 20;42 8;42 6;24 17;12 22;14 23;47
28 2 20;29 8;19 7;10 17;26 22;19 23;48
29 1 20;16 7;13 7;13 17;46 22;28 23;49
30 0 +20; 3 +7;10 7;16 18;10 22;32 23;50

Vir Leo Cnc Gem Tau Ari


198 chapter 7

* The columns for Libra, Scorpio, and Sagittarius are transcribed from f. 9v; the columns
for Capricorn and Aquarius from f. 12v; and the column for Pisces from f. 12r, for only
one entry appears on f. 12v while the rest of the column on that page is blank. In the
following list we display all entries on f. 12r, that differ from those of f. 12v. Variant
readings from f. 12r:

Cap 2 = 9; 2 Cap 6 = 10;34 Cap 8 = 11;17 Cap 12 = 12;43


Cap 14 = 13;25 Cap 15 = 13;45 Cap 18 = 14;43 Aqu 2 = 18;47
Aqu 3 = 19; 1 Aqu 4 = 19;15 Aqu 5 = 19;26 Aqu 6 = 19;38
Aqu 8 = 20; 4 Aqu 9 = 20;18 Aqu 11 = 20;36 Aqu 13 = 20;42
Aqu 14 = 20;47 Aqu 15 = 20;52 Aqu 17 = 21; 5 Aqu 18 = 21;12
Aqu 19 = 21;19 Aqu 20 = 21;26 Aqu 21 = 21;33 Aqu 23 = 21;47
Aqu 24 = 21;54 Aqu 26 = 22; 7 Aqu 27 = 22;18 Aqu 29 = 22;25

the geographical latitude for which the table was computed. It is worth noting
that the table for solar declination (f. 35v) is based on a different value for the
obliquity of the ecliptic: 23;33.
Clearly, in the Arabic copy f. 12v has the better readings; the copyist appar-
ently realized that f. 12r had many errors, and so he tried again on f. 12v. The
table of the samt discussed here is also found in the Tables of Barcelona (Mills
(1962), table 47), and suffers from the same errors as the table of al-Muqtabis in
ms Madrid 10023.
A method for recomputing the entries of the table has been suggested by
Neugebauer (cf. Kennedy and Faris, p. 24), and it consists of 3 steps:

(i) for each integer value of the argument (the longitude of the ascendant)
find its oblique ascension by means of a table for the appropriate latitude
(Neugebauer suggested using a table for 36 since no such table is known
for 35;55,48);
(ii) in a table for normed right ascensions (see ff. 48v49r; cf. al-Khwrizms
table in Suter (1914), pp. 171173), find the longitude for which its normed
right ascension equals the value obtained previously;
(iii) the declination corresponding to that longitude in a table of declinations
is the entry sought.
andalusian astronomy 199

table 5 Table for lunar eclipses

Col. 2 (d) Col. 3 (h, min, s) Col. 4 (h, min, s)


f. 52v f. 57v f. 52v f. 57v f. 52v f. 57v

12 0 55 0 54 0 6 0 0 6 0 0 0 0 0 0 0
11 1 35 1 35 0 14 0 0 54 5 0 0 0 0 0 0
10 3 50 3 50 0 50 40 1 50 40 0 0 0 0 0 0
9 4 16 4 56 1 6 50 2 6 50 0 0 0 0 0 0
8 6 47 6 47 1 43 9 2 2 9 0 0 0 0 0 0
7 8 19 8 39 1 57 50 2 36 50 0 0 0 0 0 0
6 10 29 10 29 2 0 9 3 0 20 0 0 0 0 0 0
5 12 0 11 0 2 0 50 3 0 50 0 0 48 0 0 40
4 12 0 12 0 2 0 0 3 0 0 1 2 40 0 2 40
3 12 0 12 0 2 20 0 4 20 0 1 6 48 1 6 48
2 12 0 12 0 2 25 0 4 40 0 1 8 48 1 8 43
1 12 0 12 0 2 30 0 4 50 0 1 10 12 1 10 42

d Lunar Eclipses

Tabula eclipsium lunarium (f. 52v)

The use of this table (see table 5) is described in canon 27 (f. 15vb). Columns 24
display the lunar eclipse magnitude (digits), the duration of the eclipse (hours)
and the duration of totality (hours), as functions of the argument of latitude
of the Moon (degrees). The headings are prima porta, secunda porta, tercia
porta, respectively, while that for the argument is longitudo a capite et cauda.
Among the material appearing at the end of this manuscript after the tables
associated with al-Muqtabis, folio 57v has the same table under the title: Hec
tabula est quam extraxit et composuit Alkemed, in eclipsibus lunaris in canone
suo que est extracta a canone Ebi Iusufi cognoscitur Byn Tarach, que est ualde
uerax. This author is probably to be identified with the late 8th-century astron-
omer Yacqb ibn riq, a collaborator of al-Fazr at Baghdad, and whose zij
was called the Sindhind (cf. Pingree (1968b) and (1970)). Note that Ebi Iusufi, or
Ab Ysuf, means the father of Joseph, and, in Arabic nomenclature, this nick-
name can be substituted for Jacob, who was the father of the Biblical Joseph.
Previously, Mills (1942), p. 245, had suggested that the name here was a cor-
rupt form of the name of the 11th-century astronomer Muammad ben Ysuf
ben Amad ibn Mucdh, from Jan.
200 chapter 7

The two versions of this table are displayed in table 5. The entries in both
versions of this table seem to be quite corrupt, and do not allow us to derive
the parameters underlying them. Another version of this table is found in
the Tables of Barcelona (Mills (1962), table 50), and the entries in it are also
corrupt.
Note that there is a single table for lunar eclipses, which is quite uncom-
mon, for almost all zijes have two such tables (one for minimum, and one for
maximum, lunar distance). The astronomical work of Jacob ben David Bon-
jorn, an astronomer of the 14th century from Perpignan, has only one table for
lunar eclipses, but the entries in it are unrelated to those here (cf. Chabs (1991),
p. 309).

e Color of Eclipses

Colores (f. 52v)

This table (see table 6) is arranged in 6 columns: column 1 displays the argu-
ment of lunar latitude; col. 2 gives the color of solar eclipses as a function of the
argument of lunar latitude, in degrees; col. 3 displays lunar latitude in minutes;
col. 4 gives the color of lunar eclipses in terms of lunar latitude; cols. 5 and 6 (not
shown here) display the magnitudes of solar eclipses in area digits as a function
of the magnitude of the eclipse in linear digits. Canon 27 (f. 16va) refers to the
first four columns, whereas the magnitudes of eclipses are treated on f. 15vb. All
six columns are also found in the Tables of Barcelona (Mills (1962): cols. 14
appear in table 51, and cols. 56 in table 48).
Kennedy (1956a), p. 159, has noted that Treatise viii of al-Brns Qnn
al-Masc d has a chapter on the colors of solar and lunar eclipses. On the colors
of lunar eclipses, see Goldstein (1967), pp. 234235. For Ibn al-Muthann (10th
century), the color changes during the eclipse, whereas for Ibn Ezra (Mills
(1947), p. 167) color is a function of latitude, as is the case here. Chapter 151
of Kitb alcAmal bil-Asturlb, by the Persian al-f (903986), contains a
similar list for the colors of lunar eclipses, but with different entries from those
presented here (see Kennedy and Destombes, p. 413).
Chapter 35 of the Libro de las Taulas Alfonsies, De qu color sera ell eclipsy,
concerns lunar eclipses, and gives two different rules for their colors. These
rules show similarities with those in the above-mentioned tables, but do not
fully agree with them.
andalusian astronomy 201

table 6 Table for the color of eclipses

(1) (2) (3) (4)


Arg. Solar Lat. Lunar
Lat. eclipse eclipse

1 valde niger 10 niger valde


2 niger clarus in nigredine
3 turbatus rubeus 20 niger
4 turbatus croceus cum rubedine
5 turbatus clarus 30 niger
6 turbatus cinereus cum rubedine
7 cinereus 40 niger
8 cinereus cum croceo
9 cinereus 50 turbatus
10 cinereus
11 croceus 60 cinereus
12 rubeus albus

The columns concerning the areas of eclipses (cols. 56) appear in a number
of earlier tables: the Toledan Tables (Toomer (1968), p. 113, table 76), the zij of
al-Battn (Nallino, ii, p. 89), the Almanac of Azarquiel (Mills (1950), p. 233),
and in al-Khwrizms zij (Suter (1914), p. 190, table 76, columns 67). In fact,
this table is already found in Ptolemys Almagest (vi, 8) and Handy Tables
(Stahlman (1959), p. 258). The entries in all these tables agree, except in the
Toledan Tables, where the entries for 3 and 5 (linear) digits are, respectively,
1;50 d and 3;20 d instead of 1;45 d and 3;40 d.

f Parallax in Latitude

Tabula latitudinis solis iudicate que est diuersitas respectus lune in lati-
tudine specialiter (f. 53r)

The use of this table (see table 7, below) is described in canon 27 (f. 15rb). The
table displays the adjusted parallax in latitude, p, in minutes and seconds of
arc, where adjusted parallax means the difference between the lunar and the
solar parallax (see Kennedy (1956b), p. 35). Its maximum value is 0;48,32 at 90.
The same table appears in al-Khwrizms zij (Suter (1914), pp. 191192, tables 77
and 77a, col. Diversitas respectus in latitudine), but the values tabulated there
202 chapter 7

table 7 Table for the adjusted parallax in latitude

(1) (2) (3) (4) (1) (2) (3) (4)

1 0;50 46 34;30
2 1;41 47 35;56 34;56
3 2;32 2;33 48 35; 0
4 3;23 49 35;52
5 4;13 4;14 50 36; 8 37;11
6 5; 3 5; 4
7 5;55 51 36;44
8 6;45 52 37;21
9 7;37 53 38; 7
10 8;23 8;27 8;26 54 38;59 38;39
55 39;28 38;48 39;45
11 9;16 56 39;22 39;21
12 10; 5 57 39;36
13 10;54 58 39;50
14 11;44 59 40;56 40;16
15 12;33 12;34 60 40;44 42; 2
16 13,23
17 14;12 61 40;55
18 15; 1 62 41;43
19 15;51 63 41;54 41;53
20 16;41 16;36 64 42;54 42;34 43;37
65 42;54 43;59
21 17;32 17;38 66 43;57 43;17 44;20
22 18;22 18;26 18;11 67 43;59 43;39
23 18;12 19;14 18;58 68 44; 0
24 18;58 19;58 19;44 69 44;22
25 19;44 20;44 20;31 70 44;43 45;36
26 20;31 21;17
27 21;17 22; 2 71 45; 5
28 22; 3 22;47 72 45;27
29 22;49 23;32 73 45;48
30 23;35 24;16 74 46; 2 46;39
75 46;16 46;53
31 23;35 24;22 25; 0 76 46;29 46;39 47; 6
32 24;22 25; 4 25;43 77 46;42
33 25;46 26;26 78 46;55
andalusian astronomy 203

(1) (2) (3) (4) (1) (2) (3) (4)

34 26;28 27; 8 79 47; 8


35 27;10 27;50 80 47;21 47;48
36 28;32 28;32
37 29;34 28;44 81 47;34
38 29;56 82 47;40
39 30;37 83 47; 0 48; 0
40 31;38 31;22 31;12 84 47; 5 48; 5
85 47; 9 48; 9
41 31;22 31;42 86 47;14 48;14
42 32;48 87 47;18 48;18
43 33;52 33;12 88 47;22 48;22
44 33;58 89 48;27
45 34; 4 34;19 90 48;32 48;34 48;32

differ in all cases, e.g., the entry for 90 is 0;48,45. This specific table was
discussed by Neugebauer (1962), pp. 121123.
Among the eclipse tables attributed to Yay ibn Ab Manr, and anal-
ysed by Kennedy and Faris (pp. 2038), there is a table entitled table for
the solar latitude ( jadwal ard al-shams) which coincides with this one (ms
Escorial Ar. 927, ff. 10v and 71v). Note the absurdity of the title for this table
which, in fact, deals with the latitudinal component of the adjusted parallax.
Kennedy and Faris (p. 25) give the function that underlies the entries of the
table:

[5] p = 0;48,32 sin(),

where is the solar zenith distance. They also state that the author of the
calculations necessary for this table did an extraordinarily bad job, as the
results almost never agree with the recomputed ones. These irregularities,
which also occur in this table by Ibn al-Kammd, allow us to relate it with
confidence to that of Yay.
The same table is also found in the Tables of Barcelona (Mills (1962),
table 46), and it exhibits the same inconsistencies as our text. Comparison
between the entries in al-Muqtabis and the recomputed values seem to indicate
that some shifts of entries occurred in the copying process (entries for 2330
have been shifted one place downwards, entries for 3335 have also been
204 chapter 7

shifted one place, and entries around 67 may have been shifted two places
downwards). In any case, we are far from the smoothness of the analogous table
in al-Khwrizm where the maximum value is 0;48,45.

The columns in table 7 for the adjusted parallax in latitude are the following:

(1) Solar zenith distance for each integer degree from 1 to 90.
(2) Entries in the text: adjusted parallax in latitude (p), in minutes and
seconds of arc.
(3) Variant readings in the Tables of Barcelona (ms Ripoll 21).
(4) Recomputed values by means of equation [5].

g Lunar Latitude

Tabula latitudinis lune iudicate (f. 53r)

This table is mentioned in canon 27 (f. 15rb), and it gives the lunar latitude ()
as a function of the argument of lunar latitude (). It is also found, with minor
variant readings, in al-Khwrizms zj (Suter (1914), pp. 132134, tables 2126; cf.
Neugebauer (1962), pp. 9598), where the maximum latitude is 4;30. Kennedy
and Ukashah, pp. 9596, have shown that the entries in this table were com-
puted according to the method of sines given by the formula:

[6] = 4;30 sin ().

Kennedy (1956a), p. 146, mentions that a similar table, with the same maximum
value, appears in the zij of Yay ibn Ab Manr.
The maximum value in the table on f. 53r is 4;29 instead of 4;30. This does
not suggest a different parameter; rather, it should be interpreted as a variant
reading of 4;30. The same table is also found in the Tables of Barcelona (Mills
(1962), table 44), and exhibits the same characteristics.
The table for the lunar latitude on f. 35v of this ms displays a different
maximum, 5;0, which is the value used by Ptolemy, al-Battn, Azarquiel,
among many others.
andalusian astronomy 205

h Elongation

Tabula longitudinis et dimidii sexti eius (f. 53v)

Canon 27 (f. 13rb) explains the use of this table. Two sets of entries are tabulated,
the lunar longitude (lm) and the solar longitude (ls); both sets of entries are
functions of the elongation (e) between the Moon and the Sun, given in degrees
and minutes, from 0;30 to 12;0, at intervals of 0;30. The entries lm and ls are
such that e= lm ls, where lm = 13e/12 and ls = e/12.
A similar, but more extensive table, is found in the Tables of Barcelona
(Mills (1962), table 41): the elongation is given at intervals of 0;6 instead of
0;30, and it ranges from 0 to 13;12 instead of from 0 to 12;0.

i Parallax in Longitude

Tabula diuersitatis respectus lune in longitudine (f. 53v)

This table, mentioned in canon 25 (f. 14ra), gives the longitudinal component
of adjusted parallax (p), in hours and minutes, as a function of the argument,
given in time from 0;15 h to 9 h, at intervals of 0;15 h.
The entries reach a maximum of 1;36 h, and can be easily derived from the
column with the heading Horae diversorum/diversitatis respectuum lunae
[in longitudine] in al-Khwrizms zij (Suter (1914), pp. 191192, tables 77 and
77a; cf. Neugebauer (1962), pp. 121126). However, in al-Khwrizms zij, (i) the
argument is not given in time, but in degrees, and (ii) the parallax in longitude
is given to seconds. Kennedy (1956b), pp. 4950, has shown that al-Khwrizms
table can be computed by means of the following formula:

[7] p = 1;36 sin ((t))

where

[8] t = ( sin ()),

such that , the argument (in degrees), meets the condition that 0 150.
The coefficient 1;36 = 0;4 24 = 24/15 contains the standard Hindu value for the
obliquity of the ecliptic, 24. In the case of Ibn al-Kammds table, the entries
can be recomputed by means of equation [7], where
206 chapter 7

[9] t = (24 sin ())/15,

for all 0 135; now, 135 = 9 h 15/h, and 9 h is the maximum value of the
argument, expressed in time.
Only two such adjusted longitudinal parallax tables of this kind are known:
that of al-Khwrizm and the one discovered by Kennedy (1956b, p. 48) in the
zij of Ibn al-Shir (ca. 1350), explicitly paraphrasing an early Islamic source
that has not been identified (Kennedy and Faris, pp. 3338). We can now add
to that short list the parallax table of Ibn al-Kammd and that in the Tables of
Barcelona (Mills (1962), table 45).

j Tabula eclipsium solarium ( f. 54r)


Column 2 of this table gives the magnitude of the eclipse, in (linear) digits and
minutes, as a function of the adjusted latitude (at conjunction) of the Moon
displayed in column 1, in minutes and seconds of arc, from 0;34,13 to 0. The
explanation of this table in canon 27 (f. 15va) confirms the above value for the
eclipse limit: si fuerit minus 34 minutis et 13 secundis erit eclipsis.
Columns 39 form a double argument table. The vertical argument is the
eclipse magnitude in digits (col. 2); the horizontal argument is the relative
velocity of the Moon with respect to the Sun (vm vs) in minutes and seconds
of arc per hour, from 0;27,30 /h to 0;33,30/h, at intervals of 0;1 /h.
This table, as some previous ones, seems to derive from Yay ibn Ab
Mansr (ms Escorial Ar. 927, f. 13r). Kennedy and Faris (pp. 2730), once again,
have transcribed and explained this table.
The adjusted latitude of the Moon is a linear function of the magnitude of
the eclipse, so that the graph of the function relating columns 1 and 2 should be
a straight line. However, this is not the case: there are jumps at 4;20 and 9;20
digits; and one should consider the first three entries in col. 2 and 0 d, 20 d, and
40 d, instead of 0;15 d, 0;30 d, and 0;45 d (these errors appear in this table in the
Arabic ms as well as in our Latin text).
The second part of the table displays the half-duration of the eclipse, in hours
and minutes, as a function of its magnitude and the relative velocity of the
Moon with respect to the Sun. The recomputations made by Kennedy and Faris
(p. 29) give good results, but fail to reproduce the entries of the table precisely.
andalusian astronomy 207

iv Other Tables in Al-Muqtabis

a Calendaric Tables ( ff. 27r v)


The purpose of these tables is to convert dates from the Arabic calendar to
Roman (i.e., Julian) and Egyptian calendars. The epoch of the radix given
is the Hijra: noon of July 14 ad 622. Canon 9 (f. 6ra) states that the epoch
of the radix is in the day of Mercury (Wednesday). The tables display two
correspondences:
ah0 = 932 Julian years 9 months 17;0 days from the beginning of the Seleucid
era = 9 Egyptian years 11 months 9 days from the beginning of the Yazdijird era.
These tables entirely or partially reproduce calendaric tables in the zij of
al-Khwrizm/Maslama (Suter (1914), p. 110, table 2; p. 111, table 2a; and p. 113,
table 3), and/or in pseudo-Battn (Maslama) (Nallino, ii, pp. 301, 304305).

b Trepidation ( ff. 28v, 35v)

Tabula aduenctionis puncti capitis arietis (f. 28v)

The same radix and equivalent entries for the mean motion of the vernal
point are found in Azarquiels Treatise on the motion of the fixed stars (Paris,
BnF, ms Heb. 1036; see Mills (1950), pp. 266, 324); the Liber de motu octave
sphere, attributed to Thbit ibn Qurra displays a similar table (Mills (1950),
p. 507); cf. Morelon (1987), p. xix. On the theory of trepidation see Goldstein
(1964), Dobrzycki (1965), North (1967), North (1976), vol. 3, pp. 155158, Mercier
(19761977) and Sams (1992).

Tabula directionis aduenctionis capitis arietis (f. 35v)

The entries (see table 8) display very nearly a sine function whose maximum
is 9;59 at 90; we have not succeeded in explaining the deviations from the
sine function, e.g., the entry for 30 is not half the entry for 90. Toomer (1968),
118, gives two tables, which are also sine functions, associated with the Toledan
Tables, and in fact they already appear in the Liber de motu octave sphere (Mills
(1950), pp. 507508). The radix and the mean motion of the first point in Aries
(see f. 28v and the comments on it, above) are taken from Azarquiels Treatise on
the motion of the fixed stars. The use of this table is briefly outlined in canon 12
(f. 7v), where another work by Ibn al-Kammd is explicitly mentioned: al-Amad
al al-abad. This table is also found in the Tables of Barcelona (Mills (1962),
table 20), a treatise by al-Marrkush (Sdillot (1834), p. 131), and in the tables
of Juan Gil (London, Jews College, ms Heb. 135, f. 78b).
208 chapter 7

We have tried to recompute this table in many ways, and by far the best fit
comes from formula [10] which is intended to represent Azarquiels somewhat
vaguely defined second model (Mills (1950), pp. 287289, 317318; cf. Sams
(1992), p. 230):

[10] sin (e) = r sin (i)/60

where e is the entry in the table, r = 10;24, and i is the argument. The maximum
entry in the table, 9;59, is indeed a rounded value for arcsin (10;24/60) =
9;58,54.

table 8 Table for the equation of the motion of the first


point in Aries

Degrees 0/6 1/7 2/8

1 29 0;10 5;16 8;52


2 28 0;20 5;25 8;56
3 27 0;31 5;34 9; 1
4 26 0;41 5;43 9; 5
5 25 0;53 5;52 9;10
6 24 1; 3 6; 4 9;14
7 23 1;54 6;16 9;17
8 22 1;25 6;29 9;21
9 21 1;35 6;41 9;24
10 20 1;45 7;53 9;28
11 19 1;56 7;57 9;31
12 18 2; 8 7; 2 9;35
13 17 2;19 7; 6 9;38
14 16 2;30 7;10 9;41
15 15 2;41 7;14 9;45
16 14 2;50 7;21 9;47
17 13 2;18 7;28 9;49
18 12 3; 6 7;35 9;51
19 11 3;15 7;42 9;52
20 10 3;25 7;48 9;55
21 9 3;36 7;54 9;56
22 8 3;47 8; 0 9;56
andalusian astronomy 209

Degrees 0/6 1/7 2/8

23 7 3;57 8; 7 9;57
24 6 4; 8 8;13 9;58
25 5 4;19 8;20 9;59
26 4 4;29 8;25 9;59
27 3 4;39 8;31 9;59
28 2 4;49 8;37 9;59
29 1 4;38 8;41 9;59
30 0 5; 7 8;47 9;59
5/11 4/10 3/9

Variant readings in the Tables of Barcelona (ms


Ripoll 21, f. 137r):
e (7) = 1;14 e(16) = 2;51 e(17) = 2;58 e(29) = 4;58
e(40) = 6;43 e(41) = 6;57 e(50) = 7;54 e(51) = 7;58
e(53) = 8; 6 e(55) = 8;19

figure 7.1 The geometrical model underlying


Ibn al-Kammds table for
trepidation, as reconstructed

Geometrically, we can understand the underlying model by referring to Fig-


ure 7.1: a small circle or epicycle, bce, whose center is a, lies in the plane of the
ecliptic, circle ag, and the center of the sphere is o. Note that the small circle
bce is partly inside, and partly outside, the sphere. The angle i is equal to arc
bc; radius ab = 10;24, and oa = 60. Through c, draw a line parallel to ba, reach-
ing the large circle ag at f. Draw of, and let aof be the angle e, which we seek.
To compute angle e, we drop a perpendicular fh from f to ao; then fh = cd,
210 chapter 7

table 9 Trepidation according to Ibn


al-Kammd

(1) (2) (3) (4)

10 1;45 1;43 2
20 3;25 3;24 1
30 5; 7 4;58 9
35 5;52 5;42 10
40 6;43 * 6;24 19
45 7;14 7; 2 12
50 7;48 7;38 10
60 8;47 8;38 9
70 9;28 9;22 6
80 9;55 9;50 5
90 9;59 9;59 0

* There is a textual problem with this


entry (see Table 8); hence we display
surrounding values.

and cd = r sin (i). In right triangle fho, fo = 60, and sin (e) = fh/fo; equa-
tion [10] follows. We cannot account for the remaining differences between
text and computation. The columns in table 9 for trepidation according to Ibn
al-Kammd are as follows:

(1) i: argument
(2) e: text of Ibn al-Kammd
(3) e: computed, using equation [10]
(4) the difference: T(ext)C(omp.), in minutes

The value 10;24 for r in equation [10] is made plausible by the two other
models for trepidation associated with Azarquiel and his followers. Sams
(1992), pp. 235236, describes the third model of Azarquiel and shows how
it yields a maximum value, pmax very close to 10;24 (10;23,29), based on the
equations:

[11] = r sin (i),

[12] sin (p) = sin ()/sin (23;33)


andalusian astronomy 211

table 10 Trepidation according to Ibn


al-Bann and Ibn al-Raqqm

(1) (2) (3) (4)

10 1;48 1;47,47 1;47,48


20 3;32 3;32;23 3;32,24
30 5;11 5;10,43 5;10,40
40 6;40 6;39,48 6;39,40
50 7;57 7;56,55 7;56,42
60 8;59 8;59,39 8;59,20
70 9;46 9;45,59 9;45,35
80 10;14 10;14,25 10;13,56
90 10;24 10;24 10;23,29

where p is the amount of precession corresponding to an argument i, r is 4;7,58,


and 23;33 is the value for the obliquity. Moreover, Neugebauer (1962), p. 184,
presents a formula based on Azarquiels first model:

[13] sin (p) = sin (pmax) sin (i),

If we substitute 10;24 for Pmax in equation [13], we find a set of values that
agree very well with those preserved by the late 13th-century astronomers Ibn
al-Bann (ms Escorial Ar. 909, f. 22v) and Ibn al-Raqqm (ms Kandilli 249, f. 66v)
who only tabulated these entries to degrees and minutes. However, since the
agreement is equally good using equations [11] and [12] on the one hand, and
equation [13] on the other, we cannot decide which procedure was used for
computing this table. Table 10 displays these computations, where the columns
are the following:

(1) i: argument
(2) e: text of Ibn al-Bann and Ibn al-Raqqm
(3) e: computed, using equation [13]
(4) e: computed, using equations [11] and [12]

c Mean Motion Tables ( ff. 28r34v)


The headings of the tables indicate that they were intended for the meridian
of Crdoba and calculated for Arabic years, months, etc. We have computed all
mean motions from the corresponding tabulated values for ah 720, except for
212 chapter 7

that of the lunar anomaly. The method is simply to subtract the corresponding
radix from each value for ah720, taking into account the full rotations, and then
to divide the result by the number of days elapsed from the epoch of the Hijra
calendar.

Sun (f. 28r)


The tabulated values for the mean motions are given to seconds. They do
not agree with either al-Khwrizm (Suter (1914), p. 115, table 4), or al-Battn
(Nallino, ii, p. 20). The entry for the radix is 1s 6;35,9, and the tabulated
value for ah720 is 7s 16;13,19. The difference between these two values is 6s
9;38,10 = 189;38,10. Now, in 720 Arabic years, the Sun has completed 698 full
rotations; therefore, the are length traveled by the Sun in 720 Ar. y. (= 255,144
days) is 189;38,10 + (698 360) degrees, and the daily mean motion of the
Sun resulting is 0;59,8,9,21,15, /d, which yields a year-length of 365;15,36,34,
d. The value deduced from the table is sidereal. It differs from the daily
mean motion embedded in the Toledan Tables, and attributed to Azarquiel
(0;59,8,11,28,27, /d) by an amount which is exactly equal to Azarquiels value
for the daily motion of the solar apogee (0;0,0,2,7,10,39, /d, cf. Toomer (1969),
p. 319).
The radix given here (1s 6;35,9) corresponds to the solar centrum (the
distance from a sidereally fixed apogee). To obtain the longitude of the Sun at
epoch, add the given radix to the longitude of the apogee (2s 16;45,21, f. 35r);
the result is 3s 23;20,30, a value which is close to, but not identical with, those
of Azarquiel or al-Khwrizm/Maslama (see Toomer (1968), p. 44).

Solar Apogee (f. 28v)


The daily mean motion of the solar apogee resulting from the tabulated value
for ah720 (0s 2;24,24) is 0;0,0,2,2,14,46, /d. This implies a progress of 1o in
about 299 Arabic years or in about 290 Julian years, a value which differs from
the daily motion of the apogee used by Azarquiel (0;0,0,2,7,10,39, /d), which
corresponds to a progress of 1 in about 279 Julian years. A parameter very
similar to that of Azarquiel is also found in the works of Ibn Isq, Ibn al-Bann
and Ab l-asan cAl b. Ab cAl al-Qusanayn (Mills (1950), pp. 352353;
Sams (1992), p. 212).
The recomputations show that Ibn al-Kammd used Azarquiels value for
the mean motion of the Sun, but that he incorporated a different parameter for
the mean motion of the apogee which does not appear in any known text prior
to this one.
andalusian astronomy 213

Moon (ff. 29rv)


The daily mean motion of the Moon in longitude resulting from the tabu-
lated value for 720ah (10s 8;37,1) is 13;10,34,52,46, /d. This is exactly al-
Khwrizms value (cf. Neugebauer (1962), pp. 42, 92), and very nearly that in
the Toledan Tables (cf. Toomer (1968), p. 44).
The daily mean motion of the Moon in anomaly resulting from the tabulated
value for 900 a.r. (9s 13;12) is 13;3,53,56,19, /d. This is very nearly the value
in the Toledan Tables (cf. Toomer (1968), p. 44), which differs from that of
al-Khwrizm (cf. Neugebauer (1962), p. 92).
The tabulated values for the mean motion in longitude agree almost exactly
with those in al-Khwrizm (Suter (1914), pp. 117119, table 68), and exhibit very
small differences with those in the Toledan Tables (Toomer (1968), p. 48). How-
ever, the tabulated values for the mean motion in anomaly are in agreement
with those of al-Battn (Nallino, ii, p. 20), and in the Toledan Tables (Toomer
(1968), p. 49), except for the fact that the motion in anomaly is given to seconds
there.

Planets (ff. 30r34v)


For Saturn, Jupiter, Mars and Mercury, the tables give entries for the mean
motion of the centrum (i.e., the distance from a fixed apogee) and anomaly.
For Venus, the entries display the mean motion in anomaly only, and for the
lunar node, the entries are the complement in 360 of the mean motion of
the ascending node in longitude. All entries are given to minutes. Those for
the yearly progress of the centrum of Mercury agree with those for the Sun
(f. 28r), except for the fact that the latter are given to seconds. The values for
the radices are for the Hijra (see table 11, below), and differ from those in the zij
of al-Khwrizm/Maslama and those in the Toledan Tables.
The mean motions in anomaly of the superior planets are not generally tab-
ulated in other sets of astronomical tables (and this is also the case for almost
all copies of the Toledan Tables). The recomputed values obtained here can
be compared, however, with those in one copy of the Toledan Tables (Oxford,
Bodleian Library, ms Laud. Misc. 644, cf. Toomer (1968), p. 45): 0;57,8/d (Sat-
urn), 0;54,9/d (Jupiter) and 0;27,41/d (Mars). The same values are explicitly
found in the work of abash (Debarnot (1987), p. 44), where the value for Mars
is 0;27,42/d.
All parameters computed from the tabulated entries for the mean motions
show close, although not perfect, agreement with those derived from the Tole-
dan Tables, the differences never being greater than 0;0,0,1.
214 chapter 7

table 11 Summary of the mean motions and radices

Mean motion () Radix ()

Solar longitude 0;59,8,9,21,15, 1s 6;35, 9 *


Solar apogee 0;0,0,2,2,14,46, 2s 16;45,21 **
Vernal point 0;0,0,54,56,57, 0s 3;51,11
Lunar longitude 13;10,34,52,46 4s 0;34,42 ***
Lunar anomaly 13;3,53,56,19, 3s 18;11
Double elongation 0s 14;33
Lunar node 0;3,10,46,41, 4s 6;30
Saturn (longitude) 0;2,0,25,36, 7s 26;52 *
Saturn (anomaly) 0;57, 7,44,57, 11s 27;48
Jupiter (longitude) 0;4,59,6,43, 5s 21;58 *
Jupiter (anomaly) 0;54,9,3,37, 4s 23; 2
Mars (longitude) 0;31,26,31,40, 3s 1;46 *
Mars (anomaly) 0;27,41,40,34, 8s 22;14
Venus (anomaly) 0;36,59,29,21, 1s 15;21
Mercury (longitude) 0;59,8, 11,23, 9s 4;58 *
Mercury (anomaly) 3;6,24, 7, 19, 2s 14; 1

* The values of the radices for the Sun and the planets correspond to
their centrum, i.e., their distance from apogee. Note that the ms does
not provide this information for Venus.
** This value for the longitude of the solar apogee is given on f. 35r.
However, in the table for the mean motion of the solar apogee (f. 28v)
one finds 0s 0;0,0 opposite Radix.
*** The radix for the Moon is called its centrum, here meaning
longitude.

d Spherical Astronomy

Declination (f. 35v)


This is a table giving the declination of the Sun for each integer degree; the max-
imum entry is 23;33. This value for the obliquity of the ecliptic is associated
with the zij al-Mumtaan (Vernet (1956), p. 515). The entries in this table agree,
except for minor differences, with those in the Almanac of Azarquiel (Mills
(1950), p. 174), where the argument is given only at intervals of 3. The value
found in the Toledan Tables, and attributed to Azarquiel, is 23;33,30 (Toomer
(1968), p. 30).
andalusian astronomy 215

Daylight (f. 47v)


The heading of the table mentions Crdoba, but it does not specify a value
for its latitude. The geographical table on f. 54v gives its latitude, , as 38;30,
the most prevalent value for it at the time. This table gives the half-length of
daylight as a function of the solar longitude. The maximum entry represents
half of the longest daylight (m/2), and it is 7;21 h for Cancer 0. This value follows
from the formula:

[14] tan () = cos(m/2) cot (),

where = 38;30 and = 23;33.

Normed Right Ascension (ff. 48v49r)


The same table of right ascension, beginning with Capricorn 0, is found in al-
Battn (Nallino, ii, pp. 6164), for an obliquity () of 23;35. It is also found
in the Toledan Tables (Toomer (1968), p. 34, table 17) and, with copying errors
in an abridged version, in the Almanac of Azarquiel (Mills (1950), pp. 220
221, tables 6970). The table differs from that in the Handy Tables (Stahlman
(1959), pp. 206209, table 1) and in al-Khwrizms zij (Suter (1914), pp. 171173,
tables 5959b), both calculated for higher values of the obliquity.

Oblique Ascension for Crdoba (ff. 49v51r)


The mean values between the rising times of Aries and Virgo, of Taurus and
Leo, and of Gemini and Cancer, are respectively 27;50, 29;54,30 and 32;15,30.
They are almost identical with those derived from the Toledan Tables for the
seven climates: 27;50, 29;54 and 32;16, which are the Ptolemaic values for the
right ascensions (cf. Toomer (1968), p. 42). They are also embedded in the zij of
al-Khwrizm, but differ from those in the zij of al-Battn.
When recomputing the entries, close, although not exact, agreement is
obtained with = 38;30 and = 23;51. With the two other values of the
obliquity found in our text (23;33 and 23;35), the agreement is worse.
Another column in this table represents the length of the seasonal hours.
The entry for Cancer 0 is 18;24. Now 18;24 12/15 = 14;43 h, which gives a half
daylight of 7;21,30 h, and this is quite close to the value 7;21 h for Cancer 0 (see
the table for the half-length of daylight as a function of the solar longitude on
f. 47v).
A similar table for Sal, a place in North Africa near Rabat, whose latitude is
given here as 33, is found on ff. 59v61r.
216 chapter 7

e Latitude
Moon (f. 35v)
The maximum value in the table for the latitude of the Moon is 5;0. The same
value for the inclination of the lunar orb is found in the Almagest v, 8, and in
many other texts, including the Toledan Tables and the Almanac of Azarquiel
(cf. Mills (1950), p. 173, where the argument is given only at intervals of 6). For
another table for the lunar latitude in al-Muqtabis, see Section iii. g.

Planets (f. 45rv)


This table for the latitude of the superior planets is the same as the one in the
Almagest xiii, 5, and in al-Battn (Nallino, ii, p. 140 (columns 13) and p. 141
(column 4)). The pattern of this table differs greatly from the corresponding
one in the Handy Tables. Toomer listed some mss associated with the Toledan
Tables that contain such a table, but concluded that it is not part of the original
Toledan Tables (Toomer (1968), p. 72).
In contrast to the superior planets, the latitude table for the inferior planets
does not conform to the pattern of the Almagest, the zij of al-Battn, or the
tables associated with the Toledan Tables. Rather, this table reproduces, with
variant readings, the entries which are multiples of 6 in the Handy Tables
(Stahlman (1959), pp. 331334, tables 4950, where the entries are given at
3-degree intervals). In particular, the maximum values for the mean latitude
of Mercury (3;52) agree in both sets of tables, but those for the mean latitude
of Venus differ (8;35 in our text and 8;51 in the Handy Tables). However,
canon 16 (f. 10va) gives 8;36 and 4; 18 as the values for the maximum latitude
of Venus and Mercury. Kennedy (1956a), p. 173, reports maximum values for
Venus (8;56) and Mercury (4;18), and associates the zij al-Mumtaan and Ibn
Hibint with them.
Nevertheless, the outstanding feature here is the juxtaposition of different
Ptolemaic tabular material: the Almagest for the superior planets, and the
Handy Tables for the inferior planets. The source for such a mixed approach
has not been determined.
Note that in the tables associated with al-Muqtabis no values are given for
the longitudes of the planetary nodes, although ms 10023 (f. 66r) has a list of
them.

f Equations
Moon (ff. 36r37r)
The same table is found in the Almagest v, 8, in the Handy Tables, as well
as in many medieval tables, such as the zij of Yay ben Ab Manr (Salam
and Kennedy, pp. 495496), the zij of al-Battn (Nallino, ii, pp. 7883) and
andalusian astronomy 217

the Toledan Tables (Toomer (1968), pp. 5859). The table lists columns for the
equation for mean to true apogee, an interpolation function, the increment in
the equation of center, and the equation of center. There is no column here for
the lunar latitude, which is tabulated separately (see f. 35v). The entries for the
equation of anomaly agree with those in al-Battns zij, but differ slightly from
those in the Toledan Tables (e.g., in our table the maximum of 5; 1,0 is reached
at 95, while in the Toledan Tables the value is 5;0,59, and it occurs at 9495).
On the other hand, in our table the order of the columns is the same as in the
Toledan Tables, and differs from that in al-Battns zij.

Planets (ff. 37v44v)


The tables for the equations of the five planets are essentially those found in the
Almagest xi, 11, in the Handy Tables, as well as in the zij of al-Battn (Nallino, ii,
pp. 108137) and the Toledan Tables (Toomer (1968), pp. 6068), among many
others. Although in most respects the zij of al-Battn and the Toledan Tables
agree for the planetary equations, there are some differences between them:
for instance, there is a column for the planetary stations in the Toledan Tables,
which is neither in al-Battns zij nor in our table. All we can deduce from these
tables is that Ibn al-Kammd accepted the Ptolemaic tradition, as displayed in
the Handy Tables, followed by most Muslim astronomers.
It is worth noting that in the tables of Ibn al-Kammd the maximum solar
equation (1;52,44: f. 35r) differs from the maximum equation of center for
Venus (1;59). This value for Venus is not that of the Almagest, but follows
al-Battn, etc. (cf. Goldstein and Sawyer). This indicates that Ibn al-Kammds
contribution was restricted to solar theory and that he did not introduce any
changes in planetary theory.
Values for the sidereally fixed apogees appear above the tables for the equa-
tions on ff. 37v38v (Saturn), 39r40r (Jupiter), 40v41v (Mars), 42r43r (Venus),
and 43v44v (Mercury). The apogee for Venus is that ascribed to the Sun on
f. 35r.

