Académique Documents
Professionnel Documents
Culture Documents
at Extreme
Energy Density
ETTORE MAJORANA INTERNATIONAL SCIENCE SERIES
Series Editor:
Antonino Zichichi
European Physical Society
Geneva, Switzerland
(PHYSICAL SCIENCES)
Nicola Cabibbo
University of Rome, Italy
and
Luigi Sertorio
University of Turin, Italy
The title of the book perhaps does not fully represent its
content but still is a good indication of the conceptual motiva-
tion of our Workshop.
v
PREFACE
The cosmology of the early moments after the big bang is the
milieu in which the conditions relevant to general relativity,
high-energy elementary particle physics, and high-energy
statistical mechanics naturally occur. In practice, the apprecia-
tion of this concept lay at the base of the Workshop and of its
triangular architecture: supersymmetry and supergravity -- chromo-
dynamics thermodynamics.
Nicola Cabibbo
Luigi Sertorio
CONTENTS
THEORETICAL LECTURES
vii
viii CONTENTS
Contributors 355
Index 357
ERICSON FLUCTUATIONS AND THE NEW ARGONNE DATA ON TIN SCATTERING
Steven C. Frautschi
miT
dN _ p(m) ~ e 0 (1)
dm
miT
For concreteness I shall use p(m) ~ e 0 for all quantitative
2 STEVEN C. FRAUTSCHI
R/n p (m)
miT
o
Ptheory = ce
PexperiIi1.ent
1 2 m
overlapping
resonances
m
(2 GeV) 2
We claim that the higher mass resonances are not individually de-
tectable because they overlap. Since they do not produce observable
peaks, even in individual partial waves, their existence must be
inferred more indirectly. This can be done by finding reactions
whose cross section is controlled by statistical factors associated
with the large number of overlapping resonances. Precedents for
this approach exist in nuclear physics where a similar problem of
overlapping resonances occurs.
YabiYcdi
(2)
-E +E
i
- i .!:.2
n ~ p(E)r (3)
where peE) and r now refer specifically to the density and width of
resonances with the quantum numbers B,JP, •.. appropriate to our
reaction (the main variation to keep in mind in the hadronic case is
EIT
peE) ~ eO). The typical partial width is
lyabi l 2 1
r
RJ - (4)
n
since in the elastic channel they must sum to about 1. In an
inelastic channel Ya b'Y d' may have either sign so the sum is smaller.
1 c 1
In general AJ may contain two terms:
o (7)
ERICSON FLUCTUATIONS AND ARGONNE DATA ON 1TN SCATTERING 5
and
1 -E/T
'---r.Je (8)
n
This ordinary Boltzmann factor is the Bohr expression which one sees
in poor resolution (here the averaging is provided by experimental
smearing over energy). But in good resolution one should see strong
fluctuations about the Boltzmann factor.
Even in a coherent reaction, the full rate IAc+ oAI 2 should show
fluctuations of relative height 2IoAI/IACI. Thus these "Ericson
fluctuations" are expected to occur in a broader variety of reactions
than do other statistical phenomena.
1
(10)
L
max
The energy width is evident from the argument that when the energy
is increased by r, one gets a new set of overlapping resonances with
a new set of phase relationships. The angular width follows from the
usual arguments about how fast a Legendre polynomial can oscillate.
The fluctuations in E are expected to provide the most distinctive
indication for resonances, since coherent phenomena such as diffrac-
tive peaks also exhibit rapid variations in 8.
p + Fe 56 + p + Fe 56 (11)
has been studied 8 over the range E = 9.3 to 9.6 HeV at intervals of
2 to 5 KeV at each of several angles between 63 0 and 171 0 (Fig. 3).
At fixed energy, the dependence on angle is characteristic of a
diffraction peak with minima. It is at fixed angle that the cross
section exhibits rapid Ericson fluctuations as the energy is varied.
The fluctuations are visible at all angles, but oA/A is biggest at
the larger angles where A (i.e., the coherent diffraction term) is
small. From the observed fluctuations one learns that:
iii) the resonances have roughly the level density expected from
nuclear theory (the level density is obtained from the data
by a correlation function method. In essence the method esti-
mates IAcI by smoothing over the data, and loAI by using our
result that the relative height of fluctuations is approximately
2IoAI/IAcI when Ac is present. Comparison with the expression
loAI ~ l/Ill then yields the number of overlapping resonances
n = rp from which p is readily deduced).
ERICSON FLUCTUATIONS AND ARGONNE DATA ON 1TN SCATTERING 7
N N (12)
res peaks
or as
ii) overlapping resonances, in which case
N » N (13)
res peaks
To distinguish these possibilities, one needs further information.
For example, suppose that in addition to the elastic reaction it were
possible to study
p + Fe 56 ~ p + Fe 56
~p + Fe 56 *
~D + X55
-- - - (14)
2iO
S e (15)
8 STEVEN C. FRAUTSCHI
r )
o (E) = tan-1 ( 2 nl-E) (16)
r p < 1 (19)
so that all arguments of tan- 1 are small except the nearest (Mi- E)-l.
nchannel
r p s 21f
(20)
*Some of the papers listed in Ref. 10 analyze only the single or two-
channel case, where condition (19) is not satisfied by overlapping
resonances. As a result these papers reach unduly negative conclu-
sions concerning Ericson fluctuations, which are not justified in
the general multichannel case.
ERICSON FLUCTUATIONS AND ARGONNE DATA ON 1TN SCATTERING 9
I focused attention on
(21)
where precise data with the necessary closely-spaced energy intervals
above the separable resonance region is easiest to obtain. Data of
this type already existed for 0° (Otot) and 180° scattering up to Plab
of somewhat more than 5 GeV/c. It was shown that the familiar peaks
and dips in this data, though usually interpreted as single resonan-
ces, can also be interpreted as overlapping resonances (Ericson fluc-
tuations)14,15. The peaks are, of course, much more prominent at 180°
where AC is smaller. However, the largest fluctuations (largest oA/AC)
were expected in the region around 90°, where AC is smallest (Fig. 4).
+1 cosS
(21)
pp ~ pp (22)
at intervals of order
I1E
c.m. '" 35 UeV
2
I1t 0.1 GeV
def
dt Ifixed cosS
s-n (23)
cose = +1
Previous
Experiment ---+ s
cos e = -1
NU
102 "!.
. ~
'.
do-
: 0
'\
'\\ .......
d o-x 0.01
dl ..
102
- '"t>
;;;- £P"xO.OI \
dl dt.:'0.OI ... t
.,
.0 >10 "-""
10.0 I;;;->
-I
::i..QJ
::i..'" l!)
"'-.",
l!) ".
..\·...6 CM
. . .-.. ••.•.••'ot~·
~ •• 8 =90. \ ....
.'....,
o .~
-i;1.; -I •••• • + 8c,,=80 '. -, -i;1.;
10 10
··.l8 CM=80
-2 -2
10 '. 10
t".."'.
• R e f.3 a Ref.3
103 • R e fA • R e f.5
163
• Ref, 5 • Ref. 7
• R ef.6 6 R ef.8
104 IC/
4 5 6 7 8 910 12 4 5 6 7 8 910 12 14 16 18 4 5 6 7 8 9 10 12 14 16 18 ~
s (GeV2) ( GeV2) s (GeV 2) m
<
m
Z
Fig. 6. Differential cross sections for elastic ~-p, ~+p, and pp
r>
scattering measured in the Argonne experiment 17 for some ."
::tt
representative angles near 90°. The lines are the s de- »
pendence expected on the basis of dimensional counting C
-I
rUles18 . Data from earlier experiments (Refs. 3-8 of the Ul
(")
Argonne paper) on ~-p and ~+p above 3 GeV/c are also plotted. :z::
ERICSON FLUCTUATIONS AND ARGONNE DATA ON 1TN SCATTERING 13
iv)t The total number of clearcut candidates for structures (n-p plus
n+p) is about 20. Only one or two of the n+p structures are dupli-
cated in n-p. The typical width of the structures is r ~ 100 to
200 HeV.
v)t Are the structures known resonances? Few of the structures are
likely to be known resonances from the Particle Data Tables. They
appear throughout the experimental range 2.3 GeV < Ec . m• < 4.3 GeV,
whereas previously established nN resonances are limited to Ec • m.
S 3.2 GeV. The structures with Ec .m. < 3.2 GeV are generally
narrower than known resonances and are not centered on them.
-103.1
)(10 2
10
2
"~.~<
",5
"'u 10
'""
.>
(!)
" I
\
.0
:t.
5.0
)( 2
'0
"b
'0
-I
10
-2
10
5 10 7 10 13
Fig. 7 Differential cross sections for elastic n-p and n+p measured
at constant t in the Argonne experiment 17 • The top number by
each set of data gives -t in GeV 2 /c 2 • The bottom number, if·
present, is a scale factor by which do/dt has been multiplied
for presentation. The lines are purely to guide the eye; they
are not fits.
dol
dt
cose
s
(a)
dol
dt
cose
s
(b)
2
Fig. 8 Schematic comparison of IAcI (the smoothed-out experimental
do/dt) and the purely statistical contribution to do/dt for
nN elastic scattering at a fixed angle near 90 0 :
a) as indicated by fluctuations in Argonne data 17 ;
b) as predicted on basis of statistical bootstrap model 14 •
v. Future Possibilities
iii) Study other reactions in the same spirit. For example, K-p and
~p would be interesting to compare because the latter is an exotic
channel.
Another reaction with unique features is
(24)
References
Steven C. Frautschi
Of course it has always been clear that free quarks and gluons,
at energies so high that their masses could be ignored, would give
(E) ~ T4
just like a photon gas. Thus the temperature would rise as (E)1/4
and there would be no limiting temperature.
19
20 STEVEN C. FRAUTSCHI
N
E .-_Ll (Ekin) i +
1
L Vij
ij
(1)
V = e 2 /r (2)
ei
to
V (3)
ei r
where
mc -1
.-if ~ (rscreen) (4)
r
screen
'" R·
V§gC
g22
(6)
Comparing
to
2
(V e",n).ln '" g /R (8)
2
. ).1 '" g /hc
(V e",n).ln /(E k ln (9)
2
ii) Asymptotic Freedom. As R becomes small we need to know g only
in the small r reglon where it is reliably calculated:
:l... 1 (10)
-tic '" RIn. (r /R)
1f
Thus
o (11)
(Ve~)i ~ g l
2 Rkc/g 2 3
d r p/r 4'1rg
2 Rltrc/g 2
J
o
dr rp
(V.).
e", ~
,... 31rc/2R (13)
so
(Ve",.)./(E
~
kin ).~ ,... 0(1) (14)
However this is assuming all neighbors interact with the same sign
(same charge or color state). This requires a highly ordered state.
Such a degree of ordering cannot be achieved at high T, where in
fact there are cancellations among Vij' which increase as the
effective number of neighbors rises. Thus Vet/Ekin indeed fails as
the density rises at high T. We shall come back to the possibility
of order at low T in the second part of our talk.
222
(me) gluon ,... ~ (p ) (15)
as above for electric terms but not magnetic terms. (A familiar
example: in electrodynamics, a magnetic mass implies the Meissner
effect. This is not obtained until one considers non-perturbative
effects. On the other hand, charge shielding is obtained in per-
turbation theory.) Thus we must consider the unshielded magnetic
potential separately. Fortunately the leading term at long range,
the dipole term, has the form
2
-+ -+ L
(Vmag) ij ,... g C\ . Bj"'" 3 (16)
r
which falls much faster with r than Ve~ did. Here, even i f all
pairs interacted with the same sign, we would obtain only a loga-
rithmically divergent contribution from the sum over distant
magnetic sources. In practice at high T many particles have the
opposite spin (or charge, or color) ana this logarithmic divergence
is cut off leaving a small though non-negligible term.
DENSE QUARK MATTER 23
Discussion
ii) Nothing we have said argues against the possibility that the
hadron mass spectrum has the form
Thus they do not have higher energy density than a low-mass hadron.
lihat happens in dense matter, then, is presumably that as the average
density increases past the density characteristic of an MIT hadron,
the MIT hadrons cease to exist as separate states and a phase
transition from confined quark bags to free quarks occurs, as dis-
cussed by Cabibbo and Parisi 4 .
(19)
T
/
c§'~
~
nJtee q';.,
''''I~
.Q;?
qq pcUJrA
,
\
NeUtron
Star\!
(a)
e (b) (c)
References
IN GAUGE THEORIES
F. Gliozzi
Sezione di Torino
I. INTRODUCTION
29
30 F. GLiOZZI
(1. 2) B
Thus if one knows the partition function in the low temperature
region ( f\ large) one can also describe the thermodynamic
properties at high temperature (B small).
In field theory the first example of duality transformation
was found even earlier by Dirac 9 ,10,11 in his theory of magnetic
DUALITY TRANSFORMATION AND CONFINEMENT IN GAUGE THEORIES 31
monopole (1931). Here the small and the large coupling constants
are the electric and the magnetic charges e and g, which are
related by the celebrated Dirac quantization condition
(1.4 )
which again expresses, like eq.s (1.2) and (1.3), the duality
between large and small couplings.
If one could find a duality transformation in the models
where there is a confining phase for the topological excitations,
the dually transformed phase should describe the confinement
produced on local excitations (say, electric charges or quarks)
by a highly ordered structure of topological excitations, like
a Bose condensate.
In gauge field theories there are many types of topological
excitations, like the polyakov's2 instantons, ~he de Alfaro,
Fubini and Furlan's14 merons, the Handelstam's1) gas of monopoles
or the 't Hooft 16 fluxons. These authors argue that the
quark confinement is due to a sort of condensation of these
topological excitations even if it is not yet clear which one
of the topological excitations discovered up to now, if any,
actually contributes to confinement mechanism. Horeover an
explicit duality transformation is still laking.
The solution of these problems seems to be a formidable
task.
We limit ourselves in this lecture to describe some general
property of the duality transformations for a wide class of
32 F. GLiOZZI
dynamical systems.
17-20
We shall see that the general features of the duality
transformation, and in particular the complementarity between
condensation and confinement, do not depend very much on the
dynamical details of the systems but rather on the space time
structure of the topological and local excitations (part II
and III) \vhile the duality relation between coupling constants
depends on the underlying symmetry of the model. In particular
the Jirac quantization condition of electric and magnetic
charges is the general relationship between dual coupling con-
stants associated with U(1) symmetric models 20 • Similar relations
will be found in part IV for abelian gauge theories on a lattice.
(11.2)
(11.4) --
(II.5) cL t Fp = 0
The O-form are simply the functions. For instance in d=3
case we can verify the following translation table
c:L fo grad I
r
~
cih ~ curl
d,
,f ........
t div r
r,t : ......-.
0
{ curl grad
div curl
= 0
(11.6)
Hp = ~Fpl/{for1 HP = f C PJ II BP}
In the Eucledean space In the Eucledean space
H is trivial: H = 11 HP is trivial: HP = U
p p
(I1.7)
(n.s)
for p=2 and d=4 they are just the ordinary Maxwell equations,
but one has also other interesting examples. For instance d=3
p=1 describe,s the pseudo-particle gas of polyakov 2 , d=4 p=3
is the Kalb-Ramond theory of interacting strings 20 , etc.
If fp is defined in the whole (pseudo)-euclidean space,
Hp is trivial, i.e.
(I1.9)
(I1.10)
(n.11) f fp
cP
14p (~ijP) is no longer trivial: there are topological
excitations in the theory (e.g. monopoles). Indeed we derive
from Gauss theorem that cP cannot be a boundary
DUALITY TRANSFORMATION AND CONFINEMENT IN GAUGE THEORIES 35
(rr.13)
(11.15) cl - 2
In the following table we give some examples of topological
excitations as well as the corresponding local or elementary
excitations
Theory Topological excitation Local excitation
f
Dirac- 't Hooft magnetic
q=1 electric charge 0=1
d.~4 Polyakov monopole monopole
Kalb-Ramond
string instanton q=O string 0=2
(II.16)
(II.17)
where D
1iol is an arbitrary manifold whose boundary is the
classical support Tq of the topological source, i.e.