Saturn 238;38,30
Jupiter 158;21, 0
Mars 119;41, 0
Venus 76;45,21
Mercury 198;21, 0

g Stations ( f. 46r)
Only four values are given for each planet: the positions of the first and the sec-
ond stationary points for arguments of 0 and 180 (see table 12). The tabulated
218 chapter 7

table 12 Planetary stations

Saturn Jupiter Mars Venus Mercury

1st st. at apogee 3s22;44 4s 4; 5 5s 7;28 5s15;51 4s27;14


2nd st. at apogee 8s 7;16 7s25;55 6s22;32 6s14; 9 7s 2;46
1st st. at perigee 3s25;30 4s 7;11 5s19;15 5s18;21 4s24;42
2ndt st at perigee 8s 4;30 7s22;49 6s10;45 6s11;39 7s 5;18

values for the same argument add up correctly to 360. In all cases, they coin-
cide with those found in the Toledan Tables (Toomer (1968), pp. 6068), and the
zij of al-Khwrizm (Suter (1914), pp. 138167, tables 2756), both zijes display-
ing tables for each integer degree of the argument. Nearly the same values are
also found in the zij of al-Battn (Nallino, ii, pp. 138139), which have a Ptole-
maic origin. However, Almagest xii.8 and the Handy Tables (Stahlman (1959),
pp. 335339, tables 5155) have slightly different tables for the stations: in the
latter the argument was modified, as well as the interval for the calculation of
the entries (6 in the Almagest, 3 in the Handy Tables).

h Equation of Time ( f. 46r)


Beneath the table we read: Mediatus solis in radice posita ad directionem
dierum cum noctibus: 10.23.24.50. a puncto capitis arietis. This value agrees
with that appearing in canon 11 (f. 7rb), and seems to correspond to the argu-
ment (in signs and degrees) for the minimum entry in the table. Hence it is to be
understood as 10s 23;24,50. The entries in this table, in time-degrees, rounded
to the nearest integer, may have been taken from the more precise values given
in the zij of al-Battn (Nallino, ii, pp. 6164) or in the Toledan Tables (Toomer
(1968), pp. 3435).

i Trigonometry
Functions Related to the Sine (f. 46v)
We adopt the convention that Sin = 60 sin , and similarly other capitalized
trigonometric functions are normed for r = 60 (cf. Kennedy (1956a), p. 139).
Three functions are given for each integer degree: Sin , Cos , and Vers =
r Cos . The Almanac of Azarquiel (Mills (1950), p. 229) has a table with
three functions (sine, cosine, and versine), but given at intervals of 3. Except
for copying errors, the entries agree in both tables.
andalusian astronomy 219

Cotangent Function (f. 48r)


The entries in this table represent the length of a shadow (s) projected by a
gnomon of 12 units as a function of the altitude of the Sun (h):

[15] s = 12 cotan (h),

in the tradition of al-Khwrizm (Suter (1914), p. 174, table 60) and al-Battn
(Nallino, ii, p. 60).

j Star Table ( f. 47r)


This is a list of 30 stars and it displays the following information for each
star: magnitude, name, ecliptic coordinates (longitude and latitude), and the
planets associated with it (for astrological purposes).
Kunitzsch (1966), pp. 99102, described this list under his type xv: the eclip-
tic longitude of each star is derived from that in Ptolemys star catalogue by
adding 6;38, thus indicating that the epoch of this star list is the Hijra. The
type defined by Kunitzsch only includes two versions of this list: the other ver-
sion is uniquely represented by ms Vienna 5311, f. 129v. Although the stars in
both versions are the same, the coordinates for them do not always agree, thus
suggesting a common Arabic ancestor. Kunitzsch also reports close similarity
with the star list of Ab l-asan cAl al-Marrkush (ca. 1262). We have found
additional copies of this list in some copies of the Tables of Barcelona, e.g., ms
Vatican Heb. 356, f. 65b, where the names of stars are given in Hebrew.

k Excess of Revolution
In the first table on f. 54v, the excess of revolution is given in degrees and
minutes, for 1, 2, 3, , 10, 20, , 100 years. The entry for 1 year is 92;36, which
seems to be an isolated error. To be coherent with all other entries in the table,
one should read 93;36, leading to a year-length of 365;15,36 days = 365d 6;14,24
h. The entry for 100 years is 0;5, and the resulting length of the solar year is
365;15,36,0,30 days.
In the second table on f. 54v, the excess of revolution is given in hours and
minutes of time, for 1, 2, 3, , 10, 20, , 100 years. The entry for 1 year is 6;14 h,
which corroborates the emendation above. The entry for 100 years is 23;58 h,
and the resulting length of the solar year is 365;15,35,58,30 days, a value close,
but not equal, to the parameter derived from the previous table.
These values for the solar year are to be compared with that appearing in
the canons (f. 2va): 365;15,36,19,34,12 days (cf. Mills (1942), p. 236), as well as
with the information given on f. 65v of this manuscript: Length of the solar
year according to Ibn al-Kammd:
220 chapter 7

365;15,36,19,35,32 days.

The daily mean motion of the Sun resulting from the first value is 0;59,8,9,23,44,
53 /d (the second value yields a very similar parameter: 0;59,8,9,23,44,40 /d),
which agrees quite well with the value derived above from the table for the
mean solar anomaly on f. 28r.
On f. 57v of this manuscript there is an analogous table attributed to Azar-
quiel, and for a sidereal year of 365;15,24d.
There follows a list of the various lengths of the solar year associated with
Ibn al-Kammd:

365;15,36, 0,30d (computed from the first table: f. 54v)


365;15,35,58,30d (computed from the second table: f. 54v)
365;15,36,19,34,12d (mentioned in canon 1: f. 2va)
365;15,36,19,35,32d (attributed to Ibn al-Kammd: f. 65v)

l Geographical Table ( f. 54v)


This is a list of 30 places: for each of them we are given its longitude and
latitude, in degrees and minutes. The prime meridian used here is located west
of the shore of the Western Ocean. It thus differs from that in the Toledan
Tables, where the shore of the Western Ocean seems to have been used for
most longitudes (cf. Toomer (1968), p. 136). For a general discussion of the prime
meridian in Islamic sources, see Kennedy and Kennedy, p. xi.
The entry for the latitude of Crdoba is 38;30, and that for its longitude
27;0, which is the same value given in canon 9 (f. 6ra): longitudo a circulo
occidentis ex centro Erin est gradus 27. The same values for the longitude and
latitude of Crdoba are found in some other Islamic sources, notably in a work
by Ab l-asan cAl al-Marrkush (see Kennedy and Kennedy, p. 95). We note
that in the Toledan Tables, the longitude of Crdoba is given as 9;20 and its
latitude as 38;30 (Toomer (1968), p. 134).

m Miscellaneous

Tabula directionis arcus luminis et transitus (f. 48r)

This table has 3 columns: (1) gradus longitudinis, (2) directio arcus lumi-
nis, and (3) minuta diuersitatis transitus. Col. 1 lists degrees at 3 intervals
from 3 to 90 (with some copying errors); col. 2 has entries in degrees and
minutes from 0; 14 to 1507;0; and col. 3 has entries in minutes and seconds
from 0;1,23 to 0;23,33. The structure of this table is similar to that of some
andalusian astronomy 221

tables in the Almanac of Azarquiel (Mills (1950), p. 226), and most of the
entries are the same (the entries in col. 2 agree with those in the correspond-
ing column in the Almanac of Azarquiel for arguments from 3 to 84, but
for roundings and copying errors: note that the column in the Almanac dis-
plays degrees, minutes, and seconds, rather than degrees and minutes). But
the headings for the columns in the Almanac of Azarquiel are different from
those here, and we have not succeeded in determining the purpose for which
our table (or the corresponding table in the Almanac of Azarquiel) was com-
puted.

v Other Tables in ms Madrid 10023

a. Just after the last table associated with al-Muqtabis, there are two tables
related to Azarquiels solar theory: (f. 55r) Tabula motus centri circuli exeun-
tis centrum in longitudine longiori et propinquiori a centro terre; (ff. 55v
56r) Tabula directionis composite centri circuli solis exeuntis centrum de
diuersitate centri eiusdem in longitudine propinquiori et longiori a circulo
diuersitatis motus centri morantis tempus. Toomer (1969), p. 325, has repro-
duced an excerpt and has explained the two tables.
b. (f. 56v) Tabula uisuum lunarium post occasum solis in climatibus septem
c. (f. 57r) Tabula eclipsis lune et quot digiti eclipsantur ex ea et hore dimidii
temporis eclipsis
d. One of the tables on f. 57v has already been analysed above (see Section iv,
k), in connection with a table in al-Muqtabis (f. 54v) giving the length of the
solar year. For another table on f. 57v, an eclipse table, see the comments
in Section iii, d, concerning lunar eclipses in al-Muqtabis (f. 52v). Folio 57v
has still another table for the maximum values for latitudes in the seven
climates.
e. There follow two tables that are clearly related to table 1 in the zij of al-
Khwrizm: (f. 58r) Tabula cuius est inter annos gentium et alios annos
preter illos ad inuicem; (f. 58v) Numeri dimissi per 28, 28 secundum annos
romanorum et egyptiacumi. According to Mills (1942), p. 245, they derive
from Maslama rather than from al-Khwrizm.
f. (f. 58v) Circuitus planetarum magni in sectis et divinationibus. These are
the Mighty Years of the planets: see table 13, where the entries of this table
are compared with those of Ab Macshar as preserved by al-Sijz (Pingree
(1968a), p. 64).
g. (f. 58v) Tabula dierum prouenientium in retrogradationibus planetarum et
directionibus eorum
222 chapter 7

h. (f. 59r) Tabula partis cordarum supereminentium et partis almudarat solis


et planetarum
i. Folios 59v to 61v present tables for the city of Sal: see the comments in
Section iv, d, on the table in al-Muqtabis for the oblique ascension for
Crdoba (ff. 49v51r).
j. The following table is certainly a part of al-Kawr al al-Dawr: (ff. 62v
64r) Tabula extractionis annorum quantitatis durationis creature in uentre
matris per longitudinem lune a gradu occidentis. It deals with astrological
obstetrics and has been discussed by Vernet (1949), pp. 273300.
k. (f. 64v) Tabula circuituum annorum planetarum in natiuitatibus
l. There follows information on houses, exaltations, triplicities, and signs, pre-
sented in tabular form (f. 65r), but every other entry has been left blank. The
information given agrees with that in al-Khwrizm/Maslama (Suter (1914),
p. 231, table 116). The table with the heading, Tabula terminorum egyptio-
rum (f. 65r) is the same as the fourth sub-table in al-Khwrizm/Maslama
(Suter (1914), p. 231, table 116), but some of the entries are blank here.
m. (f. 65v) Mediatus cursus solis in descensu eius ad quartas circuli secundum
probationem huius canonis. The data are for ah 480 (10871088 ad).
n. (f. 65v) Tabula terminorum ciuium Babillonie ueteris qui sunt magistri
ymaginum. Two parameters are given in this table: the length of the solar
year according to Ibn al-Kammd (365;15,36,19,35,32 days, cf. our com-
ments in Section iv, k, on the table in al-Muqtabis on f. 54v) and his length
of the lunar month (29;31,50,5,1 days).
o. (f. 66r) Tabula extracta per misilme de eo quod confirmatum extitit per
ciues huius artis yspanenses super diuisionem Yspanie per signa duodecim
12 12 12 12. Mills (1942), p. 256, suggested that misilme stands for Mas-
lama.
p. (f. 66r) Residuum ascensionum ad reuoluciones annorum solarium secun-
dum Muhad Arcadius. Mills (1942), p. 256, identified Muhad Arcadius
with Ab cAbd Allh Muammad ben Ysuf ben Amad Ibn Mucdh al-
Jayyn, from Jan. The value given for one year, 93;2,15, corresponds to
365;15,30,22 days. The same value is found in al-Khwrizm/Maslama (cf.
Neugebauer (1962), p. 132, and Goldstein (1967), pp. 143, 242). In fact, this
table attributed to Ibn Mucdh reproduces, with some scribal errors, the
table of conversion for the years of Nativity given in units of time-degrees in
al-Khwrizm/Maslama (Suter (1914), p. 230, table 115; cf. Neugebauer (1962),
p. 131).
q. (f. 66r) Tabula uisuum planetarum et absconsionum eorum sub radiis solis
r. (f. 66r) Capita draconum planetarum in anno quingentesimo et quinqua-
gesimo ab annis seductionis. Note that ah550 corresponds to 11551156ad.
andalusian astronomy 223

table 13 Mighty years of the planets

Planet ms Madrid 10023 Ab Macshar


(f. 58v)

Sun 1461 1461


Venus 1151 1151
Mercury 480 480
Moon 420 * 520
Saturn 625 * 265
Jupiter 567 * 427
Mars 684 * 284

* Note that several of the entries in ms Madrid 10023 are corrupt.

See our comments in Section iv, e, on the table of al-Muqtabis for the lati-
tude of the planets (ff. 45rv).

Note Added in Proof

We are grateful to ngel Mestres (University of Barcelona) for calling our atten-
tion to the following passage in ms Hyderabad, Andra Pradesh State Library 298
(no foliation): chap. 35 Ab l-cAbbs al-Kammd said in a horoscope he drew
in Crdoba in the year 510 Hijra (= 11161117ad).

References

Ahlwardt, W. 1893, Verzeichniss der arabischen Handschriften der Bibliothek zu Berlin,


vol. v. Berlin.
Chabs, J. 1991, The Astronomical Tables of Jacob ben David Bonjorn, Archive for the
History of Exact Sciences 42: 279314.
Chabs, J., and Goldstein, B.R. 1992, Nicholaus de Heybech and His Table for Finding
True Syzygy, Historia Mathematica 19, pp. 265289.
Debarnot, M.-T. 1987, The zj of abash al-sib: A Survey of ms Istanbul Yeni Cami
784/2, in David A. King and George Saliba (eds.), From Deferent to Equant: A Volume
of Studies in the History of Science in the Ancient and Medieval Near East in Honor of
E.S. Kennedy, Annals of the New York Academy of Sciences 500: 3569.
Dobrzycki, J. 1965, Teoria precesji w astronomii sredniowiecznej, Studia i Materialy
Dziejow Nauki Polskiej, Seria c 11: 347.
224 chapter 7

Goldstein, B.R. 1964, On the Theory of Trepidation, Centaurus 10: 232247.


Goldstein, B.R. 1967, Ibn al-Muthanns Commentary on the Astronomical Tables of
al-Khwrizm, New Haven and London.
Goldstein, B.R. 1985, Scientific Traditions in Late Medieval Jewish Communities, in
G. Dahan (ed.), Les Juifs au regard de lhistoire, Paris, pp. 235247.
Goldstein, B.R. 1992, Lunar Velocity in the Ptolemaic Tradition, in P.M. Harman and
A.E. Shapiro (eds.), An Investigation of Difficult Things: Essays on Newton and the
History of Exact Sciences, Cambridge, pp. 317.
Goldstein, B.R., and Sawyer, F.W. iii, 1977, Remarks on Ptolemys Equant Model in
Islamic Astronomy in Y. Maeyama and W.G. Saltzer (eds.) Prismata, Wiesbaden,
pp. 165181.
Goldstein, B.R., Chabs, J., and Mancha, J.L. 1994, Planetary and Lunar Velocities in the
Castilian Alfonsine Tables, Proceedings of the American Philosophical Society, 138:
6195.
Kennedy, E.S. 1956a, A Survey of Islamic Astronomical Tables, Transactions of the
American Philosophical Society ns 46.
Kennedy, E.S. 1956b, Parallax Theory in Islamic Astronomy, Isis 57: 3353; reprinted in
Kennedy, E.S., Studies in the Islamic Exact Sciences, Beirut (1983), pp. 164184.
Kennedy, E.S. 1986, The Verified Astronomical Tables for the Caliph al-Mamn by Yay
b. Ab Manr. A photographic reproduction of ms Escorial Ar. 927, with an intro-
duction. Frankfurt.
Kennedy, E.S., and Destombes, M. 1966, lntroduction to Kitb al-cAmal bil-Asturlb,
Osmania Oriental Pub., Hyderabad-Dn. (1966); reprinted in Kennedy, E.S., Studies in
the Islamic Exact Sciences, Beirut (1983), pp. 405447.
Kennedy, E.S., and Faris, N. 1970, The Solar Eclipse Technique of Yay b. Ab Manr,
Journal for the History of Astronomy 1: 2038; reprinted in Kennedy, E.S., Studies in
the Islamic Exact Sciences, Beirut (1983), pp. 185203.
Kennedy, E.S., and Kennedy, M.H. 1987, Geographical Coordinates of Localities from
Islamic Sources, Frankfurt.
Kennedy, E.S., and Ukashah, W. 1969, Al-Khwrizms Planetary Latitude Tables, Cen-
taurus 14: 8696; reprinted in Kennedy, E.S., Studies in the Islamic Exact Sciences,
Beirut (1983), pp. 125135.
Kunitzsch, P. 1966, Typen van Sternverzeichnissen in astronomischen Handschriften des
zehnten bis vierzehnten Jahrhunderts, Wiesbaden.
Mercier, R. 19761977, Studies in the Medieval Conception of Precession, Archives
internationales dhistoire des seiences 26: 197220, and 27: 3371.
Mills, J.M. 1942, Las traducciones orientales en los manuscritos de la Biblioteca Catedral
de Toledo, Madrid.
Mills, J.M. 1947, El libro de los fundamentos de las Tablas astronmicas, MadridBarce-
lona.
andalusian astronomy 225

Mills, J.M. 1950, Estudios sobre Azarquiel, MadridGranada (19431950).


Mills, J.M. 1962, Las Tablas Astronmicas del Rey Don Pedro el Ceremonioso, Madrid
Barcelona.
Morelon, R. 1987, Thbit ibn Qurra. Oeuvres dastronomie, Paris.
Nallino, C.A. 18991907, Al-Battn sive Albatenii Opus Astronomicum, 3 Vols. Milan.
Neugebauer, O. 1962, The Astronomical Tables of al-Khwrizm. Copenhagen.
North, J.D. 1967, Medieval Star Catalogues and the Movement of the Eighth Sphere,
Archives internationales dhistoire des sciences 17:7383.
North, J.D. 1976, Richard of Wallingford: An edition of his writings with introductions,
English translation and commentary, 3 Vols. Oxford.
Pingree, D. 1968a, The Thousands of Ab Macshar. London.
Pingree, D. 1968b, The Fragments of the Works of Yacqb ibn riq, Journal of Near
Eastern Studies 27: 97125.
Pingree, D. 1970, The Fragments of the Works of al-Fazr, Journal of Near Eastern
Studies 29: 103123.
Salam, H., and Kennedy, E.S. 1967, Solar and Lunar Tables in Early Islamic Astronomy,
Journal of the American Oriental Society 87: 492497.
As-Saleh, J.A. 1970, Solar and Lunar Distances and Apparent Velocities in the Astro-
nomical Tables of abash al-sib, Al-Abhath 23: 129177; reprinted in Kennedy,
E.S., Studies in the Islamic Exact Sciences, Beirut (1983), pp. 204252.
Sams, J. 1987, Al-Zarql, Alfonso x and Peter of Aragon on the Solar Equation, in David
A. King and George Saliba (eds.), From Deferent to Equant: A Volume of Studies in the
History of Science in the Ancient and Medieval Near East in Honor of E.S. Kennedy,
Annals of the New York Academy of Sciences 500: 467476.
Sams, J. 1992, Las Ciencias de los Antiguos en al-Andalus. Madrid.
Sdillot, J.-J., and Sdillot, L.-A. 1834, Trait des instruments astronomiques des Arabes.
Paris; reprinted Frankfurt 1984.
Stahlman, W.D. 1959, The Astronomical Tables of Codex Vaticanus Graecus 1291. Brown
Univ. Ph.D. Thesis.
Suter, H. 1914, Die astronomischen Tafeln des Muammad ibn Ms al-Khwrizm.
Copenhagen.
Toomer, G.J. 1968, A Survey of the Toledan Tables, Osiris 15: 5174.
Toomer, G.J. 1969, A History of Errors, Centaurus 14: 306336.
Toomer, G.J. 1987, The Solar Theory of Az-Zarql: An Epilogue in David A. King and
George Saliba (eds.), From Deferent to Equant: A Volume of Studies in the History of
Science in the Ancient and Medieval Near East in Honor of E.S. Kennedy, Annals of the
New York Academy of Sciences 500: 513519.
Vernet, J. 1949, Un tractact dobstetricia astrolgica, Boletn de la Real Academia de
Buenas Letras de Barcelona 22: 6996; reprinted in Vernet, J., Estudios sobre Historia
de la Ciencia Medieval, BarcelonaBellaterra (1979), pp. 273300.
226 chapter 7

Vernet, J. 1956, Las Tabulae Probatae, Homenaje a Mills Vallicrosa, Barcelona, ii: 501
522; reprinted in Vernet, J., Estudios sobre Historia de la Ciencia Medieval, Barcelona
Bellaterra (1979), pp. 191212.
chapter 8

Early Alfonsine Astronomy in Paris:


The Tables of John Vimond (1320)*

It has been clear for many years that medieval European astronomy in Latin
was heavily dependent on sources from the Iberian peninsula, primarily in
Arabic, but also in Hebrew, Castilian, and Catalan. The Castilian Alfonsine
Tables, compiled by Judah ben Moses ha-Cohen and Isaac ben Sid under the
patronage of Alfonso x (d. 1284), were an important vehicle for the transmission
of this body of knowledge to astronomers north of the Pyrenees, but the details
of this transmission remain elusive, in part because only the canons to these
tables survive (see Chabs and Goldstein 2003a). In this paper we build on
our preliminary studies of a figure who previously had barely been mentioned
in the recent literature on medieval astronomy (Chabs and Goldstein 2003a,
pp. 267277, and 2003b). John Vimond was active in Paris ca. 1320 and, as we
shall see, his tables have much in common with the Parisian Alfonsine Tables
(produced by a group in Paris, notably John of Murs and John of Lignres), but
differ from them in many significant ways. As far as we can tell, there is no
evidence for any interaction between Vimond and his better known Parisian
contemporaries and in our view the best hypothesis is that they all depended
on Castilian sources. As a result of our analysis, we are persuaded that Vimonds
tables are an intelligent reworking of previous astronomical material in the
Iberian peninsula to a greater extent than is the case for the Toledan Tables
(compiled in Toledo about two centuries before the Castilian Alfonsine Tables).
It is most likely that Vimonds principal source was the Castilian version of the
Alfonsine Tables.
Paris, Bibliothque nationale de France, ms lat. 7286c is a 14th-century
manuscript containing an unusual set of tables (ff. 1r8v) as well as the canons
and tables of 1322 by John of Lignres (ff. 9r58r). In a brief text at the end of
the first set of tables they are attributed to John Vimond (Iohannes Vimundus),
an astronomer who compiled them for the use of students at the University of
Paris (f. 8v):

* Suhayl 4 (2004), 207294.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_010


228 chapter 8

Et in hoc terminatur opus Iohannis Vimundi baiocensis dyocesis de dis-


posicionibus planetarum et stellarum fixarum; et cum istis sequitur de hiis
que per ipsum ordinantur ad conversionem temporum verorum et equalium
sociatorum, et de disposicionibus eclipsalibus solis et lune sibi pertinentibus,
et de aliis disposicionibus ipsorum et aliorum corporum celestium, ad utili-
tatem scolarium universitatis parisiensis et omnium aliorum.

Here ends the work by John Vimond of the diocese of Bayeux on the
dispositions of the planets and the fixed stars; () and on the dispositions
of solar and lunar eclipses and [other syzygies] corresponding to them,
and on the other dispositions of these and other celestial bodies, for the
use of students at the University of Paris and all others.

The complete set of Vimonds tables are uniquely extant in this manuscript,
and no canons for them have been identified. They are a coherent set of tables
with all the elements needed to compute the positions of the celestial bodies,
much in the tradition of the Arabic zijes and their derivatives. The exact date of
composition of Vimonds tables is not given in the text, but they were probably
produced shortly before 1320. In the paragraph preceding his tables, Vimond
tells us that they were compiled for Paris with 1320 as epoch (f. 1r: see below) and
this date is confirmed by recomputation. These tables also include a calendar
with the dates of syzygies: this strongly suggests that they were constructed
prior to the year of the calendar because the astronomical information would
no longer be of any use after the year had passed. However, the calendar poses
special problems which will be discussed below.
Vimonds only other known work is a short treatise on the construction
of an astronomical instrument, extant in Erfurt, ms ca 2 377 (ff. 21r22r),
beginning Planicelium vero componitur ex eis que sunt diversorum operum ,
and ending Explicit tractatus johannis vimundi in a manuscript containing
various works by other Parisian astronomers such as John of Murs and John of
Lignres (Thorndike and Kibre 1963, col. 1050; Saby 1987, pp. 471, 474).
John Vimond and his works were seldom mentioned by his contempo-
raries. However, in Vatican, Biblioteca Apostolica, ms Ottob. lat. 1826, we are
told that John of Spira (14th century), the author of a commentary on John
of Lignress canons (Thorndike and Kibre 1963, col. 204), composed his own
canons to several of Vimonds tables (for a description of this manuscript, espe-
cially ff. 148153, see F.S. Pedersen 2002, p. 177). This manuscript includes a
text that begins on f. 148ra ascribed to a certain M.J.C., Canon tabulae sequen-
tis quae intitulatur tabula motus diversi solis et lunae in una hora et semidi-
ametrorum secundum tabulas Alfonsi, at the end of which John Vimond is
early alfonsine astronomy in paris 229

mentioned. On the other hand, Vimond is not mentioned in Madrid, Bib-


lioteca Nacional, ms 4238, a manuscript containing a few tables that can be
attributed to him, as well as a copy of the Parisian Alfonsine Tables computed
for Morella (in the province of Valencia) for the years 1396 and 1400 (Chabs
2000).
As far as we can tell, John of Murs and John of Lignres do not refer to
Vimond at all in any of their numerous works, but it seems implausible that
they did not know him or his work which was addressed to the students at the
University of Paris. Indeed, there were not so many competent astronomers
working in Paris around 1320 and both Vimond and Murs came from the
same region, Normandy, from places about 70km apart, Bayeux and Lisieux,
respectively.
We would expect Vimond to be well known and frequently cited by practi-
tioners of astronomy, for he is named as one of the outstanding astronomers
of his time by Simon de Phares in his Recueil des plus celebres astrologues
(14941498), a chronologically ordered list with comments, edited by Boudet
(19971999, 1:467). In fact, Vimond is mentioned before John of Lignres, John
of Saxony, John of Janua, and John of Murs:

Maistre Jehan Vymond fut a Paris, homme moult singulier et grant astrolo-
gien, lequel eut en ce temps grant cours pour la science des estoilles. Entre
ses euvres, fist une verifficacion de la conjunction des lu[mi]naires, aussi des
eclipses et estoilles fixes pour plusieurs ans. Cestui predist les grans vens qui
furent en son temps et fist plusieurs beaulx jugemens, dont il acquist grant
loz et renommee en France et fut moult devost en Nostre Seigneur.

Master John Vimond lived in Paris, a most singular man and a great
astrologer, who had at that time much prestige because of (his knowledge
of) the science of the stars. Among his works is a verification of the
conjunction of the luminaries, as well as eclipses and the fixed stars, for
many years. He predicted the great winds which took place in his time
and made many fine judgments for which he acquired great praise and
renown in France and he was most devoted to our Lord.

The verifficacion de la conjunction des lu[mi]naires refers to Vimonds tables.


These tables are arranged very differently from those of his Parisian contempo-
raries and are based, in part, on parameters that probably came from the Castil-
ian Alfonsine Tables or a tradition closely associated with them. Of special
interest is the proper motion of the solar and planetary apogees, a feature pre-
viously unknown in medieval tables produced outside Spain and North Africa.
230 chapter 8

We are convinced that Vimonds tables provide an indication of the arrival in


Paris of new astronomical material coming from Castile, in the sense that they
propose new approaches to replace those based on the Toledan Tables and
developed at the end of the 13th century by astronomers working in Paris such
as Peter Nightingale, Geoffreoy of Meaux, and William of St.-Cloud. Further,
we believe that Vimonds tables are prior to, and independent of, the tabular
work developed in the early 14th century, which we call the Parisian Alfon-
sine Tables, by the group of Parisian astronomers that included John of Murs
and John of Lignres, which were also based on Castilian sources. Vimonds
tables and the Parisian Alfonsine Tables have many parameters in common
both for mean motions and equations. In principle, it is possible that one set of
tables depended on the other, but the differences between them suggest to us
that it is far more likely that they depended on a common source. Moreover, if
Vimond composed his tables prior to 1320, he did so before any datable text of
the Parisian Alfonsine Tables.
A description and analysis of Vimonds tables follow.

f. 1r The first numerical information given in this set of tables is the radix
for mean conjunctions of the Sun and the Moon: 13;54,54d. In modern ter-
minology, the initial time for a set of tables is called its epoch whereas its
radices are the positions of the Sun, Moon, planets, etc., at that time. The
medieval convention, however, is to use radix for both the time and the posi-
tion.
We are convinced that the author refers to the time, in Paris, of the mean
conjunction on March 10, 1320. The year and the place are mentioned by
Vimond himself in a short paragraph following the numerical value of the radix
(f. 1r):

Et est intelligendum quod ista radix mediarum coniunccionum sit imme-


diate post 19 secunda diei que consistunt immediate post () lucis beati
Mathie composite procedendo ab ortu solis usque ad occasum scilicet anno
domini nostri Ihesu Christi 1320 secundum numeracionem annorum roma-
norum qui incipiunt ex inicio diei circoncisionis domini nostri Ihesu Christi
et existentis ad longitudinem civitatis Parisius que distat a medio mundi per
49 g et 30 min ita quod illa civitas est in parte occidentali et etiam distat ab
illo medio per 8 min et 15s diei equalis.

Note that this radix for the mean conjunctions comes immediately after
19 seconds of a day that fall immediately after the () [space for one
word; illegible] (day)light of Saint Matthew, proceeding from sunrise to
early alfonsine astronomy in paris 231

sunset, namely, in the year of our Lord Jesus Christ 1320 according to the
count of Roman years which start from the beginning of the Day of the
Circumcision of our Lord Jesus, for the longitude of the city of Paris which
is distant from the middle of the world by 49 degrees and 30 minutes
because that city is in the western direction and distant from that middle
by 8 minutes and 15 seconds of an equinoctial day.

Madrid, Biblioteca Nacional, ms 4238, f. 66v, has a short text which is very
similar to the paragraph on f. 1r in Paris, Bibliothque nationale de France, ms
lat. 7286c:

Radix coniunccionum: dies 13 m 54 2 54


Radix opposicionum: dies 14 m 45 2 55
Hec radix est post 19 2 diei que sunt post meridiem diei Mathie anno 1320
secundum romanos. Nota quod annus in meridie diei mathie 25 dies bisexti
erit semper ultimus dies anni. Iste tabule radicum sunt facte Parisius ad
meridiem cuius cenith distat ab equinocciali 49 g 30 m vel 8 m 15 2 diei.

The numerical datum, 0;0,19d (= 0;7,36h), represents the equation of time for
that day. In the Madrid version, the radix of the opposition, 14;45,55d, is half
the length of a mean synodic month which is about 29;31,50d, and this is the
entry for the first opposition in Vimonds calendar (see below). It is clear that,
according to the version of this text in the Paris manuscript, the civil day in
the calendar includes the period of daylight, that is, the time from sunrise
to sunset, in contrast to the astronomical day that goes from noon to the
following noon. Both values given for the longitude of Paris from Arin, called
the middle of the world, are equivalent. Arin, a corruption of Ujjain (a city
in India), was thought to be halfway between the eastern and western limits
of the world (Neugebauer 1962, p. 11, n. 2). The distance from Arin to Toledo
was taken to be 61;30 and, since Paris was generally said to be 0;48h or 12 to
the east of Toledo, its longitude from Arin is 49;30, as in the passage above
(Mills 19431950, p. 49; Kremer and Dobrzycki 1998, p. 194; and F.S. Pedersen
2002, p. 431). Moreover, when a day is taken to be 360, it follows that 49;30
corresponds to 0;8,15d, for 49;30/360 = 0;8,15. The expression in the Madrid
manuscript, Parisius ad meridiem cuius cenith distat ab equinocciali is a
corrupt version of the better reading in the Paris manuscript, for it would imply
that 49;30 is the latitude of Paris, but then its equivalence to 0;8,15d would
become meaningless. We also note that, according to this text, Vimonds tables
were computed for Paris whereas in the early 1320s other Parisian astronomers
who recast the Alfonsine Tables computed them for Toledo, as is the case for the
232 chapter 8

table 1 Mean conjunctions ( f. 1r)

[years] [excess] [years] [excess] [years] [excess]


d d d

1 18;53,52 28 20; 6,54 64 12;13,41


2 8;15,53 32 6; 6,50 68 27;45,27
3 27; 9,45 36 21;38,37 72 13;45,23
4 15;31,46 40 7;38,33 76 29;17,10
8 1;31,43 44 23;10,19 152 29; 2,29
12 17; 3,29 48 9;10,15 304 28;33, 8
16 3; 3,25 52 24;42, 2 608 27;34,26
20 18;35,12 56 10;41,58 1216 25;47, 1
24 4;35, 8 60 26;13,44 2432 21;43,12

tables with epoch 1321 by John of Murs (see, e.g., Lisbon, Biblioteca de Ajuda,
ms 52-xii-35).
According to our computations based on the Parisian Alfonsine Tables, the
mean conjunction on March 10, 1320 took place in Toledo at 9;10h, civil time
(i.e., counting from midnight) which, with a correction of 0;48h, is 9;58h in Paris
(civil time), that is, 2;2h before noon. Thus, the radix for the tables on f. 1r (as
well as the radices for the planetary tables, as will be seen later) is the time of
the first mean conjunction in March 1320 (March 10, 1320, at 9;58 a.m., Paris,
or March 9, 1320, 21;58h, Paris, counting from noon). Indeed, the sexagesimal
part of the radix (0;54,54d) is exactly the sum of 12h and 9;58h. The integer
part of the radix, as will be explained later in reference to the annual calendar
presented on this same folio, is counted from the epoch of the calendar, almost
14 days before the mean conjunction of March 10, 1320, that is, February 25, 1320
or February 24b, 1320, where 24b represents the second day called February
24 in a leap year (such that the last day of February is always day 28 both in
ordinary and leap years).

f. 1r Table 1: Mean Conjunctions

The entries in this table give the instant of the first mean conjunction after a
certain number of years. We are given entries for 1, 2, 3, and 4 years; for multiples
of 4 years up to 76 (= 19 4) years; and for 152, 304, 608, 1216, and 2432 years.
The entries represent the excess of days after an integer number, n, of synodic
early alfonsine astronomy in paris 233

months have elapsed (where n = 13 for year 1, , and 30,081 for year 2432).
Madrid, Biblioteca Nacional, ms 4238, f. 66v, reproduces this table except for
the last row for 2432 years, which is missing.
The value for the mean synodic month derived from year 2432 is 29;31,50,7,
44,35d 0;0,0,0,0,4d. Thus, for year 1: 13 29;31,50,7,44,35d 365d = 18;53,52d,
in agreement with the tabulated value. In the Parisian Alfonsine Tables, the
mean synodic month is 29;31,50,7,37,27,8,25d: this value is found, for example,
in Lisbon, Biblioteca de Ajuda, ms 52-xii-35, f. 16v, containing the tables for
epoch 1321 by John of Murs. So Vimonds parameter is very similar to, but not
identical with, the parameter in the Parisian Alfonsine Tables.

Vimond Parisian Alf. t.

Mean synodic month 29;31,50,7,44,35d 29;31,50,7,37,27,8,25d

f. 1r Table 2: Annual Calendar with Syzygies

This annual calendar begins on the day of Saint Matthew (February 24) and
lists the dates associated with several saints, as well as the dates and times of 25
consecutive mean syzygies. The practice of adding the extra day in a leap year
after Feb. 24 goes back to the Roman calendar as revised by Julius Caesar, when
the additional day followed Feb. 24 and was called bis-sextus ante calendas
martias (the sixth day before the calends of March). In a leap year February
lasted 29 days, but the last day was numbered 28, for the 24th was assigned
to two consecutive days. This is what is intended in Vimonds calendar where
the year begins on that very day. We know of no other calendar in the late 13th
century or early 14th century beginning on Feb. 24; in particular, the calendars
composed by Geoffreoy of Meaux and William of St.-Cloud do not begin on
that day (Chabs and Goldstein 2003a, pp. 245247). It is worth noting that
Vimonds calendar which lists mean syzygies together with saints days is in the
tradition of these two astronomers who were active in Paris shortly before him:
they displayed planetary data in calendars and depended on the Toledan Tables
for their computations. We also note that the feast of St. Matthew is mentioned
in the canons to the Parisian Alfonsine Tables by John of Saxony as the last day
in a leap year (see Poulle 1984, p. 36, line 41). Vimond offers no explanation for
basing his calendar and his tables on syzygies; we can only conjecture that he
was being faithful to some unknown source.
234 chapter 8

table 2 Annual calendar with syzygies ( f. 1r)

(1) (2) (3)


[Saints day / No. syzygy] [date] [time since epoch]
d

Romanus February 28 4
Perpetua virgo March 7 11
1 Opposition March 10 14;45,55
Gregorius papa 12 16

2 Conjunction March 25 29;31,50

3 Opposition April 9 44;17,45

4 Conjunction April 24 59; 3,40

5 Opposition May 8 73;49,35

6 Conjunction May 23 88;35,30

7 Opposition June 7 103;21,25

8 Conjunction June 22 118; 7,21

9 Opposition July 6 132;53,16

10 Conjunction July 21 147;39,11

11 Opposition August 5 162;25, 6

12 Conjunction August 20 177;11, 1

13 Opposition September 3 191;56,56

14 Conjunction September 18 206;42,51

15 Opposition October 3 221;28,46

16 Conjunction October 18 236;14,41


early alfonsine astronomy in paris 235

(1) (2) (3)


[Saints day / No. syzygy] [date] [time since epoch]
d

17 Opposition November 2 251; 0,36

18 Conjunction November 14 265;46,31

19 Opposition November 31 280;32,26

20 Conjunction December 16 295;18,21

21 Opposition December 31 310; 4,16


Circoncisio domini Ihesu
Christi initium anni [January 1] 311

22 Conjunction January 14 324;50,11

23 Opposition January 29 339;36, 6

24 Conjunction February 13 354;22, 2


Iuliana virgo 16 357
Petrus ad cathedram 22 363
25 Opposition February 28 369; 7,57

In Table 2, columns 1 and 2 have no heading, but column 3 has the heading
days, minutes, and seconds. In the manuscript the name of the month is usu-
ally given in col. 2, and occasionally in col. 1 which has about 90 entries such
as: Annunciatio Domini, Dyonisius, Lucas Evangelista, Innocentes, etc. The syzy-
gies are numbered from 1 to 25, and they are transcribed above. The numbers
in column 3 are integers when a saints day is meant and indicate the number
of days that elapsed since the epoch (day 1) of the calendar, that is, February
25, 1320 (Julian) or what we have called February 24b, 1320. Vimond seems
to use here civil days (from midnight to midnight) rather than astronomical
days (from noon to noon), which makes sense in a calendar. When a conjunc-
tion or an opposition is indicated, we would expect the number in column 3
to refer to the accumulated time from the radix (the conjunction on March
236 chapter 8

10, 1320) in multiples of half a mean synodic month, i.e. 14;45,55d but, in fact,
we are given the accumulated time from the mean syzygyan opposition
immediately preceding the radix, which occurred on February 24, 1320. If this
was the authors intention, it is not clear how the user of these tables was to
take account of the radix given at the beginning of them. Moreover, despite
the coherence of the arithmetic in this calendar, something is seriously wrong
with it, for we find the word oppositio next to March 10, when a conjunction took
place, and the word coniunccio next to March 25, when an opposition occurred.
The same pattern is followed throughout the calendar. There is an explana-
tory note on f. 1r concerning the calendar, but we were unable to make sense
of it.
Year 1324 might be considered as an alternative date for the calendar for,
according to computations with the Parisian Alfonsine Tables, a mean oppo-
sition occurred on March 10 (counting from noon) or on March 11 (counting
from midnight). This would conform with the character of the syzygy men-
tioned in the calendar, and the computations associated with this date yield
results that are quite close to (but not exactly the same as) the information
given in the text. Indeed, in our preliminary discussion of these tables, this
near agreement mislead us to think that 1324 was the radix of Vimonds tables
(Chabs and Goldstein 2003a, p. 270). However, as indicated previously, year
1320 is specifically mentioned, and it fits much better with the radix of mean
conjunctions and with the radices for planetary positions displayed on ff. 1v and
4r.

f. 1v Radices for the argument of solar anomaly, the argument of lunar anomaly
(henceforth, solar and lunar anomaly, respectively), the solar apogee, and the
lunar ascending node:

Solar anomaly 8s 26;14,33


Lunar anomaly 1s 3; 6,14
Solar apogee 2s 29;56,15
Ascending node 10s 13;14,43

Note the use of zodiacal signs of 30, a characteristic of all tables in this set.
A short text below these parameters explains that the radices for the motion
of the solar apogee and the ascending node are counted from the beginning
of Aries on the 9th sphere, indicating that tropical coordinates are used here.
These radices were calculated for March 10, 1320, at the time of the mean con-
junction of the Sun and the Moon. According to the Parisian Alfonsine Tables
the solar apogee for March 10, 1320 is 89;23,50, a value which differs by about
early alfonsine astronomy in paris 237

half a degree from the entry in the text. Both values in turn differ from the solar
apogee for 1320 in the tables for 1322 by John of Lignres (89;24,22) that is found
by adding two values given on f. 9v of this same manuscript: the solar apogee
(81;7,15,39) and the motion of the 8th sphere at that time (8;17,6,48). The
same result, 89;24,22, can also be found in another copy of John of Lignress
tables, Erfurt, ms ca q 362, f. 21ra. For the rest of the radices, recomputation with
the Parisian Alfonsine Tables for the epoch, March 9, 1320, at 21;10h in Toledo
(counting from noon), yields results which are very close to the values in the
text, especially for the Moon:

Vimond Parisian Alf. t.