(II.18)
(II.21)
(II.22)
(II.23)
p-1
where V f is a manifold bounded by a small cycle Clinked
to Dq.
Eq. (11.23) is the generalization of the Dirac quantization
condition for the theories defined by Eq. (11.7,8).
Eq. (II.23) could be considered as a consistency condition
on the coupling constant e d~e to the presence of a topological
excitation. Another important effect is produced when there are
many topological excitations in the system. He shall see that
if their number is sufficiently large they produce the
confinement of the elementary excitations (electric charges).
The best parameter to study the confinement of the "electric"
charges is the Hilson factor given in eq. (II.20). The signal
of confinement is that the expectation value of the Wilson
exponent is proportional to the volume 'tYP of a manifold which
has V,-I as boundary. This is the so called area law for the
Wilson integral.
38 F. GLIOZZI
(III.3)
where P
w( '. I •• 'AI) is the probability of the system of N
charges in the configuration specified by the coordinates
C1 , C2 ···CN•
The flux cp varies from 0 to 2-D' dep'ending on the mutual
pos1t1on 0 t e ~·tL charge and the V 11"" man1fold.
. . f h • It i s
approximately zero when their distance is large with respect to
the size of VptJ • According to Dirac's quantization condition
cf>. would reach its maximum value 211' if V'" were closed
c'l'a
V'·'; 0 ) and l i nked to C ~ • Since ? V"'* 0 this
DUALITY TRANSFORMATION AND CONFINEMENT IN GAUGE THEORIES 39
(IV.1)
where ~ is the coupling constant. Note that the sum over the
orientations makes S real.
The object of interest is the partition function. It reads
S
't I
(IV.2) ~ e
{'t 1
i
where
the variables 'e.
indicates the sum over all the possible values o£
associated with the links of L.
One can also restrict the set {~t J to any discrete
subgroup o£ U(1).
A particularly interesting choic~ is ZeN), i.e. the discrete
, '"11"
group formed by the N elements e w (n=1,2, ••• N), because
it can be considered as the center o£ a fully broken SU(N)
gauge theory (say by Higgs mechanism making all gluons massive)
which seems to be 16 one o£ the most important ingredients to
understand quark confinement.
The generalization o£ eq. (IV.1) to a gauge theory of type
p is straightforward: S is simply the sum over all the (p+1)-cells
of the products
. r;r ,., ,,/
(0". o£ the elements 'cl associated
,.,
. c.0.,(
Wl. th the p-cells borderl.ng G • Gauss theorem shows that the
operator T;1 ,., has the role o£ the exterior derivative d
actl.ng on where
~ I 0.,
(IV.3) -= e c.'
It is an useful and amusing exercise to verify the
following translation table from the exterior forms language
DUALITY TRANSFORMATION AND CONFINEMENT IN GAUGE THEORIES 43
Translation table
a)
<
b)
~ ,. , ~ ,. =1
(', f: 'J
n'It' () ,pIt)) n
G
c)
A~a,*~~
J( ~, cf ,. ,/+1)
in b) the C/ are all the p-cells bordering c/·'
while
in c) the product r-T is the one over all the (p+1)-cells
. . , J
hav~ng a s~de, ~n common.
(IV.4)
s ":: f'l.!. ~ ~f
or~ented
plaquesses
(IV.G) Z ((I,) = Z 2. T1 e -1
{'t1 .( "lIfel f pI
44 F. GLiOZZI
•
The idea is now to integrate first over the ~e ~. To this end
we have to transform the Boltzmann term into the product of
factors, each of which refers to only one plaquette.
It is then easy to verify that the factor with the Lagrange
multiplier can be rewritten in the form
(IV.7)
--
where the product II Q.f refers to the four links *e
of It L
bordering the plaquette .tr l dual to P.
multiplier ~*'
becomes the character
.
ide have not yet specified the range of values of the Lagrange
'ile may normalize
Ul#l
Of~!
h.*e
such that eq.(IV.7)
i.e.
(IV.S)
~J..~~~l) : ~tJ..¥l
This choice has many advantages. First of all inserting
(IV. 8) in (IV. C) vie see that the duality transformation is
directly related to the harmonic or Fourier analysis of the
Poltzmann factor, ,.,hich is just the historical way to do the
duality transformation in lattice theories. Moreover the set of
the indices k*"! of the characters of an abelian group G
forms itself a group ~ G called the dual of G. Thus under a
duality transformation a theory with a symmetry group G becomes
one ,.,i th another symmetry JIf G.
Ue have the follovling table,
which exhausts all the possible abelian groups
Table
G
U(~) compact
Z (w) ~ 2 (N)
*" * G- G
can now answer the follovling question: what are the necessary
\:Ie
and sufficient criteria satisfied by a p-gauge theory on a
DUALITY TRANSFORMATION AND CONFINEMENT IN GAUGE THEORIES 45
2 P = tL-2
ii) G=" G
iii) The functional form of the action is invariant under the
characters expansion.
Condition i) tells us that for d=4 only the ordinary p=1
gauge theory might be self-dual. From condition ii) we see that
the simplest self-dual theories are those associated to a non-
-compact U(1) or to Z(N), although there are other possible
choices made with a suitable combination of the groups of the
above table such that one has again G ~ ~ (;. .
For a non-compact U(1) it is well known that the only
functional form of the Boltzmann factor which is invariant
under the Fourier analysis is the gaussian, where the action
may be considered as the quadratic approximation of the action
-
(IV.4).
Ide have
S 'L
- P
(IV.9) e J dp e e
= -Ob
( IV. 10 ) f B = (G IT )_1
'L
(IV. 11)
f .-. eJ. "
Hence we find, as we expected, that the duality transformation
on the lattice version of a U(1) gauge theory yields again the
46 F. GLIOZZI
i: ..f 89~(~/
Boltzmann factor.
(IV. 15)
B-=f
It is then easy to verify the following table
Z (4) (\ _ ~(It\ll.)
c-
24, 25
15 16
The works of Handelstam and It Hooft have emphasized the
relevance of ZeN) gauge fields in understanding the quark
confinement in SU(N~ gauge theories. Indeed one can construct
classical solutions 6 of SU(N) gauge theory which represent
non-Abelian magnetic flux tubes (fluxons).
In the lattice version these solutions become just the
topological excitations of the ZeN) gauge theory25.
The phase where quarks are permanently confined is the one
in which the fluxons condense. Such a condensation is just the
phase transition described by the duality transformation. Thus
it is possible to have a crude extimate 2 5 of the critical
coupling constants in the SU(N) models caused by this
condensation.
REFERENCES
R. Hagedorn
1. Montvay
J. Rafelski
ABSTRACT
49
SO R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
1. INTRODUCTION
where the sum runs over all the sets with j elements (clusters)
wi th b. e {l, ... , b - l}, such that the conservation of baryon num-
~
(1. 2)
expresses most clearly the fact that the cluster described by p(m)
consists of an arbitrary number of clusters of smaller mass, each
of which is in turn made of an arbitrary number of smaller clusters,
. .
etc., as can b e seen ~terat~ng h
tea b ove .
equat~on
*) •
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 53
Section 3 The mass spectrum and its Laplace transform are used
to obtain a thermodynamic description of the system.
*) At this point the reader may have the uneasy feeling that we
are doubly (in fact more than doubly) counting states. This
problem has been discussed in detail in references 1) and 2), to
which we refer the reader. Here we must ask him to simply be-
lieve that our counting is correct, if clusters exist (see also
Subsection 3.2).
54 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
n = c = k (Boltzmann constant) = I
the only dimensional unit is I GeV
Metric : a • b =a ~
b
~
= aobo -+-+
- a • b.
~o6;(E) L00
X
(.21)3 :=: ~~r O\.!
.
)
(2.2)
Then with
(2.3)
(2.6)
The next important idea (see also Subsection 3.2) was to ad-
mit other particles than just pions, and in particular resonant
states of pions, just as if they were stable particles 6 ). Not
knowing which ones should be admitted and how many there are, we
might put them in a mass spectrum of admissible input particles
p. (m); the pion contributes to p. (m) a 0 function 0 (m-m1f ) ,
1n 1n
resonances contribute smeared-out 0 functions. For the moment,
p. (m) is a function which represents our (incomplete) knowledge
1n
of the true mass spectrum p(m).
DO tk I)t
etc.) are now possible and neither the total particle number nor
that of any of the various components is fixed.
58 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(2.9)
)( 03C tl pi dill! 3 l
f .""
-1
Now the left-hand side has dimension m and represents the irtter-
na1 density of states of a system of mass m at rest in its own
confining volume vo; this density begins with o(m - m) and has
TI
a continuum for m > 2m • It therefore might be considered as an
TI
averaged mass spectrum (the true one is not yet continuous at
m~ 2mTI ), which asymptotically becomes physically equivalent to
p(m). In the first paper on statistical bootstrap8), the postulate
was accordingly that the logarithms of the densities of states (i.e.,
60 R. HAGEDORN, I. MONTVAY AND J. RAFElSKI
(2.10)
- 1 (2.na)
and
(2.l1b)
(2.12)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 61
J
d"Pi ~ (pf) = d3Pi L{""~)J.lp:-.: )dfu~ dpo, i
= d"Pi S
1:'(-1) /IIIi d ...
V~2+ _,2.
(2.13)
(2.14)
I
)
62 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(2.l5a)
(2.l5b)
For a dilute gas we know that the first expression is correct, the
second wrong. Furthermore, (2.l5a) can be derived in a straight-
forward way from the elementary rule that the available number of
states in {p, d 3 p} of one particle enclosed in a volume V is
given by
(2.17)
while (2.l5b) cannot be justified that way. For more detail see
Ref. 17). Since in the absence of interaction p(m) + oem - m ),
1T
the density of states a(p2, P'V) should reduce to that of a free
pion gas. Therefore, we conclude that SBM should be formulated in
64 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
IPS (Touschek measure) as was done in the first papers and not in
IMS as has become customary since.
= (2.18)
(2.19)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 6S
(2.20)
>
""
O"(V-r. Vi )
;:1'1 'A lp~)
1=1
(2.21)
with arbitrary given A(p2); we will not pursue this possibility.
(2.22)
(2.24)
(2.25)
V(~)= A Mi-
(2.26)
S(~) = 2Arm,'/(1:CC)3
where A is some fundamental parameter of our mode1*) • We can esti-
mate its order of magnitude in various ways.
(2.27)
Af"-
'-
(2.28)
- A . mt. (2.29)
q.. (120 MeV)" ~
~. 2. . 10 3 [4eV -If]
(2.30)
loS.1f [MeV. ~-3J
which corresponds to 11 = 9 m7f in our previous estimate. We
shall consider A as a free parameter limited to 10 3 to 10~
-~
GeV
With Eq. (2.26) satisfied, the o~ functions for the volumes
factor out and we obtain the reduced BE
(2.33)
. (2.34)
I J
(2.35a)
(2.35b)
(2.36)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 71
(2.37)
(2.38)
Define G('IJ by
(2.39)
(2.40)
G (4')
1.2 ----..--_...
----
--- .............
08
"
,---------------
,
G(%)=ln 2 =0693
G
(a) (b)
= (2.41)
(2.42)
(2.43)
(2.44)
A more familiar expression for the mass spectrum is p(m) (cf. the
text and the equations from (2.7) to (2.11)), related to T(m 2 ) by
T(m 2 ) dm 2 = p(m) dm:
(2.45)
rv (2.46)
For later we note that, had we not absorbed the p2 factor into
T(p2), but defined T(p2) to be the mass spectrum, we would have
obtained
74 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(2.47)
(2.48)
t/) -1
Rewriting this expansion in terms of S and applying ~ to it
is straightforward but tedious. The result is found in Ref. 14).
While this recipe yields still only an asymptotic expression for
T(m 2 ) -- which is, however, rather good down to almost m -- an
1T
exact expression for small p2 = m2 has been given by Yellin l3 ) in
terms of a series expansion of G(,) at ~ = 0:
(2.49)
WI
C(>(~) -
(2.50)
(2.51)
U -1
Applying N to Eq. (2.49) yields therefore
(2.52)
The 1MS integrals are well-known functions, for which powerful com-
puter programs exist. Therefore, Eq. (2.52) is very useful at not
too large p2, since the sum has actually only a finite number
76 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(2.53)
while now
(2.54)
One sees that the rapidly decreasing lIn! has been replaced by the
(exponentially increasing!) g. Thus the Q in Eq. (2.53) have
n n
been multiplied by n!g, which is the total number of possible
n
ways to cluster n pions; this factor represents the whole boot-
strap dynamics.
(2.55)
(2.56)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 77
(J -
(flk -
(2.57)
Given Eqs. (2.48), (2.49) and (2.57) the bootstrap equation (2.33)
can be considered as solved.
3. THERMODYNAMICS
(3.1)
78 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(3.2)
r
-
(3.3)
.
}
(3.4)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 79
(0)
ZA (~,V) (3.5)
(3.6)
*) We keep the bad habit of physicists to use the same symbol for
mathematically different functions; the information is con-
tained in the arguments.
80 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
-~ ~ A.,Z(9)(A VA)
V 'b~ \-, , energy density
pressure (3.8)
'A- (3.9)
(3.10)
82 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(3.11)
(3.12)
Obviously
(3.13)
(3.14)
(3.15)
is a critical temperature.
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 83
We do not specify now what happens at To, because that will de-
pend on the power of m in front of the exponential of ,(m 2 )
(see Eqs (2.44)-(2.47) and Table 3.1). We shall come back to this
in Subsection 3.3.
(3.16)
obeys, as before, the BE, but where T(m 2 ) is taken to be the mass
spectrum. Because of the identity (3.16), ¢(B) is the same func-
tion as before:
(3.17)
but now the partition function
(3.18)
1 == ) Oo(p~M4't)dltU.t.
in the integrals of Eqs (3.17) and (3.18) and integrating over p:
(3.19)
84 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(3.20)
(3.21)
(3.22)
Comparison of Eqs. (3.22) and (3.14) shows how important the proper
definition of T(m 2 ) is (cf. discussion after Eq. (2.31». Indeed,
Eqs. (3.22) and (3.14) lead to different physical behaviour near the
critical temperature To. We recall that GCY) = $(S) has a square
root singularity: $(S) = $0 -const·(S - So)%. Therefore, Zl(S) ~
~ const·(S - So)-%, Zl ~ const - const·(S- So)3/2. The correspond-
ing energy diverges in the first case: E(S) ~ const·(B - Bo)-3/2
I
and remains finite in the second: E(B) = Eo - const·(B _Bo)1.
We now return to the discussion of the thermodynamics follow-
ing from the choice of T(m) made in Section 2, Eq. (2.32): from
Zl(S,V) we find the N fireball partition function
-I
N'
(3.23)
(3.24)
tn(~,VrA) -~ 1.'R,u Z l~ V A)
V ra" 'I
Here, however, n(B,V,A) is the average number of fireballs present.
For this n(B,V,A) we have the ideal gas equation (due to the
linearity of In Z in A)
P = m·T (3.26)
(3.27)
i) the mass spectrum is split into a Bose part and a Fermi part
(3.28)
""Z( ~,V,,,) ~
(3.30)
ii) the integral over the rest of the sum (n =2 ••• 00) is
finite even if the first term becomes infinite at So.
.(3.31)
(3.32)
(3.33)
we obtain
(3.34)
~-~.= ¥ o
~ C tr;-T) = CAT (3.35)
+C
(3.36)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 89
(3.37)
Thus
(3.38)
(3.39 )
I
C C C C + CL\T3/4 C CL\T
1
L\T3!2 L\T3!2 L\T 5 !2 L\T7!2
An important new feature for SBM is our Eq. (2.26) which tells
us that the energy density in fireballs,
:: ..!. (3.40)
A
is finite, constant and of the order of the rest-energy density of
a proton. Therefore it occurs to us (in agreement with the contri-
bution of W. Nahm to this volume) that it is not reasonable to ap-
ply the thermodynamics derived from the BE beyond the point where
the energy density E{T) becomes larger than l/A (~ 10- 3 GeV~).
Furthermore, one may argue that the choice of T{m 2 ) as the mass
spectrum (cf. Eqs (3.16) - (3.22» is physically more consistent
than our conservative choice L{m 2 ), because T{m2 ) leads to
a =4 and therefore to a finite energy density at T ~ To (see
Table 3.1) in accordance with the above Eq. (3.40).