Solar anomaly 266;14,33 266;47, 0


Lunar anomaly 33; 6,14 33; 6,28
Solar apogee 89;56,15 89;23,50
Ascending node 313;14,43 312;54,39

The solar longitude is the sum of the solar anomaly and the solar apogee:

Vimond Parisian Alf. t.

Solar longitude 356;10,48 356;10,50

and again the agreement is very good. Since this is the time of a mean con-
junction, the mean lunar longitude will be equal to the mean solar longitude.
According to the Parisian Alfonsine Tables, the mean lunar longitude at this
epoch was 356;11,3, i.e., it differed from the mean solar longitude by only 0;0,13
(note that the Moon travels this distance in about 20 seconds of time which is
below the accuracy of 1 minute for the time of mean conjunction). Hence the
absence of a radix for lunar mean motion simply reflects the fact for the epoch
of Vimonds tables the mean longitude of the Moon is the same as the mean
longitude of the Sun.
The agreement for the radix of lunar anomaly to the minute is particularly
impressive since the motion in lunar anomaly is about 0;30/h. Lunar anomaly
is not subject to precession and it is independent of solar motion. (We use the
term precession for a constant motion of the eighth sphere, and trepidation for a
238 chapter 8

table 3 Yearly radices ( f. 1v)

Year 1 Year 2 Year 3


s () s () s ()

Solar anomaly 0 18;22, 3 0 7;37,47 0 25;59,49


Lunar anomaly 11 5;37, 8 9 15;25,15 8 21; 2,22
Solar apogee 0 0; 1,12 0 0; 2,18 0 0; 3,30
Ascending node 11 9;43,23 10 20;58,27 10 0;40,50

Year 8 Year 608


s () s ()

Solar anomaly 0 1;24,48 0 20; 6, 6


Lunar anomaly 1 5;51,58 4 8;22, 2
Solar apogee 0 0; 9, 7 0 11;32,36
Ascending node 6 25;27,27 4 29;26,44

variable motion of the eighth sphere.) So, even though Vimond and the authors
of the Parisian version of the Alfonsine Tables differ on matters of definition
and made slight changes in mean motions, it is unlikely that either of them
would change the motion in anomaly significantly from what it had been in
their common source.

f. 1v Table 3: Yearly Radices

This table displays the radices for the solar anomaly, the lunar anomaly, the
solar apogee, and the lunar ascending node for 1, 2, 3, and 4 years; for multiples
of 4 years up to 76 years; and then for 152, 304, 608, 1216, and 2432 years, as in
Table 1. A selection of the entries is displayed in Table 3.
The entries for year 1 represent the progress made by the Sun, the Moon,
the solar apogee, and the lunar node in a year of 13 mean syzygies of the same
kind (henceforth lunations) of 29;31,50,7,44,35d. To be sure, the difference
between the solar anomalies for year 2 and year 1 is 349;15,44, meaning that
year 2 contains 12 lunations, for 349;15,44/ (29;31,50 0;59, 8, ) = 12, whereas
the difference between year 3 and year 2 is 378;22,2 (the same value as that
early alfonsine astronomy in paris 239

for year 1), indicating that year 3 contains 13 lunations (378;22,2/ (29;31,50
0;59, 8, ) = 13), as is the case with year 1. With this procedure, we see that 50
is the total number of lunations in the first 4 years, 99 in the first 8 years, ,
7,521 in the first 608 years, and so on. That is, for Vimond 1 year is equivalent
to 13 mean lunations, 2 years is equivalent to 25 mean lunations, etc. Where
possible, we have derived the associated mean motions from the entries for
year 608, because those for years 1216 and 2432 are not completely legible in
the manuscript.
The mean motion in solar anomaly resulting from the entry for year 608
(0s 20;6,6), that is, after 7,521 lunations, and the length of the synodic month
obtained before (29;31,50,7,44,35d), is 0;59,8,8,23,30/d, for

(608 360 + 20;6,6)/(7521 29;31,50,7,44,35d)


= 0;59,8,8,23,30/d.

This daily mean motion implies a year length of 365;15,42,32d which is side-
real. In the Parisian Alfonsine Tables, however, the length of a sidereal year
is variable, and the fixed length of the tropical year is 365;14,33,9,57, d (=
360/0;59,8,19,37,19,13,56/d).
Similarly, the mean motion in lunar anomaly can be computed from the
entry corresponding to 7,521 lunations (year 608: 4s 8;22,2), for 7521 luna-
tions corresponds to 8060 complete revolutions in anomaly with an excess of
about 120 (computed with approximate values for the appropriate parame-
ters). Hence, with the data in the text, the mean motion in lunar anomaly
is

(8060 360 + 128;22,2)/(7521 29;31,50,7,44,35d)


= 13;3,53,57,27,11/d,

in very good agreement with the corresponding value in the Parisian Alfonsine
Tables (13;3,53,57,30,21/d); the difference only accumulates to 1 in well over
10,000 years.
As for the motion of the solar apogee derived from the entry corresponding
to 7,521 lunations (year 608: 0s 11;32,36), we find 0;0,0,11,13,35/d. By the same
reasoning, the mean motion of the lunar ascending node resulting from the
entry of year 8 in the table (6s 25;27,27) is 0;3,10,18,6,48/d, in contrast to the
value found in the Parisian Alfonsine Tables (0;3,10,38,7,14,49,10/d). In this
case, the entry in the manuscript for 608 years is corrupt.
We note that Vimonds value for the motion of the solar apogee includes
precession as well as its proper motion for, if we add the value for the mean
240 chapter 8

Vimond Parisian Alf. t.

Year length 365;15,42,32d 365;14,33, 9,57d


Solar anomaly 0;59, 8, 8,23,30/d 0;59, 8,19,37,19/d
Lunar anomaly 13; 3,53,57,27,11 13; 3,53,57,30,21
Solar apogee 0; 0, 0,11,13,35
Ascending node 0; 3,10,18, 6,48 0; 3,10,38, 7,14

motion in solar anomaly (which is sidereal) to the motion of the solar apogee,
we find 0;59,8,19,37,4/d, in close agreement with the corresponding value
of the mean motion in solar longitude (tropical) in the Parisian Alfonsine
Tables. In the Almagest, the planetary apogees are sidereally fixed whereas
the solar apogee is tropically fixed. In the 9th century astronomers in Bagh-
dad fixed the solar apogee sidereally so that it too was subject to precession
(or trepidation). But in the 11th century Azarquiel realized that the solar
apogee had a proper motion in addition to precession, and fixed its amount
as 1 in 279 Julian years or about 0;0,0,2/d (Chabs and Goldstein 1994, p.
28). In one Andalusian tradition, this proper motion of the solar apogee was
applied to the planetary apogees as well (see Sams and Mills 1998, p. 269;
cf. Mestres 1996, pp. 394395). If we take al-Battns value for precession of
1 in 66 years or about 0;0,0,9/d and add it to the proper motion of the
solar apogee, the result is about 0;0,0,11/d. There is no hint of this proper
motion for either the solar apogee or the planetary apogees in the Parisian
Alfonsine Tables where these apogees are all sidereally fixed and, instead of
precession, the Parisian Alfonsine Tables have tables for trepidation; hence,
there is nothing in those tables with which to compare directly the motion
of the solar apogee in Vimonds tables. We see, then, that the parameters in
Vimonds tables are not identical with those in the Parisian Alfonsine Tables,
and some of these parameters (e.g., the length of the solar year) are defined
differently.

ff. 1v2r Table 4: Monthly Radices

This table displays the radices for the solar anomaly, the lunar anomaly, the
solar apogee, and the lunar ascending node for 25 consecutive syzygies after
the corresponding integer numbers of semi-lunations have elapsed. An excerpt
is shown in Table 4.
early alfonsine astronomy in paris 241

table 4 Monthly radices ( ff. 1v2r)

Syzygy 1 Syzygy 2 Syzygy 25


s () s () s ()

Solar anomaly 0 14;33, 9 0 29; 6,19 0 3;48,53


Lunar anomaly 6 12;54,30 0 25;49, 1 4 22;42,27
Solar apogee 0 0; 0, 3 0 0; 0, 6 0 0; 1, 9
Ascending node 11 29;13,10 11 28;26,20 11 10;29,13

The entries represent the progress made by the Sun, the Moon, the solar
apogee, and the lunar node in 1, 2, , 25 mean semi-lunations of 29;31,50,7,44,
35d/ 2 = 14;45,55,3,52,17d. The entries in this table agree with those in Table 3,
for in each case the value for 26 consecutive semi-lunations (the sum of the
entries for Syzygy 1 and Sygygy 25 in Table 4) equals the value for 13 lunations
(year 1 in Table 3).

f. 2r Table 5: Sun

This table in 5 columns is original in presentation. Column 1 gives the argument


(argumentum) at 3-intervals in signs and degrees from 0s 3 to 12s 0; this
is the mean solar anomaly. Column 2 displays the true solar anomaly (motus
completus) in signs, degrees, and minutes. Column 3 (motus gradus) displays
the increment in true anomaly per degree of the argument. Column 4 gives
the solar velocity, in units of minutes and seconds of arc in a minute of a day
(minutum diei), i.e., in a sixtieth of a day. Column 5 displays the time (also called
argumentum), in days, with sexagesimal fractions of a day, that the Sun takes
to complete the arc indicated in column 1.
To obtain an entry in column 5 multiply the corresponding entry in column 1
by the daily mean motion in solar anomaly; the entry for 360 (365;15,42d)
represents the length of the sidereal year, in good agreement with the value
deduced from 99 mean synodic months in Table 3.
As shown in Table 5a, the difference between the argument (col. 1) and the
true anomaly (col. 2) represents the solar equation, with a maximum of 2;10
as in the Parisian Alfonsine Tables. To emphasize the solar equation, we have
added a third column for the differences between entries in columns ii and i,
labeled ii i.
242 chapter 8

table 5 Sun ( f. 2r)

(1) (2) (3) (4) (5)


argum. motus c. motus g. min. diei argum.
s () s () min. min. d

0 3 0 2;54 57;51 0;57 3; 2,38


0 6 0 5;47 57;51 0;57 6; 5,16

2 27 2 24;51 59;47 0;59 88;16,18


3 0 2 27;50 59;59 0;59 91;18,55
3 3 3 0;50 60; 3 0;59 94;21,33
3 6 3 3;50 60;16 0;59 97;24,11
3 9 3 6;51 60;19 0;59 100;26,49

5 27 5 26;53 62;24 1; 1 179;35,13


6 0 6 0; 0 62;24 1; 1 182;37,51
6 3 6 3; 7 62;22 1; 1 185;40,29

8 21 8 23; 9 60;16 0;59 264;48,53


8 24 8 26;10 60; 3 0;59 267;51,31
8 27 8 29;10 59;59 0;59 270;54, 9
9 0 9 2;10 59;47 0;59 273;56,46
9 3 9 5; 9 59;43 0;59 276;59,24

11 27 11 27; 6 57;51 0;57 362;13, 4


12 0 12 0; 0 57;51 0;57 365;15,42

table 5a The solar equation embedded in Table 5

i ii ii i
argumentum motus completus
s () s () ()

0 3 0 2;54 0; 6
0 6 0 5;47 0;13

2 27 2 24;51 2; 9
3 0 2 27;50 2;10
early alfonsine astronomy in paris 243

i ii ii i
argumentum motus completus
s () s () ()

3 3 3 0;50 2;10
3 6 3 3;50 2;10
3 9 3 6;51 2; 9

5 27 5 26;53 0; 7
6 0 6 0; 0 0; 0
6 3 6 3; 7 0; 7

8 21 8 23; 9 2; 9
8 24 8 26;10 2;10
8 27 8 29;10 2;10
9 0 9 2;10 2;10
9 3 9 5; 9 2; 9

11 27 11 27; 6 0; 6
12 0 12 0; 0 0; 0

The entries for the solar equation are not explicit in Vimonds table; they can
be graphed as a smooth curve but they do not allow us to decide which specific
table for the solar equation he used. The reason is that Vimonds entries are
only given to minutes in contrast to most other tables in which the maximum
equation is 2;10,0 where entries are given to seconds, and rounding those
values produces Vimonds entries.

ff. 2v3r Table 6: Moon

This table has the same format as Table 5. An excerpt is displayed in Table 6.
Column 1 gives the argument (argumentum) at 1-intervals in signs and
degrees from 0s 1 to 6s 0 and its complement in 360 from 6s 0 to 11s
29. For columns 2, 3, and 4, one enters with the mean argument of lunar
anomaly, whereas for column 5 one enters with the argument of lunar latitude.
Column 2 displays the lunar equation of center (motus completus) in degrees
and minutes with a maximum of 4;56 as in the Parisian Alfonsine Tables.
244 chapter 8

table 6 Moon ( ff. 2v3r)

(1) (2) (3) (4) (5)


argumentum motus c. motus min. min. diei latitud.
s () s () () sec. min. ()

0 1 11 29 0; 5 5 12; 9 0; 5,13
0 2 11 28 0;10 5 12; 9 0;10,27

2 29 9 1 4;54 0 13; 4 4;59,58


3 0 9 0 4;55 0 13; 5 4; 0, 0*
3 1 8 29 4;55 0 13; 6 4;59,58
3 2 8 28 4;56 0 13; 8 4;59,50
3 3 8 27 4;56 0 13; 9 4;59,35
3 4 8 26 4;56 0 13; 9 4;59,15
3 5 8 25 4;56 0 13;11 4;58,51
3 6 8 24 4;56 0 13;13 4;58,21
3 7 8 23 4;56 0 13;14 4;57,45
3 8 8 22 4;55 0 13;15 4;57, 4

5 29 6 1 0; 6 6 14;25 0; 5,13
6 0 6 0 0; 0 6 14;25 0; 0, 0

* Sic, instead of 5;0,0.

Column 3 (motus minuti) displays the line-by-line differences in column 2


divided by 60 (for purposes of interpolation). Column 4 gives the lunar velocity,
in minutes and seconds, in a minute of a day (minutum diei). The minimum
corresponds to 0;30,23/h and the maximum to 0;36,3/h: for a comparison with
other tables for lunar velocity, see Goldstein 1996. Column 5 displays the lunar
latitude, with a maximum of 5;0,0 as in the Parisian Alfonsine Tables and the
Almagest. It is surprising that the expression motus completus is used here for
the lunar equation of center, whereas in Table 5 it was used for the true solar
anomaly; clearly, it has a range of meanings and cannot be translated by a single
expression.
early alfonsine astronomy in paris 245

table 7(1) True syzygies ( f. 3r)

* 11 [diff.] 12 [diff.] 13 [diff.] 14


() () () () ()

1 1;31 7 1;24 7 1;17 5 1;12


2 3; 2 15 2;47 13 2;34 11 2;23
3 4;33 23 4;10 19 3;51 16 3;35
4 6; 4 31 5;33 25 5; 8 22 4;46
5 7;35 38 6;57 32 6;25 28 5;57
6 9; 6 45 8;20 38 7;42 33 7; 9
7 10;36 53 9;43 45 8;58 38 8;20

* In the ms, gradus velocitatis appears above this column but it refers to the headings
of the other columns, labeled: 11, 12, 13, 14.

f. 3r Table 7: True Syzygies

There are two subtables for computing the time from mean to true syzygy:
see Tables 7(1) and 7(2). The first subtable is a double-argument table where,
on analogy with the other subtable, the vertical argument seems to be the
elongation between the Sun and the Moon (at 1-intervals from 1 to 7) and the
horizontal argument, the velocity in elongation (i.e., the difference between the
lunar and the solar velocities) in degrees per minute of a day (only four values
for the velocity in elongation are given: 11, 12, 13, and 14).
An entry, e, in this subtable was derived by means of the formula (expressed
in modern notation)

e = 16;40 / [vm(t) vs(t)]

where is the true elongation at mean conjunction (or the result after sub-
tracting 180 at mean opposition), and the velocity in elongation, vm(t) vs(t),
is the difference between the daily velocities of the Moon and the Sun at the
time of mean syzygy. We cannot give a satisfactory explanation for the factor
16;40 (= 100/6) or for the headings of the columns indicating that the entries
are in degrees and minutes (rather than in units of time). Between these four
columns, one finds the differences, in minutes (but labeled seconds), between
two consecutive entries in the same row, to facilitate interpolation.
The second subtable is also a double-argument table giving the time in days
as a function of the elongation (at intervals of 0;1 from 0;1 to 1, or 60 minutes)
246 chapter 8

table 7(2) True syzygies ( f. 3r)

min.* 11 [diff.] 12 [diff.] 13 [diff.] 14


min. sec. min. sec. min. sec. min.

1 5;27 27 5; 0 23 4;37 20 4;17


2 10;55 55 10; 0 46 9;14 40 8;34

9 49; 5 245 45; 0 208 41;32 178 38;34

days min. days min. days min. days

10 0;55 5 0;50 4 0;46 3 0;43

59 5;22 27 4;55 23 4;32 19 4;13


60 5;27 27 5; 0 24 4;37 20 4;17

* In the ms, gradus velocitatis appears above this column but it refers to the headings
of the other columns, labeled: 11, 12, 13, 14.

and the velocity in elongation in degrees per minute of a day (again, only 4
values for the velocity in elongation are given: 11, 12, 13, and 14). Between these
four columns, one finds the differences, in minutes of a day, between successive
entries in the same row, to facilitate interpolation. Some selected rows of this
subtable are displayed in Table 7(2).
The entries in this subtable were computed by means of the formula (ex-
pressed in modern notation)

t = / [vm(t) vs(t)]

where t is the time interval between mean and true syzygy, is the true elon-
gation, and the velocity in elongation, vm(t) vs(t), is the difference between
the daily velocities of the Moon and the Sun at the time of mean syzygy. This
approach to the problem of finding true syzygy was followed by a number of
medieval astronomers and differs from that presented by Ptolemy in Almagest
vi.4 (Chabs and Goldstein 1997, pp. 9396; cf. Kremer 2003, pp. 305329).
Madrid, Biblioteca Nacional, ms 4238, f. 67r, reproduces both subtables ex-
cept that the last row of Table 7(2) corresponds to the argument of 9 min.
early alfonsine astronomy in paris 247

table 8 Correction of the lunar position for each day between syzygies ( ff. 3v4r)

Day 1 Day 2 Day 14


motus c. min motus c. min motus c. min
s () s () s () s ()

0 12 0 10;57 11;53 0 22;50 11;55 6 5;37 14;37


0 24 0 10; 7 12; 1 0 22; 8 12; 6 6 6;41 14;26
1 6 0 9;24 12;11 0 21;35 12;19 6 7;38 14;10

2 0 0 21; 4 12;53

2 24 0 8;10 13;18 6 9;27 13;56

5 18 0 13;37 14;50 0 28;27 14;52 6 5;25 11;49


6 0 0 14;45 14;45 0 29;30 14;41 6 4;27 11;51
6 12 0 15;47 14;35 1 0;22 14;26 6 3;30 11;55

7 18 1 1;38 12;52

8 12 0 18;11 13; 7

9 6 5 29;30 13;28

11 6 0 13;46 11;49 0 25;35 11;41 6 2;17 14;44


11 18 0 12;48 11;48 0 24;36 11;42 6 3;27 14;46
12 0 0 11;51 11;49 0 23;40 11;47 6 4;29 14;45

ff. 3v4r Table 8: Correction of the Lunar Position for Each Day
between Syzygies

This double-argument table displays two columns for each day, from day 1 to
day 14. The days in the horizontal argument refer to the time from conjunction
to opposition. The vertical argument is given at intervals of 12, from 0s 12
to 12s 0. The heading calls it elongatio lune ab auge epicicli and it represents
the mean lunar anomaly at mean syzygy. For each day, the first column gives
the increment in lunar longitude, here called motus completus, in signs and
degrees, to be added to the mean lunar longitude at the preceding mean syzygy,
whereas the second column displays one sixtieth of the differences between
248 chapter 8

successive entries in the same row, here called motus ad minutum diei, and
given in arc-minutes. The entries in the second column thus represent the true
lunar velocity in a minute of a day for that particular day.
As mentioned above, John of Spira composed canons to some of Vimonds
tables. In particular, the canon in Vatican, Biblioteca Apostolica, ms Ottob. lat.
1826, ff. 152v153r, describes the use of Table 8, here entitled Tabula veri loci
lune ad dies datos post mediam coniunccionem vel opposicionem solis et lune.
The canon ends with an explicit reference to John Vimond working in Paris:

Explicit canon tabule sequentis que est una tabularum quas composuit
Magister Johannes Vimondi. Iste autem canon est undecimus canonum
quos composuit magister Johannes de Spira supra tabulas predicti magistri
Johannis Parisius.

On ff. 153v155v we find a copy of Table 8, but in this case the entries in the
second column (the true lunar velocity in a minute of a day) are given to one
sexagesimal place.
We know of only a few similar tables for the same purpose, but the entries in
them differ from those given by Vimond. Erfurt, ms ca 2 388, is a 15th-century
manuscript which, according to Poulle (1973), contains one of the rare copies
of John of Lignres Tabule magne. On ff. 30r32v, there is an expanded version
of Table 8, with the same structure and the same columns. In this case, the
horizontal argument runs from day 1 to day 15 and the column for velocity
gives entries in minutes and seconds per hour which result from the entries
in Vimonds Table 8 by multiplying them by 2;30 (= 60/24) for conversion from
arc to time. Another example is furnished by Levi ben Gerson (d. 1344) who
compiled a double-argument table, based on his own model, for finding the
lunar position between syzygies as a function of the number of days since
syzygy from 1 to 14 and the mean lunar anomaly at 10-intervals from 0 to 350
(Goldstein 1974, pp. 148149, 246254). Yet another such table is found in an
anonymous zij in Hebrew for year 1400: this double-argument table shares the
same structure, but the anomaly is given at intervals of the daily increment
in mean lunar anomaly from day 0 to day 27 (cf. Goldstein 2003, p. 166). The
zij of Judah ben Verga (ca. 1470) also includes a table with the same structure
(Goldstein 2001, pp. 247, 269270).
early alfonsine astronomy in paris 249

f. 4r Radices for the Planets

In a small table, the text gives the following values for the radices of the
planets:

Mercury 11s 16; 6,10


Venus 3s 3;46,55
Mars 3s 15; 9,42
Jupiter 6s 4; 8, 5
Saturn 9s 14; 0,46

When recomputed for the instant of the mean conjunction on March 10, 1320,
these radices that depend on the mean longitudes or mean arguments of
anomaly (henceforth, simply anomaly) confirm the use of this date as epoch.
In the case of the superior planets the radix can be represented by the following
formula:

Rx(planet) = 0 a(Sun)

where 0 is the mean longitude of the planet at epoch, and a(Sun) is the apogee
of the Sun at that time. According to the Parisian version of the Alfonsine
Tables, the mean motions for the superior planets on that day, in Toledo at 9;10
a.m. (= 9;58 a.m. in Paris), counting from midnight, are:

Saturn 13;57, 1
Jupiter 274; 4,20
Mars 195; 5,57

If we subtract the value of the solar apogee for this epoch (89;56,15) given by
Vimond (f. 2r), we obtain:

Saturn 284; 0,46


Jupiter 184; 8, 5
Mars 105; 9,42

in perfect agreement with the radices given in the text. Note that using the
standard Alfonsine value for the solar apogee at that time (89;23,50) yields no
agreement, confirming the authors preference for his value, 89;56,15. The rea-
son for subtracting the solar apogee is that for Vimond the planetary apogees
partake in the motion of the solar apogee.
250 chapter 8

For Venus and Mercury, Vimonds radices can be obtained by adding the
planets anomaly and the solar longitude and subtracting from the sum the
value for the solar apogee at epoch. For Venus we compute according to the
Parisian Alfonsine Tables at Vimonds epoch:

Rx(Venus) = t v + 0(Venus) + 0(Sun) a(Sun)


= 24754;52,55d 0;36,59,27,23,59,31/d + 45;45,55,19 + 356;10,50 89;56,15
= 93;46,58

where t, the time from epoch Alfonso to epoch Vimond, is 24754;52,55d; v, the
mean motion in anomaly for Venus, is 0;36,59,27,23,59,31/d; 0(Venus), the
radix for Venuss mean anomaly at Alfonsos time, is 45;45,55,19; 0(Sun), the
mean longitude of the Sun at Vimonds epoch, is 356;10,50; and a(Sun), the
solar apogee at Vimonds epoch, is 89;56,15. This result, 93;46,58, differs from
the radix in Vimonds text by only 0;0,3.
For Mercury, we compute according to the Parisian Alfonsine Tables at
Vimonds epoch, as for Venus, where 0(Sun) a(Sun) = 266;14,35:

Rx(Mercury) = 24754;52,55d 3;6,24,7,42,40,52/d + 213;48,38,56 + 266;


14,35 = 346;6,15

whereas Vimonds text has 11s 16;6,10 (= 346;6,10), in excellent agreement with
our recomputation.
A short text below these radices tells us that we should add two quantities,
the radix for the planet and the solar apogee. For Vimond the solar apogee
and each of the planetary apogees share the same motion; hence the differ-
ence between them is always the same. In particular, since Venuss apogee is
always the same as that of the Sun, nothing is given for Venus. The text then
displays values for each planet of the distance of its apogee from the solar
apogee:

Saturn 5s 12 = 162
Jupiter 2s 22 = 82
Mars 1s 14 = 44
Venus
Mercury 3s 29 = 119

These values agree closely with those of Ibn Isq (early 13th century) [Mestres
1996, p. 395]. They are used as shifts in subsequent tables for the planets,
and can be derived from the radices used in the Parisian Alfonsine Tables by
early alfonsine astronomy in paris 251

table 9 Yearly radices ( f. 4r v)

Year 1 Year 2 Year 8


s () s () s ()

Mercury 4 11; 1,24 4 21;11,55 2 23;56,47


Venus 8 15; 2,47 3 12;46,53 0 3;48,53
Mars 6 1;10, 7 0 26;51,41 3 1;58,32
Jupiter 1 1;53,32 2 1;19,53 8 2;52,21
Saturn 0 12;50,22 0 24;41,28 3 7;46,36

subtracting the solar apogee for the time of Alfonso from the radix of the apogee
for each planet (see, e.g., the editio princeps of the Alfonsine Tables printed
by Ratdolt (1483), c8d1; note that the signs used there are physical signs of
60):

Saturn 4, 2;35,20,41 1,20;37,0 = 161;58,20


Jupiter 2,42;48,38,41 1,20;37,0 = 82;11,38
Mars 2, 4;23,51,41 1,20;37,0 = 43;46,51
Venus 1,20;37, 0 1,20;37,0 = 0
Mercury 3,19;51,11,41 1,20;37,0 = 119;14,11

These results, when rounded to the nearest degree, are in perfect agreement
with Vimonds data. Therefore, the conclusion is that Vimond started with the
same planetary apogees as those in the Parisian Alfonsine Tables.

f. 4rv Table 9: Yearly Radices

Table 9 displays selected entries.


This table displays the radices for the five planets for 1, 2, 3, and 4 years;
for multiples of 4 years up to 76 years; and then for 152, 304, 608, 1216, and
2432 years, as in Table 3. As was the case for the radices for the Sun and the
Moon, 1 year is equivalent to 13 mean lunations, 2 years is equivalent to 25 mean
lunations, , 8 years is equivalent to 99 mean lunations, etc.
The mean daily motion in longitude resulting from the entries for year 8
(computed in the same way that was used for finding the mean motions in
Table 3) are shown below under the heading Vimond. If we add the daily
motion of the apogees (0;0,0,11,13,35/d), as we did in the case of the Sun, we
252 chapter 8

obtain the entries displayed in the second column, in good agreement with the
values for the mean motions in longitude in the Parisian Alfonsine Tables (see
Ratdolt 1483).

Including the motion


Vimond of the apogee Parisian Alf. t.

Saturn 0; 2, 0,24, 3,56/d 0; 2, 0,35,17,31/d 0; 2, 0,35,17,40/d


Jupiter 0; 4,59, 4, 1,19 0; 4,59,15,14,54 0; 4,59,15,27, 7
Mars 0;31,26,27,26,34 0;31,26,38,40, 9 0;31,26,38,40, 5
Venus 1;36, 7,35,47,21 1;36, 7,47, 0,56 1;36, 7,47, 1,19
Merc. 4; 5,32,16, 5,55 4; 5,32,27,19,30 4; 5,32,27,20, 0

It is most unusual for the mean motions of Venus and Mercury to be the sum of
their mean motions in anomaly and the solar mean motion, but there can be no
doubt that this is what Vimond did, as is confirmed by the note on f. 4vb. In fact,
we know of no other medieval astronomer writing in Latin who presented the
mean motions of the inferior planets in this way. For purposes of comparison,
the entries for Venus and Mercury under Parisian Alfonsine Tables are the
sum of the mean motions in anomaly and the solar mean motion: for Venus
0;36,59,27,24,0/d and 0;59,8,19,37,19/d, and for Mercury 3;6,24,7,42,41/d and
0;59,8,19,37,19/d. Note that in Ptolemys models the solar mean motion is also
the mean argument of center for Venus and Mercury.

ff. 4v5r Table 10: Monthly Radices

Table 10 displays selected entries.


This table displays the radices for the five planets for 25 consecutive syzygies.
The entries in this table are based on the same motions as those embedded in
the previous table. As was the case for the monthly radices in Table 4, for each
planet the entries for Syzygy 1 and Syzygy 25 add up to the entry corresponding
to Year 1 in the previous table (except for 1 for Mercury, Mars, and Jupiter).
For Venus and Mercury the mean motions extracted from Table 9 give exact
agreement, confirming the interpretation given above. Thus, in the cases of
Venus and Mercury one has obtained the sum of the solar anomaly and their
mean anomalies, respectively, at any syzygy (see Figure 8.1). This quantity is not
the argument in the table of equations (see Tables 12 and 15, below), and it is
early alfonsine astronomy in paris 253

table 10 Monthly radices ( ff. 4v5r)

Syzygy 1 Syzygy 2 Syzygy 25


s () s () s ()

Mercury 2 0;25,26 4 0;50,53 2 10;35,57


Venus 0 23;39,20 1 17;18,41 7 21;23,27
Mars 0 7;44,14 0 15;28,28 6 13;25,52
Jupiter 0 1;13,36 0 2;27,12 1 0;29,57*
Saturn 0 0;29,38 0 0;59,16 0 12;20,44

* Sic, instead of 1s 0;39,57.

figure 8.1 A geometric interpretation


of Vimonds tables for the mean
motion for Venus

not clear that there is any advantage to this method as against computing the
mean anomaly directly.

Tables 11 (Mercury, f. 5r), 14 (Venus, f. 5v), 17 (Mars, f. 6r), 20 (Jupiter,


f. 6v), and 23 (Saturn, f. 7r): Equation of Center and First Station

Tables 11 and 12 are to be used together to compute the true longitude of a planet
from its mean longitude. In most zijes in the Ptolemaic tradition, there is only
one such table for each planet, but Vimond has separated those functions that
depend on the mean argument of center from those that depend on the mean
anomaly and put them in different tables. A similar idea is already found in
the zij of Ibn Isq, described in Mestres 1996. Ibn Isqs parameters for the
maximum equations of center for Mars and Mercury are those of al-Battn,
254 chapter 8

but for Saturn, Jupiter, and Venus they are not; rather, they are 5;48 for Saturn,
5;41 for Jupiter, and 1;51 for Venus. The tables for planetary equations () are
divided into two groups: the first group contains the tables for the equation of
centre and the interpolation function. () The second group (two tables for
each planet) contains the tables for the equations of anomaly at apogee and
perigee and for the middle position (Mestres 1999, p. 234). So, the arrangement
of Vimonds tables bears a similarity to an Andalusian/Maghribi tradition that
is not otherwise attested in Latin.
However, it is not uncommon to find later sets of tables associated with
the Parisian Alfonsine Tables where the planetary equations are split into two
tables for each planet: see, e.g., Erfurt, ms ca q 362, ff. 28r36r, where the entries
are displayed at intervals of 1 and the radices are given for Paris (1320) as well
as for London and Brugge (1366).
Besides offering two tables for the equations of each planet, Vimonds tables
give additional information arranged in a presentation which is certainly pecu-
liar, as explained below.

table 11 Equation of center and first station of Mercury ( f. 5r)

(1) (2) (3) (4) (5) (6)


s () s () min min min s ()

0 6 0 8;51 61; 0 60 60 4 24;30


0 12 0 14;57 60;30 60 59 4 24;30
0 18 0 21; 0 60;20 59 58 4 24;32
0 24 0 27; 2 59;40 59 57 4 24;35
1 0 1 3; 0 59;30 59 54 4 24;38
1 6 1 8;57 59; 0 58 51 4 24;44
1 12 1 14;51 58;50 58 48 4 24;50
1 18 1 20;44 58;20 57 44 4 24;56
1 24 1 26;34 58;10 57 40 4 25; 7
2 0 2 2;23 58;10 57 35 4 25;20
2 6 2 8;12 57;50 57 29 4 25;37
2 12 2 13;59 57;40 57 24 4 25;53
2 18 2 19;45 57;30 57 19 4 26; 9
2 24 2 25;30 57;30 57 14 4 26;24
3 0 3 1;15 57;30 57 10 4 26;38
3 6 3 7; 0 57;30 57 6 4 26;50
3 12 3 12;45 57;30 57 4 4 27; 2
3 18 3 18;30 57;30 57 2 4 27; 9
early alfonsine astronomy in paris 255

(1) (2) (3) (4) (5) (6)


s () s () min min min s ()

3 24 3 24;15 57; 0 56 1 4 27;13


4 0 3 29;59 57;10 56 0 4 27;14
4 6 4 5;40 57;30 57 1 4 27;12
4 12 4 11;25 57;30 57 3 4 27; 7
4 18 4 17;10 57;30 57 5 4 26;59
4 24 4 22;55 57;30 57 7 4 26;46
5 0 4 28;40 57;30 57 11 4 26;34
5 6 5 4;25 57;30 57 16 4 26;19
5 12 5 10:10 57;40 57 21 4 26; 4
5 18 5 15:56 58; 0 57 26 4 25;48
5 24 5 21;44 58;10 57 31 4 25;30
6 0 5 27;33 58;10 57 37 4 25;16
6 6 6 3;22 58;30 58 41 4 25; 3
6 12 6 9;13 59; 0 58 45 4 24;54
6 18 6 15; 7 59;10 58 49 4 24;48
6 24 6 21; 2 59;30 59 52 4 24;42
7 0 6 26;59 59;50 59 55 4 24;37
7 6 7 2;58 60;30 60 57 4 24;34
7 12 7 9; 1 60;40 60 59 4 24;31
7 18 7 15; 5 61; 0 60 60 4 24;30
7 24 7 21;11 61;40 61 60 4 24;29
8 0 7 27;21 61;50 61 60 4 24;29
8 6 8 3;32 62; 0 61 59 4 24;29
8 12 8 9;44 62;10 61 59 4 24;30
8 18 8 15;57 62;20 61 58 4 24;32
8 24 8 21;11 62;30 62 57 4 24;34
9 0 8 28;26 62;40 62 56 4 24;36
9 6 9 4;44 63; 0 62 55 4 24;38
9 12 9 11; 2 63;10 62 54 4 24;39
9 18 9 17;21 63;10 62 54 4 24;40
9 24 9 23;44 63;20 62 53 4 24;41
10 0 10 0; 3 63;10 62 53 4 24;42
10 6 10 6;22 63;10 62 53 4 24;41
10 12 10 12;41 63;10 62 54 4 24;40
10 18 10 19; 0 63; 0 62 54 4 24;39
10 24 10 25;18 62;50 62 55 4 24;37
256 chapter 8

table 11 Equation of center and first station of Mercury ( f. 5r) (cont.)

(1) (2) (3) (4) (5) (6)


s () s () min min min s ()

11 0 11 1;35 62;40 62 56 4 24;35


11 6 11 7;51 62;40 62 57 4 24;33
11 12 11 14; 7 62;10 61 58 4 24;31
11 18 11 20;20 62; 0 61 59 4 24;30
11 24 11 26;32 61;50 61 60 4 24;29
12 0 12 2;43 61;20 60 60 4 24;29

table 14 Equation of center and first station of Venus ( f. 5v)

(1) (2) (3) (4) (5) (6)


s () s () min min min s ()

0 6 0 5;47 57;50 57 0 5 15;52


0 12 0 11;34 57;50 57 0 5 15;54

2 24 2 21;51 59;50 59 27 5 17; 2


3 0 2 27;50 60; 0 59 31 5 17;11
3 6 3 4;50 * 60;20 59 33 5 17;17
3 12 3 9;52 60;30 60 36 5 17;23

8 18 8 20; 8 60;20 59 36 5 17;23


8 24 8 26;10 60; 0 59 33 5 17;17
9 0 9 2;10 59;50 59 31 5 17;11
9 6 9 8; 9 59;30 59 27 5 17; 2

11 24 11 24;13 57;50 57 0 5 15;52


12 0 12 0; 0 57;50 57 0 5 15;50

* Sic, instead of 3s 3;50.

The table for the equation of center of each of the five planets has six columns.
Column 1 gives the argument (argumentum) at 6-intervals in signs and degrees
from 0s 6 to 12s 0. Column 2 displays the entry in col. 1 corrected for the equa-
tion of center (motus completus), in signs, degrees, and minutes. The author
early alfonsine astronomy in paris 257

table 17 Equation of center and first station of Mars ( f. 6r)

(1) (2) (3) (4) (5) (6)


s () s () min min min s ()

0 6 0 12;31 50;50 26 6 5 8;41


0 12 0 17;36 50; 0 26 4 5 8;21

1 12 1 12;22 49; 0 26 0 5 7;29


1 18 1 17;16 49;10 25 0 5 7;31

4 18 4 6;36 60;30 31 32 5 13;46

7 12 7 11;33 73;30 38 60 5 19;14


7 18 7 18;54 73;10 38 59 5 19;13

10 12 10 23;24 58;50 30 31 5 13;36

11 24 12 2;13 51;50 27 10 5 9;31


12 0 12 7;24 51;10 26 8 5 9; 6

follows here the same pattern as that for the true solar anomaly (see Table 5).
Column 3 (motus gradus) gives the increment of the true argument per degree
of the argument, in minutes and seconds. Most entries in this column are gen-
erated by dividing by 6 the differences between two successive entries in col. 2
and thus were probably intended for interpolation in col. 2. Column 4 (motus
diei) displays the velocity in minutes of arc per day, and the range of values
for each planet is the same as in the column labeled motus centri or motus
puncti (that only depends on the argument of center) in the table for planetary
velocities associated with the Toledan Tables and the Castilian Alfonsine Tables
(Chabs and Goldstein 2003a, pp. 170182); for the other component of the plan-
etary velocity, see Tables 12, 15, 18, 21, and 23, col. 4, below. So, the entries in this
column are only one component of the planets velocity. Column 5 is intended
to provide minutes of interpolation and is headed diametri (perhaps to distin-
guish these linear minutes from minutes of an hour, minutes of a day, and
minutes of a degree). Finally, column 6 lists the first station in signs, degrees,
and minutes.
But for a shift of the entries, the equations of center for Mercury, Mars, and
Saturn that can be derived from cols. 1 and 2 are basically the same (with
258 chapter 8

table 20 Equation of center and first station of Jupiter ( f. 6v)

(1) (2) (3) (4) (5) (6)


s () s () min min min s ()

0 6 0 11;43 58; 0 5 23 4 5;19


0 12 0 17;31 57;20 5 20 4 5; 9

2 12 2 12;59 44;10 4 0 4 4; 6
2 18 2 18;23 44; 0 4 0 4 4; 5
2 24 2 23;48 44; 0 4 0 4 4; 5
3 0 2 29;12 44;10 4 0 4 4; 6

5 24 5 18; 3 60;10 4 32 4 5;44

8 12 8 10;55 66;20 6 59 4 7;10


8 18 8 17;33 66;30 6 60 4 7;11
8 24 8 14;14 66;20 6 60 4 7;11
9 0 9 0;52 66;20 6 60 4 7;10

11 18 11 23;57 59;50 5 32 4 5;47


11 24 11 29;56 59;10 5 29 4 5;39
12 0 12 0;51 * 58;40 5 26 4 5;29

* Sic, instead of 5;51.

minor variants) as in the zij of al-Battn (Nallino 19031907, 2:110137) and


the Toledan Tables (Toomer 1968, pp. 6068; F.S. Pedersen 2002, pp. 1259
1308).
The maximum value for Mercury (3;2) occurs at about 0s 24 and 7s 6, that
of Mars (11;24) at 4s 18 and 10s 12, and that of Saturn (6;31) at 2s 12 and 8s
12. However, for the other two planets the entries differ systematically from
those in the above-mentioned zijes: for Venus the maximum value is 2;10 at
3s 0 and 3s 6, and 8s 24 and 9s 0; and for Jupiter the maximum value is
5;57 at 5s 24 and 11s 18. The entries for Mercury, Mars, Jupiter, and Saturn
are shifted by about 119, 44, 82, and 162, respectively, in relation to those
in the zij of al-Battn and the Toledan Tables. No such shift appears in the
table for Venus. As mentioned above, these shifts result from the difference
between the apogee of each of the planets and that of the Sun. Because of
these shifts, for the superior planets one enters these tables in col. 1 directly
early alfonsine astronomy in paris 259

table 23 Equation of center and first station of Saturn ( f. 7r)

(1) (2) (3) (4) (5) (6)


s () s () min min min s ()

0 6 0 8;46 66;20 2 57 3 25;22


0 12 0 15;24 66; 0 2 55 3 25;19

2 12 2 18;31 59;20 2 31 3 24;11

5 6 5 6;39 53;30 2 0 3 22;45


5 12 5 12; 0 53;30 2 0 3 22;44
5 18 5 17;21 53;40 2 0 3 22;45

8 12 8 5;29 60;10 2 31 3 24;11

11 6 11 12; 0 67;10 2 60 3 25;30


11 12 11 18;43 67;10 2 60 3 25;28
11 24 11 25;25 67; 0 2 59 3 25;27
12 0 12 2; 7 66;30 2 58 3 25;25

with their mean motions for a given syzygy (the radix plus the motion in
years and semi-lunations); for Venus and Mercury one enters with the solar
anomaly for a given syzygy. Clearly, Vimond intended to make this table more
user-friendly than the standard version of the table for the equation of cen-
ter.
Vimond has a double motion of the solar apogee: precession and proper
motion. The planetary apogees are fixed with respect to the solar apogee (i.e.,
they are subject to both precession and the proper motion of the solar apogee).
If we add the solar apogee (about 90) to the values for the shifts listed above,
we find that the planetary apogees are 209 for Mercury, 90 for Venus, 135 for
Mars, 172 for Jupiter, and 252 for Saturn. In the Toledan Tables, the apogees of
the Sun and of Venus are both 77;50 (Toomer 1968, p. 45), that is, about 12 less
than 90. Adding this difference to the planetary apogees in the Toledan Tables
rounded to degrees, we find the following:
260 chapter 8

Apogees
v (from the shifts) v (from the radices) tt + 12

Mercury 209 209 210


Venus 90 90 90
Mars 135 134 134
Jupiter 172 172 176
Saturn 252 252 252

table 11a Maximum values for the equation of center

al-Battn Toledan t. Vimond Parisian Alf. t.

Mercury 3; 2 3; 2 3; 2 3; 2
Venus 1;59 1;59 2;10 2;10
Mars 11;24 11;24 11;24 11;24
Jupiter 5;15 5;15 5;57 5;57
Saturn 6;31 6;31 6;31 6;31

The agreement of Vimonds data with the apogees in the Toledan Tables
shows that Vimond has included the motion of the solar apogee in the motions
of the planetary apogees, thus following a theory for which there was no
previous evidence outside al-Andalus and the Maghrib (Sams and Mills 1998,
pp. 268270). We know of no other set of planetary equation tables arranged in
this way. See also Table 27 (equation of access and recess), below, for yet another
shift in Vimonds tables.
The maximum values for the equation of center in Vimonds planetary
tables are the same as in the editio princeps of the Alfonsine Tables (see Table
11a).
Despite their agreement for the values of the maximum equations, the
structure of Vimonds tables is very different from that of the Parisian Alfonsine
Tables and would seem to be independent of it. Moreover, it is significant that
the maximum equation of center for Jupiter in both cases is 5;57, for this value
is not known in any text prior to the Parisian Alfonsine Tables, indicating a
strong connection between the tables of Vimond and the work of his Parisian
contemporaries. The origin or derivation of this parameter for Jupiter is not
described in any extant text, and it is likely that this value was simply taken
early alfonsine astronomy in paris 261

from an earlier work: the most reasonable candidate is the Alfonsine Tables as
they existed in Castile.
For all planets, except Mercury, an entry, c, in column 5 can be computed,
but for shifts, from the modern formula

c = 60 (1 cos )/2,

where is the mean argument of center. The same approach is found in Levis
lunar theory (Goldstein 1974, table 35, col. ii: see p. 54).
The entries for Mercury in col. 5 do not follow the same pattern as that for the
rest of the planets. The entries can be recomputed, approximately, according to
the following formula:

[1] c5() = [d r()] / [d d]

where d is the maximum distance of the center of the epicycle from the
observer, d is the minimum distance, and r() is the distance as a function of
the mean argument of center, .
A similar formula for interpolation was already used by abash in the 9th
century (as-Saleh 1970, pp. 137138). In Ptolemys model for Mercury d is
69 for argument 0, d is 55;34 for an argument close to 120, and r(180)
is 57 (O. Pedersen 1974, pp. 313324). Hence, formula [1] can be replaced
by

[2] c5() = [69 r()] / 13;26.

In general, the computation of r() is a difficult and lengthy procedure, and it


is likely that Vimond (or his source) used approximations (if this, indeed, was
the formula he had in mind). We computed the distances from the observer
to the center of Mercurys epicycle according to formulas in modern terms
given by O. Pedersen (1974, p. 320, equations 10.34 and 10.35), and then used
them in equation [2], above. A comparison of our results for c5() with the
entries in Vimonds table is displayed in Table 11b. Col. ii has the values for
c5() that depend on the distances computed according to the formulas given
by O. Pedersen and equ. [2], above; col. iii has the arguments in Vimonds
table (with the shift); and col. iv has the entries in Vimonds Table 11, col. 5.
Although the agreement is not exact between col. ii and col. iv, the trend is
clear. Vimonds value for 180 (col. iii), 37, has the poorest agreement, but this
entry should probably be corrected to 36, judging from the surrounding
values.
262 chapter 8

figure 8.2 Vimonds equation of center, col. 5, for Venus, Mars, Jupiter, and Saturn

figure 8.3 Vimonds equation of center, col. 5, for Mercury


early alfonsine astronomy in paris 263

table 11b A comparison of column 5 for


Mercury with recomputation

i ii iii iv
c5( ) (Vimond) c5( ): Vimond

0 0; 0 120 0
30 10;50 150 11
54 29;24 174 31
60 34;15 180 37
66 38;53 186 41
90 53;33 210 55
120 60; 0 240 60
150 56;38 270 56
180 53;34 300 53

It may be of interest that in Copernicuss table for the equations of Mercury


(Copernicus 1543, ff. 177v178r), his col. 4 (for interpolation) shows the same
trend as Vimonds col. 5. We are convinced that column 5 in Vimonds tables for
the equation of center is intended to be used for interpolation with column 5
in the tables for the equation of anomaly, and this is analogous to Copernicuss
use of his col. 4 (see below). Indeed, Vimonds col. 5 serves much the same
purpose as col. 8 in Ptolemys tables for the planetary equations (Almagest,
xi.11) but, since the definitions for the columns that yield the equation of
anomaly are different, so is the function for interpolation. Moreover, in contrast
to the geometric methods in the Almagest used for computing the coefficients
of interpolation for each of the four planets (Venus, Mars, Jupiter and Saturn),
Vimond has approximated the results that would be derived from the geometry
of the models by introducing a single trigonometric function in those cases.
In Almagest, xi.11 (Toomer 1984, pp. 549553), col. 8 in the planetary equa-
tion tables is intended for interpolation as a function of , the mean argument
of center, and the entries are given to minutes and seconds (for Ptolemys
method of computation and a graph of the entries in his col. 8, see Neugebauer
1975, pp. 184186, 1267). A similar set of values, given only to minutes, is found in
al-Battns zij in the tables for the planetary equations, col. iv (Nallino 1903
1907, 2:110137), and in corresponding tables in the Parisian Alfonsine Tables,
col. 3 (Ratdolt 1483, e7rg5v).
As for the entries for the first station of each planet, they are essentially
the same as in previous tables of the same kind (Almagest, Handy Tables,
264 chapter 8

al-Khwrizm, al-Battn, and the Toledan Tables) with the same shifts that we
noted above.

Tables 12 (Mercury, f. 5r), 15 (Venus, f. 5v), 18 (Mars, f. 6rv), 21


(Jupiter, f. 7r), and 24 (Saturn, f. 7v): Equation of Anomaly

The tables for the equation of anomaly for each of the five planets have seven
columns. Table 12 displays a selection of values for the equation of anomaly for
Mercury.
Column 1 gives the mean argument of anomaly (argumentum) at 6-intervals
(at 3-intervals for Mars and Venus) from 0s 6 to 6s 0 and its complement
in 360 from 6s 0 to 11s 24. Column 2 displays the correction due to the
argument of anomaly at maximum distance (motus completus) in degrees and
minutes and represents the difference between the equation of anomaly and
the correction for maximum distance (cf. Almagest, xi.11, columns 6 and 5; and
Neugebauer 1975, pp. 183184). The only other text of which we are aware that
treats the equation of anomaly in this way is the zij of Ibn al-Bann (d. 1321)
where this presentation is applied in his tables for Saturn and Jupiter but not
in those for the other planets (see Sams and Mills 1998, pp. 278285). The
extremal values in col. 2 that appear in the text are shown below; they are
followed by the corresponding entries for col. vi and col. v in the zij of al-Battn
(Nallino 19031907, 2:109137):

Vimond al-Battn

Mercury 19; 1 (= 21;59 2;58) at 3s 18


Venus 44;49 (= 45;59 1;10) at 4s 15
Mars 36;44 (= 40;58 4;14) at 4s 6
Jupiter 10;34 (= 11; 3 0;29) at 3s 12
Saturn 5;53 (= 6;12 0;19) at 3s 0 and
(= 6;13 0;20) at 3s 6

These corrections agree with those that follow from the Almagest as well as
the zij of al-Battn, the Toledan Tables, and the editio princeps of the Alfonsine
Tables (with minor variants: 40;59 rather than 40;58 for Mars; 6;12 and 0;19
correspond to 3s 0 rather than 3s 1 for Saturn), and this means that Ptolemys
eccentricities underlie them even though, in the case of Venus and Jupiter,
early alfonsine astronomy in paris 265

table 12 Equation of anomaly for Mercury ( f. 5r)

(1) (2) (3) (4) (5) (6) (7)


s () s () () min min min sec sec

0 6 11 24 1;28 15 45 0;18 2 8
0 12 11 18 2;56 15 45 0;33 2 8
0 18 11 12 4;24 14 44 0;48 2 8
0 24 11 6 5;50 14 44 1; 3 2 8
1 0 11 0 7;15 14 42 1;18 2 8
1 6 10 24 8;37 13 42 1;33 2 8
1 12 10 18 9;58 13 40 1;48 2 8
1 18 10 12 11;15 12 39 2; 0 2 8
1 24 10 6 12;30 11 36 2;18 3 9
2 0 10 0 13;39 10 34 2;35 3 9
2 6 9 24 14;44 10 31 2;53 3 9
2 12 9 18 15;44 9 28 3;10 4 8
2 18 9 12 16;38 8 25 4;14 4 8
2 24 9 6 17;25 7 19 4;30 4 8
3 0 9 0 18; 4 5 15 4;45 4 8
3 6 9 24 18;34 3 9 4;57 4 8
3 12 8 18 18;53 1 3 5; 5 4 8
3 18 8 12 19; 1 1 3 5;10 3 6
3 24 8 6 18;56 3 9 5;13 2 4
4 0 8 0 18;39 6 18 5; 6 1 3
4 6 7 24 18; 4 9 27 4;55 1 1
4 12 7 18 17;12 11 35 4;29 0 1
4 18 7 12 16; 4 15 45 4;55 2 6
4 24 7 6 14;36 18 55 4;12 4 13
5 0 7 0 12;49 21 66 4;29 6 18
5 6 6 24 10;42 24 74 3;55 7 22
5 12 6 18 8;18 26 81 3;12 9 29
5 18 6 12 5;42 28 87 2;15 11 34
5 24 6 6 2;53 29 89 1;10 12 36
6 0 6 0 0; 0 29 89 0; 0 12 36

the eccentricities were modified for computing the equation of center (cf.
North 1976, 3:196). Similarly, in the tables of Ibn al-Bann the eccentricities
underlying the equations of anomaly are taken from the Almagest, but his
266 chapter 8

maximum equations of center for Venus and Jupiter are not those of either
Ptolemy or of Vimond (Sams and Mills 1998, p. 276).
Column 3 (motus gradus) gives the increment of the motus completus in
col. 2 per degree of the argument in minutes: in most cases the entry results
from taking the difference between successive entries in col. 2 and dividing
that difference by 6 (or by 3 for Mars and Venus); the purpose of this column
is facilitate interpolation. Column 4 (motus diei) displays the velocity in min-
utes of arc per day, and the range of values for each planet is the same as in
the column labeled motus argumenti (that only depends on the argument of
anomaly) in the table for planetary velocities; see the comments to the Castil-
ian Alfonsine Tables, chapter 27 (Chabs and Goldstein 2003a, pp. 170182). So,
an entry in this column is the second component of the planets velocity and it
complements the first component already displayed in Tables 11, 14, 17, 20, and
23, above. The entries in column 5 (minutum diametri), in minutes and seconds,
actually represent degrees and minutes, and result from adding the correction
for maximum distance to the correction for minimum distance (columns c5
and c7 in Almagest, xi.11). For the extremal values in col. 5 in the text see below;
they are followed by the corresponding entries for col. v and col. vii in the zij
of al-Battn (Nallino 19031907, 2:109137):

Vimond al-Battn

Mercury 5;13 4;56 *


Venus 3;34 (= 1;42 + 1;52) at 5s 12
Mars 13;37 (= 5;34 + 8; 3) at 5s 9
Jupiter 1; 3 (= 0;30 + 0;33) at 3s 24
Saturn 0;46 (= 0;21 + 0;25) at 3s 12

* In al-Battns zij for 3s 24 we find 3;4 + 1;52 =


4;56, whereas for 3s 24 in Vimonds table we find
3;12 + 2;1 = 5;13, al-Battns maximum which
occurs at 4s 104s 12.

In the absence of instructions by Vimond it is not easy to decide how the


correction to the planets mean longitude is to be computed, but it seems likely
that one component of this correction is to be computed by adding an entry in
col. 2 to an interpolation factor times an entry in column 5, as is the case with
the tables of Ibn al-Bann for Saturn and Jupiter. The most likely candidate for
this interpolation factor is col. 5 in Table 11, for it depends on the argument of
early alfonsine astronomy in paris 267

table 15 Equation of anomaly for Venus ( f. 5v)


(The entries from 5s 18 to 6s 12 are given at 2-intervals, rather than at 3-intervals
as in the rest of the table.)

(1) (2) (3) (4) (5) (6) (7)


s () s () () min min min sec sec

0 3 11 27 1;15 25 15 0; 2 0 0
0 6 11 24 2;30 25 15 0; 3 1 0

4 12 7 18 44;44 2 1 2;18 2 1
4 15 7 15 44;49 2 1 2;25 2 1
4 18 7 12 44;44 6 4 2;32 3 2

5 12 6 18 33;25 74 46 3;34 2 1
5 15 6 15 29;43 89 55 3;27 4 2
5 18 6 12 25;25 104 64 3;14 7 4

5 28 6 2 4;48 144 89 0;45 22 14


6 0 6 0 0; 0 144 89 0; 0 22 14

table 18 Equation of anomaly for Mars ( f. 6r v)


(The entries from 5s 18 to 6s 12 are given at 2-intervals, rather than at 3-intervals
as in the rest of the table.)

(1) (2) (3) (4) (5) (6) (7)


s () s () () min min min sec sec

0 3 11 27 1; 8 23 11 0; 8 3 1
0 6 11 24 2;16 23 11 0;17 3 1

4 3 7 27 36;40 1 1 8;53 9 4
4 6 7 24 36;44 0 0 9;19 9 4
4 9 7 21 36;43 3 1 9;46 9 4

5 6 6 24 28;15 46 21 13;30 0 0
5 9 6 21 25;56 53 25 13;37 6 2
5 12 6 18 23;17 62 29 13;19 13 6

268 chapter 8

table 18 Equation of anomaly for Mars ( f. 6r v) (cont.)

(1) (2) (3) (4) (5) (6) (7)


s () s () () min min min sec sec

5 28 6 2 3; 1 90 42 2;29 74 35
6 0 6 0 0; 0 90 42 0; 0 74 35

table 21 Equation of anomaly for Jupiter ( f. 7r)

(1) (2) (3) (4) (5) (6) (7)


s () s () () min min min sec sec

0 6 11 24 0;57 9 8 0; 4 1 1
0 12 11 18 1;52 9 8 0; 8 1 1

3 6 8 24 10;33 0 0 0;59 0 0
3 12 8 18 10;34 1 1 1; 1 0 0
3 18 8 12 10;29 2 2 1; 2 0 0
3 24 8 6 10;15 3 3 1; 3 0 0
4 0 8 0 9;54 5 4 1; 2 0 0

5 24 6 6 1;21 13 12 0; 9 1 1
6 0 6 0 0; 0 13 12 0; 0 1 1

table 24 Equation of anomaly for Saturn ( f. 7v)

(1) (2) (3) (4) (5) (6) (7)


s () s () () min min min sec sec

0 6 11 24 0;34 5 5 0; 3 1 1
0 12 11 18 1; 7 5 5 0; 7 1 1

2 24 9 6 5;46 1 1 0;41 0 0
3 0 9 0 5;53 0 0 0;42 0 0
3 6 8 24 5;53 0 0 0;44 0 0
3 12 8 18 5;51 1 1 0;46 0 0

early alfonsine astronomy in paris 269

(1) (2) (3) (4) (5) (6) (7)


s () s () () min min min sec sec

5 24 6 6 0;42 7 7 0; 7 1 1
6 0 6 0 0; 0 7 7 0; 0 1 1

figure 8.4 Ptolemys model for the three


superior planets and Venus (not to
scale)

center as it should (see Sams and Mills 1998). Column 6 (motus gradus) seems
to be the increment per degree of argument of the entries in col. 5: in many
cases the entry in col. 6 results from taking the difference between successive
entries in col. 5 and dividing it by 6 (or by 3 for Mars and Venus), and it is for
purposes of interpolation. The entries in col. 6 are given in seconds. The entries
in column 7 (motus diei) are also given in seconds; they are probably associated
with those in the preceding column, for in all cases columns 6 and 7 have their
extremal values for the same arguments, but we have failed to identify their
specific purpose.
Figure 8.4 displays Ptolemys model for the three superior planets and Venus.
o is the observer, d is the center of the deferent circle rac, and e is the equant
point, such that the eccentricity, e = od = de. a is the planets apogee, and =
angle aec, the mean argument of center, is measured from it to the center of the
epicycle about point e. Angle gcp is the mean argument of anomaly, , and the
planet is at point p. Angle hcg is the equation of center and it is also applied to
correct the mean argument of anomaly to yield the true argument of anomaly,
= angle hcp. In the case of the superior planets, cp, the direction from the
270 chapter 8

center of the epicycle to the planet, is always parallel to s, the direction from
the observer to the mean Sun. In the case of Venus, ec is parallel to the direction
from the observer to the mean Sun. The goal is to find the direction from the
observer to the planet, i.e., angle rop is the longitude of the planet, and r is in
the direction to Aries 0.
The mean argument of center, , is angle aec, and the true argument of
anomaly, , is angle hcp. With these arguments, and , we can determine
the equation of anomaly, c(), with Vimonds tables and compare the result
with computations based on the Parisian Alfonsine Tables. According to our
understanding of Vimonds procedure,

c() = c2() + c5() c5(),

where ci refers to the i-th column in the table. Note that c5() is taken from the
table for the equation of center (with the shifts), and c5() is taken from the
table for the equation of anomaly. For instance, for Venus, when = 120 and
= 135

c() = c2(135) + c5(120) c5(135)


c() = 44;49 + 0;45 2;25
c() = 46;38.

With the same arguments for Venus in the Parisian Alfonsine Tables, we find

c() = c5() + c3() c6()


c() = c5(135) + c3(120) c6(135)
c() = 45;59 + 0;31 1;15
c() = 46;38

and this is exactly what resulted from Vimonds tables. In the tables for the
planetary equations in Almagest xi.11 and its derivatives in al-Battn and in the
Parisian Alfonsine Tables (among others), the rules for computing the equation
of anomaly require careful attention to algebraic signs. Vimond simplified the
rules for this computation, making his tables more user-friendly. A similar
procedure is described by Copernicus for using his planetary tables in De
revolutionibus, v.23, to compute the equation of anomaly (Copernicus 1543,
ff. 173v179r; cf. Swerdlow and Neugebauer 1984, p. 453).
early alfonsine astronomy in paris 271

Tables 13 (Mercury, f. 5r), 16 (Venus, f. 6r), 19 (Mars, f. 6v), 22 (Jupiter,


f. 7r), and 25 (Saturn, f. 7v): Planetary Latitudes

The tables for the planetary latitudes, both for the superior and the inferior
planets, are in the style of Almagest xiii.5, the zij of al-Battn (Nallino 1903
1907, 2:140141), and some tables associated with the Toledan Tables (Toomer
1968, pp. 7172; F.S. Pedersen 2002, pp. 13221326), as opposed to those in the
Handy Tables and those in the zij of al-Khwrizm.
The table for the planetary latitudes of Mercury has seven columns; the table
for Venus lacks the seventh; and the tables for the superior planets have only
five columns (i.e., cols. 1, 3, 4, 5, and 6).
In all cases column 1 displays the argument (argumentum) at 12-intervals
from 0s 12 to 12s 0. Column 2 is only found in the tables for the inferior planets
and the entries are given in minutes. The heading is radix meridionalis in the
case of Mercury and radix septentrionalis in that for Venus. This column is for
determining the deviation, otherwise called the third component of latitude,
that is, the inclination of the plane of the deferent with respect to the ecliptic.
The entries for deviation can be derived from:

3 = 0;45 c5 for Mercury


3 = +0;10 c5 for Venus

where c5 is the column for the minutes of proportion in the table for planetary
latitude in Almagest xiii.5 (given there in minutes and seconds). As will be
seen, column 5 for Venus in Table 16, given only to minutes, corresponds to
c5 in Almagest xiii.5. It is noteworthy that column 2 for Mercury is shifted
downwards about 119 whereas there is no shift in the case of Venus. This is
exactly the same feature we noticed in the tables for the equation of center and
the amount of the shift is the same. The column for deviation is certainly not
a common feature in medieval tables (for a survey of the few that have them,
see Goldstein and Chabs 2004), and Vimonds is the earliest set of tables in the
West we know to display such columns.
For the inferior planets, columns 3 and 5 (diametri) give the minutes of
proportion for the inclination and the slant, respectively. We note that columns
3 and 5 for Mercury also exhibit a shift of less than 120, and that no shifts appear
in the case of Venus. We also note that column 5 for Venus lists the rounded
values in the column for the sixtieths found in the corresponding table in the
Almagest xiii.5, the zij of al-Battn, etc.
For the superior planets, columns 3 and 5 give the minutes of proportion for
the northern and southern latitudes, respectively, of the planets. Only half of
272 chapter 8

table 13 Latitude of Mercury ( f. 5r)

(1) (2) (3) (4) (5) (6) (7)


s () min min min * min min * sec

0 12 13 57 1;44 17 0;12 1
0 24 4 60 1;40 5 0;44 4
1 6 5 60 1;39 7 1; 6 7
1 18 14 57 1;16 19 1;26 9
2 0 23 51 0;59 31 1;44 10
2 12 31 44 0;38 41 2; 0 12
2 24 37 34 0;16 49 2;14 13
3 6 41 23 0;15 55 2;25 14
3 18 44 11 0;48 59 2;29 15
4 0 45 0 1;25 60 2;29 15
4 12 43 13 2; 6 58 2;20 14
4 24 40 25 2;47 54 2; 0 13
5 6 36 36 3;26 48 1;29 9
5 18 29 45 3;54 39 0;48 5
6 0 22 52 4; 5 29 0; 0 0
6 12 13 57 3;54 17 0;48 5
6 24 4 60 3;26 5 1;29 9
7 6 5 60 2;47 7 2; 0 12
7 18 14 52 2; 6 19 2;20 14
8 0 23 51 1;25 31 2;29 15
8 12 31 44 0;48 41 2;29 15
8 24 37 34 0;15 49 2;25 14
9 6 41 23 0;16 55 2;10 13
9 18 44 11 0;38 59 2; 0 12
10 0 45 1 0;59 60 1;44 10
10 12 43 13 1;16 58 1;26 9
10 24 40 25 1;30 54 1; 6 7
11 6 36 36 1;40 48 0;44 4
11 18 29 45 1;44 39 0;12 1
12 0 22 52 1;46 29 0; 0 0

* Despite the headings, these columns display degrees and minutes.


early alfonsine astronomy in paris 273

table 16 Latitude of Venus ( f. 6r)

(1) (2) (3) (4) (5) (6)


s () min min min * min min *

0 12 10 12 1; 1 59 0;16
0 24 9 24 0;59 55 0;33
1 6 8 35 0;55 48 0;49
1 18 7 44 0;46 40 1; 5
2 0 5 52 0;35 30 1;20
2 12 3 57 0;29 18 1;35
2 24 1 60 0;18 6 1;50
3 6 1 60 0;10 6 2; 3
3 18 3 52 0;32 18 2;15
4 0 5 44 0;59 30 2;25
4 12 7 35 1;38 40 2;30
4 24 8 24 2;23 48 2;28
5 6 9 12 3;44 55 2;12
5 18 10 0 5;13 59 1;27
6 0 10 12 7;12 60 0; 0
6 12 10 24 5;13 59 1;12
6 24 9 35 3;44 55 2;28
7 6 8 44 2;23 48 2;30
7 18 7 52 1;38 40 2;25
8 0 5 57 0;59 30 2;15
8 12 3 60 0;32 18 2; 3
8 24 1 60 0;10 6 1;50
9 6 1 57 0;19 6 1;35
9 18 3 52 0;29 18 1;20
10 0 5 44 0;35 30 1; 5
10 12 7 44 0;46 40 0; 5 **
10 24 8 35 0;55 48 0;49
11 6 9 24 0;59 55 0;33
11 18 10 12 1; 1 59 0;16
12 0 10 0 1; 3 60 0; 0

* Despite the headings, these columns display degrees and minutes.


** Sic.
274 chapter 8

table 19 Latitude of Mars ( f. 6v)

(1) (3) (4) (5) (6)


s () min min * min min *

0 12 50 0; 9 m 0; 4
0 24 55 0;13 m 0; 6
1 6 59 0;16 m 0; 9
1 18 59 0;21 m 0;15

4 12 8 2; 1 m 2;10
4 24 s 2;34 10 2;56

6 0 s 4;21 43 7;30

10 12 s 0;21 2 0;15
10 24 10 0;16 [blank] 0; 9
11 6 22 0;13 m 0; 6
11 18 33 0; 9 m 0; 4
12 0 43 0; 6 m 0; 2

* Despite the headings, these columns display degrees and minutes.

the columns are filled with numbers, the others have capital letters indicat-
ing North [s] and South [m]. Column 3 is shifted about 45 (Mars), about
100 (Jupiter), and about 110 (Saturn) in relation to the corresponding columns
in the Almagest, whereas the shifts for column 5 are increased by 180 in
each case. These shifts are totally consistent with those found for the equa-
tion of center (about 44, 82, and 162 for Mars, Jupiter, and Saturn, respec-
tively). Indeed, subtracting these numbers for each planet, we find 0 (Mars),
about 20 (Jupiter), and +50 (Saturn), in perfect agreement with the dif-
ferences given by Ptolemy in Almagest xiii.6 between the northern limits
on the deferent and the apogees of each superior planet, respectively. Thus,
it is quite clear that the compiler of Vimonds tables, whether Vimond or
not, had a good understanding of this difficult issue as it is presented in the
Almagest.
Columns 4 and 6 display the inclination (declinatio minuti diametri) and
the slant (reflexio minuti diametri) for the inferior planets, and the entries are
given in degrees and minutes, despite the headings, which read minutes and
seconds. For the superior planets, these two columns display the northern and
early alfonsine astronomy in paris 275

table 22 Latitude of Jupiter ( f. 7r)

(1) (3) (4) (5) (6)


s () min min * min min *

0 12 10 1; 8 0 1; 6
0 24 12 1; 9 m 1; 7

3 6 60 1;33 m 1;33
3 18 60 1;39 m 1;39

6 0 12 2; 5 m 2; 8
6 12 0 2; 3 0 2; 6
6 24 s 2; 0 12 2; 3

9 6 s 1;27 60 1;26
9 18 s 1;21 60 1;21

11 18 s 1; 8 34 1; 6
12 0 s 1; 6 12 1; 5

* Despite the headings, these columns display


degrees and minutes.

southern limits (both labeled latitudo minuti diametri) and are given in degrees
and minutes.
The extremal values of columns 4 and 6 in the text are shown below:

Mercury 4; 5 (for 6s 0) 2;29 (for 3s 184s 0 and 8s 08s 12)


Venus 7;12 (for 6s 0) 2;30 (for 4s 12 and 7s 6)
Mars 4;21 (for 6s 0) 7;30 (for 6s 0)
Jupiter 2; 5 (for 6s 0) 2; 8 (for 6s 0)
Saturn 3; 2 (for 6s 0) 3; 5 (for 6s 0)

These extremal values in Vimonds tables agree with those in the Toledan
Tables with two exceptions, one of which is a trivial variant for Mercury. But,
as far as we know, the maximum value for the inclination of Venus in Vimonds
table is not attested in any other previous text. It is probably significant that
this value later appeared in the editio princeps of the Alfonsine Tables (1483), as
indicated in Table 13a.
276 chapter 8

table 25 Latitude of Saturn ( f. 7v)

(1) (3) (4) (5) (6)


s () min min * min min *

0 12 [blank] 2; 5 9 2; 3
0 24 2 2; 7 [blank] 2; 4
1 6 14 2;10 s 2; 7

3 18 60 2;39 s 2;39

5 18 33 3; 1 s 3; 3
6 0 22 3; 2 s 3; 5
6 12 10 3; 1 s 3; 3
6 24 n 2;59 2 3; 0

9 18 n 2;21 60 2;21

11 18 n 2; 5 33 2; 3
12 0 n 2; 3 22 2; 2

* Despite the headings, these columns display degrees and minutes.

table 13a Extremal planetary latitudes

Almagest al-Battn Toledan t. Vimond Paris. Alf. t.

Mercury 4; 5 4; 5 4; 5 4; 5 4; 5
2;30 2;30 2;30 2;29 2;30
Venus 6;22 6;22 7;24 7;12 7;12
2;30 2;30 2;30 2;30 2;30
Mars 4;21 4;21 4;21 4;21 4;21
7; 7 7; 7 7;30 7;30 7;30
Jupiter 2; 4 2; 4 2; 5 2; 5 2; 8
2; 8 2; 8 2; 8 2; 8 2; 8
Saturn 3; 2 3; 2 3; 2 3; 2 3; 3
3; 5 3; 5 3; 5 3; 5 3; 5
early alfonsine astronomy in paris 277

table 26 Yearly radices ( f. 7v)

s () s ()

years 76 years 608

mean motion 0 0;33,32 mean motion 0 4;28, 3


argument 0 3;54,46 argument 1 1;16,21

years 152 years 1216

mean motion 0 1; 7, 2 mean motion 0 8;56, 4


argument 0 7;49,16 argument 2 2;32,27

years 304 years 2432

mean motion 0 2;14, 3 mean motion 0 17;52, 5


argument 0 15;38,18 argument 4 5; 4,38

For the inferior planets, between columns 2 and 3 and between columns
5 and 6 we are also given some indications (North and South) to help the
user.
Column 7 appears only in Table 13 (Mercury), and it seems to be outside
the general framework of the table. Its entries are given in seconds and result
from dividing the corresponding entries in column 6 by 10. This probably cor-
responds to the instructions given by Ptolemy in Almagest xiii.6: to compute
the true minutes of proportion for the slant, add 110 when the argument lies
between 90 and 270, or subtract 110 when the argument lies between 0 and
90 or 270 and 360. Whether tabulated or not, these instructions are rarely
found in the medieval Latin literature on the planets (Goldstein and Chabs
2004).

f. 7v Table 26: Yearly Radices

This table displays the radices for the mean motion (motus) and argument
(argumentum) of the fixed stars for intervals of 76, 152, 304, 608, 1216, and
2432 years. Vimond does not give a radix for a specific year but perhaps this
information was in the canons that we have not found. As we shall argue (see
278 chapter 8

Table 27, below), it is likely that the epoch of this table was also 1320 or a date
close to it, that is, the epoch is consistent with our dating of the other radices.
In 76 years the value in the text for the mean motion of the fixed stars is
0;33,32 and in 2432 years it is 17;52,5, corresponding to 0;0,0,4,20,56/d and
0;0,0,4,20,42/d, respectively. These values are equivalent to 48,954 years and
48,999 years, respectively, to complete one revolution, or 1 in about 136 years,
as in the linear term in the standard Alfonsine model for trepidation which
is based on one revolution in exactly 49,000 years. These differences in the
periods depend on the seconds in the entries in Vimonds table and have no
astronomical significance. However, they indicate that Vimond is not using
the standard table for mean motion of the apogees and the fixed stars in the
Parisian Alfonsine Tables (Ratdolt 1483, f. d4v).
In 76 years the value in the text for the mean motion of the argument for
the fixed stars is 3;54,46 and in 2432 years it is 4s 5;4,38, corresponding to
0;0,0,30,26,47/d and 0;0,0,30,24,52/d, respectively. These values are equiva-
lent to 6,992 years and 7,000 years, respectively, to complete one revolution.
The periodic term in the standard Alfonsine model for trepidation is based on
one revolution in exactly 7,000 years, and it corresponds to 0;0,0,30,24,49/d.
These differences have no astronomical significance, but indicate that, once
again, Vimond is not using the standard table for mean motion of access and
recess in the Parisian Alfonsine Tables (Ratdolt 1483, f. d4r).
In fact, an entry for the mean motion of the argument is 7 times the corre-
sponding entry for the mean motion of the linear term.
As in the Parisian Alfonsine Tables, Vimond separates two terms for trepida-
tion: a linear term which corresponds to the difference between the calendar
year of 365;15 days and a fixed tropical year, and a periodic term which corre-
sponds to the difference between a variable sidereal year and the calendar year
of 365;15 days. But in his other tables Vimond has used a fixed sidereal year: we
are unable to account for this inconsistency. To be sure, Vimonds canons may
have explained what he intended.

f. 7v Table 27: Motion of the Fixed Stars

The argument is given at 6-intervals from 0s 6 to 12s 0 and the equation of


access and recess (here called motus) is given in degrees and rounded to min-
utes. In Table 27, below, the editors have supplied a minus sign in a few entries,
where appropriate. The table has a maximum of 17;17 for argument 204 and
a minimum of 0;43 for argument 24. These extremal values are 18 apart (=
17;17 + 0;43); hence the amplitude of the sinusoidal curve corresponding to
early alfonsine astronomy in paris 279

table 27 Motion of the fixed stars ( f. 7v)

argumentum motus argumentum motus


() () () ()

0 6 0;19 6 6 16;53
0 12 0;33 6 12 17; 7
0 18 0;41 6 18 17;15
0 24 0;43 6 24 17;17
1 0 0;39 7 0 17;13
1 6 0;29 7 6 17; 3
1 12 0;13 7 12 16;47
1 18 0; 8 7 18 16;26
1 24 0;35 7 14 15;59
2 0 1; 6 8 0 15;28
2 6 1;43 8 6 14;51
2 12 2;24 8 12 14;10
2 18 3; 8 8 18 13;26
2 24 3;56 8 24 12;38
3 0 4;47 9 0 11;47
3 6 5;40 9 6 10;54
3 12 6;35 9 12 9;58
3 18 7;30 9 18 9; 4
3 24 8;26 9 24 8; 8
4 0 9;22 10 0 7;11
4 6 10;18 10 6 6;16
4 12 11;12 10 12 5;22
4 18 12; 4 10 18 4;30
4 24 12;54 11 24 3;40
5 0 13;41 12 0 2;53
5 6 14;24 11 6 2;10
5 12 15; 4 11 12 1;30
5 18 15;39 11 18 0;55
5 24 16;15 11 24 0;19
6 0 16;34 12 0 0; 0

Vimonds table is 9. This is indeed the characteristic parameter of the table


for the equation of access and recess in the Parisian Alfonsine Tables, whose
maximum is 9 for argument 90.
280 chapter 8

Comparison of the entries in both tables shows that the curve representing
Vimonds table is the same as that used by other Parisian astronomers of his
time but shifted in two ways: 247 on the x-axis and 8;17 on the y-axis. In
fact, the entries in Vimonds table can be derived from those in the Parisian
Alfonsine Tables by taking an argument and its corresponding equation in the
latter (where they are given to seconds) and then adding 113 to the argument
and 8;17 to the equation.
Vimonds table begins at a point that in the Parisian Alfonsine Tables cor-
responds to a value of the equation of 8;17 and an argument of about 247.
The value for the equation of access and recess that Vimond thought correct
for his time was 8;17, and he shifted the curve (i.e., the entries in the table)
accordingly; indeed, calculation of the periodic term in trepidation with the
parameters for 1320 in the Parisian Alfonsine Tables yields 8;17 exactly:

1320 0;3,5,8,34,17/y 67;53


radix Incarnation 359;13

Total 67; 6

and

9 sin 67;6 = 8;17.