(4.1)
Equation (4.1) is not a single BE, but a member (with baryon number
b) of an infinite set of coupled integral equations, each having
its own input term. The (k!)-l is necessary for correct count-
ing. The non-vanishing pion and nucleon mass ensure that for any
finite p2 the set (4.1) has only a finite number of equations:
Ibmax I -< Ip2/m.
p Therefore the solutions for any finite p2 can
(in principle) be built up iteratively by starting with
4m~ ~ p2 ~ m~ and by increasing stepwise this interval to include
higher and higher Ibl. This, incidentally, also allows to prove
that for any p,V,b, Equations (4.1) have a physical solution.
Q., + III .,. ... + Q,.. ~ "" .,. Q'I. .,. ...
U'., of Ll'
Cor. .
(4.2)
(4.3)
V(~·. ,I..)
, = A 1M., (4.4)
(4.5)
(4.6)
(4.7a)
(4.7b)
The bootstrap equation (4.8) is much richer than that for the
pion gas; we have allowed the presence of arbitrarily complicated
clusters characterized by the baryonic number b.. For b = 0 we
1
have a description of meson - fireballs - but in order to under-
stand these fireballs properly, especially when baryon-antibaryon
clusters are among their constituents we have to obtain a solution
for the function -r for all values of b.
l [{('-)1: = 2:
J.-= -00
-
'A 4 ~ (' 1 - (4.9)
(4.l0a)
(4.l0b)
where
we find that the bootstrap equation takes the form of the pion
bootstrap equation, however, with a much more involved input func-
tion LOA:
(4.11)
c
We need to consider the Laplace transform in p of the function LA
in order to proceed further (compare Eqs. (2.35a), (2.35b)):
98 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(4.13)
(4.14)
These two equations define two basic quantities we will often refer
to further below. In complete analogy to the case of pionic boot-
strap we find now the usual bootstrap equation (see Eq. (2.37»:
(4.15)
(4.l6a)
(4.l6b)
The quantity BbTo in Eq. (4.l6b) is the input term of Eq. (4.6).
The relation between ~ and ~b is given by the inverse L trans-
form:
and similarly C
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 99
(4.17b)
(4.19)
q>(A
\.- t )
r. )
t
= &c.tt-1 (4.20)
(4.21)
(4.22)
fVL· (4.23)
J
This expansion considers in each term "£" objects consisting of
groups of n. "elementary" clusters with baryon number j, such
J
that the total baryon number is conserved. This power series ex-
pansion in "non-interacting elementary clusters" <f b (B,j) has the
same physical meaning as the expansion of an interacting field in
free-particles -- it is a series of products of free objects
102 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
and only the expansion coefficients contain the deeper physical in-
formation. The behaviour of this power expansion illustrates the
various physical regimes contained in the bootstrap equation, which
describes a loosely interacting system at relatively low tempera~
tures and densities where the expansion (4.21) works well, further
a strongly interacting system with dominant clustering, where the
series (4.21) converges badly and finally a qualitatively different
regime, where the Yellin expansion (4.21) does not exist.
(4.24)
(4.26)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 103
(5.1)
the only change being the dependence of the mass spectrum on the
.
baryon~c number b *) •
When n such clusters are present, but each with the same
b, we find for the n cluster function the usual expression"
cf. Eq. (3.23):,
M.
(5.3)
(5.4)
(5.5)
(5.6)
r
.0 110. ot
Z(~(V/))= J= -~ L
0.:0
~~ [AaZ1f,.) (~\Vla)]
= 41 V ,S ) }
or
(5.10)
106 R. HAGEDORN,I. MONTVAY AND J. RAFELSKI
(5.11)
(5.13)
(5.14)
(5.15)
(5.16)
(5.17)
where
,
m,
7T
m..
1'1
are the pion and nucleon masses, respectively.
Through Eq. (5.17) we have implemented explicitly the assumption
that the mass of a ground state cluster is proportional to the
baryonic number (this assumption might be given up in more refined
models). We now find for the grand canonical partition function
108 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
(5.18)
(5.20)
(5.22)
We insert Eqs. (5.22), (5.21) (5.20) into Eq. (5.18) to obtain the
final result
(5.23)
On first sight it might seem that the pion and nuclear contribu-
tions to In Z are additive -- however, when we recall Eq. (4.22)
we realize that cj>b (S,O) contains already a great deal of nuclear
contributions; the curly brackets describe that part of In Z
which is not generated via baryon-antibaryon creation but enforced
onto the system by a given fugacity A ~ 1. We note that the
multiple integrals in Eq. (5.23) may be reduced using:
(5.24)
~(~,V/').)
(5.28)
"(~IV, ~)
and the relative fluctuations of the energy density and the baryon
number density
(5.29)
!
~ (& V \):
~ "t Y2=
7
'I: ( - )
'\) (I' ~'2. (5.30)
possible to cross this border line (in contrast to the border line
e = Bo whose crossing would require infinite energy). We there-
fore expect that the line ec = f(A)c is a critical curve separa-
ting two different phases of nuclear matter.
(5.31)
(5.32)
lO
= ~.(,. (~11) + contributions of remaining subsets
(5.33)
One easily sees that the above subset of the sum (4.22) yields
(5.34)
- (5.35)
(5.36)
1 (5.37)
but the input term that describes only "raw" pions and nucleons
takes the form
(5.38)
(5.39)
114 R. HAGEDORN,I. MONTVAY AND J. RAFELSKI
(5.40)
(5.41)
it "1
As in Section 3, Eq. (3.9), we introduce the chemical potential II
by
'). = e ~r (5.42)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 115
~ [GeV]
1.0
0.8
0.6
III
0.4
0.2
o 50 100 150
T[MeV]
Fig. 5.1 The critical curve ~c = ~c(Tc) in the ~T plane se-
parating the gaseous phase (I) from the "liquid" phase
(II). The dot-dashed line would be the critical curve
if pions were excluded. Region (III) is inaccessible
(T > To): infinite energy density. For T = 0 the
critical chemical potential equals the nucleon mass;
note that this is not its maximum value.
116 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
.1L e-i.
K'Yl.(~) = \JJf [
1 + -q. ",,2:., + -.. ]
(5.43)
2~ g i.
We find, using Eqs. (5.41)-(5.43), the relation for large Sc :
(5.44)
is exponentially small at 13
c
~ 11m1T ~ lITo we expect that the
limiting temperature is little changed from that of pionic boot-
strap. The change of To induced by the possible baryon produc-
tion is obtained by expanding Eq. (5.41) around So. We find
-4 -I
As ~S is positive, ~ 4.2 x 10 MeV , the change of To IS
(5.47)
(5.50)
(5.51)
*) All figures have been computed using the A value of Eq. (2.27).
The value given in Eq. (5.49) would yield To around 200 HeV.
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 121
(5.52)
The first term is the only one remaining in the absence of pions
and is shown as a dashed-dotted curve in Fig. 5.2. Since for
T ~ To we have ~/T» 1, the asymptotic form (5.43) for the
Bessel function in Gr N can be used to determine v. Therefore
we find
c.;,. ",;n~
( 0) ( b)
Fig. 5.3 (figure caption on following page).
~
-
Co)
....,
.".
VNV VNV
./ II
,T=20 MeV II 1.1.= 0.94 GeV /././
O.B T=50 MeV ./ O_B
T=BO MeV ./ I.I.=O.B-GeV .,.",./
./ ,I
0.6 I 0.6
/
0.4 0.4
P ~0-5 P~ :s:
o
Z
(c) (d) --I
»<
Fig. 5.3 Baryon number per nucleon volume VN in various representations up to the critical curve. -<
(a) against chemical potential with isotherms; (b) against the temperature with V as »
z
parameter; (c) against the pressure with isotherms; (d) against the pressure with V o
~
as parameter. Po = A-I ~ proton rest energy density ("internal proton pressure"). The :::0
dash-dotted line is the critical curve, region (II) the liquid phase. The white lower »"T1
right corner in (c) is due to the impossibility of having no baryons at high temperature m
r
(unsymmetry of our input term). en
A
NUCLEAR MAnER AND THE STATISTICAL BOOTSTRAP MODEL 125
(5.55)
dropping henceforth the lower index \I. The results are shown in
Fig. 5.4. Here the isotherms T = constant are shown for VNE
as a function of VNV. We record the nearly linear behaviour (in the
gas phase) of the energy density: E ~ C1 + C2 v with temperature
dependent constants Cl. C2 • We reca1l that for very sma1l V(T) our
neglect of antibaryons is not justified. But above VNV = 0.1 and
T $ 120 MeV our results should be in~ependent of this approximation.