Note that 67;6 + 180 = 247;6 or about 247, and 360 247 = 113 which is
the phase angle of the shift introduced by Vimond.
This table establishes a strong connection between Vimond and the Parisian
Alfonsine Tables, for this theory of trepidation is not found in any previous text.
But again, since the mean motions are different (see Table 26), we see no reason
to assume that Vimond based his theory on the Parisian Alfonsine Tables.
Rather, Vimond may have depended on an Andalusian or Castilian tradition
that was closely related to (but distinct from) the Castilian Alfonsine Tables,
for there is no hint of phase shifts in the Castilian canons.

f. 8rv Table 28: Fixed Stars

This table displays the longitude, the latitude, and the magnitude of 225 stars
and nebulae but, in general, their names are omitted. The list is too long to be
related to an astronomical instrument, and the absence of star names makes
us wonder what purpose it was intended to serve. Both coordinates are given
early alfonsine astronomy in paris 281

to minutes. The stars are divided into three groups, in turn divided into sev-
eral subgroups according to the associated planets, a feature which is certainly
not common. Group i has 137 stars that belong to the zodiacal constellations
arranged in 52 subgroups, group ii has 44 stars in northern constellations
(19 subgroups), and group iii has 44 stars in southern constellations (19 sub-
groups); the total number of subgroups is thus 90. We note the balanced repre-
sentation of the stars on both sides of the zodiac.
We have found the same table in an early 14th-century copy, Cambridge,
Gonville and Caius College, ms 141/191, pp. 377382 (for an excerpt, see F.S. Ped-
ersen 2002, pp. 15071508); as well as in Segovia, Catedral, ms 84, ff. 46r51v; and
in Paris, BnF, ms 7482; ff. 61v69v. There are some cases where an entry in one
copy does not agree with the value in, or derived from, Ptolemys treatises in
contrast to the other copy, but there are also examples where entries in both
copies do not agree with those in Ptolemy. On the other hand, in all cases where
there is a blank entry in one copy, it is filled in the other copy.
In the Paris copy only 18 star names are given whereas in the Cambridge
copy this number is reduced to 15. The star names in these copies are generally
not identical, and they are not always ascribed to the same stars. For instance,
the names almalak and almalac are attributed, respectively, to the star in
the 20th subgroup (ms Paris) and to the first star of the 8th subgroup (ms
Cambridge). The star list does not bear a general title in the Paris copy but the
Cambridge copy reads tabula de dispositionibus stellarum fixarum existentibus
ad terminum complementi radicis mediarum coniunctionum solis et lunae quae
alibi signantur. Et primo de dispositionibus illarum stellarum quae sunt prope
viam solis. (Here begins the table on the groups of the fixed stars as they were
at the point of completion [the epoch?] of the radix of the mean conjunctions
of the Sun and the Moon specified elsewhere. First come the groups of those
stars close to the zodiac [lit.: the path of the Sun].)
The first sentence serves as a general title for the table, and the second sen-
tence is a heading for the groups in the zodiacal constellations, corresponding
to the headings in both manuscripts for the groups in the northern and south-
ern constellations. The expression the radix of the mean conjunctions seems
to refer to the radix given on f. 1r, 13;54,54d, which we identified with March
10, 1320. But we do not understand the expression at the end of the comple-
ment.
Madrid, Biblioteca Nacional, ms 4238, ff. 65v66v, reproduces the same star
list except that the signs used here are of 60, contrary to the other manuscripts
containing this list.
We are grateful to Paul Kunitzsch for information on two additional copies
of the same star list: Erfurt, Universittsbibliothek, ms Amplon. 2395, ff. 104v
282 chapter 8

105v; and Munich, Bayerische Staatsbibliothek, ms Clm 26667, ff. 46v47v (cf.
Kunitzsch 1986a, p. 96, n. 10, and p. 98, n. 44). In both manuscripts the list is
anonymous, but in the Erfurt copy a marginal note (in the same hand as the
list) reads: Notandum istas stellarum tabulas fuisse equatas ad annum domini
1338 (f. 105v). As Kunitzsch suggested to us (in a private communication), this
marginal note may have been added by the copyist and not belong to the orig-
inal list; no date appears in the other three manuscripts.
In fact, the list in the Erfurt ms has two extra stars: one is added to the
northern constellations, in subgroup 7 (Bootes), and the other to the southern
constellations, in subgroup 6 (Eridanus). We also note that in the list for the
southern constellations the stars in subgroup 19 (Ara) are located in the Erfurt
ms between subgroups 4 and 5 in the manuscripts in Paris and Cambridge (we
have not seen the manuscript in Munich). Another special feature of the Erfurt
ms is that the subgroups are not numbered; rather, most are given the name of
a star belonging to them or even a generic name. But its main distinguishing
characteristic is that the subgroups have no associated planets, in contrast to
the copies in Paris and Cambridge.
It may be of interest that the 5 manuscripts of which we are aware that
contain this star list are spread all over Europe: 2 in Germany, 1 each in England,
France, and Spain.
The order and the grouping of the stars in this list is peculiar, for they
do not follow the pattern of the catalogue in Ptolemys Almagest that was
generally adopted in medieval star lists and catalogues. Rather, this list is
organized according to Ptolemys Tetrabiblos, a handbook on astrology writ-
ten by Ptolemy after the Almagest. It was translated several times from Arabic
into Latin: in 1138 by Plato of Tivoli, in 1206 anonymously, and in 1256 via
Castilian at the court of Alfonso x by Egidius de Tebaldis (Chabs and Gold-
stein 2003a, p. 232), and was known as the Quadripartitum. In Tetrabiblos
i.9, Ptolemy grouped the stars into three main categories (zodiacal, north-
ern, and southern constellations), following an order differing from that in
the Almagest where the northern constellations precede the zodiacal constel-
lations, and grouped the stars within each category according to their asso-
ciated planets. As an example, we reproduce a passage of Tetrabiblos i.9
corresponding to the stars in the constellation of Aries (Robbins 1940, p.
47):

The stars in the head of Aries, then, have an effect like the power of
Mars and Saturn, mingled; those in the mouth like Mercurys power and
moderately like Saturns; those in the hind foot like that of Mars, and those
in the tail like that of Venus.
early alfonsine astronomy in paris 283

As is readily seen, the order, the subgroups, and the planets associated with
the stars in Aries in Vimonds list perfectly match those in Ptolemys Tetrabiblos.
And this is indeed the case for almost all stars in the 90 subgroups displayed in
Vimonds list.
The star positions generally agree with those in Gerard of Cremonas version
of Ptolemys star catalogue in the Almagest with an increment in longitude of
17;52 for precession, a value otherwise unattested. If the rate of precession was
taken to be 1 in 66 years, 17;52 would correspond to about 1179 years and,
if we add it to 137ad (the date of the star catalogue in the Almagest), we get
1316ad. But it is not clear that this date had any significance for the author. We
have compared this list to that in the Libro de las estrellas de la ochaua espera
(Madrid, Universidad Complutense, ms 156; see also Rico Sinobas 18631867,
vol. 1, pp. 5145), also known as Libro de las xlviii figuras de la viii spera or even
as Libro de las estrellas fixas. This is an adaptation of the star catalogue for 964
ad by the Persian astronomer al-f (903986) which in turn depended on the
star catalogue in Ptolemys Almagest (see Comes 1990). This work, where the
total precession is 17;8, was compiled in 1256 by Judah ben Moses ha-Cohen,
one of the most distinguished collaborators of Alfonso x. The presentation of
the star data in this Alfonsine text differs substantially from that of a typical
star list although the data themselves are what one would expect, namely,
for each star we are given its name, longitude, latitude, and magnitude. The
associated planets are also given for each star, often adding an indication of
their relative strength, showing that the Alfonsine Libro ultimately relied on
Ptolemys Quadripartitum. However, after comparing the data in the Libro
with those of Vimond, we see no evidence to suggest that the star list found
among Vimonds tables is systematically related to this Alfonsine book. As
Kunitzsch informed us, there is a star list by John of Lignres containing data
for 276 stars, but the longitudes are Alfonsine, i.e., Ptolemys values plus 17;8:
Paris, Bibliothque nationale de France, ms lat. 10264, ff. 36v38v, and Florence,
Biblioteca Nazionale Centrale, ms Conv. soppr. j.4.20, fols. 214v216r. This list
was extracted from the star table that later appeared in the editio princeps of
the Alfonsine Tables (1483), and sheds no additional light on the list included
in Vimonds tables.
Moreover, in the course of examining the star names in the five manuscripts
containing this list, Kunitzsch noticed that the author drew upon a variety of
Latin sources, mainly the translations of the Tetrabiblos but also sources not in
the Tetrabiblos tradition (some of which cannot be identified). Thus, Vimonds
list is dependent on Ptolemy in two ways: the choice of the stars, their order and
grouping, as well as the associated planets, are borrowed from the Quadriparti-
tum; and the numerical data are taken from the Latin version of the Almagest.
284 chapter 8

In sum, we believe that the star list attributed to Vimond in the Paris ms, and
that is anonymous in the Cambridge, Erfurt, Madrid, and Munich mss, derives
from an unknown archetype; we know of no similar star list in Latin in the 14th
century or in the previous Arabic literature with which to compare it.
In Table 28 we present in the first 3 columns a complete transcription of the
Paris copy with translations of the headings and the names of the associated
planets in each case. For the latitudes north is indicated by an abbreviation
of the term septentrionalis, and south by an abbreviation of meridionalis; we
have replaced them with the modern designations + and . Column iv gives the
few star names found in the Paris copy, which were added in interstitial spaces
within the table (some of the star names are partly hidden in the gutter of the
manuscript and cannot be read completely); column v lists the modern star
designation; column vi gives the standard number assigned to each of the 1028
stars in Ptolemys catalogue; column vii offers comparisons and comments,
together with variants in the Cambridge copy; and column viii provides the
identification of the star names.
table 28 Star list ( f. 8r v)

[Constellation] Associated planets


i ii iii iv v vi vii viii
Longitude Latitude Magn. Name Modern Number Comparisons Identification of
(sign, degrees) (degrees) designation (P.-K.) and comments star names

[Zodiacal constellations]

1 [Aries] Mars, Saturn


0 24;32 +7;20 3 Ari 362
0 25;32 +8;20 3 Ari 363
early alfonsine astronomy in paris

2 [Aries] Mercury, Saturn


0 28;52 +7;40 5 Ari 364 C(iii): blank
C(iv): flamai? Unidentified
0 29;22 +6; 0 5 Ari 365 C(iii): blank
C(iv): hercules Unidentified:
see Gem, below
3 [Aries] Mars
1 2;52 5; 5 4 Cet 374 C: +5;15, G: 5;15
1 5;52 1;30 5 Ari 373 C: +1;30
1 7;32 1;20 5 Ari 372 C: +1;20
G: 1;10, +1;10
285
table 28 Star list ( f. 8r v) (cont.)
286

i ii iii iv v vi vii viii

4 [Aries] Venus
1 9;12 +4;50 5 Ari 368
1 11;42 +1;40 4 Ari 369
1 13;12 +2;30 4 Ari 370
1 14;52 +1;50 4 Ari 371
5 [Taurus] Venus, Jupiter C: Moon
1 17;32 9;30 5 30(e) Tau 384
1 21;32 8; 0 3 Tau 385 C: +8; 0
6 [Taurus, The Pleiades] Moon, Mars
1 20; 2 +4;30 5 19 Tau 409
1 20;22 +4;40 5 23 Tau 410
1 20;32 +5; 5 5 27 Tau* 412
1 21;32 +5;20 5 bsc 1188* 411
7 [Taurus] Mars
2 0;32 5;10 1 aldebaran? Tau 393 C(iv): aldebaran G, p. 89 n. 10, etc.
chapter 8
i ii iii iv v vi vii viii

8 [Taurus] Saturn, Mercury


1 26;52 5;45 3 Tau 390 C(iv): almalac If this is a corruption of
Arabic al-malik (the king),
it should designate Leo
(Regulus). See G, p. 101
n. 12.
1 28;42 5;50 3 1 Tau 392 C: +5;50
G: 5;50, 0;50
early alfonsine astronomy in paris

1 29;42 3; 0 3 Tau 394 C: +3; 0


2 3;32 4; 0 4 Tau 399 C: +4; 0
G: 4; 0, +4; 0
9 [Taurus] Mars
2 7;52 3;30 5 106(l) Tau 397
2 8;12 5; 0 5 104(m) Tau 396 C: +5; 0
2 13;32 +5; 0 5 Tau 230/400 G: 3
2 15; 2 2;30 3 Tau 398 G + 17;52: 15;32
10 [Gemini] Mercury, Venus
2 24;22 1;30 4 Gem 437 C: +1;30
2 26; 2 1;15 4 Gem 438 C: +1;15
2 28; 2 3;30 4 Gem 439 C: +3;30
287
table 28 Star list ( f. 8r v) (cont.)
288

i ii iii iv v vi vii viii

2 29;52 7;30 3 Gem 440 C: +7;30


3 2;32 10;30 4 Gem 441 C: +10;30
11 [Gemini] Saturn
3 9;32 5;30 3 Gem 435
12 [Gemini] Mars
3 11;12 +9;40 2 ( )annai? Gem 424
13 [Gemini] Mars
3 14;32 [..]6;15 [..] hercules? Gem 425 C(ii): +6;15. C(iii): 2 R, p. 48: Herakles
C(iv): almueredan K1959, p. 127: Vir is
called almuredin
14 [Cancer] Mercury, Mars
3 20;32 +1; 0 5 Cnc 456
3 25; 2 7;30 4 Cnc 457
15 [Cancer] Saturn, Mercury
3 26;12 +11;50 4 Cnc 455
4 4;22 +5;30 4 Cnc 454 G: 5;30
16 [Cancer] Moon, Mars
3 28;12 +0;40 n meollef? gc 2632 449 C: 2. C(iv): mellef? P, f. 15va: meelef
Galaxy m 44
chapter 8
i ii iii iv v vi vii viii

17 [Cancer] Mars, Sun


3 28;12 +2;40 4 assinis? Cnc 452 C(iv): asini G
3 29;12 +0;10 4 Cnc 453 G: 0;10
18 [Leo] Saturn, Mars
4 12; 2 +9;30 3 Leo 465
4 12; 2 +12; 0 3 Leo 464 G + 17;52: 12;12
19 [Leo] Saturn, Mars C: Mercury
4 18; 2 +11; 0 3 Leo 466
4 18;32 +4;30 3 Leo 468
early alfonsine astronomy in paris

4 20; 2 +7;30 2 Leo 467 G: +8;30


20 [Leo] Mars, Jupiter
4 20;22 +0;10 1 almalak? Leo 469 G, p. 101 n. 12.
21 [Leo] Venus, Saturn
4 29;12 +13;15 5 60(b) Leo 480 C, G: +12;15
5 2; 2 +13;40 2 Leo 481
5 2;12 +11;30 5 81 Leo* 482
5 4;12 + 9;40 3 Leo 483
5 12;22 +12;50 1 ?? Leo 488 G: +11;50
22 [Leo] Venus, Mercury
5 8;12 +5;50 3 Leo 484
5 8;22 3; 0 5 Leo 487
289
table 28 Star list ( f. 8r v) (cont.)
290

i ii iii iv v vi vii viii

5 9;32 +0;50 4 Leo 486


5 9;32 +1;15 4 Leo 485
23 [Virgo] Mercury, Mars
5 14;12 +4;35 5 Vir 497
5 14;52 +5;40 5 Vir 498
5 17; 2 +6; 0 3 Vir 501 G + 17;52: 16;52
5 18; 2 +5;30 5 Vir 500
24 [Virgo] Mercury, Venus
5 26; 7 +1;10 3 Vir 502
6 1; 2 +2;50 3 Vir 503
25 [Virgo] Saturn, Mercury
6 0; 2 +15;10 3 Vir 509
26 [Virgo] Venus, Mercury
6 14;32 2; 0 1 almure? Vir 510 C(iv): alcimech G: ascimech
27 [Virgo] Mercury, Mars
6 24;32 +7;30 4 Vir 518
6 25;12 +2;40 4 Vir 519
6 7;52 +0;30 4 Vir 521 C, G + 17;52: 27;52
7 0;32 +9;50 4 Vir 522
chapter 8
i ii iii iv v vi vii viii

28 [Libra] Jupiter, Mercury


7 5;52 +0;40 2 Lib 529
7 10; 2 +8;30 2 Lib 531
29 [Libra] Saturn, Mercury
7 9;12 [..]1;15 [..] Lib 534 C(ii): +1;15. C(iii): 4
7 11;52 [..]1;40 [..] Lib 533 C(ii): +2;40. C(iii): 4
7 15;22 [..]3;45 [..] Lib 535 C(ii): +3;45. C(iii): 4
7 20;52 [..]4;30 [..] Lib 536 C(ii): +4;30. C(iii): 4
30 [Scorpius] Mars, Saturn
early alfonsine astronomy in paris

7 23;32 1;40 3 Sco 547


7 23;32 5; 0 3 Sco 548 C: +5; 0
7 24;12 +1;20 3 Sco 546
31 [Scorpius] Mars, Jupiter
8 0;29 4; 0 2 Sco 553 G + 17;52: 0;32
32 [Scorpius] Saturn, Venus
8 5;52 15; 0 4 1 + 2 Sco 558 G + 17;52: 6;42
8 11; 2 19;30 3 Sco 561 C: +19;30
8 16; 2 18;50 3 Sco 562 C: +18;50
8 15;52 15;10 3 Sco 564 G + 17;52: 16;52
C(ii): +16;10
8 18;22 16;40 3 1 Sco 563 C: +16;40
291
292

table 28 Star list ( f. 8r v) (cont.)

i ii iii iv v vi vii viii

33 [Scorpius] Mercury, Mars, Moon C: [blank]


8 15;52 23;30 4 Sco 566 C, G + 17;52: 14;52
C, G: 13;30
8 14;22 13;20 3 Sco 565 C, G + 17;52: 15;22
C(ii): +13;20
34 [Scorpius] Mars, Moon
8 19; 2 13;15 n G Sco* 567 C: 2
+ CGlo 6441
35 [Sagittarius] Saturn, Moon
8 22;22 6;30 3 Sgr 570
9 0;52 3;50 4 Sgr 576 C: +3;50
36 [Sagittarius] Jupiter, Mars
8 24;32 +2; 7 4 Sgr 574
8 26;52 1;30 3 Sgr 573
37 [Sagittarius] Mercury, Jupiter, Sun, Mars C: Moon
9 3; 2 7;45 n 1 + 2 Sgr 577 G: 0;45. C(iv): 2
38 [Sagittarius] Jupiter, Mercury
9 4;12 6;45 3 Sgr 591
9 5;32 2;30 4 Sgr 590 C: +2;30, G: 4;30
9 7;52 2;30 5 Sgr 589 C: +2;30
chapter 8
i ii iii iv v vi vii viii

39 [Sagittarius] Jupiter, Saturn


9 4;52 18; 0 2 Sgr 593
9 5;32 23; 0 2 1 + 2 Sgr 592 C: +23; 0
9 14;32 13; 0 3 Sgr 594 G + 17;52: 24;32
C(ii): +13; 0
40 [Sagittarius] Venus, Saturn
9 16;32 5;50 5 59(b) Sgr 599 C, G + 17;52: 16;22
9 16;32 4;50 5 60(a) Sgr 598 G + 17;52: 15;32
C(ii): +4;50
early alfonsine astronomy in paris

9 16;42 4;50 5 Sgr 597 C: +4;50


9 17;32 6;30 5 62(c) Sgr 600 C: +6;30
41 [Capricornus] Mars, Venus
9 25;12 +2;20 3 1 + 2 Cap 601
9 25;12 +5; 0 3 Cap 603
9 26;42 +1;30 6 Cap 607
9 26;52 +0;45 6 Cap 605
42 [Capricornus] Mars, Mercury
9 29;32 8;40 4 Cap 612
10 4;32 7;40 4 24(a) Cap 613 C: +7;40
10 8; 2 6;50 4 Cap 614 C: +6;50
10 8;12 6; 0 5 36(b) Cap 615 C: +6; 0
293
table 28 Star list ( f. 8r v) (cont.)
294

i ii iii iv v vi vii viii

43 [Capricornus] Saturn, Mercury


10 12;42 +2;10 3 Cap 623 G: +2;10, 2;10
10 14;12 +2; 0 3 Cap 624 G: +2; 0, 2; 0
10 14;42 0;20 4 42(d) Cap 625 G: +0;20
10 15;32 2;50 5 Cap 627 C, G: +2;50
44 [Aquarius] Saturn, Mercury
10 2;32 +8;40 3 Aqr 636
10 4; 2 +8; 0 4 Aqr 635
10 14;22 +8;50 2 Aqr 632
10 24;12 +11;15 4 Aqr 630 G: +11;0. G: 3
45 [Aquarius] Mercury, Saturn
10 19;12 5; 0 4 Aqr 647 G + 17;52: 29;12
10 19;32 7;30 3 Aqr 646 G + 17;52: 29;32
C, G: +7;30
10 22;32 5;40 5 53(f) Aqr 648 C: +5;40
46 [Aquarius] Saturn, Jupiter
11 5;32 1; 0 4 83(h) Aqr 653
11 6;52 7;30 4 1 Aqr 656 C: +7;30, G: 8;30
11 7;52 0;30 4 Aqr 654 C: +7;30
11 8;12 1;40 4 a[n]phora Aqr 655 C: +1;40 See note 1.
chapter 8
i ii iii iv v vi vii viii

47 [Pisces] Mercury, Saturn


11 9;32 +9;15 4 Psc 674
11 12; 2 +7;30 4 Psc 675
11 13;52 +9;20 4 7(b) Psc 676
48 [Pisces] Jupiter, Mercury
11 13;52 +4;30 4 Psc 679
11 17;32 +2;30 4 Psc 680
49 [Pisces] Saturn, Mercury
11 23;52 +6;20 4 Psc 681
early alfonsine astronomy in paris

11 28;52 +5;45 6 41(d) Psc 682


50 [Pisces] Jupiter, Venus
0 17;12 +15;20 4 Psc 706 G + 17;52: 17;22
0 20; 2 +17; 0 4 Psc 705
51 [Pisces] Saturn, Jupiter
0 13;32 +14;20 4 1 Psc 702
0 14;12 +13; 0 4 2 Psc 703
0 15;32 +12; 0 4 Psc* 704
52 [Pisces] Mars, Mercury
0 20;22 8;30 3 Psc 692
295
table 28 Star list ( f. 8r v) (cont.)
296

i ii iii iv v vi vii viii

[Title:] Then follow the constellations (dispositio) of the other fixed stars in the northern part.

[Northern constellations: all latitudes are positive]

1 [Ursa Minor] Saturn, Venus


4 5; 2 72;50 2 aliedin UMi 6 C(iv): aliedim See note 2.
4 14; 2 74;50 2 alforcami UMi 7 C(iv): alfoza K1961, p. 58: al-farqadn
(+ UMi)
2 [Ursa Maior] Moon, Venus
5 0; 2 53;30 2 UMa 33 G: 13;30
5 5;52 55;40 2 benezna UMa 34 G: 15;40 K1966, p. 42, no. 23:
benenaz ( UMa)
5 17;42 54; 0 2 UMa 35 G: 14; 0
3 [Draco] Saturn, Mars
5 26;22 84;50 3 Dra 67
5 27;52 88; 0 3 Dra 68 G: 78; 0
4 [Cepheus] Saturn, Jupiter
0 4;32 69; 0 3 Cep 78
0 25;22 71;10 4 Cep 77
11 27;12 72; 0 4 Cep 79
chapter 8
i ii iii iv v vi vii viii

5 [Hercules] Saturn, Mars


7 28; 2 53;30 4 Her 130 G: 13;30
7 29;42 54;10 3 Her 129 G + 17;52: 21;42
G: 16;10
8 1;52 59;50 3 Her 133 G: 19;50, 59;50
8 3;12 60;20 4 69(e) Her 134
6 [Corona Borealis] Venus, Mercury
6 29;32 46;10 4 CrB 112
7 2;32 44;30 2 alfeca CrB 111 C(iv): alfeca G
early alfonsine astronomy in paris

7 5; 2 44;45 4 CrB 115


7 7; 2 44;50 4 CrB 116
7 [Bootes] Mercury
6 9;12 28; 0 3 Boo 107
8 [Lyra] Venus, Mercury
9 5;12 62; 0 1 lilurah Lyr 149 C(iv): lulurach G: allore
9 [Perseus] Saturn, Jupiter
1 17;29 23; 0 2 eiumezuz? Per 202 G + 17;52: 17;32 Unidentified
10 [Perseus] Mars, Mercury
1 22;42 30; 0 2 Per 197
11 [Auriga] Mars, Mercury
2 12;52 22;30 1 alhaioch Aur 222 C(iv): alhaioch G
297
table 28 Star list ( f. 8r v) (cont.)
298

i ii iii iv v vi vii viii

12 [Ophiuchus] Saturn, Venus


8 12;42 36; 0 3 alhanue Oph 234 C(iv): alhanue K1966, p. 55, no. 33:
alhaue, alhane
13 [Serpens] Saturn, Mars
7 12;12 25;30 3 Ser 271 G: 25;20
7 12;42 36;30 4 Ser 270 C, G: 26;30
7 14;12 24; 0 3 Ser 272
7 16;32 16;30 4 Ser 273
14 [Sagitta] Mars, Venus
9 24;32 39;10 6 Sge 282
9 28; 2 39;20 4 Sge 281
15 [Aquila] Jupiter, Mars
9 21;42 29;10 2 vultur Aql 288 C: 19;10 G
C(iv): vultur
16 [Delphinus] Saturn, Mars
10 6;22 32; 0 3 Del 304
10 8; 2 33;50 3 Del 305
10 9;12 32; 0 3 Del 306
10 11;22 32;10 3 Del 307 G: 33;10
chapter 8
i ii iii iv v vi vii viii

17 [Pegasus] Mars, Mercury


0 15; 2 12;31 3 Peg 316 G + 17;52: 10; 2
G: 12;30. C, G: 2
11 20; 2 31; 0 2 Peg 317
18 [Andromeda] Mars, Venus
0 13;32 15; 7 3 And 345
0 19;42 30; 0 3 And 347
0 19;52 32;30 3 And 348
0 25;42 26;20 3 And 346 G + 17;52: 21;42
early alfonsine astronomy in paris

C(ii): +16;20
1 4;42 23; 0 3 And 349
19 [Triangulum] Mercury
0 28;52 16;30 3 Tri 358
1 3;52 20;40 3 Tri 359

[Title:] Then follow the constellations (dispositio) of the other stars in the southern part.

[Southern constellations: all latitudes are negative]

1 [Piscis Austrinus] Mars, Venus, Mercury


10 9;42 16;30 4 PsA 1020
299
300

table 28 Star list ( f. 8r v) (cont.)

i ii iii iv v vi vii viii

10 16; 2 15; 0 4 PsA 1019 C, G + 17;52: 13; 2


10 16;42 14; 0 4 PsA 1018 G: 14;40
10 18;32 20;20 4 PsA 1012
2 [Cetus] Saturn
0 12;52 20; 0 2 Cet 725
3 [Orion] Mars, Mercury
2 19;52 17; 0 1 Ori 735
4 [Orion] Jupiter, Saturn
2 7;42 31;30 1 Ori 768
2 13;12 24;10 2 Ori 759
2 15;12 24;50 2 Ori 760
2 16; 2 25;40 2 Ori 761
5 [Eridanus] Jupiter
0 18; 2 53;30 1 Eri 805 C(i): 16; 2, G: 13;30
6 [Eridanus] [..] C: Saturn
2 5;12 31;50 4 Eri 772 G + 17;52: 6;12
7 [Lepus] Saturn, Mars
2 12;42 44;20 3 Lep 813
2 13;22 41;30 3 Lep 812
chapter 8
i ii iii iv v vi vii viii

8 [Canis Maior] Venus


2 13;52 57;40 2 Col 845
2 16;52 59;40 2 Col 844 C: 57;40
9 [Canis Maior] Jupiter, Mars
3 5;32 39;10 1 CMa 818
3 7;32 35; 0 4 CMa 819
10 [Canis Minor] Mercury, Mars
3 13;22 14; 0 4 CMi 847 G + 17;52: 12;52
3 17; 2 16;10 1 CMi 848
early alfonsine astronomy in paris

11 [Hydra] Saturn, Jupiter


4 17;52 20;30 2 Hya 905
4 23;52 26;30 4 Hya 906
4 26;32 26; 0 4 1 Hya 907
12 [Crater] Venus, Mercury
5 17;52 18; 0 4 Crt 923
5 17;52 18;30 4 Crt 924 G + 17;52: 24;52
5 20;22 19;30 4 Crt 922
13 [Corvus] Saturn, Mercury
6 2;12 19;40 3 Crv 929
6 6;22 14;50 3 Crv 931 C: 6;12
301
302

table 28 Star list ( f. 8r v) (cont.)

i ii iii iv v vi vii viii

14 [Argo] Saturn, Jupiter


3 5; 2 69; 0 1 Car 892 G: 29; 0
15 [Centaurus] Mars, Venus
6 24; 2 25;40 3 Cen 939
7 3;32 22;30 3 Cen 940 C: 2
16 [Centaurus] Venus, Jupiter
6 26;13 41;10 1 Cen 969 G + 17;52: 26;12
C: 26;12
6 27;52 51;10 2 Cru 965
6 29; 2 55;20 2 Cru 968
7 3;12 51;40 2 Cru 966
7 12; 2 45;20 2 Cen 970 C: 45;[..]
17 [Lupus] Venus, Mars
7 13;42 29;10 3 Lup* 973
7 15;52 24;10 3 Lup 972 G: 24;50
18 [Corona Australis] Saturn, Mercury
9 4;22 15;20 4 Cr a 1005 G: 15;10
9 4;42 16; 0 4 Cr a 1004
9 5;12 17;10 4 Cr a 1003 G + 17;52: 4;52
chapter 8
i ii iii iv v vi vii viii

19 [Ara] Jupiter, Mercury


8 8;32 30;20 4 1 Ara 994 G: 5
8 12;52 33;20 4 Ara 996
8 13; 2 34;10 4 Ara 995
early alfonsine astronomy in paris
303
304 chapter 8

Col. i. The number of the zodiacal sign is not repeated in col. vii where variants
are listed; in all cases reported in that column only the degrees and minutes
differed from the entry in the Paris manuscript.

Col. iii: n means nebulous.

Col. iv: In the manuscript the names of the stars are not presented in a column.

Col. v: The entries in this column have been taken from Toomer 1984.
* indicates that Kunitzsch 1986 and Kunitzsch 1991, pp. 187200, give a
different modern designation.

Col. vi: These numbers are taken from Peters and Knobel 1915 (ultimately from
Baily 1843), and they are also used in Kunitzsch 1986 and 1990.

Col. vii: C refers to Cambridge, Gonville and Caius College, ms 141/191; in certain
cases, it is followed by a column number in Roman numerals. G refers to
Gerard of Cremonas version of Ptolemys star catalogue (Kunitzsch 1990). We
underline entries in Vimonds table for which there is a variant reading. The
entries for longitudes in both copies generally agree with those in G with an
increment of 17;52 for precession; those cases where they differ have been
noted.

Col. viii: G refers to Gerard of Cremonas version of Ptolemys star catalogue


(Kunitzsch 1990); K1959 refers to Kunitzsch 1959; K1966 refers to Kunitzsch
1966; P refers to Plato of Tivolis Latin version of the Tetrabiblos (ed. 1493); and
R refers to Robbins 1940.

Note 1. We are informed by Kunitzsch that anphora is not a proper name; rather,
it is a noun used in the description of the stars position: where the water flows
out from the vessel; Erfurt, Universittsbibliothek, Amplon. 2395, f. 105r, in
decursu aque ab anphora; p, f. 16va: In aque vero decursu collocate (without
anphora).

Note 2. As Kunitzsch informed us, aliedim, apparently renders the Arabic al-
jady (the kid), an old Arabic name for UMi (Kunitzsch 1961, p. 62). It is uncer-
tain where the compiler of this list might have found it. In the Tetrabiblos tra-
dition this name never occurs.
early alfonsine astronomy in paris 305

Acknowledgments

We thank Paul Kunitzsch and Beatriz Porres for assistance with the Latin texts
cited in this article, and Fritz S. Pedersen, John D. North, and Julio Sams for
detailed comments on a preliminary version of this paper.

References

as-Saleh 1970. See Saleh, J.A. as- 1970.


Baily, F. 1843. The Catalogues of Ptolemy, Ulugh Beigh, Tycho Brah, Halley, Hevelius.
Memoirs of the Royal Astronomical Society, vol. 13. London.
bsc. See Hoffleit (ed.) 1964.
Boudet, J.-P. 19971999. Le Recueil des plus celebres astrologues de Simon de Phares, 2
vols. Paris.
Casulleras, J. and J. Sams (eds.) 1996. From Baghdad to Barcelona: Studies in the Islamic
Exact Sciences in Honour of Prof. Juan Vernet. Barcelona.
Chabs, J. 2000. Astronoma alfons en Morella a finales del siglo xiv, Cronos: Cuader-
nos Valencianos de Historia de la Medicina y de la Ciencia, 3:381391.
Chabs, J. and B.R. Goldstein 1994. Andalusian Astronomy: al-Zj al-Muqtabis of Ibn
al-Kammd, Archive for History of Exact Sciences, 48:141.
Chabs, J. and B.R. Goldstein 1997. Computational Astronomy: Five centuries of Find-
ing True Syzygy, Journal for the History of Astronomy, 28:93105.
Chabs, J. and B.R. Goldstein 2003a. The Alfonsine Tables of Toledo. Dordrecht.
Chabs, J. and B.R. Goldstein 2003b. John Vimond and the Alfonsine Trepidation
Model, Journal for the History of Astronomy, 34:163170.
Comes, M. 1990. Al-f como fuente del libro de la Ochaua Espera de Alfonso x, in
Comes, Mielgo, and Sams (eds.) 1990, pp. 11113.
Comes, M., H. Mielgo, and J. Sams (eds.) 1990. Ochava espera y astrofsica. Barce-
lona.
Copernicus, N. 1543. De revolutionibus. Nuremberg.
Goldstein, B.R. 1974. The Astronomical Tables of Levi ben Gerson. Transactions of the
Connecticut Academy of Arts and Sciences, 45. New Haven.
Goldstein, B.R. 1996. Lunar Velocity in the Middle Ages: A Comparative Study, in
Casulleras and Sams (eds.) 1996, pp. 181194.
Goldstein, B.R. 2001. The Astronomical Tables of Judah ben Verga, Suhayl, 2:227289.
Goldstein, B.R. 2003. An Anonymous Zij in Hebrew for 1400ad: A Preliminary Report,
Archive for History of the Exact Sciences, 57:151171.
Goldstein, B.R. and J. Chabs 2004. Ptolemy, Bianchini, and Copernicus: Tables for
Planetary Latitudes, Archive for History of Exact Sciences, 58:453473.
306 chapter 8

Hoffleit, D. (ed.) 1964. Catalogue of Bright Stars. Yale University Observatory. New
Haven.
King, D.A. and M.H. Kennedy (eds.) 1983. Studies in the Islamic Exact Sciences. Beirut.
Kremer, R.L. 2003. Wenzel Fabers Table for Finding True Syzygy, Centaurus, 45:305
329.
Kremer, R.L. and J. Dobrzycki 1998. Alfonsine meridians: Tradition versus experience
in astronomical practice c. 1500, Journal for the History of Astronomy, 29:187199.
Kunitzsch, P. 1961. Untersuchungen zur Sternnomenklatur der Araber. Wiesbaden.
Kunitzsch, P. 1966. Typen von Sternverzeichnissen in astronomischen Handschriften des
zehnten bis vierzehnten Jahrhunderts. Wiesbaden.
Kunitzsch, P. 1986a. The star catalogue commonly appended to the Alfonsine Tables,
Journal for the History of Astronomy, 17:8998.
Kunitzsch, P. (ed. and tr.) 1986b. Claudius Ptolemus, Der Sternkatalog des Almagest: Die
arabisch-mittelalterliche Tradition. i: Die arabischen bersetzungen. Wiesbaden.
Kunitzsch, P. (ed.) 1990. Claudius Ptolemus, Der Sternkatalog des Almagest: Die ara-
bisch-mittelalterliche Tradition. ii: Die lateinische bersetzung Gerhards von Cre-
mona. Wiesbaden.
Kunitzsch, P. 1991. Claudius Ptolemus, Der Sternkatalog des Almagest: Die arabisch-
mittelalterliche Tradition. iii: Gesamtkonkordanz der Stern-koordinaten. Wiesba-
den.
Mestres, A. 1996. Maghrib Astronomy in the 13th Century: a Description of Manuscript
Hyderabad Andra Pradesh State Library 298, in Casulleras and Sams (eds.) 1996,
pp. 383443.
Mestres, A. 1999. Materials Andalusins en el Zj dIbn Isq al-Tnis. Doctoral Thesis,
University of Barcelona.
Mills, J.M. 19431950. Estudios sobre Azarquiel. MadridGranada.
Nallino, C.A. 19031907. Al-Battn sive Albatenii Opus Astronomicum, 2 vols. Milan.
Neugebauer, O. 1962. The Astronomical Tables of al-Khwrizm. Copenhagen.
Neugebauer, O. 1975. A History of Ancient Mathematical Astronomy. Berlin.
North, J.D. 1976. Richard of Wallingford, 3 vols. Oxford.
Pedersen, F.S. 2002. The Toledan Tables: A review of the manuscripts and the textual
versions with an edition. Copenhagen.
Pedersen, O. 1974. A Survey of the Almagest. Odense.
Peters, C.H.F. and E.B. Knobel 1915. Ptolemys Catalogue of Stars: A Revision of the
Almagest. Washington.
Poulle, E. 1973. John of Lignres, in The Dictionary of Scientific Biography, 7:122128.
New York.
Poulle, E. 1984. Les tables alphonsines avec les canons de Jean de Saxe. Paris.
Ptolemy. Quadripartitum. See G. Salio (ed.) 1493.
Ratdolt, E. (ed.) 1483. Tabule astronomice illustrissimi Alfontij regis castelle. Venice.
early alfonsine astronomy in paris 307

Rico Sinobas, M. 18631867. Libros del Saber de Astronoma del Rey D. Alfonso x de
Castilla, 5 vols. Madrid.
Robbins, F.E. (ed. and trans.) 1940. Ptolemy: Tetrabiblos. London.
Saby, M.-M. 1987. Les canons de Jean de Lignres sur les tables astro-nomiques de 1321.
Unpublished thesis: Ecole Nationale des Chartes, Paris. A summary appeared as:
Les canons de Jean de Lignres sur les tables astronomiques de 1321, cole Nationale
des Chartes: Positions des thses, pp. 183190.
Saleh, J.A. as- 1970. Solar and Lunar Distances and Apparent Velocities in the Astro-
nomical Tables of abash al-sib, Al-Abhath, 23:129176. Reprinted in King and
Kennedy (eds.) 1983.
Salio, G. (ed.) 1493. Liber quadripartiti Ptholemei; Venice.
Sams, J. and E. Mills 1998. The Computation of Planetary Longitudes in the Zj of Ibn
al-Bann, Arabic Sciences and Philosophy, 8:259286.
Sdillot, J.-J. and L.-A. Sdillot 1834. Trait des instruments astronomiques des Arabes.
Paris. Reprinted Frankfurt a/M (1984).
Swerdlow, N.M. and O. Neugebauer 1984. Mathematical Astronomy in Copenricuss De
revolutionibus. New York and Berlin.
Thorndike, L. and P. Kibre 1963. A catalogue of incipits of mediaeval scientific writings in
Latin. London.
Toomer, G.J. 1968. A Survey of the Toledan Tables, Osiris, 15:5174.
Toomer, G.J. 1984. Ptolemys Almagest. New York and Berlin.
chapter 9

John of Murss Tables of 1321*


To John D. North (19342008), in memoriam


John of Murs (fl. 13201340), a scholar active in Paris in the first half of the four-
teenth century, was a key figure in the history of astronomy in addition to mak-
ing contributions to music and mathematics. Indeed, his work on astronomy
played a decisive role in the transmission of scientific ideas in the late Middle
Ages.1 He is largely responsible for the introduction of Alfonsine astronomy into
the Parisian milieu, notably a set of astronomical tables originally elaborated
in Toledo by the astronomers in the service of King Alfonso of Castile (d. 1284).

* Journal for the History of Astronomy, 40 (2009), 297320.


1 L. Gushee, New Sources for the Biography of Johannes de Muris, Journal of the American
Musicological Society, xxii (1969), 326; E. Poulle, John of Murs, in The Dictionary of Scientific
Biography (16 vols, New York, 19701980), vii (1973), 128133; G. Beaujouan, Observations et
calculs astronomiques de Jean de Murs (13211344), in Proceedings of the xivth International
Congress of the History of Science (TokyoKyoto 1974) (Tokyo, 1975), ii, 2730, reprinted in idem,
Par raison des nombres: Lart du calcul et les savoirs scientifiques mdivaux (Aldershot, 1991),
Essay vii; J.D. North, The Alfonsine Tables in England, in Y. Maeyama and W.G. Salzer
(eds), Prismata: Festschrift fr Willy Hartner (Wiesbaden, 1977), 269301; G. lHuiller, Aspects
nouveaux de la biographie de Jean de Murs, Archives dhistoire doctrinale et littraire du
moyen ge, xlvii (1980), 272276; C. Schabel, John of Murs and Firmin of Beauvals Letter
and Treatise on the Calendar Reform for Clement vi, Cahiers de lInstitut du moyen-ge
grec et latin, lxvi (1996), 187215; J. Chabs and B.R. Goldstein, The Alfonsine Tables of Toledo
(Dordrecht and Boston, 2003); M. Lejbowicz, Prsentation de Jean de Murs observateur et
calculateur sagace et laborieux , in C. Grelland (ed.), Mthodes et statut des sciences la
fin du Moyen ge (Villeneuve d Ascq, 2004), 159180; R.L. Kremer, John of Murs, Wenzel
Faber and the Computation of True Syzygy in the Fourteenth and Fifteenth Centuries, in
J.W. Dauben et. al. (eds), Mathematics Celestial and Terrestrial: Festschrift fr Menso Folkerts
zum 65. Geburtstag (Halle [Saale], 2008), 147160.