(5.56)
shown in Fig. 5.5. For small temperatures (T < 30 MeV) and den-
sities this should be just the usual 3/2 T, which we actually
find for T = 20 MeV. For higher temperatures, as we can see in
Fig. 5.5b, this is the lower limit of the thermic and interaction
energy E~r. For T = 50 MeV and higher, we have a large pion
component; thus the energy per baryon (total energy divided by to-
tal baryon number), which also includes the energy of the pions,
stays high above the lower limit 3/2 T. We note that our interac-
tion energy is, by definition, positive -- our nuclear mass ~
for the input nucleon should, in principle, include all the. binding
effects at saturation, thus be really - EB• Therefore, at den- ~
sities lower than the saturation density in the gaseous phase, the
thermal energy 3/2 T is the lower limit on the energy per baryon.
Furthermore, we note that within our model the thermic energy domi-
nates the picture between ~ 20 and ~ 60 MeV, at which point the
onset of pion and resonances excitation becomes important.
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 127
1.6 II
T=120 MeV
1.2
T= 80 MeV
0.8 T=5O MeV
=20MeV
0.4
e: ~'Kie~ e: ~r [Ge'll
10 1.0
\
\ ,
\
,I,
\
T t/n ..... _~ 0.8
8 \
,,
\
\ (b)
,, '1 \
,, \
6 (a) 0.6
T=140Me~
,,
\ I \ , II
,,
4 \ '-J II 0.4
:JJ
:r::
»Gl
T=50 MeV m
2 0.2 o
o
:JJ
Z
T=20MeV :-
I ~ -- , s:
I I I) I ..
o 0.2 0.4 0.6 0.8 VNv 0' 0~2 0'4 0'.6' 0~8 VNv .. o
Z
-l
Fig. 5.5 Energy per baryon (minus rest mass) as a function of baryon number per nucleon volume. <
»
(a) from 0 to 10 GeV - (b) from 0 to 1 GeV with isotherms up to the critical curve sepa- -<
rating gas (I) from fluid (II). Note that energy per baryon is not the energy per carried
»
z
by a baryon but : total energy of the system divided by total baryon number minus rest o
<-
mass : E/b - mN. At very low temperatures (T = 20 MeV) it is ~ 3/2 T. :JJ
»"T1
m
r
en
A
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 129
£= - -\? (5.57)
PUO-5 pJ
T=145 MeV
20
T=140 MeV
15
II
10
T=120 MeV
5
T=100 MeV
T=80 MeV
T=5O Me
T=20MeV
0 2 3 4 5 6
llV N V
E gr [GeV]
T=120 MeV
T=100 MeV
II
2 8
the case of the input term (pion and nucleon) defined by Eq. (5.38)
can be written,
(6.1)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 133
(6.2)
at)
(~)
(ql}<f.,..1:=
(IL) (6.4)
c; ". . (If.) <ft ):=
o
From the definition in Eq. (6.4) it follows immediately that
'Oatlr.} ('"et,)
I(
'bf,
. - Ci, q
, (Il)
Ie.. , (CP,) eft )
the one in Ref. 26), where the Bose-Einstein condensation was con-
sidered in the grand-microcanonical ensemble.) We choose the canon-
ical description which is simpler here (but the grand-microcanonical
description is also possible and equivalent).
(6.6)
(6.9)
(6.10)
p~ =- VI -6- )
Using Eq. (6.9) in the gas phase and adding, of course, the pion
gas contribution coming from Z , the same equations of state can be
1f
derived for the gas phase as in the grand canonical ensemble (cf.
Subsections 5.5 to 5.10).
(6.12)
For V > V (13) (i.e., in the "liquid phase") the saddle point dis-
appears. *As Eq. (2.48) shows, G(<p) has a square root cut in the
complex A.. plane extending from 'lo to 00 This cut generates a
cut in the A-plane extending from AO(13) to 00
IrnA
Re A
~
£.--+0
a (Jr.)
k (AiLE.) = (6.13)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 137
(6.16)
>c elcrt ~ [o(N J./~h,% ).f ".,lf -H~ I~, !PI!' >JJ )( lf
(6.17)
therefore
AI
Z'lfN(~IVI~ )= "N(~1~,r
rD d.~ .&.,.'(')., "VN )
i ~ ): e... "
~o(~) (6.18)
(6.20)
(6.22)
(6.23)
(6.24)
140 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
1m }.,
-----4------4------±~--;_~--~~--------~Re}.,
We know from the dominance of the saddle point in the gas phase
that in the thermodynamic limit
S(... )
CUI
/ ~
.s "«1
( ... ) --7> 0 6.25)
Here and below ( ... ) stands for the integrand ln Eq. (6.21) belonging
to the appropriate density (va or v). On the other hand, the above
explicit calculations show (see Eqs. (6.9) and (6.19) that in the
/~
case cr(S) = AO(S)-l we have
( ... ) 1 (6.26)
S~
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 141
-- (6.27)
Cv Sv
This proves the dominance of the tip of the cut in the "liquid
phase".
'Q(>c) : (6.30)
NUCLEAR MATTER AND THE STATISTICAL BOOTSTRAP MODEL 143
was also used which actually gives the average energy per particle
in a relativistic ideal (Bo1tzman~ gas by mR(mS).
(6.32)
The large cluster's phase ddes not contribute to the pressure, hence
P is constant along the isotherms (and equal to the pressure of
the critical gas). Hence, P*(S) is the vapour pressure when the
nuclear gas and its condensate co-exist. What we have called
"liquid" turns out to be really a co-existence of two phases. We
understand fully the pure gaseous phase and the state of two co-
existing phases, but not yet the pure "liquid". For this the volume
of the clusters must be introduced also on the thermodynamical level.
The property P = P* and the coincidence of the isotherms with the
A = constant curves in the "liquid" phase (A = AO(S» is also
• • d •
true f or t h e Bose-E1nste1n con ensat10n 26) • This shows a structural
144 R. HAGEDORN, I. MONTVAY AND J. RAFELSKI
REFERENCES
11) C.J. Hamer and S. Frautschi, Nuovo Cimento 13A (1973) 645.
15) E.M. Ilgenfritz and J. Kripfganz, Nuclear Phys. B56 (1973) 241.
18) A. Chodos, R.L. Jaffe, K. Johnson, C.B. Thorn and V.F. Weisskopf,
Phys. Rev. D9 (1974) 3471.
22) I.S. Gradstein and I.M. Ryzhik, Table of Integrals, Series and
Products, Acad. Press, N.Y. (1965).
Julius Kuti*
1. INTRODUCTION
149
150 JULIUS KUTJ
Callan, Dashen and Gross observe (3) that the response of the
pseudoparticle vacuum against color fields can be treated as the
four-dimensional version of the magnetostatics of a polar medium
provided the density of the pseudoparticles is small and they don't
overlap. Following their reasoning, I will often refer to the re-
sults of Appendix B. The reader is advised to go through Appendix
B first.
F~v)
z= J[dA ] exp[- __1__
~ 4l f Tr(F
~v
d 4x] (2.1)
'41 J Tr (F ~vF ~v ) d 4x •
Da (2.2)
).1\1
£ (P , R) (2.3)
The interaction energy, £(P,R), depends upon the size of the in-
stanton and its orientation R in the color space.
1 a
'2 e: llvaS F as
is the microscopic magnetic field whose sources are the external
sources and the dipoles.
(2.4)
Ba II Ha
llV llV
exp[e:(p,R)]
P (p ,R) (2.5)
f [dR] exp [e: (p ,R)]
where e:(p,R) is taken from Equation (2.4). The mean dipole moment
is
(2.7)
llD.G. 1 + 47T2X
D.G.
1+47T2 Jp5dp D(p) 87T 2
g2p,p)
£.4
8
, (2.8)
or
I
00
7 -x
llD. G. 1 + const. dx x e (2.9)
x
c
where the critical lower limit x of the integration is about x ~ 14
at which point the instanton den~ity becomes large. The corres~on
ding value of llD.G. is about 11, indicating a strongly polar medium.
a (2.11)
F~v
156 JULIUS KUTI
(2.12 )
For not very strong fields, the instanton medium can be treated as
linear.
F (n,E) = F
o + F.ln t - !2 ED (2.13)
Our medium is very similar to a van der Waals gas with a repulsive
core and an interaction energy proportional to -1/r4.
THE BAG MODEL AND QUANTUM CHROMODYNAMICS 157
.,- ,
I
,.
~,
,,"
--
,.., ,."
Figure 2. The interaction energy between an instanton and
anti-instanton. The hard core at R serves to
leave out overlapping pseudoparticles from the
partition function.
(2.15 )
Callan, Dashen and Gross solve Equations (2.14) -(2.16) with the
choice R = 2.2p and calculate n(p), ~ and D as a function of E3.
Their result for D(E) is plotted in Figure 3.
4
tr'flfi~I." ~•
.Ie"". ~ •••
1 , ,
Figure 3. First-order phase transition in external
chromoelectric field.
E '" 61-. 2
c
There are two phases in equilibrium with this value of the field.
p (2.17)
THE BAG MODEL AND QUANTUM CHROMODYNAMICS 159
p = 18 A4 (2.18)
(3.1)
I (3.2)
d '" 0.4 I
__
0_ = 1 _ E:
(3.4)
o
vac
Inside the hadron we want 0 '" 0, and therefore, E: '" 1. Outside the
hadron, 0 = 0 , and E: = O.
vac
The phenomenological Lagrangian density is assumed to be in
Minkowski space (1):
1 a ajJv 1 - jJ+;;+ - 1 jJ a
~ = - '4 € F jJV F + '2 \jJ (y d)1 + im) \jJ - g\jJ '2 Aa y \jJ AjJ
U (0
vac
) o (3.6)
and
U(O)=p. (3.7)
H = HQCD(E) + pV + SA (3.8)
(3.10)
The volume energy and surface energy will be added to the Hamil-
tonian at the end of the calculation.
and
a
D plays the role of the canonical momentum of the Yang-Mills fields
cgnjugate to Aa •
]l
THE BAG MODEL AND QUANTUM CHROMODYNAMICS 163
P
ij = ~ ", A-I "
Yij - o~u OJ (3.11)
h
were t"-1,~s t h e ~nverse
' 0f th I
e Lap ace tor, we can d e f'~ne the
opera
transverse part of A~ and D~ by
AT, p" A ,
a~ ~J aJ
and
T
D , = p, , D ,
a~ ~J aJ
respectively. The longitudinal part of A~ is set to zero in Coul-
omb gauge. The vafiables Aa, DO, and di t,,-loj Daj can be elimin-
ated in terms o·f D , , AT" 1/J0 anda~ which become our independent
dynamic variables.a~ a~
T
{ AT,
a~, ~j }* 0,
{ DT , T }* 0,
ai Dbj
(3.12)
{ T
Da~" ~j }* Poo,
~J
and the rest of the brackets are zero. On the right-hand side of
the third and fourth brackets in Equation (3.12), a 0 function, is
understood but not written explicitly.
T T l T T
Bai = - £ijk OJ Aak + 2 g £ijk fabc Abj ACk '
and the Coulomb Green's function Gab (x) will be used in the Hamil-
tonian. Gab (~, y.) is defined as the inverse of the operator
H= f 3{ 1
~
TTl
al. - al. - -
T
al. -
T
d x -2e:() D ,(x) D ,(x) + -2 e:(x) B ,(x) B ,(x) +
al.-
".+
'f'
(~) (-iaV
++ + 8m) ljI(~) + gljl + (~) 2
1 Aa -+
a ljI (~) -+T
Aa (~) +
al.,e: (x)
(3.13)
T -
9 D ,(x)
al. -
e: 2 (~)
where
p a = ljI + 1 Aa
(x) -2 ljI (x) + f b A.T , (x) DT , (x)
- - a c --bl. - Cl.-
div -+E = / 4
'3 (3.14)
THE BAG MODEL AND QUANTUM CHROMODYNAMICS 165
The minimum energy shape of the bag can be determined from the
equation
1 +2 1 1
-2 E + S (- + - )
= P (3.16)
Rl R2
at the boundary. Here, l/R, and 1/R2 are the principal curvatures
in two orthogonal direction§ at a given point of the static surface.
2
L = 0.2 (3.17)
3 'IT
220 MeV
s (3.18)
fermi 2
d =1 fermi.
VC1')
.2 (GeV]
The light quarks have a very small mass parameter in the QCD
Lagrangian (the value of m in Equation (3.10». However, ins tan-
tons generate an effective four-fermion interaction which breaks
the chiral symmetry of the vacuum. This picture as calculated in
Hartree-Pock approximation (2,20) in the instanton vacuum suggests
an effective quark mass m(p) which depends on the momentum.
(4.1)
a
E: = 1 -
a
vac
i
2
1jJ (yfl a+
].l
+ i (m+fa)) l/J
- 21 (d].l 0)
2
- u(a) ,
''.C,: 0
$.: .... .
: .. .
~f-+---->
.. 1
(4.3)
From the solution of the Dirac equation for the point quark
with the scalar self-mass term a/4z, it follows (1) that the quark
is repelled from the E = 0 phase.
When the two nucleons are very close to each other, there is
a repulsive soft core in the interaction. It is generated by the
chromomagnetic gluon force between quarks. At large cluster sep-
aration, we have a deformed bag with increasing separation energy.
The individual nucleons as represented by three-quark clusters can
be easily recognized in this configuration.
ACKNOWLEDGMENTS
2
H = E.:.
2m
+ Vex)
For an approximate quantum ground state, we may construct two
Gaussian wave functions 1/JG(x-a), peaked at x = a, and 1/JG (x + a)
peaked at x = -a. The width of the Gaussians is determined by the
THE BAG MODEL AND QUANTUM CHROMODYNAMICS 171
N f [dx] e~ S
1
--
en HT \ x.>
~
(A.l)
and [dx] designates integration over all functions x(t) with the
boundary conditions x(-T/2) = Xi and x(T/2) = x f . For large T
asymptotically
N f [dx] (A.3)
172 JULIUS KUTI
(A.5)
E = ~m (~~) 2 _ V(x)
(A.7)
-v
We want to compute
-HT
<al e- HT la> <-a I e I-a> (A.8)
and
A
x(t) = a tanh a rm t (A.10)
where the particle starts at the top of the left hill at time -T/2
and moves to the top of the right hill at time T/2.
L
solution vanishes in the large T limit. The action of an instanton,
+a
So = I dt [~ (:)2 +
m V(i)] = dx/2V. (A. H)
1
;::;2 wT
(A.12)
in the T+oo limit. The term Kn in Equation (A.12) comes from the
location of the n widely separated pseudoparticles. The exponential
term with its multiplier in front comes from the fact that V" = w2
everywhere except at the locations of instantons and anti-instantons.
174 JULIUS KUTI
X.
'l
_T
.t -1.4 r ~
I.~ t
) it\S~"",i.,,, QI1 l i i u-l....tort
7
-q,
\.. .z.
1
<a I ~fi HTI a> (~) ~e -~wT \'
1ft'! L
(A.13)
n even
since n must be even if we want to get back at a for the orbits of
the functional integral. For <al exp(-(l/~)HT) I-a>, we get the
same expression as Equation (A.13) except that the sum runs over
odd values of n.
1
1 -r S
2~w± hKe l l 0 (A.14)
Periodic Potential
V{~)
inS
I
n
£ 1/I G (x-an) (A. IS )
-I/(X)
. on-n-n +n (A.l6)
+ -
where n is the number of instantons and n the number of anti-ins tan-
tons. It is easy to find the continuum of energy eigenstates
labeled by the parameter S, and with the energy band of a Bloch
wave:
F
a~'V
a~ Aa\) - a
\)
A
a~
+ g fabc ~~ A
. c\)
(B.2)
1 ijk
2 E: Fajk
+ +
where E is the chromoelectric field of the theory. B is the
chromom~gnetic field. We shall adopt the guage A~ = O~ temporarily.