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_011


john of murss tables of 1321 309

Together with John of Lignres and John of Saxony, he recast these tables into
what are now called the Parisian Alfonsine Tables.
Although in 1317 John of Murs was a convinced defender of the Toledan
Tables, as attested in his earliest known work, beginning Auctores kalendarii ,
a few years later he adhered to Alfonsine astronomy and remained faithful to
it thereafter. In particular, in his Expositio intentionis regis Alfonsii circa tabulas
eius (probably composed in 1321) he explained parameters and models already
found in these tables and described many of the features of what he called the
tables of Alfonso, but remained silent on the way he got access to them. In any
case, the Alfonsine material was available to him (and others) in Paris by 1321.2
Reconstructing the transmission of astronomical ideas is a complex task
that is especially difficult for a period when scholars rarely mentioned the
names of their contemporaries or near contemporaries on whose work they
depended, as is the case in the late Middle Ages. In his Expositio John of
Murs mentions only one of his predecessors in Paris, William of Saint-Cloud
(end of the thirteenth century), and he even reproduces parts of Saint-Clouds
Almanach Planetarum of 1292 almost word for word (without marking any of
these passages as quotations).3 Yet some parameters of Alfonsine origin that
John of Murs incorporated into his own tables (such as the maximum value
of the solar equation, 2;10) had already reached Paris, for we find them in the
tables of John Vimond for 1320.4 It seems unlikely that John of Murs did not
know about John Vimond since both astronomers came from the same region,
Normandy, and both worked on planetary tables in Paris at the same time.
John Vimond composed his tables for the use of students at the University
of Paris and was therefore a known participant in the Parisian astronomical
community.
The texts under consideration in this paper have not previously been studied
in detail,5 but we claim that they are of great importance for understanding
the transmission of Alfonsine astronomy from Toledo to Paris. In fact, there
were at least five texts produced in Paris in 1320 and shortly thereafter that
bear on this transmission: (1) John Vimonds tables of 1320, (2) John of Murss

2 Chabs and Goldstein, Alfonsine Tables of Toledo (ref. 1), 277281.


3 See E. Poulle, Jean de Murs et les tables alphonsines, Archives dhistoire doctrinale et littraire
du moyen ge, xlvii (1980), 241271, especially pp. 261265; Chabs and Goldstein, Alfonsine
Tables of Toledo (ref. 1), 263264, 279280.
4 See B.R. Goldstein and J. Chabs, The Maximum Solar Equation in the Alfonsine Tables,
Journal for the History of Astronomy, xxxii (2001), 345348; J. Chabs and B.R. Goldstein, Early
Alfonsine Astronomy in Paris: The Tables of John Vimond (1320), Suhayl, iv (2004), 207294.
5 For a brief account, see North, Alfonsine Tables in England (ref. 1), 284285.
310 chapter 9

Expositio, (3) John of Murss tables of 1321, (4) John of Murss Patefit, and (5)
John of Lignress canons and tables of 1322. We have previously discussed the
evidence in items (1), (2), and (5), and in this paper we focus on items (3) and
(4).6 As we noted in an article published in 1994 (with Jos Luis Mancha), one
lengthy chapter on planetary velocity in the Latin canons to the tables of the
Parisian astronomer, John of Lignres (fl. 13201335), is almost identical with
a chapter in the Castilian canons to the Alfonsine Tables of Toledo, composed
some 50 years earlier.7 As we demonstrate in this paper, John of Murss tables of
1321 and the tables for syzygies in his Patefit are based on the same models and
parameters that underlie the Parisian Alfonsine Tables (although the formats
for these tables are entirely different), indicating that, for matters other than
presentation, John of Murss contribution to Alfonsine astronomy was made all
at once. Some of these parameters are not known from any text or table prior
to those of John Vimond and John of Murs, although others can be discerned
in Castilian material of the late 13th century.8 Hence, what we learn from items
(3) and (4) confirms our results, based on items (1), (2), and (5). In other words,
despite the fact that the Alfonsine Tables of Castile are not extant, the evidence
we present strongly supports the claim that the Parisian material produced in
the 1320s relied on a Castilian tradition associated with King Alfonso x, as John
of Murs himself asserts.9
John of Murs compiled several sets of tables. They have not yet been thor-
oughly examined, except the set called Tabulae permanentes, which is re-
stricted to the computation of the time from mean to true syzygy.10 The tables
of 1321 are his first and most extensive set, and they are entirely devoted to the
planets and the two luminaries, the Sun and the Moon, that is, matters related

6 See, e.g., Chabs and Goldstein, Alfonsine Tables of Toledo (ref. 1), 266284; Chabs and
Goldstein, John Vimond and the Alfonsine Trepidation Model, Journal for the History of
Astronomy, xxxiv (2003), 163170; Chabs and Goldstein, John Vimond (ref. 4).
7 B.R. Goldstein, J. Chabs, and J.L. Mancha, Planetary and Lunar Velocities in the Castilian
Alfonsine Tables, Proceedings of the American Philosophical Society, cxxxviii (1994), 6195.
8 See J. Chabs, Were the Alfonsine Tables of Toledo First Used by Their Authors?, Centau-
rus, xlv (2003), 142150.
9 For the evidence in John of Murss Expositio, see Goldstein and Chabs, Maximum Solar
Equation (ref. 4), 347 n. 2. For contrary views, see E. Poulle, The Alfonsine Tables and
Alfonso x of Castille, Journal for the History of Astronomy, xix (1988), 97113; and idem,
Les astronomes parisiens au xive sicle et l astronomie alphonsine, Histoire littraire de
la France, xliii (2005), 154.
10 B. Porres and J. Chabs, John of Murss Tabulae permanentes for finding true syzygies,
Journal for the History of Astronomy, xxxii (2001), 6372.
john of murss tables of 1321 311

to the daily rotation and trigonometry are not mentioned.11 That John of Murs is
the author of these tables is attested by a note accompanying the text that refers
to them as istas tabulas magistri Jo. de muris (Oxford, Bodleian Library, ms
Can. Misc. 501, f. 106v). A close examination of this material throws light on the
transmission of Alfonsine parameters, although the author does not address
this issue at all.
As far as we know, John of Murss Tables of 1321 are extant in only two
manuscripts: Lisbon, ms Ajuda 52-xii-35 (henceforth ms L); and Oxford, Bod-
leian Library, ms Can. Misc. 501 (henceforth ms O). They are accompanied
by a text (L 65r66v, O 103r105v) consisting of about three pages and begin-
ning Si vera loca planetarum per presentes tabulas volueris invenire a tempore
incarnationis domini dato perfecto deme 1320 After roughly one page we read
Expliciunt canones super revolutiones planetarum. These very short canons
give some indication of the way to use the tables, but the text is too condensed
to be meaningful for anyone without previous familiarity with them, that is
John of Murs was either writing for himself or for a very select audience. This
is consistent with a sentence where he refers to his tables: in hac arte nulli sci-
entifico ignotum est. Several commentsor notesare appended concerning
tables for conjunctions with the Sun, mostly on the periods of anomaly of the
Moon and the planets. At the end of these comments there is another explicit:
Explicit compositio tabularum de certis revolutionibus planetarum.
In this text John of Murs avoided the type of canons which explain at length
how to use a set of tables, such as those written by John of Saxony in 1327 for the
Parisian Alfonsine Tables, filling 30 pages in the editio princeps of the Alfonsine
Tables,12 or the Castilian canons of the Alfonsine Tables of Toledo, completed
no later than 1272 which explain the use of these tables in 54 chapters.13
Unfortunately, as mentioned above, the original tables that corresponded to
these canons are not extant.
A description of the tables follows, based on both manuscripts. We note that
physical signs of 60 are generally, although not systematically, used in both
manuscripts (here incorporated into sexagesimal notation such that a circle
contains 6,0 = 6 physical signs).

11 For a brief account of their contents, see Poulle, Les astronomes parisiens (ref. 9), 2426,
where these tables are dated no earlier than 1325.
12 Tabule astronomice illustrissimi Alfontij regis castelle, edited by E. Ratdolt (Venice, 1483), ff.
a2rb8v.
13 For a transcription of these canons, see Chabs and Goldstein, Alfonsine Tables of Toledo
(ref. 1), 1994.
312 chapter 9

table 1 Mean conjunctions of the Sun and Saturn (excerpt)

Year Day Month Time Motus Centrum Verus locus


(h) () () ()

1320 100 Apr 10 15;17 0,27;14 2,15;51 0,22;59


1 113 Apr 23 17;29 0,39;54 2,28;31 0,36;43
2 126 May 6 19;41 0,52;34 2,41;10 0,50;35

59 102 Apr 11 20;53 0,29; 8 2,17;15 0,24;59

add 1 5;36 1;54 1;24

1 Tables in L and O

L 26rv and O 55rv display a table for the mean conjunctions of the Sun and
Saturn (Table 1). It is called tabula principalis in its heading and lists the dates
of 58 successive conjunctions of the Sun and Saturn from 1320 to 1359 (where
no such conjunction took place in years 21 and 50 after the radix), as well as
the corresponding motus (mean motion), centrum (mean argument of center),
and verus locus (true longitude) of the planet. The title of this table is: Tabula
medie coniunctionis principalis solis et saturni secundum radices per Alfonsum
regem Castele (L adds ultimo) verificatas super Toletum (L: Tollectum) distans
a Parisius in occidente per 48 m. hore (L: 48 hore). It is highly significant that the
title in both manuscripts indicates that the radices, i.e., the values correspond-
ing to the initial time, were verified (verificatas) for Toledo by Alfonso x, king
of Castile. The title also specifies that Toledo is located 48 minutes of an hour
west of Paris (L erroneously reads 48 hours). The same reference to Toledo and
to King Alfonso is also found in all the tables discussed below with radices for
the rest of the celestial bodies, a clear sign that John of Murs took the initial
values from Alfonsine astronomy as computed in Toledo and then converted
these data to the meridian of Paris. And this is indeed the case, as we shall
see.
The first column in Table 1 displays the years from 1320 to 1359; the second
column displays the number of days in that year (counted from the last day
of the previous year) that have elapsed; the third column replaces the number
of days displayed in the second column by the date. Below the table are the
amounts to be added to the first entry in a given column to arrive at the final
entry in that column.
john of murss tables of 1321 313

As derived from the entries, the time between two successive mean conjunc-
tions of the Sun and Saturn, is 1 year, 13 days, and 2;12h (L 65v and O 103v give a
more precise value: 378 days 2;12,13,12h). Thus, in a little over 59 years, 57 mean
conjunctions of the Sun and Saturn occur.
According to our calculations, using a spreadsheet to compute astronomical
positions with the standard Alfonsine Tables,14 a mean conjunction of the Sun
and Saturn took place in Toledo on 10 April 1321 at 15;17h at a mean longitude
of 27;14,25 for both celestial bodies, when the true longitude of Saturn was
22;57,59 and its mean argument of center was 135;51,34, in good agreement
with the values corresponding to the radix. It follows that the epoch 1320
has to be understood as 1320 completed, that is, 1321 current, which is our
usual reckoning.15 It also indicates that these entries were computed using a
solar model with a maximum value of the solar equation of 2;10 and a mean
motion of 0;59,8,19,37,19,13,56/d, as well as a model for the motion of Saturn
with a maximum value of the equation of center of 6;31, a maximum value
of the equation of anomaly of 6;13 and a mean motion of 0;2,0,35,17,40,21/d,
corresponding to the basic parameters of the standard Alfonsine Tables used
in the aforementioned programme for the Sun and Saturn.

14 For recomputations according to the editio princeps of the Alfonsine Tables (1483), we have
used a spreadsheet provided to us by Richard L. Kremer (Dartmouth College, usa), which
was prepared by Lars Gisln (Lund University, Sweden). By the standard Alfonsine Tables
we mean the collection of tables found in the editio princeps, edited by E. Ratdolt (ref. 12)
that, by and large, goes back to a compilation made in Paris in about 1327 with canons by
John of Saxony. Among many others, it includes: tables for the differences between the
eras, tables to transform dates of various eras, a set of radices for various eras, a table for
the movement of the 8th sphere (with a maximum of 9;0), tables of the mean motions
(presented as 60 consecutives multiples of the daily mean motions), equations of the lumi-
naries (with maximum values of 2;10 and 4;56 for the Sun and the Moon), equations of
the planets, etc. There is no modern edition of these tables which would have to be based
on the vast number of extant manuscripts, but we have examined many manuscripts con-
taining them, and none has exactly the same collection as the first edition. However, for
example, Madrid, Biblioteca Nacional, ms 10002, and Vienna, Nationalbibliothek, ms 2288,
share most of the characteristics of the editio princeps although, in contrast to it, they do
not have a star table.
15 There were two conventions for dates counted from the Incarnation in the Middle Ages:
(1) the number of years that have been completed, and (2) the current year which has not
been completed. So for a date as we reckon it in current years, e.g., noon, 15 Feb. 1321, one
might say equivalently that 1320 years have been completed plus 1 month (January) plus
14 days.
314 chapter 9

table 2 Correction for Saturn when not in conjunction with the Sun (excerpt)

w d Days Hours 0,0 hd 0,12 2,12 hd 2,24 5,36 hd 5,48 hd


() () () () () () () () () ()

1 5 12 14;48 1;29 34 0;21 2;52 28 1;55 3;51 34 2;44 37


3 4 25 4;57 2;56 35 1;46 1;16 29 0;17 5;19 34 4;11 37

27 0 189 1; 6 5;35 43 4;10 1;13 38 2;29 8;30 42 7; 5 45

52 1 365 11;43 9;53 33 8;46 7;23 33 8;29 12;17 35 11; 8 37


54 0 378 2;12 11;24 33 10;18 9; 7 33 10;13 13;44 35 12;34 35

In order to find the true longitude of Saturn when the planet is not in con-
junction with the Sun, another table, called contratabula (Table 2), is needed. It
appears on O 56r58v and on three folios in L, labelled successively 27rv, 26rv,
and 27rv, for the manuscript has two folios numbered 26 and two numbered
27.
In the heading of Table 2, w stands for week, d for day of the week,
and hd for half the difference (in minutes of arc) between two successive
values of the mean argument of center of the planet. In this double argument
table of 34 columns and 30 rows, the first four columns are for the time after
a mean conjunction, that is, the age of Saturn (given as a number of days,
expressed in weeks and days within a week, and hours). The rows are evenly
separated, and the last one corresponds to 54 weeks (378 days or 1 year and
13 days) and 2;12h. As indicated above, this is roughly the time between two
successive mean conjunctions of Saturn and the Sun, and consequently it is
equal to the time between the radix and the first mean conjunction listed in
the table called principalis (see Table 1, above). The headings for the rest of
the columns represent values of the mean argument of center of Saturn, from
0,0 to 5,48 at intervals of 12. The entries give, in degrees and minutes, the
correction to be added, or subtracted, to the mean motion of the planet to
obtain its true longitude.
In Figure 9.1 we represent three rows of this double argument table, showing
the entries (in minutes of arc on the y-axis) for three selected times (12 days,
189 days, and 378 days) after a mean conjunction and for the mean argument
of center from 0 to 360 on the x-axis. As seen at a glance, for a given value of
the age of the planet, the entries are distributed along a sinusoidal curve with
a period of 360.
john of murss tables of 1321 315

figure 9.1 Correction for Saturn as a function of its mean argument of center (the curves, from
lowest to highest, correspond to 12 days, 189 days, and 378 days, respectively)

To illustrate how the contratabula works, let us derive the true longitude of
Saturn exactly one mean conjunction after the radix (text: 36;43; see Table 1,
second row), from the mean values of the preceding conjunction. One begins
by entering Table 2 with the mean argument of center (2,15;51) already given in
the Table 1. After interpolation between the values 2,12 and 2,24 for the mean
argument of center corresponding to a time of 54 weeks and 2;12h, one finds
that 9;29 is the positive correction to be applied. By adding this amount to the
mean motion of the previous conjunction, one finds 36;43 (= 27;14 + 9;29),
which is exactly the entry found in Table 1. The same procedure applies to any
other time after a mean conjunction. Direct recomputation with the standard
Alfonsine Tables confirms this result.
L 28r50r and O 59r80r display tables for the conjunctions with the Sun
(tabula principalis) and corrections (contratabula) for each of the other planets.
In all cases, the titles indicate that the tables were computed for Toledo and
use the radices given by King Alfonso. The times between successive mean
conjunctions with the Sun can be derived from the tables, but the text on
L 65r66v and O 103r105r gives more precise data:

Saturn 378 days 2;12,13,12h


Jupiter 398 days 21;12, 8,24h
Mars 779 days 22;22,34h
Venus 583 days 22;14, 3h
Mercury 115 days 21; 5, 2, 7h
316 chapter 9

table 3 Mean conjunctions of the Sun and the


Moon (excerpt)

Year Days Time motus argum. lune


(h) () ()

1 18 2; 3 5, 5;38 4,26;53
2 7 10;51 4,54;55 3,36;41

75 10 13;49 4,58;15 2, 9;12


76 28 11;22 5,16;39 1, 4;49 *

* L: 1,44;49.

These values for the periods of anomaly of the planets are not difficult to
recompute. For the three superior planets, they result from dividing 360 by
the difference between the daily mean motions of the Sun and each planet,
whereas for the inferior planets they are obtained by dividing 360 by the daily
mean motions in anomaly of each planet. All these periods are consistent with
the standard mean motions of the planets in Alfonsine astronomy, as they
appear, for example, in the editio princeps of 1483. As was the case for the
radices of Saturn, the entries corresponding to those of the other planets can
be recomputed from the standard Alfonsine Tables. Therefore, all the values for
the mean motions of the planets that were later used in the Parisian Alfonsine
Tables are already embedded in the Tables of 1321 by John of Murs.
Next we find tables for the conjunctions of the Sun and the Moon (L 50v57v,
O 81r87r). They share the same format as those for the five planets, and
have the same references to King Alfonso and Toledo in the title. The first
table (Table 3), also called tabula principalis, lists the dates of the first mean
conjunction of the Sun and the Moon for each year in 76 consecutive years, as
well as the corresponding motus (mean position) and argumentum lune (mean
lunar anomaly).
Although no date is specified in Table 3, the first row corresponds to the
first conjunction of 1322. Indeed, recomputation with the standard Alfonsine
Tables indicates that a mean conjunction between the Sun and the Moon took
place at Toledo on 18 January 1322 at 2;3h, when the mean longitude of the Sun
and the Moon, here called motus, was 305;38, in perfect agreement with the
tabulated data. The rest of the rows correspond to the first conjunction of the
luminaries (occuring in January) of each year after 1322. This table thus covers
76 years, i.e., the least common multiple of a solar cycle of 4 years and a lunar
john of murss tables of 1321 317

cycle of 19 years. Contrary to the tables for the planets, this table begins in 1322,
and it is possible that the row for the radix (1321, that is, 1320 completed) is
missing in both manuscripts. In that case, one should expect to find 28 January
at 17;14h (time), 5,16;21 (mean longitude), and 5,17;5 (mean lunar argument
of anomaly), according to our computation for Toledo using the standard
Alfonsine Tables. This is confirmed by the table for mean conjunctions given
in the Patefit (see Table 14, below).
L 51r and O 81r display three short tables related to the Moon and they contain
the following data for one lunation:

Time 29 days 12;44, 3, 3h *


Mean motion 29; 6,24,12
Mean lunar anomaly 25;49, 6,30
Mean argument of lunar latitude 30;40,13,48
Daily mean motion of the Moon 0;13,10,35, 0/d

* O: 29 days 12;44, [blank], 3h.

Once again, the title refers to King Alfonso and indicates that his radices are
not included in the mean motion: Tabula mediarum coniunctionum solis et
lune infrascriptarum absque radice per Alfonsum. All these data later became
characteristic of the Parisian Alfonsine Tables.
As was the case for the planets, in order to find the true longitude of the
Moon when not in conjunction with the Sun, another table, again called con-
tratabula, is needed. It is displayed on L 51v57r and O 81v87r (Table 4).
In this double argument table the entries are presented in 60 columns and
30 rows. The vertical argument is the time after a mean conjunction, that is,
the age of the Moon (given here in days) and the horizontal argument is the
mean lunar argument of anomaly from 0,6 to 6,0 at intervals of 6. Again, hd
stands for half the differences (in minutes of arc) between two successive values
of the lunar argument for a given time. For example, 0;14 is half the difference,
rounded to minutes, between 0,11;24 and 0,10;57, corresponding to 0,6 and
0,12, respectively, for day 1; hence the entry 14 in the first column labelled
hd in that row. The entries, in degrees and minutes, give the correction to
be added to the mean motion of the Moon to obtain its true longitude, as
can be seen in the following example. Consider the mean conjunction for 18
January 1322, which occurred at 2;3h at Toledo; the mean motion and the mean
lunar argument of anomaly at that time were 5,5;38 and 4,26;53, respectively
(see Table 3). If we wish to know the position of the Moon one or more
days after that time, we enter in Table 4, with the mean lunar argument of
318 chapter 9

table 4 Correction for the Moon when not in conjunction with the Sun (excerpt)

Days 0,6 hd 0,12 3,0 5,54 hd 0,0 hd


() () () () () () () ()

1 0,11;24 14 0,10;57 0,14;45 0,12;19 14 0,11;51 13


2 0,23;14 12 0,22;50 0,29;30 0,24; 7 13 0,23;40 13

15 3,19;45 15 3,20;14 3,16;18 3,18;41 16 3,19;14 16

20 4,30;10 8 4,29;56 4,16;35 4,30;28 3 4,30;21 5

29 0,20;20 * 13 ** 0,19;53 0,23;47 0,21;15 15 0,20;46 13


30 0,32;26 12 0,32; 3 0,38; 6 0,33;14 13 0,32;52 12

* L: 0,20;6.
** L: 11.

anomaly. By interpolation between the values 4,26 and 4,30 we find, for
exactly, say, one day after mean conjunction, 17;56 as the correction to be
applied. Adding this value to the mean motion at conjunction, we find 5,23;34;
recomputation based on the standard Alfonsine Tables yields 5,23;34,41, in
very good agreement with it.
Figure 9.2 shows the behavior of the correction for the Moon (in degrees) as
a function of the mean lunar argument of anomaly for a given time of the syn-
odic month or age (taken here as 20 days after a mean conjunction), whereas
Figure 9.3 displays the correction of the Moon as a function of the age of the
Moon for a given value of the mean lunar argument of anomaly (taken here as
4,0).
As was the case for the planets, the values for the mean motion of the
Moon as well as the maximum lunar equation (4;56) and the maximum solar
equation (2;10), all parameters later used in the Parisian Alfonsine Tables, are
already embedded in John of Murss tables. In his Patefit, John of Murs also
compiled other tables for true syzygies, and they will be examined in section 4,
below.
It is very interesting to contrast John of Murs tables for the mean conjunc-
tions of the luminaries and the position of the Moon (Tables 3 and 4) with
those for the same purpose by John Vimond.16 Vimonds tables are only extant

16 Chabs and Goldstein, John Vimond (ref. 4), 213 and 229, Tables 1 and 8.
john of murss tables of 1321 319

figure 9.2 Correction for the Moon for a given age (20 days)

figure 9.3 Correction for the Moon for a given value of mean lunar argument of anomaly (4,0)

in one manuscript, Paris, Bibliothque nationale de France, ms lat. 7286c,


and the same manuscript also contains the canons and tables for 1322 by
John of Lignres. In Vimonds tables the epoch of the first mean conjunction
is 10 March 1320 (computed for Paris), whereas in John of Murss table it is
18 January 1322 (computed for Toledo), indicating that John Vimonds tables
preceded those by John of Murs. Both astronomers use values of the mean
synodic month that are almost identical: the value in John Vimonds tables
is not explicit, but it can be derived from them, namely, 29 days 12;44,3,6h
320 chapter 9

(= 29;31,50,7,44,35d), whereas that in John of Murss table is given as 29 days


12;44,3,3h. The difference amounts to 3 sixtieths of a second of an hour. It also
happens that both use double argument tables to find the true position of
the Moon between syzygies, and the tables look very similaralthough the
vertical and horizontal arguments are switchedand the entries are given
to the minutes in both cases. However, John of Murss table is larger, for the
argument of anomaly is taken at 6-intervals in contrast to 12-intervals in the
tables of John Vimond, and covers 30 days whereas that of Vimond covers only
14 days; hence, it has more than four times as many entries (60 30 for John of
Murs and 30 14 for John Vimond). We note that in John of Murss tables the
entries for half the differences between the corrections corresponding to two
consecutive values of the argument of anomaly are given to one place (minute
of arc), whereas in John Vimonds tables the entries for interpolation represent
another quantity, the difference between two consecutive corrections for the
same value of the argument of anomaly, and are displayed to two places (second
of arc). That is, we are offered two interpolation schemes that complement
each other: that of John de Murs applies for a fixed day, whereas that of John
Vimond is for a fixed argument of anomaly.
The 420 entries for the corrections common to both tables agree, except for
copyists errors (see Table 5 for a comparison between two selected columns in
both tables).
This does not happen by chance. We are thus faced with two possibilities: the
most likely is that John of Murs used the principle established by John Vimond
and expanded the table, but it could also be that they both depended on a table
in an otherwise unknown prior text.
The approach for the motion of the lunar nodes (L 59v, O 87v) parallels that
for the Sun. In this table we are given the true position of the node for 1320
(4,57;14) and then at four year intervals to 1392, adding a final row for 1393,
thus using the standard period of 93 years used by astronomers in the Alfonsine
tradition.17 At the bottom of this page is a short table for the increments in years
1, 2, and 3 within the four-year calendaric cycle (for year 1 the entry is 5,4;40
in both manuscripts, whereas they should read 5,40;40). Another sub-table
lists the true motion of the node for the days of the year (given by the month,
beginning in January, and the day within the month), at intervals of 6 days.
Note that the entry corresponding to 31 December is correctly given as 5,40;40.

17 E.g. Abraham Zacut: see J. Chabs and B.R. Goldstein, Astronomy in the Iberian Peninsula:
Abraham Zacut and the Transition from Manuscript to Print (Philadelphia, 2000), 60 and
117.
john of murss tables of 1321 321

table 5 Comparison of extracts ( for mean arguments of anomaly 12


and 180) of the tables of John Vimond and John de Murs for
the correction of the Moon

J. of Murs J. Vimond J. of Murs J. Vimond


Days 0,12 0s12 3,0 6s 0
() (s) () () (s) ()

1 0,10;57 0 10;57 0,14;45 0 14;45


2 0,22;50 0 22;50 0,29;30 0 29;30
3 0,34;45 1 4;45 0,44; 5 * 1 14;11
4 0,46;48 1 16;48 0,58;38 1 28;38
5 0,59; 3 1 29; 3 1,12;44 2 12;44
6 1,11;46 2 11;46 1,26;20 2 26;20
7 1,24;47 2 24;47 1,39;36 3 9;36
8 1,38; 5 3 8; 5 1,52;36 3 22;36
9 1,51;53 3 21;53 2, 5;11 4 5;11
10 2; 6;12 4 6;12 2,17;22 4 17;22
11 2,20;57 4 20;57 2,29;14 4 29;14
12 2,35;55 5 5;55 2,40;56 5 10;56
13 2,50;51 5 20;51 2,52;39 5 22;39
14 3, 5;37 6 5;37 3, 4;27 6 4;27

* L mg.: al. 11 (i.e., other manuscripts read 0,44;11); O: 0,44;11.

The daily motion of the node resulting from the table is 0;3,10,38,11,34/d,
in good agreement with the same parameter in the Parisian Alfonsine Tables
(0;3,10,38,7,14/d), and quite far from other parameters historically used for
the motion of the nodes. As was the case with the previous tables, the title
indicates that the radix was derived from the tables compiled by Alfonso x,
King of Castile, for Toledo. And it is indeed so: our recomputation, using the
standard Alfonsine Tables for noon of the first day of year 1321, indicates that
the true longitude in Toledo was 4,57;12,12, in close agreement with the entry
in the text.
Then come tables for the planets: Saturn (L 60rv, O 88rv), Jupiter (L 60r and
61r, O 88r and 89r),18 Mars (L 61v62r, O 89v90r), Venus (L 62v63r, O 90rv), and

18 The columns for the arguments, the equation of center, and the stations of Saturn and
322 chapter 9

table 6 Argument of anomaly table 7 Equation of center


and argument of center of and stations of Saturn
Saturn (excerpt) (excerpt)

Anom. Center Mean center Eq. center Station


Days () () () () () ()

12 0,12 0;25 * 0, 6 5,54 0;40 22;45


25 0,24 0;51 0,12 5,48 1;17 22;47
38 0,36 1;16 0,18 5,42 1;55 22;49

189 3, 0 6;20 1,30 4,30 6;31 24;11

353 5,36 11;49 2,48 3,12 1;25 25;27


365 5,48 12;15 2,54 3, 6 0;43 25;28
378 0, 0 12;40 3, 0 3, 0 0; 0 25;30

* L: 0;15.

Mercury (L 63v64r, O 91v92r). In addition to the tables for the conjunctions


of the planets with the Sun described above, for each of the five planets we are
given three tables: one for the argument of anomaly and the argument of center,
a second for the equation of center and the stations, and a third for latitude.
Excerpts of the three tables for Saturn follow (Tables 6, 7, and 8).
In Table 6 the first column is for the argument, expressed in days within a
period of anomaly (378 days 2;12,13,12h), as indicated above. The second column
represents the mean motion of the argument of anomaly, in degrees, and the
third column displays the mean motion of the argument of center, in degrees
and minutes.
In Table 7 the first two columns represent the mean argument of center, in
degrees. The third column is for the equation of center of Saturn, in degrees
and minutes. Its maximum, 6;31, agrees with that in the Almagest, the zij of
al-Battn, and the Toledan Tables, as well as the Parisian Alfonsine Tables. The
fourth column represents the first station, in degrees and minutes, but we note
that the zodiacal signs (required for the entries to be meaningful) are not given,
a feature shared by both manuscripts that will be addressed below.

Jupiter are displayed on a single page, whereas the latitudes of these two planets are
presented separately.
john of murss tables of 1321 323

The tables for the planets other than Saturn present the same characteristics.
The maximum values for the equation of center of the other planets are 5;57
(Jupiter), 11;24 (Mars), 2;10 (Venus), and 3;2 (Mercury). This indicates that
John of Murss Tables of 1321 adhere to the tradition represented by the Toledan
Tables for all planets except for Jupiter and Venus, where the latter have 5;15
and 1;59, respectively. However, all 5 parameters used by John of Murs were
already in Paris at the time, for they are found, once again, in the tables of John
Vimond, including the new ones for Jupiter and Venus.19 For the stations of
Saturn and Jupiter there is no column for the signs, in contrast to each of the
other three planets. This is most peculiar in the case of Saturn, for which we
are given entries ranging from 22;45 (for 0,6 = 6) to 25;30 (for 3,0 = 180), as
shown in Table 7. Now, in the Toledan Tables, which use zodiacal signs of 30,
the entries range from 3s 22;44 (for 0s 0) to 3s 25;30 (for 6s 0 = 180), with 3s
22;45 for 0s 6. It would therefore seem that John of Murs, who used physical
signs of 60 in this very same table, took the entries for Saturn from a table,
such as that in the Toledan Tables, using zodiacal signs of 30, and omitted the
signs. In sexagesimal notation the values displayed would range from 1,52;45 to
1,55;30. We mentioned the Toledan Tables as an example only, for other tables
with the same values, using zodiacal signs, were available in Paris at the time:
see, e.g., those of John Vimond which display the same entries, although shifted
differently depending on the planet.20
The tables for planetary latitudes are of great interest because they are
also presented as double argument tables, a feature for which we know of
no precedent. As is often the case, superior and inferior planets are treated
differently. Table 8 reproduces an excerpt of the table for Saturn.
For the superior planets, the entries, in degrees and minutes, are presented
in 7 columns and 31 rows (L 60v61r and 62r; O 88v89r and 90r). The vertical
argument is the argument of center of the planet, shifted +50 in the case of
Saturn, 20 for Jupiter, and with no shift for Mars, in accordance with the
instructions given in Almagest xiii.6. In the heading, hd stands for half the
difference (in minutes of arc) between the entries of two successive columns
for a fixed argument of center. The extremal values appear in the column
corresponding to 180:

Saturn +3; 2 (North) 3; 5 (South)


Jupiter +2; 5 (North) 2; 8 (South)
Mars +4;21 (North) 6;30 (South)

19 Chabs and Goldstein, John Vimond (ref. 4), 244.


20 Chabs and Goldstein, John Vimond (ref. 4), 242.
324 chapter 9

table 8 Latitude of Saturn (excerpt)

Center 0/360 hd 30/330 150/210 hd 180/180


() () () () () () ()

310 310 2; 3 3 2; 8 2;57 3 3; 2


316 304 2; 1 3 2; 7 2;55 3 3; 0

34 226 0;13 0 0;14 0;19 0 0;19


40 220 0; 0 0 0; 0 0; 0 0 0; 0
46 214 0;13 0 0;13 0;19 0 0;20

124 136 2; 0 2 2; 3 2;55 4 3; 3


130 130 2; 1 2 2; 5 2;57 4 3; 5

We note that the northern (positive) and southern (negative) limits for
Saturn agree with the tradition (Almagest, al-Battn, Toledan Tables, and also
Vimond) but differ from the values that were to become part of the Parisian
Alfonsine Tables (+3;3 and 3;5, respectively). As for Jupiter, the northern
limit agrees with that in the Toledan Tables and the tables of Vimond, and
departs slightly from that in the Almagest and the zij of al-Battn (+2;4) and
in the Parisian Alfonsine Tables (+2;8), whereas the southern limit agrees with
all the sets of tables mentioned so far. This is also the case for the northern limit
of Mars, but not for its southern limit (Almagest and al-Battn: 7;7; Toledan
Tables, Vimonds tables, and Parisian Alfonsine Tables: 6;30).
For each of the inferior planets, we are given two double-argument tables
for latitude (L 63r and 64r O 91r and 92r): one for the inclination (here called
declinatio) and one for the slant (reflexio). The entries, in degrees and minutes,
are displayed in 7 columns and 16 rows. Tables 9 and 10 display excerpts for the
latitude of Venus.
In all four tables (two for each of the inferior planets), the first four columns
are for the argument, which is shifted +60 in the table for the inclination of
Venus and +90 in the table for the inclination of Mercury. They are small
double argument tables, with 7 columns and 16 rows, where the maximum
value for Venus is 7;22 (cf. Almagest and the zij of al-Battn: 6;22; Toledan
Tables: 7;24; and Vimonds tables and the editio princeps of the Alfonsine
Tables: 7;12) and that for Mercury, 4;5 (in agreement with the values in the
Almagest, the zij of al-Battn, the Toledan Tables, and Vimonds tables, as
well as in the editio princeps of the Alfonsine Tables). Most notable is the
john of murss tables of 1321 325

table 9 Latitude of Venus, inclination (excerpt)

Center 0/360 hd 30/330 150/210 hd 180/180


() () () () () () ()

300 300 120 120 1; 3 3 0;57 3; 3 129 7;22


306 244 114 126 1; 3 3 0;57 3; 2 127 7;16

342 258 78 162 0;47 3 0;42 2;15 95 5;25

24 216 36 204 0; 7 0 0; 6 0;18 15 0;47


30 210 30 210 0; 0 0 0; 0 0; 0 0 0; 0

presentation of a column for the 3rd component of latitude, or deviation (called


3 latitudo in L), in the tables for the slants of both inferior planets. This is
indeed a very unusual feature in medieval tables, for there are not many which
list values for the deviation. These values were mentioned, but not tabulated,
in Almagest xiii.6, with a limit of 0;10 (north) for Venus and a limit of 0;45
(south) for Mercury. It is significant that the canons to the Castilian Alfonsine
Tables explicitly addressed this particular problem, giving instructions to take
into account the third component of latitude appearing in the tables they
describe.21 To this we can add that John Vimond included the deviation of
Venus and Mercury in his own tables and used the values mentioned in the
Almagest.22 As can be seen in Table 10, John of Murs gave 0;10 (septentrionalis)
for the maximum deviation of Venus, but for that of Mercury he used 0;23
(meridionalis), a value about half that of Ptolemy and all others who worked in
the Ptolemaic tradition. There are no tables for the third component of latitude
in the Parisian Alfonsine Tables.
The last table in both manuscripts is for the days in a year (L 64v, O 102v),
where each day of the year is assigned an ordinal number from 1 to 365.
In addition to the contents of John of Murss Tables of 1321 that appear in
both manuscripts, there are a few other tables closely related to them that are
only found in one of the two manuscripts.

21 Chabs and Goldstein, Alfonsine Tables of Toledo (ref. 1), 4243.


22 Chabs and Goldstein, John Vimond (ref. 4), 257258.
326 chapter 9

table 10 Latitude of Venus, slant (excerpt)

Center 0/360 hd 30/330 150/210 hd 180/180 3rd lat.


() () () () () () () ()

0 360 180 180 0; 0 20 0;41 2;22 70 0; 0 10


6 354 174 186 0; 0 20 0;41 2;20 70 0; 0 10

42 318 138 222 0; 0 15 0;30 1;46 53 0; 0 8

84 276 96 264 0; 0 2 0; 4 0;15 7 0; 0 1


90 270 90 270 0; 0 0 0; 0 0; 0 0 0; 0 0

2 Tables in L but not in O

On L 57v58v there is a table for the true positions of the Sun for each day in
a year: see Table 11. The table is also called principalis, and in the title we are
again told that the radices were computed by King Alfonso of Castile for the
city of Toledo. The entries are given in physical signs, degrees, minutes, and
seconds. The date for which the table is valid is not indicated, but it corresponds
to 1321, as shown from the recomputations displayed in Table 11, where the true
positions of a sample of entries are compared with recomputations for 1321.
Note that the values for text and computation do not differ by more than 7
seconds.
Reinforcing our claim that Vimonds tables were composed prior to those by
John of Murs is the fact that Vimond displays a one-year calendar with syzygies
valid for 1320. It would seem odd for either astronomer to compile a one-year
table for the (recent) past. It is far more likely that such a table was produced
at (or near) the beginning of the year in question.
The Sun has also its contratabula (L 59r), of which an excerpt is given in
Table 12, showing the extremal values of the various columns. Note that this
contratabula differs from those reviewed above, among other things for the fact
that it is not a double argument table, although its purpose is substantially the
same.
The argument in Table 12 is the day of the year given at intervals of 6 days,
beginning in January. The second column lists the correction to be subtracted
from the true position of the Sun for 1321, found in the preceding table, in order
to determine its true position for dates corresponding to one, two, or three
years after 1321. For example, let us consider the true longitude of the Sun for
john of murss tables of 1321 327

table 11 True solar positions for 1321 (excerpt)

tc
Text (t) Computation (c) (in seconds)

Jan. 1 4,49;47, 5 4,49;47, 6 1


Feb. 1 5,21;17,33 5,21;17;33 0
Mar. 1 5,48;19,14 * 5,49;19,13 +1
Apr. 1 0,19;46,14 0,19;46,18 4
May 1 0,48;43, 1 0,48;43, 1 0
June 1 1,18;16,45 1,18;16,45 0
July 1 1,46;47,35 1,46;47,36 1
Aug. 1 2,16;24,35 2,16;24,35 0
Sep. 1 2,46;20,24 ** 2,46;26,24 0
Oct. 1 3,16; 2, 2 3,16; 2, 2 0
Nov. 1 3,47;12,50 3,47;12,50 0
Dec. 1 4,17;47,50 4,17;47,57 7

* Instead of 5,49;19,14.
** Instead of 2,46;26,24.

6 January 1322. The procedure is to note the true solar longitude exactly one
year before, 6 January 1321, which is given in Table 11 as 294;53,39. Then we
have to subtract from it the difference for one year displayed in the column
labelled correction for that date in Table 12, and it is 0;14;52; the result is
294;38,47, in good agreement with our recomputation based on the standard
Alfonsine Tables (294;38,27). Note that the column for the correction exhibits
a minimum value of 0;13,48 (31 May30 June) and a maximum value of 0;14;56
(618 December). The approach to finding the position of the Sun in these two
tables is strongly reminiscent of an almanac, in the sense that we are given the
true position of the Sun at noon for all days in a year (tabula principalis) and a
simple procedure to find its true position at noon for each day in the next three
years (contratabula), with no indication of the underlying parameters and no
rule for computing true positions beyond this 4-year period.
The third column gives the hourly velocity of the Sun in minutes and seconds
per hour, from 0;2,23/h to 0;2,34/h, for purposes of interpolation between the
entries in the previous column. These are the same limits as in the Parisian
Alfonsine Tables.23

23 See, e.g., Tabule astronomice illustrissimi Alfontij (ref. 12), ff. g6rg7r.
328 chapter 9

table 12 Contratabula for the Sun (excerpt)

Month Day Correction Velocity Eq. of time


() (/h) (min)

Jan 6 14;52 2;33 5; 2

Jan 31 14;43 2;32 0; 0

May 6 13;55 2;24 20;58

Jun 18 13;48 2;23 15;30

Jul 24 13;55 2;24 12; 0

Oct 24 14;43 2;32 32;38

Dec 12 14;56 2;34 16;53

In the fourth column we find the equation of time. There are not many
medieval astronomical tables where the argument is expressed in days of the
year and the entries are given in time, as is the case for this table. In fact, we
only know of two of them, one in Abraham Zacuts ibbur and another in the
Tabule Verificate for Salamanca.24 In Table 12, the extremal values of the entries
are the following:

Min: 0; 0h 24 Jan12 Feb


max: 0;20,58h 6 May
min: 0;12, 0h 1824 July
Max: 0;32,38h 24 Oct

When converted into time-degrees, these extremal values correspond to 0,


5;14,30, 3;0, and 8;9,30. As far as we know, these values for the equation of
time are unprecedented in the astronomical literature. In particular, this table
differs both in format and in content from the table of Peter of St. Omer for

24 See Chabs and Goldstein, Abraham Zacut (ref. 17), 108109.


john of murss tables of 1321 329

12921293, another astronomer working in Paris shortly before John of Murs,


and reproduced by John of Lignres, with the following extremal values:25

Min: 0; 0 Aqr 1825


max: 5;21 Tau 2527
min: 2;49 Leo 5
Max: 7;57 Sco 89

However, all other parameters for the Sun in these tables are those that are
found in the Parisian Alfonsine Tables.

3 Tables in O but not in L

On O 92v98r there are five tables for the conjunctions of the planets with the
Sun that are very similar to those described in section 1. However, the entries are
different, for they correspond, as indicated in the titles, to 1452 perfecto (i.e.,
1453 in the usual reckoning) and were computed for Paris, not Toledo. The titles
add that the radices are those of Alfonso x, king of Castile. For each planet we
are only given the tabula principalis, on the assumption that the corresponding
contratabula to be used is the one already found among John of Murss Tables
of 1321. Note also that the column for the true position has been eliminated and
the accuracy of the entries has been improved. Table 13 displays an excerpt of
the table for the mean conjunctions of the Sun and Saturn, beginning in 1453
(1452 completed).

table 13 Mean conjunctions of the Sun and Saturn for 1453 (excerpt)

Year Day Month Time med. motus centrum med.


of the year and day (h) () ()

1452 283 Oct 10 10; 9,14 3,28;20,24 5,15;37,31


1 296 Oct 23 12;21,27 3,41; 0,17 5,28;16,49
2 309 Nov 6 14;33,40 3,53;40,10 5,40;56, 7

59 285 Oct 11 15;45,42 3,30;14, 9 5,16;58, 0

25 See Chabs and Goldstein, Alfonsine Tables of Toledo (ref. 1), 186.
330 chapter 9

It is clear that these tables do not belong to John of Murss Tables of 1321 and
are simply an extension of them made more than a century later, indicating
that his table was taken as a model by at least one subsequent user.
The same can be said about the tables on O 98r99v concerning the con-
junctions of the Sun and the Moon beginning in 1453 (1452 completed), which
follow the pattern of Table 3, above.

4 Syzygies in the Patefit

In addition to the Tables of 1321, John of Murs composed another work con-
taining tables, associated with the canon beginning Patefit ex Ptolomei disci-
plines in libro suo , and traditionally dated 1321.26 The Patefit survives in very
few copies. Three manuscripts have previously been noted: London, British
Library, Royal ms 12.c.xvii; Erfurt, Biblioteca Amploniana, mss 4 360; and
Erfurt, Biblioteca Amploniana, 4 371.27 We can now add to this short list Vati-
can, Biblioteca Apostolica, ms lat. 3116. We have also found other manuscripts
containing extracts of this work, as is the case with the Lisbon manuscript itself
(L 1r22v), described above in relation to the Tables of 1321. Among the tables
associated with the Patefit there are five directly concerning syzygies, two of
which are also double argument tables. For the description of the tables, below,
we have followed the London manuscript (henceforth called B), where zodia-
cal signs are used (except in one case), in contrast to the Tables of 1321 and the
Lisbon manuscript where physical signs are used.
The first is a table for the mean conjunctions and oppositions for a period
of 76 years from 1321 to 1396 (expressed in current years), covering 1880 (=
940 2) successive syzygies. The table has no title and, under the headings
conjunctions and oppositions, four quantities are given in each case: time
in days, hours, and minutes; mean motion of the Sun in zodiacal signs, degrees,
and minutes; mean lunar anomaly in zodiacal signs, degrees, and minutes;
mean argument of lunar latitude in zodiacal signs, degrees, and minutes (see
Table 14, an excerpt of B 155v168r).
The entries in Table 3, above (limited to mean conjunctions), agree with
the corresponding entries in Table 14 of the Patefit, except for the minutes in
a few cases. We also note that in this table, and in general in the Patefit, years

26 According to Poulle, this work was probably composed in the late 1320s: cf. Poulle, Les
astronomes parisiens, (ref. 9), 2627.
27 See, e.g., Kremer, John of Murs, Wenzel Faber (ref. 1), 148.
john of murss tables of 1321 331

table 14 Mean conjunctions and oppositions for 13211396 (excerpt)

1321 Conjunctions 1321 Oppositions


Day Time motus arg. lune arg. lat. Day Time motus arg. lune arg. lat.
(h) (s) () (s) () (s) () (h) (s) () (s) () (s) ()

Jan 28 17;14 10 16;21 10 17; 5 0 20;40 13 22;52 10 1;48 4 4; 9 6 5;20


Feb 27 5;58 11 15;28 11 12;54 1 21;20 12 11;36 11 0;54 4 29;59 7 6; 1
Mar 28 18;42 0 14;34 0 8;43 2 22; 1 14 0;20 0 0; 1 5 25;48 8 6;41

Oct 21 11;51 7 8;19 6 9;26 9 26;42 6 17;28 6 23;46 11 26;31 3 11;22


Nov 20 0;35 8 7;25 7 5;15 10 27;22 5 6;13 7 22;52 0 22;20 4 12; 2
Dec 19 13;19 9 6;32 8 1; 4 11 28; 2 4 18;57 8 21;58 1 18; 9 5 12;42

1322 Conjunctions 1322 Oppositions


Day Time motus arg. lune arg. lat. Day Time motus arg. lune arg. lat.
(h) (s) () (s) () (s) () (h) (s) () (s) () (s) ()

Jan 18 2; 3 10 5;38 8 26;53 0 28;42 3 7;41 9 21; 5 2 13;58 6 13;22


Feb 16 14;47 11 4;44 * 9 22;42 * 1 29;22 * 1 20;25 10 20;11 3 9;47 7 14; 3
Mar 18 3;31 0 3;51 10 18;31 3 0; 3 3 9; 9 11 19;18 4 5;36 8 14;43

* In L these entries are given as 5,34;45 (= 11s 4;45), 4,52;43 (= 9s 22;43), and 1,0;22 (= 2s
0;22), respectively.

1396 Conjunctions 1396 Oppositions


Day Time motus arg. lune arg. lat. Day Time motus arg. lune arg. lat.
(h) (s) () (s) () (s) () (h) (s) () (s) () (s) ()

Jan 10 13;49 9 28;16 4 9;15 0 12;15 25 8;11 10 12;49 10 22;10 6 27;33

Dec 0 9;53 8 18;27 1 23;14 11 19;34 15 15;32 8 3;53 7 10;20 5 4;15


Dec 29 22;48 * 9 17;33 2 19; 5 0 20; ** 15 4;16 9 3; 0 8 6; 9 6 4;55

* In L this entry is given as 22;38h.


** Vatican, ms lat. 3116: 0s 20;14. The column for the zodiacal signs corresponding to the lunar
anomaly for oppositions is missing in this manuscript in most cases.
332 chapter 9

refer to the current year. Thus, Table 14 was also computed for the meridian
of Toledo, following the radices set up by King Alfonso, a feature not specified
in the title. The same table is found in Vatican, lat. 3116, ff. 11r23v, and all the
entries displayed above agree with those in B.28 In L the information contained
in this table is split into four different tables: the times of 940 successive mean
conjunctions for the 76-year period, 13211396; the mean motion of the Moon
for 235 successive conjunctions (that is, a lunar cycle of 19 years); the mean
lunar anomaly for 251 successive conjunctions (that is, the cycle according to
which 251 synodic months = 269 returns in lunar anomaly); and the mean
argument of lunar latitude for 223 successive conjunctions (that is, the Saros
cycle of about 18 years): for these cycles see, e.g., Almagest iv.2.29
The second table displays the times of true conjunctions and oppositions for
a period of 76 years (13211396), given in days, hours, and minutes (see Table 15,
an excerpt of B 168r172r). The title, Tabula continens veras coniunctiones et
oppositiones ad Tholetum per 48 m. hore distante a Parisius, indicates that the
entries were computed for Toledo, which is distant by 48 minutes of an hour
from Paris. This table is not in Vat. lat. 3116. In the heading of Table 15, m stands
for month. As expected, the entries in this table were computed with Alfonsine
models and parameters for the Sun and the Moon.
The third table deals with true conjunctions although in the three manu-
scripts we have examined (London, Vatican, and Lisbon) the title refers both to
conjunctions and to oppositions: Tabula vere coniunctionis et opposicionis solis
et lune. Five quantities are given: true longitude of the Moon at mean conjunc-
tion, in physical signs, degrees, and minutes; lunar time correction, in hours
and minutes; hourly lunar velocity, in minutes and seconds; true longitude of
the Sun at mean conjunction, in physical signs, degrees, and minutes; solar time
correction, in hours and minutes (see Table 16, an excerpt of B 172v174r). The
columns dealing with the Moon cover 251 conjunctions, whereas those for the
Sun only cover 235 conjunctions. It is noteworthy that in this table physical
signs of 60 are used, contrary to the other tables in the Patefit. Also in L 9r12v
physical signs are used. The same table is found in Vatican, lat. 3116, ff. 24r25v.
Note that the time correction for the Moon and the time correction for the Sun
are the two terms in which the time from mean to true syzygy, t, is divided.

28 On L 8rv there is another table containing much the same information with the title,
Tabula medie coniunctionis solis et lune in annis ad meridiem Tholeti secundum radices
Alfonsii regis castelle. We are only given entries for the last conjunction of each of the 76
years, but with a higher accuracy, both for time (to seconds of an hour) and for the three
other quantities (to seconds of an arc).
29 G.J. Toomer, Ptolemys Almagest (New York and Berlin, 1984), 174176.
john of murss tables of 1321 333

table 15 Time of true conjunctions and oppositions for 13211396 (excerpt)

1321 1322 1396


m. Time Conj. Time Opp. m. Time Conj. Time Opp.
(d) (h) (d) (h) (d) (h) (d) (h)

1 28 13;58 14 9;16 1 18 18;45 3 18;47


2 27 7;29 12 20;16 2 16 9;54 2 9;35
3 29 0;36 14 5; 8 3 18 1;36 3 21;26

10 21 7;12 6 12;54 10 10 23;24 25 7;22


11 19 17;29 5 5;17 11 9 9;41 24 2; 7
12 19 5; 6 5 1;32 12 9 0; 2 23 21;17

To illustrate this, consider the mean conjunction for 28 January 1321, occurring
at 17;14h after noon (see Table 14). The true conjunction occurs at 13;58h (see
Table 15). Thus, true conjunction precedes mean conjunction, and t = 3;16h.
As readily seem in Table 16, the sum of the lunar correction (6;40h) and the
solar correction (+3;24h) is 3;16h. Computing the time from mean to true
syzygy is an issue that interested many medieval astronomers and they offered
a variety of methods to give a proper answer.30 In a recent paper Kremer
analysed this table and argued convincingly (1) that cols. 2 and 5 are based on
the parameters for the lunar and solar corrections that later appeared in the
Parisian Alfonsine Tables; (2) that col. 4 is based on the lunar velocity table in
the Toledan Tables (which is identical with the corresponding table in the zij
of al-Battn); and (3) that cols. 3 and 6 are based on a computational scheme
similar to the one that underlies the table by Nicholaus de Heybech (fl. 1400)
for finding the time from mean to true syzygy, separating the solar and lunar
components.31
The fourth table has the heading Tabula veri motus lune ad dimidiam luna-
tionem. Argumentum lune ad coniunctionem mediam inventum (see Table 17, an
excerpt of B 177r182v). It is a double argument table where the horizontal scale
is for the argument of lunar anomaly, ranging from 0s 3 to 12s 0 at 3 intervals,
and the vertical scale is the number of days after mean conjunction or oppo-

30 See J. Chabs and B.R. Goldstein, Computational Astronomy: Five Centuries of Finding
True Syzygy, Journal for the History of Astronomy, xxviii (1997), 93105.
31 For additional details, see Kremer, John of Murs, Wenzel Faber (ref. 1), 148155.
334 chapter 9

table 16 True conjunctions: positions and velocities (excerpt)*

True lunar Time correction Hourly lunar True solar Time correction
position for the Moon veloc. position for the Moon
() (h) (/h) () (h)

1 5,19;30 6;40 30;52 5,17;58 3;24


2 5,46;45 2;50 30;24 5,47;35 4;31
3 0,13;49 1;29 30;20 0,16; 8 4;25

235 4,51;42 8;17 34;20 4,48; 2 1;20


236 5,21;22 9;44 32;59 0, 0; 0 0; 0

250 0, 8;51 9;44 32;56 0, 0; 0 0; 0


251 0,37;29 9; 5 31;46 0, 0; 0 0; 0

* We have added a minus sign () to the entries under deme (subtract) in the table.

sition, up to 16 days. The same table is found in Vatican, lat. 3116, ff. 29r34v.
An extract is also found in Brussels, Bibliothque Royale, ms 10861115, f. 29v.
This table is essentially the same as Table 4 but has twice as many columns and
about half as many rows. Another difference is the use of zodiacal signs rather
than physical signs. And yet the most striking difference is that the entries com-
mon to both tables differ slightly in a systematic way. In Table 18 we compare
excerpts of these two tables of John de Murs, one belonging to the Tables of 1321
(Table 4) and other to the Patefit (Table 17).

table 17 True lunar positions for each day between successive syzygies (excerpt)

Days 0s 3 diff. 0s 6 3s 0 11s 27 diff. 12s 0 diff.


(s) () () (s) () (s) () (s) () () (s) () ()

1 0 11;36 14 0 11;22 0 8;10 0 12; 4 14 0 11;50 14


2 0 23;24 13 0 23;11 0 21;38 0 23;51 13 0 23;38 14
3 1 5;12 11 1 5; 1 1 5;26 1 5;36 12 1 5;24 12

15 6 19;30 * 16 6 19;44 6 22;19 6 18;57 17 6 19;14 16


16 7 4; 7 * 13 7 4;20 7 4;54 7 3;39 14 7 3;53 16

* Vatican, ms lat. 3116 has mistakenly 19;16 and 4;13, respectively.


john of murss tables of 1321 335

table 18 Comparison of extracts ( for mean arguments of anomaly 12


and 180) of the two tables of John of Murs for the correction of
the Moon

t. of 1321 Patefit t. of 1321 Patefit


Days 0,12 0s12 3,0 6s 0
() (s) () () (s) ()

1 0,10;57 0 10;55 0,14;45 0 14;47


2 0,22;50 0 22;47 0,29;30 0 29;33
3 0,34;45 1 4;41 0,44; 5 * 1 14;15
4 0,46;48 1 16;43 0,58;38 1 28;43
5 0,59; 3 1 28;58 1,12;44 2 12;49
6 1,11;46 2 11;41 1,26;20 2 26;25
7 1,24;47 2 24;42 1,39;36 3 9;41
8 1,38; 5 3 8; 0 1,52;36 3 22;41
9 1,51;53 3 21;48 2, 5;11 4 5;16
10 2; 6;12 4 6; 8 2,17;22 4 17;26
11 2,20;57 4 20;54 2,29;14 4 29;17
12 2,35;55 5 5;53 2,40;56 5 10;58
13 2,50;51 5 20;51 2,52;39 5 22;40
14 3, 5;37 6 5;38 3, 4;27 6 4;27
15 3,20;14 6 20;15 3,16;18 6 16;18
16 3,34;38 7 4;44 3,28;14 6 28;11

* L mg.: al. 11; O: 0,44;11.

As is readily seen, the corresponding entries in these tables differ systematically


in the minutes, and the differences range from 0;5 to +0,5. We note, however,
that when computing the true position of the Moon exactly one day after
the mean conjunction of 18 January 1322 at Toledo we found it to be 5,23;34,
using the correction deduced from Table 4 (17;56) and in good agreement with
recomputation, 323;34,41. Had we used Table 17, the correction to be applied
would have been 18;1, and thus the true position of the Moon 5,23;39, in worse
agreement with recomputation.
There is a fifth table whose heading is Tabula invencionis veri loci lune incip-
iendo a coniunctione eius a sole (B 183r188v). Again, it is a double argument
table, for 30 days and the argument of lunar anomaly, and very similar to
Tables 17 and 4, the latter of which is also for 30 days. This table is not in Vat.
lat. 3116.
336 chapter 9

5 Conclusions

John of Murss Tables of 1321 are exclusively concerned with the luminaries and
the planets. Their most significant feature is the organizational principle: the
mean motions of the planets are presented in tables for the mean conjunctions
of each planet with the Sun, and their equations are given in double argument
tables. This approach meant that astronomers could avoid a lot of cumbersome
computations to determine true planetary positions, compared with using
tables previously available in Latin. In setting up his tables that way, John of
Murs took advantage for the planets of the pattern used for the Moon in its
conjunctions with the Sun. We know of no other example of this presentation
for the planets in the Alfonsine corpus.
On the other hand, as far as we know, double argument tables were a novelty
in Europe. North noted that they had been used by Ibn Ynus (tenth century,
Cairo) and al-Baghdd (thirteenth century, Baghdad), but we are not aware of
any double argument tables produced in the Iberian Peninsula prior to 1320.32
In the tables devoted to syzygies in the Patefit, John of Murs split the time
from mean to true syzygy, t, into two separate terms, one for the Sun and
one for the Moon, and was probably the first to introduce this approach to a
complicated problem. The tables he compiled, among them double argument
tables, helped practitioners of astronomy in completing their tasks by reducing
the number of required computations to reach a certain result.
John of Murs compiled his tables with material already in place in Paris
and, in particular, there is evidence that he shared parameters and approaches
with his contemporary, John Vimond. Indeed, we have shown that there is a
stronger relationship between John Vimonds work and John of Murss than
previously thought. As repeatedly indicated in the titles of the tables, John of
Murs used radices for Toledo which he attributed to Alfonso, king of Castile.
No wonder, for John of Murs had a thorough knowledge of what he called
the tables of Alfonso that he had described in his Expositio intentionis regis
Alfonsii circa tabulas eius. Indeed, in his Tables of 1321 John of Murs used
parameters that occur in that text. As Lejbowicz put it regarding John of Murs,
lappropriation des hritages fournit aux novateurs l appui ncessaire leur
travail.33 Moreover, it is also clear that all basic parameters for the planets
and the two luminaries (for mean motions and equations) that some years

32 North, Alfonsine Tables in England (ref. 1), 279 and 293; Chabs and Goldstein, Finding
True Syzygy (ref. 30), 93 and 104 n. 3.
33 Lejbowicz, Prsentation de Jean de Murs (ref. 1), 175.
john of murss tables of 1321 337

later formed the core of the Parisian Alfonsine Tables are already embedded
in his Tables of 1321. Hence credit should properly be given to John of Murs for
these innovations that played a decisive role in the transmission of Alfonsine
astronomy.

Acknowledgements

We thank Richard L. Kremer (Dartmouth College) for his helpful comments on


a draft of this paper.
chapter 10

Isaac Ibn al-adib and Flavius Mithridates:


The Diffusion of an Iberian Astronomical
Tradition in the Late Middle Ages*

Isaac Ibn al-adib (or al-Adab) first appears in the literature, when he was in
Castile in the 1370s, as a student of Judah ben Asher ii (then resident in Burgos),
the great-grandson of Asher ben Yeiel of Cologne (d. c. 1328) who became
chief rabbi of Toledo in 1305. Judah ben Asher ii (d. 1391) composed a set of
astronomical tables that are poorly preserved in a unique copy;1 he was killed
during the anti-Jewish riots that took place all over Spain beginning in 1391 and,
as a result of these riots, many Jews left Spain around that time. Ibn al-adib
was a member of a prominent Jewish family in Castile and arrived in Sicily no
later than 13962 when the island was ruled by Joan i (d. 1396), King of Aragon
and eldest son of Pere iii of the house of Barcelona. Both Pere and Joan were
keen on astronomy and had Jewish scholars at the royal court.3 Ibn al-adibs
main astronomical work was a set of tables in Hebrew for conjunctions and
oppositions of the Sun and the Moon, called Ora selulah (the paved way:
cf. Prov. 15:19). This text is mentioned by several later astronomers, notably
by Abraham Zacut in chapter 5 of his Great composition (ha-ibbur ha-gadol),
composed in 1478.4 Ibn al-adib died in Sicily around 1426.
Flavius Mithridates was a name assumed by William Raymond of Moncada,
a shadowy figure who was active in Italy in the late 15th century.5 Mithridates

* Journal for the History of Astronomy, 37 (2006), 147172.


1 B.R. Goldstein, Abraham Zacut and the Medieval Hebrew Astronomical Tradition, Journal
for the History of Astronomy, xxix (1998), 177186, p. 179.
2 M. Steinschneider, Mathematik bei den Juden, 2nd edn (Hildesheim, 1964), 168; B.R. Goldstein,
Descriptions of Astronomical Instruments in Hebrew, in Essays in Honor of E.S. Kennedy, ed.
by D.A. King and G. Saliba, Annals of the New York Academy of Science, d (1987), 105141, p. 128.
3 A. Rubi i Lluch, Documents per lhistria de la cultura mig-eval (Barcelona, 19081921); J.M.
Mills, Las Tablas Astronmicas del Rey Don Pedro el Ceremonioso (MadridBarcelona, 1962);
J. Chabs, with the collaboration of A. Roca and X. Rodrguez, Lastronomia de Jacob ben David
Bonjorn (Barcelona, 1992).
4 See, e.g., Oxford, Bodleian Library, ms Opp. Add. 8 42, f. 17b. Cf. F. Cantera Burgos, El judo
salmantino Abraham Zacut, Revista de la Academia de Ciencias Exactas, Fsico-Qumicas y
Naturales de Madrid, xxvii (1931), 63398, espec. pp. 113, 171.
5 See K. Lippincott and D. Pingree, Ibn al-tim on the Talismans of the Lunar Mansions,

koninklijke brill nv, leiden, 2015 | doi: 10.1163/9789004281752_012


isaac ibn al-adib and flavius mithridates 339

reports that he was the son of a Sicilian rabbi, Nissim Ab l-Faraj (Rabi nissim
abu ilfaragh: Vatican, ms Urb. lat. 1384, f. 3a).6 It has been established that he
converted to Christianity in the 1460s and that before his conversion his name
was Samuel ben Nissim.7 His new name was presumably chosen in honour
of his first patron, William Raymond of Moncada, Count of Adern (that is,
Adrano, in Sicily). A document dated 1473 indicates that Guillelmo Raymundo
de Moncata was a student at the University of Naples, and some evidence shows
that by 1477 Mithridates had reached Rome where he attracted the attention of
Federico of Montefeltro, Duke of Urbino (14221482), among others.8 Federico
played a major role in political and military affairs in Italy and he was also a
leading patron of the arts during the Renaissance. Mithridates is best known as
an adviser to the great humanist, Giovanni Pico della Mirandola (d. 1494), for
whom he translated many kabbalistic texts from Hebrew into Latin around the
year 1486.9 We will focus on Mithridatess interest in astronomy, rather than
kabbalah, and on his astronomical tables, uniquely preserved in Vatican, ms
Urb. lat. 1384. Samuels father was a student of Isaac Ibn al-adib10 and, as we
shall see, there is a strong connection between the astronomical tables of Ibn
al-adib and those of Mithridates.
The main purpose of Ibn al-adibs tables is to compute the time and posi-
tion of true syzygy, and the circumstances for solar and lunar eclipses.11 Some of

Journal of the Warburg and Courtauld Institutes, l (1987), 5781, pp. 5859. Although Wil-
liam Raymond of Moncada did not take Flavius Mithridates as his nom de plume until the
1480s, we generally refer to him by the name by which he is best known.
6 Cf. R. Starrabba, Ricerche storiche su Guglielmo Raimondo Moncada, ebreo convertito
siciliano del xv, Archivio Storico Siciliano, iii (1878), 1591, p. 87.
7 A. Scandaliato, Le radici familiari culturali di Guglielmo Raimondo Moncada, ebreo
convertito del rinascimento, nella Sicilia del. sec. xv, in Una Manna Buona per Mantova.
Man Tov le-Man Tovah. Studi in onore di Vittore Colorni, ed. by M. Perani (Florence, 2004),
203240; E. Engel, A Palaeographical Analysis of Mithridates Hebrew Autographs, in
Guglielmo Raimondo Moncada alias Flavio Mitradate: Un ebreo converso siciliano. Atti del
Convegno Internazionale, Caltabellotta (Agrigento), 2324 ottobre 2004, ed. by M. Perani
(Palermo, 2008), 201223.
8 Starrabba Guglielmo Raimondo Moncada (ref. 6), 39, 41, and 4748; M. Steinschneider,
Die hebraeischen bersetzungen des Mittelalters (Berlin, 1893), 986987.
9 G. Busi, et al., The Great Parchment: Flavius Mithridates Latin Translation, the Hebrew
Text, and an English Version (Turin, 2004) pp. 1617; C. Wirszubski, Pico della Mirandolas
Encounter with Jewish Mysticism (Cambridge (Mass.) and London, 1989).
10 Munich, Staatsbibliothek, ms Heb. 246, f. 83a; see M. Steinschneider, Die hebraeischen
Handschriften der K. Hof- und Staatsbibliothek in Muenchen (Munich, 1895), 120.
11 For other approaches to this problem, see J. Chabs and B.R. Goldstein, Computational
340 chapter 10

his tables are based on previous identifiable astronomical material, originally


compiled in the Iberian peninsula and used there in the fourteenth century.
Indeed, there was a rich astronomical tradition in Hebrew in the fourteenth
and fifteenth centuries both in the Iberian peninsula and in southern France,
and a great many sets of tables in Hebrew were composed at that time, each
with its own special characteristics, by Levi ben Gerson, Jacob ben David Bon-
jorn of Perpignan (known as ha-Poel), Immanuel Bonfils of Tarascon, Abraham
Zacut, and Judah ben Verga, among others.12 It is quite unusual to find astro-
nomical tables in Latin compiled in the fifteenth century that do not depend
on either the Toledan Tables or the Parisian Alfonsine Tables. In fact, most
sets of tables in Latin that are independent of these two families derive from
the astronomical tradition in Hebrew; this applies to Mithridatess set, for it
depends primarily on the zij of Ibn al-adib. To be sure, the various tables com-
posed in this period all depend on the Ptolemaic tradition as it was elaborated
in al-Andalus (i.e., Muslim Spain) and then diffused to the Jewish and Christian
communities in the Iberian peninsula, later spreading to other parts of Europe.
Ibn al-adibs tables are preserved in about 20 Hebrew manuscripts of which
we have inspected the following: Vatican, ms Heb. 379; Paris, Bibliothque
nationale de France, ms Heb. 1086; Munich, Staatsbibliothek, ms Heb. 343. We
know of no Latin version of these tables, but they were preserved in Greek:
Venice, ms Marc. gr. 326 (ff. 135139); and Mount Athos, Vatopedi, ms 188
(ff. 113116v). According to Tihon and Mercier,13 Matthew Camariotes (d. 1490/1)

Astronomy: Five centuries of Finding True Syzygy, Journal for the History of Astronomy,
xxviii (1997), 93105. On Ibn al-Kammd (early 12th century), see J. Chabs and B.R. Gold-
stein, Andalusian Astronomy: al-Zj al-Muqtabis of Ibn al-Kammd, Archive for History of
Exact Sciences, xxxxviii (1994), 141; and on Ibn al-Raqqm (d. 1315), see, E.S. Kennedy, The
Astronomical Tables of Ibn al-Raqqm a Scientist of Granada, Zeitschrift fr Geschichte der
Arabisch-Islamischen Wissenschaften, xi (1997), 3572.
12 See, e.g., Immanuel ben Jacob Bonfils, Sefer shesh kenafayim [Hebrew edition of The Six
Wings, bound with Isaac ben Solomon, Sefer or ha-levanah] (Zhitomir, 1872); Mills, Las
Tablas Astronmicas (ref. 3); J. Chabs, The Astronomical Tables of Jacob ben David
Bonjorn, Archive for History of Exact Sciences, xlii (1991), 279314; Chabs, Lastronomia
(ref. 3); B.R. Goldstein, Levi ben Gersons Astronomical Tables (New Haven, 1974); idem,
Astronomy in the Medieval Spanish Jewish Community, in Between Demonstration and
Imagination: Essays in the History of Science and Philosophy Presented to John D. North,
ed. by L. Nauta and A. Vanderjagt (Leiden, 1999), 225241; J. Chabs and B.R. Goldstein,
Astronomy in the Iberian Peninsula: Abraham Zacut and the Transition from Manuscript to
Print (Philadelphia, 2000); B.R. Goldstein, The Astronomical Tables of Judah ben Verga,
Suhayl, iv (2001), 227289.
13 A. Tihon and R. Mercier, Georges Gmiste Plthon: Manuel dastronomie (Louvain-la-
Neuve, 1998), 12.
isaac ibn al-adib and flavius mithridates 341

composed an adaptation of the canons, extant in Leiden, ms bpg 74e, which


does not include the tables.
Vatican, Biblioteca Apostolica, ms Urb. lat. 1384, is a manuscript of 88 folios,
probably copied in 14801481, containing three items: (1) De imaginibus coe-
lestibus, a translation into Latin of a work in Arabic attributed to Ibn al-tim;
(2) canons and tables on eclipses; and (3) a translation from Arabic into Latin
of the Quran.14 In item 2, the canons appear on ff. 30r43v, and its title includes
a short dedication to Federico of Montefeltro, Duke of Urbino, together with
the name of the author, given here as Guillelmus Raymundus de Moncata,
i.e., William Raymond of Moncada. There are similar dedications for the two
other texts that begin on ff. 1r and 65r. On f. 31r the epoch of the tables is given
explicitly as Sunday, January 8, 1475. We do not think this date has any spe-
cial significance other than the fact that it is the date of a mean conjunction
of the Sun and the Moon, and that the civil day January 7 (having 12 hours
in common with the astronomical day January 8 that begins at noon on Jan-
uary 7 in the civil calendar) is Saint Raymonds day in the Christian calen-
dar.
Several Greek and Muslim authorities are mentioned in the canons: Aristo-
tle, Ptolemy, and Ibn al-tim on f. 39r; Ibn al-tim on f. 42v; and Ibn Sina
and Ptolemy on f. 43v. On f. 42v we are told that Ibn al-tim observed a solar
eclipse in al-Andalus on July 19, 939.15 Of particular interest is that three other
astronomers are cited on f. 32v: al-Battn, Ibn al-Raqqm, and Ibn al-Kammd:
() apud Il Bactani et Ibn il raccam nec non et Ibn il chimadi. These are the same
three astronomers mentioned by Ibn al-adib in the introduction to his Ora
selulah.16 We are not aware of any other Latin text of the fourteenth or fifteenth
centuries in which Ibn al-Raqqms name appears.
Throughout the canons there are a few references to tables, and there is
even a complete table for the possibility of an eclipse (f. 35v), where the limits
for the distances from a lunar node are 15 for lunar eclipses and 10;16 for
solar eclipses (cf. Vatican, ms Heb. 379, f. 4a). According to Ptolemy the limit for
the possibility of a lunar eclipse corresponds to distances from a lunar node of
12;12; whereas for solar eclipses, the limits are 17;41 to the north and 8;22

14 For a detailed table of contents and discussion of items 1 and 3, see Lippincott and Pingree,
Ibn al-tim (ref. 5); K. Lippincott, More on Ibn al-tim, Journal of the Warburg and
Courtauld Institutes, li (1988), 188190.
15 Cf. Lippincott and Pingree, Ibn al-tim (ref. 5), 58.
16 Vatican, ms Heb. 379, f. 2a.
342 chapter 10

to the south.17 As Pedersen explains, an eclipse may be possible even if at true


syzygy the true Sun is 3 from the place of the mean Sun at mean syzygy.18 This
means that a lunar eclipse may be possible when the mean Sun at syzygy is
about 15 from a lunar node. In the tables on ff. 47v and 52v, below, the lunar
eclipse limits are 12 from the lunar node. On f. 31v, lines 69, the text refers
to the movement on the 8th sphere, which is called motus proprie et per se,
whereas the movement on the 9th sphere is called motus naturalis. This is
the same terminology as that used in previous astronomical texts in Castilian
and in Arabic.19 On f. 35r, the difference between the 8th and 9th spheres, that
is the difference between sidereal longitudes and tropical longitudes, is said to
be 12;30 without an explanation of the way this value was determined (see also
Tabula gradus solaris, f. 56v, below).
A commentary on Ibn al-adibs tables by Abraham Gascon (Cairo, mid-
sixteenth century) contains a worked example for the solar eclipse of August
11, 1542 (New York, Jewish Theological Seminary of America [jtsa], ms 2571,
ff. 1a8b), based on Ibn al-adibs zij. In this manuscript the authors name
is consistently given as Ibn al-adib and not Ibn al-Adab as in some other
manuscripts.20 In Gascons commentary (f. 3a) the difference between the 8th
and 9th spheres, which he calls the motion in access (tenucat ha-haqbalah), is
taken to be 12, without any explanation of the way this value was determined.
Ibn al-adib does not address this issue in the canons to his tables except to
say that his tables are arranged for 8th sphere, i.e., his coordinates are sidereal
(Vatican, ms Heb. 379, f. 2b). However, in a commentary on these tables by Ibn
al-adibs son, Jacob, the motion in access is said to be 12 (New York, jtsa,
2571, f. 16a), and this may be the source for Gascons remark.21
The tables of Mithridates are on ff. 44r61v, following the canons. We offer a
brief description of them and their relationship to those of Ibn al-adib, with
special attention to the parameters embedded in them, in order to identify lines
of transmission of the astronomical material. The first four tables, together with
two other tables on f. 61rv, list the solar and lunar equations at syzygy, and it is
worth noting that these tables come before the tables for mean motions (see

17 Almagest vi.5; G.J. Toomer, Ptolemys Almagest (New York, 1984), 286287.
18 O. Pedersen, A Survey of the Almagest (Odense, 1974), 229230.
19 See J. Chabs and B.R. Goldstein, The Alfonsine Tables of Toledo (Dordrecht and Boston,
2003), 217218.
20 On Gascon, see B.R. Goldstein, The Hebrew Astronomical Tradition: New Sources, Isis,
lxxii (1981), 237251.
21 Another copy of Jacobs commentary is extant in London, British, Library, Or. 2806, ff. 20b
39b.
isaac ibn al-adib and flavius mithridates 343

ff. 54r55r and 56v57r), whereas the standard order in medieval sets of tables
is the opposite.

f. 44r. Prime adequationes temporis in .xij. signis

This table displays the correction for the solar position (in hours, minutes, and
seconds) at syzygy as a function of the solar anomaly given for each degree. The
maximum is 3;54,22h at anomaly 3s 1 (see Table a). The entries are the same
as those in the column for the correction for the time in Ibn al-adibs tables
in Vatican, ms Heb. 379, ff. 7a8a (see Table aa), where the same maximum
value is found. The letter s at the top left-hand margin of f. 44r means that the
entries are to be subtracted between arguments 0s 1 and 5s 30, and the letter
a at the bottom left-hand margin indicates that the entries are to be added
between 6s 0 and 11s 29. It is worth noting that the maximum occurs at 91; we
are not aware of any other table for the solar equation with a maximum value
at that argument. After inspection of the surrounding entries, it seems that
this particular entry is an isolated error and that the intended maximum was
3;54,20h at an argument of 3s 2. To determine the underlying maximum solar
equation, we have compared Table a with various tables for the solar equation:
Ibn al-Kammds table where the maximum is 1;52,44;22 al-Battns where the
maximum is 1;59,10;23 and the Parisian Alfonsine Tables where the maximum
is 2;10,0.24 Dividing an entry in each of these tables by the corresponding entry
in Table a leads to the result that the best fit is with the entries in al-Battns
table.

table a The solar equation in time at syzygy (Mithridates, f. 44r)

s 0[s] 1[s] 2[s] 3[s] 4[s] 5[s]

1 0; 4, 1h 1;57,10h 3;21,11h 3;54,22h 3;23, 0h 1;56,55h


2 0; 7,56 2; 0,16 3;23,15 3;54,20 3;21,55 1;53,13
3 0;11,52 2; 4, 0 3;25,17 3;54,18 3;20, 8 1;49,25
4 0;15,48 2; 7,22 3;27,15 3;54, 8 3;18, 1 1;45,35

22 Chabs and Goldstein, Ibn al-Kammd (ref. 11), 6 ff.


23 C.A. Nallino, Al-Battn sive Albatenii Opus Astronomicum (2 vols, Milan, 19031907), ii,
78 ff.
24 Tabule astronomice illustrissimi Alfontij regis castelle ed. by E. Ratdolt (Venice, 1483), e2v
e4r.
344 chapter 10

table a The solar equation in time at syzygy (Mithridates, f. 44r) (cont.)

s 0[s] 1[s] 2[s] 3[s] 4[s] 5[s]

5 0;19,41 2;10,43 3;29, 9 3;53,51 3;15,51 1;41,45

10 0;39,20 2;26,41 3;37,28 3;51,56 3; 3,44 1;22,21

20 1;16,35 2;55,32 3;49,21 3;41,15 2;33, 0 0;42, 5

30 1;53,52 3;19, 6 3;54, 2 3;24,40 2; 0,41 0; 0, 0

a 11[s] 10[s] 9[s] 8[s] 7[s] 6[s]

0s 20. Read: 1;17,35.

table aa Table for the solar anomaly (Ibn al-adib, Vatican, ms Heb.
379, f. 7a), the first of 6 similar sub-tables

0[s]
Subtract

Solar Corr. for Corr. for Corr. for


Anomaly * the Lunar Anomaly the Time the Position

1 0; 2,47 0; 4, 1h 0; 2,12
2 0; 5,32 0; 7,56 0; 4,20
3 0; 8,11 0;11,52 0; 6,29
4 0;11,12 0;15,48 0; 8,40
5 0;13, 3 0;19,41 0;10,49

10 0;27,15 0;39,20 0;21,36

20 0;55,25 1;16,35 0;42,37

30 1;19,32 1;53,42 1; 2,24

11[s]
Add
isaac ibn al-adib and flavius mithridates 345

* In some of the sub-tables in Vatican, ms Heb. 379, the


manuscript reads Lunar Anomaly, but in Paris, ms Heb.
1086, 7a8a, the heading is consistently Solar Anomaly.

f. 44v. Secunde adequationes temporis in totidem signis

This table displays the correction for the lunar position (in hours, minutes, and
seconds) at syzygy as a function of the lunar anomaly given for each degree.
The letters a and s in the margins have the same meaning as in the previous
table, but their placement is interchanged. The maximum is 9;42,6h at anomaly
3s 6 (see Table b); comparison with the surrounding entries indicates that
the entries around the maximum value are mistaken. The entries in this table
are the same as those in the column for the correction for the time in the
corresponding table in Vatican, ms Heb. 379, ff. 8b9b (see Table bb), where the
same maximum value is found. Again, when we divide an entry in each of the
tables of other authors for the lunar equation by the corresponding entry in
Table b we find that the best fit is with the entries in al-Battns table, where
the maximum lunar equation is 5;1,0. But an underlying value of 4;56 cannot
be definitely excluded.

table b The lunar equation in time at syzygy (Mithridates, f. 44v)

a 0[s] 1[s] 2[s] 3[s] 4[s] 5[s]

1 0; 9,20h 4;38,25h 8; 9, 5h 9;40, 1h 8;39,46h 5; 5,30h


2 0;18,40 4;46,52 8;14,25 9;41,35 8;36,15 4;55,47
3 0;28, 2 4;55, 0 8;18,36 9;41,54 8;31,18 4;46,20
4 0;37,20 5; 3, 6 8;22,46 9;42, 4 8;26, 9 4;36,37
5 0;46,40 5;11,12 8;27,50 9;42, 5 8;20,49 4;26,53
6 0;56, 0 5;19,19 8;32, 2 9;42, 6 8;15,14 4;17, 3

10 1;33, 0 5;50,32 8;49,41 9;40,11 7;51,29 3;36,47

20 3; 3,49 7; 1,20 9;22,55 9;21,53 6;40,13 1;50,30

30 4;30,15 8; 1,44 9;40,11 8;45,42 5;14,26 0; 0, 0

s 11[s] 10[s] 9[s] 8[s] 7[s] 6[s]


346 chapter 10

3s 1. Probably a mistake for 9;41,0h; the same mistake is found in Vatican, ms Heb. 379.
In Hebrew, using alphabetical notation for numbers, 41 is ma (with no space between
the letters) and 40,1 is m a (with a space between the letters).

table bb Table for the lunar anomaly (Ibn al-adib, Vatican, ms Heb.
379, f. 8b), the first of 6 similar sub-tables

0[s]
Add

Lunar Corr. for Corr. for Corr. for


Anomaly the Lunar Anomaly the Time the Position

1 0; 5, 5 0; 9,20h 0; 0,23
2 0;10,10 0;18,40 0; 0,46
3 0;15,15 0;28, 2 0; 1, 9
4 0;20,20 0;37,20 0; 1,32
5 0;25,25 0;46,40 0; 1,55

10 0;50,48 1;33, 0 0; 3,49

20 1;40, 4 3; 3,49 0; 7,32

30 2;28, 7 4;30,15 0;11,14

11[s]
Subtract

Thus, the two preceding tables (a and b) were intended for computing a first
approximation of the time between mean and true syzygy and, presumably,
were based on tables for the solar and lunar equations with maxima of 1;59,10
and 5;1,0, respectively, both of which are well represented in the astronomical
literature in the Iberian peninsula. On the other hand, the next two tables (c
and d) are intended for computing a first approximation of the difference in
longitude between mean and true syzygy. These tables treat the effect of each
luminary separately and differ substantially from those compiled for the same
purpose at about the same time.25

25 Cf. Chabs and Goldstein, True Syzygy (ref. 11).


isaac ibn al-adib and flavius mithridates 347

f. 45r. Prime adequationes motus draconis et loci solis in .xij. signis

This table displays the correction for the solar position (in degrees, minutes,
and seconds) as a function of the solar anomaly given for each degree. The
letters s and a in the margins have the same meaning as in the previous
tables. The maximum is 2;8,51 at anomaly 3s 1 (see Table c). The entries are
the same as those in the column for the correction for the position in the
corresponding table in Vatican, ms Heb. 379, ff. 7a8a (see Table aa), where the
same maximum value is found.

table c The solar equation at syzygy (Mithridates, f. 45r)

s 0[s] 1[s] 2[s] 3[s] 4[s] 5[s]

1 0; 2,12 1; 4,22 1;50,38 2; 8,51 1;51,34 1; 4,18


2 0; 4,20 1; 6, 5 1;51,23 2; 8,47 1;49,58 1; 2,11
3 0; 6,29 1; 8, 9 1;52,49 2; 8,45 1;49,39 1; 0, 7
4 0; 8,40 1;10, 0 1;53,53 2; 8,40 1;48,49 0;58, 9
5 0;10,49 1;11,50 1;54,56 2; 8,39 1;47,38 0;55,55

10 0;21,36 1;20,36 1;59,31 2; 7,28 1;41,30 0;45,14

20 0;42,37 1;35,59 2; 6, 1 2; 1,35 1;26, 8 0;23, 7

30 1; 2,29 1;49,26 2; 8,44 1;52,22 1; 6,20 0; 0, 0

a 11[s] 10[s] 9[s] 8[s] 7[s] 6[s]

The entries in this table, here given in degrees, are equivalent to those in Table
a, given in hours. If ai and ci are the corresponding entries in Tables a and c,
then the expression

ci = 0;32,56 ai,

where 0;32,56/h, the mean lunar velocity, yields fairly good results. This
seems to mean that an entry in Table c is the distance the Moon travels at its
mean velocity corresponding to the time computed in Table a. We note that
0;32,56/h is an approximation of the lunar mean motion in longitude,
13;10,35/d.
348 chapter 10

f. 45v. Secunde adequationes motus draconis et loci solis in totidem


signis

This table displays the correction for the lunar position (in degrees, minutes,
and seconds) as a function of the lunar anomaly given for each degree. There
is no indication in the table or the margins concerning the addition or sub-
traction of the entries; we have introduced a and s in accordance with the
presentation in Ibn al-adibs corresponding table. The maximum is 0;23,55
at anomaly 3s 4 (Table d). The entries are the same as those in the col-
umn for the correction for the position in the corresponding table in Vati-
can, ms Heb. 379, ff. 8b9b (see Table bb), where the same maximum value is
found.

table d The lunar equation at syzygy (Mithridates, f. 45v)

[a] 0[s] 1[s] 2[s] 3[s] 4[s] 5[s]

1 0; 0,23 0;11,25 0;20, 6 0;23,52 0;21,21 0;12,32


2 0; 0,46 0;11,46 0;20,21 0;23,53 0;21,13 0;12, 9
3 0; 1, 9 0;12, 7 0;20,30 0;23,54 0;21, 0 0;11,44
4 0; 1,32 0;12,26 0;20,40 0;23,55 0;20,50 0;11,21
5 0; 1,55 0;12,46 0;20,52 0;23,54 0;20,35 0;10,58

10 0; 3,49 0;14,23 0;21,46 0;23,50 0;19,27 0; 8,56

20 0; 7,32 0;17,18 0;23, 8 0;23,20 0;16,27 0; 4,31

30 0;11, 4 0;19,47 0;23,50 0;21,36 0;12,50 0; 0, 0

[s] 11[s] 10[s] 9[s] 8[s] 7[s] 6[s]

The entries in this table, here given in degrees, are equivalent to those in Table
b, given in hours. If bi and di are the corresponding entries in Tables b and d,
then the expression

di = 0;2,28 bi,

where 0;2,28/h, the mean solar velocity, yields good results. This seems to
mean that an entry in Table d is the distance the Sun travels at its mean velocity
isaac ibn al-adib and flavius mithridates 349

corresponding to the time computed in Table b. We note that 0;2,28/h is an


approximation of the mean motion of solar longitude, 0;59,8/d.

Two tables on f. 61rv, with the same format as Tables a, b, c, and d, are described
below.

f. 46r. Tabula multiplicationis numerorum

In this multiplication table we find the numbers 1 to 10 across the top and the
numbers from 1 to 10 and then multiples of 5 from 15 to 60 at the beginning of
each row. However, Vatican, ms Heb. 379, f. 15r, displays a multiplication table
with the numbers 1 to 10 across the top and the numbers 1 to 30 (at intervals of
1) at the beginning of each row. On f. 15v the table continues with 1 to 10 across
the top and 31 to 60 (at intervals of 1) at the beginning of each row. Mithridates
kept the heading for the columns, but reduced the number of rows.

f. 46v47r. Tabula horarum et minutorum meridierum in his


latitudinibus

The entries in this table correspond to the time of half-daylight (in hours and
minutes) for various latitudes (0;0, 8;28, 16;50, 24;0, 30;20, 36;0, 40;30,
45;0, 48;30, 51;30, 59;45, 63;0, 64;45, and 66;25) as a function of the solar
longitude. The argument is given at 10-intervals, and the columns for the last 4
latitudes are not completely filled in. This table is similar to that of Ibn al-adib
(Vatican, ms Heb. 379, f. 10a), but the argument there is given at 5-intervals and
for 9 latitudes that are slightly different from those in Mithridatess table.

The table for lunar eclipses on f. 47v and the two tables for lunar eclipses on f. 48r
that follow agree with the corresponding tables in Wing 4 of the Six Wings
by Immanuel Bonfils of Tarascon (Provence, Southern France), composed in
Hebrew c. 1365.

f. 47v. Quot puncta globi lunaris deficient

This table displays the digits of a lunar eclipse (in digits and minutes of a
digit), where 1 digit is a twelfth of the lunar diameter, as a function of the
argument of latitude (from 0 to 12, for each integer degree) and the lunar
350 chapter 10

anomaly (from 0s to 11s, for each integer sign). The maximum value is 21;36
digits, corresponding to an argument of latitude of 0s 0. The entries are the
same as those in the columns for the digits of the diameter in Ibn al-adibs
tables for lunar eclipses: see Vatican, ms Heb. 379, f. 11a. The maximum in this
table agrees with the maximum in Wing 4 in the tables of Immanuel Bonfils.26
This maximum is also found in al-Battns tables (Nallino 19031907, 2:90). In
the Almagest the maximum is called entire following an entry of 21 digits;
however, Neugebauer shows that the parameters in the Almagest lead to a
maximum of 21;36 digits.27

f. 47v. Tabula colorum

The heading for this table refers to the colour of eclipses. Only the frame and the
arguments are given; no other entries are displayed. The argument runs from
1 to 12 digits, for each integer digit. A table for colours of solar eclipses with
these arguments appears in Vatican, ms Heb. 379, f. 13b.28 The text preceding
the tables mentions a table for the colours of eclipses (f. 38v, line 2).

f. 47v. Alia secundum il chimadi

The heading for this table also refers to the colour of eclipses. As in the previ-
ous table, the frame and the arguments are given without any other entries.
The argument runs from 1 to 24 digits, for each integer digit, and there are
two successive numbers in each of the 12 rows. The heading mentions Ibn
al-Kammd.29 There are two tables for the same purpose on f. 53v. These argu-
ments probably refer to digits of a lunar eclipse.

26 See Bonfils 1872, pp. 3840; cf. P. Solon, The Hexapterygon of Michael Chrysococces. (Ph.D.
dissertation, Brown University, 1968); and idem, The Six Wings of Immanuel Bonfils and
Michael Chrysokokkes, Centaurus, xv (1970), 120, p. 7.
27 Toomer, Almagest (ref. 17), 307; O. Neugebauer, A History of Ancient Mathematical Astron-
omy (Berlin and New York, 1975), 136.
28 B.R. Goldstein, Colors of Eclipses in Medieval Hebrew Astronomical Tables, Aleph, v
(2005), 1134.
29 For colours of eclipses according to Ibn al-Kammd, see Chabs and Goldstein, Ibn
al-Kammd (ref. 11), 1819.
isaac ibn al-adib and flavius mithridates 351

f. 48r. Quot horis durabit medius defectus

This table displays the time of the half-duration of lunar eclipses (in hours and
minutes) as a function of the argument of latitude (from 0 to 12, for each
integer degree) and the lunar anomaly (from 0s to 11s, for each integer sign).
The maximum value is 2;2h, corresponding to an argument of latitude of 0s 0
and a lunar anomaly of 0s. The entries are the same as those in the columns
for the half-duration in Ibn al-adibs tables in Vatican, ms Heb. 379, f. 11a.
We note that the term defectus is used for eclipse; on f. 42v the term ekleipsis
is written in Greek characters. This maximum for the half-duration of a lunar
eclipse is also found in the tables of Immanuel Bonfils where the time from
first contact to the beginning of totality is given as 1;8h and the half-duration of
totality is given as 0;54h, for 1;8h + 0;54h = 2;2h.30

f. 48r. Numeri medie obscuritatis

This table displays the time of the half-duration of totality of lunar eclipses (in
hours and minutes) as a function of the argument of latitude (from 0 to 12,
for each integer degree) and the lunar anomaly (from 0s to 11s, for each integer
sign). The maximum value is 0;54h, corresponding to an argument of latitude
of 0s 0 and a lunar anomaly of 0s. The entries are the same as those in the
columns for the half-duration of totality in Ibn al-adibs tables: see Vatican,
ms Heb. 379, f. 11a. They also agree with the values in Wing 4 of the Six Wings by
Immanuel Bonfils.31

f. 48r. Numeri defectus s[eptentrionalis] aut m[eridionalis] fuerint

In this table only the frame is given with headings for the columns, but there
are no other entries. It has the same pattern as the previous tables for the digits
of the diameter, the half-duration, and the half-duration of totality.

30 Bonfils 1872, pp. 3840; cf. Solon, The Six Wings (ref. 26), 7.
31 Ibid.
352 chapter 10

f. 48v. Diversus aspectus in primo climate ad latitudinem xvi

The entries in this table for the longitudinal and latitudinal components of the
adjusted parallax are almost identical with those in Ptolemys Handy Tables
for latitude 16;27 and in al-Battns zij for latitude 16;32, according to the
heading.32

f. 49r. Diversus aspectus in secundo climate ad latitudinem xxiiii

The entries in this table for the longitudinal and latitudinal components of the
adjusted parallax are almost identical with those in Ptolemys Handy Tables for
latitude 23;51 and in al-Battns zij for latitude 24;0, according to the heading.

f. 49v. Diversus aspectus in tertio climate ad latitudinem xxx

The entries in this table for the longitudinal and latitudinal components of the
adjusted parallax are almost identical with those in Ptolemys Handy Tables
for latitude 30;22 and in al-Battns zij for latitude 30;40, according to the
heading.33 This table appears in Vatican, Heb. 379, f. 12a, with the heading: Table
for parallax in longitude and latitude for [geographical] latitude 30.

f. 50r. Diversus aspectus in quarto climate ad latitudinem xxxvi

The entries in this table for the longitudinal and latitudinal components of the
adjusted parallax are almost identical with those in Ptolemys Handy Tables
for latitude 36;0 and in al-Battns zij for latitude 36;22, according to the
heading.34

32 See W.D. Stahlman, The Astronomical Tables of Codex Vaticanus Graecus 1291 (Ph.D. disser-
tation, Brown University, 1959; University Microfilms, No. 625761), 268269; and Nallino,
Al-Battn (ref. 23), ii, 95.
33 See Stahlman, Astronomical Tables (ref. 32), 272273; and Nallino, Al-Battn (ref. 23), ii,
97.
34 See Stahlman, Astronomical Tables (ref. 32), 274275; and Nallino, Al-Battn (ref. 23), ii,
98.
isaac ibn al-adib and flavius mithridates 353

f. 50v. Diversus aspectus in quinto climate ad latitudinem xxxx

The entries in this table for the longitudinal and latitudinal components of
the adjusted parallax are almost identical with those in Ptolemys Handy Tables
for latitude 40;56 and in al-Battns zij for latitude 41;15, according to the
heading.35 This table appears in Vatican, Heb. 379, f. 12b, with the head-
ing: Table for parallax in longitude and latitude for [geographical] latitude
40.

f. 51r. Diversus aspectus in sexto climate ad latitudinem xxxxv

The entries in this table for the longitudinal and latitudinal components of the
adjusted parallax are almost identical with those in Ptolemys Handy Tables
for latitude 45;1 and in al-Battns zij for latitude 45;22, according to the
heading.36

f. 51v

This table has no title and very few entries, but it was certainly intended for
parallax for a latitude where the longest daylight (from sunrise to sunset) is
16;30h.

f. 52r

This table has no title and the entries are given in minutes (see Table e). The
argument reads gradus motus differenctia, here meaning the lunar anomaly,
and the entries represent minutes of proportion. Its purpose is to correct
the parallax when the Moon is not at its mean distance at syzygy. Parallax
varies inversely with the lunar distance from the Earth (and lunar distance is a
function of lunar anomaly) but the parallax tables assume that the Moon is at
its mean distance at syzygy. In Ptolemys lunar model for syzygy, the radius of

35 See Stahlman, Astronomical Tables (ref. 32), 276277; and Nallino, Al-Battn (ref. 23), ii,
99.
36 See Stahlman, Astronomical Tables (ref. 32), 278279; and Nallino, Al-Battn (ref. 23), ii,
100.
354 chapter 10

table e Table for correcting parallax

0[s] 1[s] 2[s] 3[s] 4[s] 5[s]

2 5 4 3 0 3 5 28
4 5 4 3 0 3 5 26
6 5 4 3 0 3 5 24
8 5 4 2 1 3 5 22
10 5 4 2 1 3 5 20
12 5 4 2 1 3 5 18
14 5 4 2 1 4 5 16
16 5 4 2 1 4 5 14
18 5 4 1 1 4 5 12
20 5 4 1 2 4 5 10
22 5 4 1 2 4 5 8
24 5 4 1 2 4 5 6
26 5 3 1 2 4 5 4
28 5 3 1 2 4 5 2
30 5 3 1 2 4 5 0

11[s] 10[s] 9[s] 8[s] 7[s] 6[s]

the epicycle is 5;15.37 Hence, the maximum lunar distance is 65;15 at apogee (0
anomaly), the minimum distance is 54;45 at perigee (180 anomaly), where the
mean distance is 60 (near 90 anomaly). Therefore, the correction for parallax
is negative near apogee and positive near perigee and, in general,

p() = p + e() p,

where p is the value in one of the parallax tables, e() is the entry in Table e for
an anomaly , and p() is the corrected value for parallax as a function of .
This table also appears in Vatican, ms Heb. 379, f. 13b, where the heading is:
Table for correcting parallax. Above the columns for 0s, 1s, and 2s, is the word
for subtract, and above the columns for 3s, 4s, and 5s, is the word for add. The
argument is labelled: degrees of [lunar] anomaly [oq]. Similar tables are found

37 In addition to the value 5;15 for this parameter, there are variants 5;13 and 5;14: see Almagest
iv.6; trans. Toomer, Almagest (ref. 17), 197, 202, and 209.
isaac ibn al-adib and flavius mithridates 355

in the astronomical literature in the Iberian peninsula: in an expanded version


(to two sexagesimal places) in the Tabule Verificate for Salamanca (c. 1461), and
in shorter versions in Abraham Zacuts ha-ibbur ha-gadol as well as in its Latin
version, the Almanach Perpetuum.38

f. 52rv

This table has no title, and its entries represent the time between noon and
nonagesimal, here called medium celum (midheaven), given in hours and min-
utes, for each zodiacal sign. The argument is the geographical latitude (from
30 to 51, for each integer degree). The entries are the same as those in the
corresponding table in Vatican, ms Heb. 379, f. 11b, although the horizontal and
vertical arguments are interchanged. In Ibn al-adibs table the geographical
latitude varies from 30 to 45, at intervals of one degree.

f. 52v

This table has no title, and the entries (in degrees, minutes, and seconds)
represent the lunar latitude as a function of the argument of lunar latitude from
0 to 12 in steps of 1. The maximum value, 1;2,16, is reached at argument 12.
The entries are the same as those in the first two columns of a table in Vatican,
ms Heb. 379, f. 11a which has the heading: Table for eclipses of the Moon and
its latitude at the time of an eclipse. The argument of 12 represents the limit
for the Moons distance from the lunar node at the time of a lunar eclipse and,
with a maximum lunar latitude of 5, the value for 12 of argument would be
1;2,22, which is close to the value here (cf. the table on f. 47v).

f. 53r

Again, this table has no title. The entries represent the velocity of the Moon rel-
ative to that of the Sun, in minutes and seconds of arc per hour, from 0;27,30/h
to 0;32,30/h, as a function of lunar anomaly, given in degrees for every other
integer degree. The relative velocity increases monotonically by 0;0,4/h for
arguments between 0s 0 and 6s 0, and then decreases monotonically by

38 Chabs and Goldstein, Zacut (ref. 12), 3334, 62, 118122.


356 chapter 10

0;0,4/h for arguments between 6s 0 and 12s 0. The entries are the same as
those in the corresponding table in Vatican, ms Heb. 379, f. 11b.

f. 53r

As was the case in several of the previous tables, this one is untitled. The entries
give the fraction of the solar diameter that is darkened in a solar eclipse (in
digits and minutes), where 1 digit is a twelfth of the solar diameter, as a function
of the true lunar latitude (in minutes and seconds of arc): see Table f. It is
noteworthy that the entries for the latitude are not rounded whereas the digits
of the diameter are such that the successive differences between them are 0;20
digits or 0;15 digits. The entries are the same as those in the column for digits
of the diameter in the corresponding table in Vatican, ms Heb. 379, f. 13a. We
note, however, that in Mithridatess table the argument (true latitude) increases
downward, but in Ibn al-adibs table it decreases.

table f Digits of a solar eclipse

True latitude Digits True latitude Digits

0 0 11 40 16 19 6 0
1 1 11 20 17 15 5 40
1 12 11 0 18 12 5 20
2 44 10 40 19 8 5 0
3 31 10 20 20 4 4 40
4 48 10 0 21 1 4 20
5 8 9 40 22 57 4 0
6 4 9 20 23 53 3 40
7 56 9 0 24 50 3 20
8 54 8 40 25 47 3 0
9 51 8 20 26 44 2 40
10 46 8 0 27 39 2 20
11 45 7 40 28 35 2 0
12 35 7 20 29 31 1 40
13 30 7 0 30 28 1 20
14 21 6 40 31 23 1 0
15 20 6 20 32 20 0 45
33 17 0 30
34 13 0 15
35 0 0 0
isaac ibn al-adib and flavius mithridates 357

f. 53v

This is a double argument table for the half-duration of a solar eclipse and it
is mentioned in the text (f. 42r). The vertical argument is the corrected lunar
latitude, in degrees, at 2-intervals, and the horizontal argument is the velocity
of the Moon relative to that of the Sun, in minutes and seconds of arc per
hour, from 0;27,30/h to 0;32,30/h. The entries give the half-duration of a solar
eclipse, and they are the same as those in the corresponding table in Vatican,
ms Heb. 379, f. 13a. As was the case with the table on f. 53r, the vertical argument
(corrected lunar latitude) increases downward, whereas in Ibn al-adibs table
it decreases. This table is identical (but for minor variants) to those in Ibn
al-Kammds zij (Madrid, Biblioteca Nacional, ms 10023, f. 54r) and in the Tables
of Barcelona (Table 49).39

f. 53v

There are three frames for small tables, also untitled, but two of them have the
word Colores in the heading, as was the case with the two tables on f. 47v whose
entries are missing.

The following tables yield mean positions and times for conjunctions. In the
canons to Ibn al-adibs tables, chapter 2 (Vatican, ms Heb. 379, f. 2b), we are
told that these motions are computed for the western extremity (not, e.g., for
Toledo) and each day begins at noon of the day preceding it, e.g., Sunday begins
at noon on Saturday and ends at noon on Sunday. In other respects the tables
are arranged according to the Jewish calendar with its 19-year cycle, beginning
with Molad Tohu, the conjunction of Tishri, year 1 Anno mundi, i.e., the era of
creation. According to the Jewish tradition, the conjunction of creation took
place on Monday (day 2) at 5h 204 elaqim (where 1h = 1080 elaqim) counting
from sunset,40 at a location whose distance from the western extremity is
75;43,45 in the view of Ibn al-adib. The epoch of the Jewish calendar, Tishri

39 See Chabs and Goldstein, Ibn al-Kammd (ref. 11), 23; Mills, Las Tablas Astronmicas
(ref. 3), p. 237; and J. Chabs, Astronoma andalus en Catalua: Las Tablas de Barcelona,
in From Baghdad to Barcelona: Studies in the Islamic Exact Sciences in Honour of Prof. Juan
Vernet, edited by J. Casulleras and J. Sams (Barcelona, 1996), 477525, pp. 511512.
40 See, e.g., Maimonides: Sanctification of the New Moon, trans. by S. Gandz, introduction by
J. Obermann, astronomical commentary by O. Neugebauer (New Haven, 1956), 116.
358 chapter 10

1, 1 Anno mundi, is equivalent to Oct. 7, 3761bc (jdn 347998). The radices for
the conjunction of creation are given as the entries for the first month, Tishri.
As we will see, the parameters for the mean motion tables are the same for
Mithridates as they are for Ibn al-adib, but the presentation is different and
the initial values are different. In the case of Mithridates there are places where
radices were to be given but they are filled with names which do not correspond
to any numerical values. As we learn from the worked example in Gascons
commentary, Ibn al-adibs mean motion tables are set up such that the entries
in them refer to the current 19-year cycle, the current year within the cycle, and
the conjunction at the beginning of each month.

f. 54r. Tabula argumenti solis

This table is arranged for conjunctions and it consists of 3 sub-tables, giving


the solar anomaly (in signs, degrees, minutes, and seconds) for 13 consecutive
months, 19 consecutive years, and groups of 19-year cycles (for cycles 1, 2, 3, , 10,
20, 30, , 100, , 700), respectively. This table also displays a numerical value, 0s
14;33,10, together with the word FRAEDERICO, referring to Federico, Duke of
Urbino. The same value is found for the solar anomaly in Ibn al-adibs tables in
Vatican, ms Heb. 379, f. 6b, and it represents the motion of the solar anomaly in
half a mean synodic month, i.e., the time between successive syzygies. Indeed,
it is so labelled in Vat. 379, f. 6b, and is intended for making the tables usable
for oppositions.
The entries for two successive 19-year cycles show a constant difference of
359;44,50, and this is also the entry for 1 cycle. This line-by-line difference is the
same as in the column for the solar anomaly in Ibn al-adibs table in Vatican,
ms Heb. 379, f. 6a, but that table is presented differently: the information is only
given for cycles 273 to 280 (11s 20;54,8 to 11s 19;7,58), where cycle 273 refers
to the beginning of cycle 273, i.e., 272 cycles have been completed or 272 19
= 5168 years have passed since the creation according to the Jewish tradition.
For an explanation of the way the entries in this column were computed, see
the comments on Tabula gradus solaris (f. 56v), below. The constant difference
of 359;44,50 means that in one 19-year cycle the Sun progresses 6839;44,50.
Now, the first entry in the table for the time on f. 55r, below, associates 2d
16;33,3,20h with a 19-year cycle. (This is also identical to the value underlying the
column for time in Ibn al-adibs table in Vatican, ms Heb. 379, f. 6a.) This time
corresponds to the excess over an integer number of weeks of the time between
two consecutive 19-year cycles, which is taken here as 991 weeks + 2d 16;33,3,20h.
Thus, the time between two consecutive 19-year cycles is 6939;41,22,38,20d.
isaac ibn al-adib and flavius mithridates 359

Dividing the amount progressed by the Sun by this time we obtain a value for
the solar mean motion of 0;59,8,9,16/d and it is a sidereal value. The closest
value of which we are aware is that of Ibn al-Kammd: 0;59,8,9,21,15/d.
In the sub-table for single years the entry for year 1 is 11s 20;41,15, whereas in
the tables of Ibn al-adib the entry for year 1 is 0;0. In the sub-table for months
the entry for month 1 is 9s 7;14,49,42, whereas for Ibn al-adib it is 3s 15;20,10
which, according to Ibn al-adibs canons, is the radix for the conjunction of
creation. We have no explanation for these differences but is seems likely that
Mithridates has changed the epoch.

f. 54v. Tabula motus diuersi

This table consists of 3 sub-tables giving the lunar anomaly (in signs, degrees,
minutes, and seconds) for 13 consecutive months, 19 consecutive years, and
groups of 19-year cycles (for cycles 1, 2, 3, , 10, 20, 30, , 100, , 700), respec-
tively, as in the previous table. This table also displays a numerical value, 6s
12;54,30, together with the words DUCI URBINI, referring to Federico, Duke of
Urbino. The same value for the lunar anomaly is found in Ibn al-adibs tables
in Vatican, ms Heb. 379, f. 6b, and it represents the lunar motion in anomaly in
half a mean synodic month.
The entries for two successive 19-year cycles show a constant difference of
306;54,52, as in the column for the lunar anomaly in Ibn al-adibs table in
Vatican, ms Heb. 379, f. 6a, where this information is presented differently. The
entry for cycle 1 is 10s 6;54,52. In Ibn al-adibs tables the entry for 273 refers
to the beginning of the cycle (as we learn from Gascons commentary). Since
the radices for this epoch are given by Ibn al-adib as entries for the month
Tishri, the table for cycles should not have radices embedded in them. Ibn
al-adibs entry for cycle 273 is 320;42,12, and it corresponds to the motion
in anomaly in 272 completed cycles of 19 years. To derive this value from the
constant difference of 306;54,52, we first compute:

272 306;54,52 = 83480;43,44 = 320;43,44.

Now let us take 83480;42,12 and divide it by 272; the result is 306;54,51,40,
which rounds to 306;54,52. This indicates that Ibn al-adibs table was con-
structed from a basic parameter of 306;54,51,40 rather than 306;54,52.
The constant difference of 306;54,52 means that in one cycle the Sun pro-
gresses (251 360) + 306;54,52 = 90666;54,52. Dividing the Moons progress by
the time between two consecutive 19-year cycles, 6939;41,22,38,20d, we obtain a
360 chapter 10

value for the mean motion in lunar anomaly of 13;3,53,55,56,18/d. This is a clas-
sical parameter, to be compared, for example, with those in Ptolemys Almagest
(13;3,53,56,17,52/d) and the Toledan Tables (13;3,53,56,17,57 /d).
In the sub-table for single years the entry for year 1 is 11s 15;38,0, whereas in
the tables of Ibn al-adib the entry for year 1 is 0;0. In the sub-table for months
the entry for month 1 is 2s 12;53,26,31, whereas for Ibn al-adib it is 8s 21;51,25
which, according to Ibn al-adibs canons, is the radix for the conjunction of
creation. We have no explanation for these differences but it seems likely that
Mithridates has changed the epoch.

f. 55r. Tabula temporum

This table consists of 3 sub-tables giving the excess over an integer number
of weeks (in days, hours, minutes, seconds, and thirds) for 13 consecutive
months, 19 consecutive years, and groups of 19-year cycles (for cycles 1, 2, 3,
, 10, 20, 30, , 100, , 700), respectively, as in the two previous tables. This
table also displays a numerical value, 7d 18;22,1,50h, together with the word
GUILLELMUS, referring to William Raymond of Moncada. Almost the same
value is found for the time in Ibn al-adibs tables in Vatican, ms Heb. 379, f. 6b,
where we have 0d 18;22,2h. It represents the duration of half a mean synodic
month in days (mod 7) and hours. We note that Mithridatess value has a higher
precision than that in Vatican, ms Heb. 379. We also note that 18;22,1,50h =
0;45,55,4,35d and twice this amount is 1;31,50,9,10d, i.e., the excess of a mean
synodic month over 28 days (= 4 weeks).
The entries for two successive 19-year cycles show a constant difference
of 2d 16;33,3,20h (= 2;41,22,38,20d), which is also the value for the first entry.
This is the same value that can be derived from the column for time in
Ibn al-adibs table in Vatican, ms Heb. 379, f. 6a, where this information
is presented differently and with a lower precision (the entries are rounded
to seconds of time). The entry for cycle 273 is 3d 13;51,7h (= 3;34,37,47,30d)
which implies that the entry for cycle 0 would be 4d 7;26,57h = 4;18,37,22,30d
(for 4;18,37,22,30d + 273 2;41,22,38,20d = 4;18,37,22,30d + 6;16,0,25d (mod 7) =
3;34,37,47,30d (mod 7) = 3d 13;51,7h). We have no explanation for Ibn al-adibs
initial value. The value for the constant different in a 19-year cycle is equivalent
to 2d 16h and 595 elaqim and this parameter is exactly that given, for example,
in Maimonidess Sanctification of the New Moon.41

41 Neugebauer, Maimonides (ref. 40), 115.


isaac ibn al-adib and flavius mithridates 361

In the sub-table for single years the entry for year 1 is 1d 21;30,10,0h, whereas
in the tables of Ibn al-adib the entry for year 1 is 0d 0;0h. In the sub-table
for months the entry for month 1 is 2d 1;43,6,51h, whereas for Ibn al-adib it
is 2d 6;8,25h which, according to Ibn al-adibs canons, is the radix for the
conjunction of creation. We have no explanation for these differences but it
seems likely that Mithridates has changed the epoch.

f. 55v. In quoto die mensium coniungium vel dissidium futurum sit

This table displays the day of the (Christian) month in a 19-year cycle when
a conjunction takes place (entries range from 1 to 31). The letter p is written
opposite years 2, 5, 8, 10, 13, 16, and 18 in the 19-year cycle, indicating a leap
year, i.e., a year of 13 lunar months.

f. 56r

There are 4 small tables concerning weekday numbers. In one of them the
words Federico, Duke, Count, William, Raymond, Moncata, and Urbino are
assigned to the first day of the month, and in another, numbers between 1 to 7
are assigned to these 7 words. In the two other tables, a number between 0 and
6 is ascribed to each month in a year beginning in January, whether a common
year or a leap year.

f. 56v Tabula gradus solaris

This table consists of 3 sub-tables giving the mean longitude of the two lumi-
naries at mean conjunction (in signs, degrees, minutes, and seconds) for 13 con-
secutive months, 19 consecutive years, and groups of 19-year cycles (for cycles
1, 2, 3, , 10, 20, 30, , 100, , 700), respectively, as in previous tables. This table
also displays three numerical values, 6s 0;0,0, 0s 12;30,0, and 0s 14;33,9,59.
The first indicates the amount to the added to the lunar longitude at opposi-
tion; the second is a parameter already mentioned in the text (f. 35r, lines 710),
to be used to change from sidereal to tropical coordinates; and the third repre-
sents the motion of the Sun in longitude in half a mean synodic month. These
values are presented together with the word RAYMUNDUS, referring to William
Raymond of Moncada. Almost the same value for the solar motion for the posi-
tion occurs in Ibn al-adibs tables in Vatican, ms Heb. 379, f. 6b where, for the
362 chapter 10

third value, we find 0s 14;33,10. We note that, as was the case in the analogous
table on f. 55r, Mithridatess value has a higher precision than that in Vatican,
ms Heb. 379, although elsewhere Mithridates gives it as 0s 14;33,10 (f. 54r: see
above).
The entries for two successive 19-year cycles show a constant difference
of 359;48,50, as in the column for the position in Ibn al-adibs table in
Vatican, ms Heb. 379, f. 6a, where this information is presented differently.
This means that in one cycle the Moon progresses (253 360) + 359;48,50 =
91439;48,50. Dividing the Moons progress by the time between two consec-
utive 19-year cycles, 6939;41,22,38,20d, we obtain a value for the lunar mean
motion in longitude of 13;10,34,52,40,30/d. This parameter may be compared
with the value in Ptolemys Almagest (13;10,34,58,33,30,30/d) and those in
the zij of al-Khwrizm, the zij of Ibn al-Kammd, and the Toledan Tables
(13;10,34,52,46/d).
In the sub-table for single years the entry for year 1 is 11s 20;43,15, whereas in
the tables of Ibn al-adib the entry for year 1 is 0;0. In the sub-table for months
the entry for month 1 is 5s 7;3,48,40, whereas for Ibn al-adib it is 5s 16;41,31
which, according to Ibn al-adibs canons, is the radix for the conjunction of
creation. We have no explanation for these differences but it seems likely that
Mithridates has changed the epoch.
We can also deduce from the text the solar mean motion in longitude: (18
360 + 359;48,50)/6939;41,22,38,20d = 0;59,8,11,20/d. The difference between
the solar mean motion in longitude and the motion of the solar anomaly is the
proper motion of the solar apogee: 0;59,8,11,20/d 0;59,8,9,16/d = 0;0,0,2,4/d
(or 1 in about 286 Julian years). Standard values for the proper motion of
the solar apogee are close to this amount, e.g., for Ibn al-Kammd it is 1 in
about 290 Julian years (Chabs and Goldstein 1994, p. 28). If we subtract the
mean solar anomaly at epoch (the conjunction of creation) according to Ibn
al-adib from his value for the mean solar position at epoch, the result is the
solar apogee at epoch: 166;41,31 105;20,10 = 61;21,21.
To verify that a sidereal solar position derived from Ibn al-adibs tables
is approximately correct, we compare the value based on his tables for the
solar eclipse that took place on Aug. 11, 1542 with the sidereal solar position
according to the Toledan Tables: in both cases the result is about 135 (see
below). Hence, the sum of the entries for the solar position in Ibn al-adibs
tables is acceptable (by the standards of his day). But there is a problem:
we expect the radix to be shown in the table for months opposite Tishri,
as explained in Ibn al-adibs canons, corresponding to the conjunction of
creation. But there is another radix embedded in the table for the mean
motions. First, let us compute the mean solar position for Aug. 11, 1542, the mean
isaac ibn al-adib and flavius mithridates 363

conjunction of Elul [month 12], 5302am, which took place on Ab [month 11]
29, in year 1 of cycle 280 (note that the entry for year 1 in Ibn al-adibs table is
0).

cycle 280: 0s 7;43,58


Elul: 4s 6;51,11

sum: 4s 14;35, 9

The entry for Elul can be computed by adding 11 times the increase in longitude
in a mean synodic month (= 29;6,20: Vatican, ms 379, f. 6b) to the radix (the
entry for Tishri):

11 29;6,20 + 5s 16;41,31 = 320;9,40 + 166;41,31 = 126;51,11.

On analogy with the computation of the mean motion in lunar anomaly (see
Tabula motus diuersi, f. 54v), one expects the entry for the beginning of cycle
280 to be 279 times the line-by-line differences in the column for solar position
in the table for cycles, i.e.,

279 359;48,50 = 308;4,30.

But the entry for cycle 280 is 7;43,58, i.e., it exceeds the expected value by
59;39,28. Now the sum we computed for Elul, cycle 280, can be considered to
have 4 components:

308;4,30 59;39,28 + 320;9,40 + 166;41,31 = 126;51,11.

The components 308;4,30 and 320;9,40 are just the result of multiplying
the mean motion parameter, and so they are fixed. But it is unclear why Ibn
al-adib did not consider as radix 107;2,3 (= 166;41,31 59;39,28), combining
the two other components. The only reason that comes to mind is that the
creation, according to the Jewish tradition, took place at the beginning of the
month Tishri which is associated with the autumnal equinox. Moreover, the
purpose of the 19-year cycle is to keep each month in the same season. A radix
of 107;2,3 entered in the month Tishri would mean that the creation took place
closer to summer solstice than to the autumnal equinox.
The same procedure also works for the mean motion in solar anomaly, that
is, we find that the shift in the entries for cycles is the same amount, 59;39,28.
This is appropriate since the difference between the solar position and the solar
364 chapter 10

anomaly is the longitude of the solar apogee; hence, a shift in one requires a
shift in the other.

f. 57r. Tabula motus draconis

This table consists of 3 sub-tables giving the argument of latitude (in signs,
degrees, minutes, and seconds) for 13 consecutive months, 19 consecutive years,
and groups of 19-year cycles (for cycles 1, 2, 3, , 10, 20, 30, , 100, , 700), respec-
tively, as in previous tables. It also displays a numerical value, 0s 15;20,7,1,
together with the words DE MONCATHA, referring to William Raymond of
Moncada. Almost the same value for the argument of lunar latitude occurs in
Ibn al-adibs tables in Vatican, ms Heb. 379, f. 6b, where we find 0s 15;20,7. It
represents the motion of the argument of lunar latitude in half a mean synodic
month. As was the case in previous tables, Mithridatess table exhibits a higher
precision than Vatican, ms Heb. 379.
The entries for two successive 19-year cycles show a constant difference of
7;35,20, as in the column for the argument of latitude in Ibn al-adibs table
in Vatican, ms Heb. 379, f. 6a, where this information is presented differently.
This means that in one cycle the Moon progresses (255 360) + 7;35,20 =
91807;35,20. When we divide this amount by the time between two consec-
utive 19-year cycles, 6939;41,22,38,20d, we obtain a value for the mean motion
of the argument of latitude of 13;13,45,39,47,10/d Subtracting the mean motion
in argument of latitude from the lunar mean motion, we find 0;3,10,47,6,40/d.
Again, this value for the mean motion of the node is close to those of al-
Khwrizm and Ibn al-Kammd, whereas the value for the mean motion of the
argument of latitude is close to Ptolemys 13;13,45,39,48,56,37/d.
In the sub-table for single years the entry for year 1 is 5s 14;26,9, whereas in
the tables of Ibn al-adib the entry for year 1 is 0;0. In the sub-table for months
the entry for month 1 is 4s 29;29,28,46, whereas for Ibn al-adib it is 6s 3;40,27
which, according to Ibn al-adibs canons, is the radix for the conjunction of
creation. We have no explanation for these differences but it seems likely that
Mithridates has changed the epoch.

ff. 57v60r. Quarundam urbium longitudines & latitudines

This is a list of the geographical coordinates (longitude and latitude) of 435


localities, of which many are in Sicily, ordered from west to east. The coordi-
nates are mostly given in degrees and minutes, e. g., for the city called Agrigen-
isaac ibn al-adib and flavius mithridates 365

tum we are given 38;30 (longitude) and 36;24 (latitude). However, those for
Toledo are 10 (longitude) and 42 (latitude). The prime meridian for geograph-
ical longitude is the one used by Ptolemy (and many other medieval tables such
as the Toledan Tables), passing through the Canary Islands, which he referred
to as the Fortunate Islands. It differs from the so-called meridian of water,
usually considered to be 17;30 to the west of Ptolemys prime meridian, and
used in some Andalusian zijes.42

f. 60v

This table is for transforming time-degrees (for each integer degree from 1 to
180) into hours (from 0;4h to 12;0h).

The following two tables have the same format as those on ff. 44r45v.

f. 61r. Prime adequationes motus diversi in .xij. signis

This table displays the correction for the lunar anomaly (in degrees, minutes,
and seconds) as a function of the solar anomaly, given for each degree. The
letters s and a in the margins have the same meaning as in Tables a, b, and
c. The maximum is 2;43,35 at anomaly 3s 2 (see Table g). The entries are the
same as those in the column labelled correction for the lunar anomaly in the
corresponding table in Vatican, ms Heb. 379, ff. 7a8a (see Table aa), where
the same maximum value is found. We have not succeeded in accounting for
the entries in this table despite the fact that its entries differ from those in
Tables a and c by factors of proportion, and despite our expectation that the
explanation for Table g should be analogous to that for Table h (see below).

f. 61v. Secunde adequationes motus diversi in totidem signis

This table displays the correction for the lunar anomaly (in degrees, minutes,
and seconds) as a function of the lunar anomaly, given for each degree. There
is no indication in the table or the margins concerning the addition or sub-

42 M. Comes, The Meridian of Water in the Tables of Geographical Coordinates of al-


Andalus and North Africa, Journal for the History of Arabic Science, x (19921994), 4151.
366 chapter 10

table g First correction for the lunar anomaly

s 0[s] 1[s] 2[s] 3[s] 4[s] 5[s]

1 0; 2,47 1;21,47 2;22,28 2;43,31 2;24,30 1;22,10


2 0; 5,32 1;24, 4 2;24,45 2;43,35 2;23,30 1;20, 1
3 0; 8,18 1;26,42 2;25,36 2;43,34 2;21,53 1;17,40
4 0;11,12 1;29, 8 2;26,30 2;43,28 2;20,29 1;13,10
5 0;13,50 1;31,33 2;27,42 2;43,19 2;18,40 1;12,30

10 0;27,15 1;4