+
The wave functional ~[A] in field-diagonal represent~tion de-
pends upon the dynamic variables A of the Yang-Mills theory.
a
The Schrodinger equation is
+
E ~ [A] (B.3)
A = -ig Aa Aa
~ 2 ~
where g is the gauge coupling constant and the A's are Gell~ann's
matrices. The matrix-valued field-strength tensor F is defined
~\)
by
F
~\)
=a ~
A -
\)
a\)
A + [A , A ] •
~ ~ \)
(B.4)
(B.5)
(B.6)
(B.7)
-+
Al(y) is pure gauge, the corresponding field-strengths and class-
ical energy are zero, though we shall see that this configuration
is separated by a barrier from A~ = O. It is another local minimum
of the field potential energy which must be kept in the construction
of the true ground state.
(B.9)
THE BAG MODEL AND QUANTUM CHROMODYNAMICS 179
+ \' in8 +
'¥ e [A]. = L e '¥ [A] • (B.10)
n n
(B .12)
1 1
g2 kT
180 JULIUS KUTI
S = 2:...2 1:.4
g
I Tr (F
Jl\l
FJl\l) d 4 x (B.13)
d P - [dA ll ]
z
In the semiclassical approximation, we may evaluate the func-
tional integral in Equation (B.ll) by saddlepoint integration. The
saddlepoints are solutions to the Euclidean Yang-Mills field equa-
tions:
oS = 0 (B. 15)
oAll
in close analogy with Equation (A.4).
1
nall \I = E
oall\l
+
2
E
abc
E
bCll\l
The solution in Equation (B.16) is called an lnstanton, or pseudo-
particle localized around x = 0 in Euclidean space. There is also
an anti-instanton by the replacement
- 1
n = £ - - £ E
all\l oall \I 2 abc bCll\l
THE BAG MODEL AND QUANTUM CHROMODYNAMICS 181
O. o. ·
• 0 • • •O. 00
0.0 O.
• 0
Figure 14. The pseudoparticle vacuum in four-dimensional
Euclidean space-time.
+
dP~ +
Z ~
D(p~) , (B.17)
(p~) 5
~
11 2 (B.19)
1 + 8"ii7 g R-n].lp
1
11 R-n IP (B.20)
HT _8TI2/g2
<8 I e - I8> -+ E (8) = EO - 2VD cos 8 e (B. 22)
It
P
f(p) = n 2 p4 2 n(p) dp
o
Introducing the notation x = 8n 2 jg2 and using Equation (B.20)
in the asymptotically free regime (small p), we can write:
00
'V
f (x) D (x) (B. 23)
x
0.0f. If#.)~
o.o't
0.02,
'----'-_....i...._-'---+x
10 11. ,+ ,(,
REFERENCES
10. See, for example, Hasenfratz, P., Kuti, J. and Szalay, A. S.,
Proceedings X th Recontre de Moriond, Meribel, Vol. 2, Page
209, March, 1975; Gnadig, P., Hasenfratz, P., Kuti, J. and
Szalay, A. S., Proceedings of the Neutrino '75 IUPAP Conference,
Vol. 2, Page 251, 1975; Physics Letters, 64B, Page 62, 1976.
In References 1 and 9, the notation B was-USed for the volume
energy which I changed for p as a reminder of vacuum pressure
(see, also, References 4 and 5). I also changed the notation
for surface tension from cr to S.
14. Belavin, A. A., et al., Physics Letters, 59B, Page 85, 1975.
15. 't Hooft, G., Physics Review, D14, Page 3432, 1976.
21. Chodos, A. and Thorn, C. B., Physics Review, D12, Page 2733,
1975.
22. Creutz, .'1., Physics Review, D10, Page 1749, 1974; D12,
Page 443, 1975.
24. For a few, see, for example: 't Hooft, G., Nuclear Physics,
B138, Page 1, 1978; Yoneya, T., Nuclear Physics, B144,
Page 195, 1978; FOerster, D., Physics Letters, 77B, Page 211,
1978; Nielsen, N. K. and Olesen, P., NBI-HE-78-24; Preprint;
Nielsen, B. H. and Ninomiya, M., NBI-HE-78-39 Preprint.
25. Glimm, J. and Jaffe, A., Nuclear Physics, B149, Page 49, 1979.
QU_~ AND FERMIONIC GEOMETRY
Jerzy Lukierski
INTRODUCTION
The supersymmetric generalization of Einsteins
gravity was firstly obtained in the framework of four-
dimensional Minkowski Q]'T (1,2] .Recently it appears
however rather evident that the geometric interpreta-
tion of supertransformations requires the introduction
of additional anticommuting variables, which describe
together with four Minkowski coordinates the suoer-
space as a basic geometric manifold
..
where Tit denote anticommuting flat translations of Q,
/4/
/15/
=
where Z" (<t", 1:)'
We see that
a/ In the limi t R~ Co we obtain flat fermionic Cit,.. space,
with the group of motions described by inhomogeneous
U(',~) group with anticommuting flat supertranslations
. q" -. !~~E~(derived from the Grassmann rotations /9/
J.f ""It. -. ..)
*
b/ For R 0 the supertranslations are "curved", and
the nonlinear generators A.,.. S", in the parametrization
/14/ are given by the formulae
A...=t(~..-!~I'~..)+H~+~j1tl''') /17/
/28/
QUARKS AND FERMIONIC GEOMETRY 195
5 = tTiQ.
Introducing 2-dimensional Euclidean Dirac matrices ~,~&
one can write the equations /31/ in a Dirac form
196 JERZY LUKtERSKI
where
/33/
or in Dirac form
...
5"~ l~ 'X.c. = 0 Xci.. == (~~) /35/
w~ere _{lltl ,'XI'!'\= 0 _ Using the complex variables ~. f1+
+"j,..\ =~.. - i. SJ. one can write the general solution
of /351 as follows:
~!~!t(\) t;=2~(i) /36/
We see therefore that the fermionic QT instanton and
antiinstanton solutions satisfy the free two-dimensio-
nal Dirac equation, and are described by the formulae
/36/-
Abelian gluons and W-" _
In this Section we described only the model with
The quaternionic 6lT'string
model as well as the generalization to arbitrary
can be constructed in analogous way.
HADRONIZATION OF QT FERMIONIC COORDINATES
REFERENCES
1. D.Z. Freedman, P. van Nieuvenhuizen and S. Ferrara,
Phys. Rev. D~), 3214 /1976/
2. J. Wess and • Zumino, Phys. Lett. 62B, )35 /1976/
). R. Haag, J. Lopuszanski and M. SohnIUS, Nucl. Phys.
B88, 257 /19751
4. S:-Ferrara, M. Kaku, P. van Nieuvenhuizen and P. K.
Townsend, Nucl. Phys. Bj29, 125 /1977/
5. J. Wess and B. Zumino, ucl. Phys. f70, )9 /1974/
6. R. Penrose, Journ. Math. Phys. 8, ) 5 /1967/
7. F.A. Berezin and G.J. Kac, Math7 Sbornik, 82, )4)
/1970/
8. C. Fronsdal, Lett. Math. Phys. 1, 165 /1976/
9. F. Mans o1,l.ri , Journ. Phys,. ,18, 52 /1977/
10. J. Lukierski, ICTP, preprim IC/78/82, July 1978
submitted to Journ. Math. Phys.
11. V. L~ Golo and A.M. Perelomov, ITEP-62 Moscow Uni-
versity preprint
12. H. Eichenherr, Freiburg University preprint, 1978
1). D'Adda, P. diVecchia and M. Lfischer, Nordita pre-
print, 1978
14. P.G.O. Freund and I. Kaplansky, Journ. Math. Phys.
17, 228 /1976/
15. A7 Ferber, Nucl. Phys. Bl)2, 55 /1978/ .
16. J. Lukierski, Stony Brook preprint ITP SB 78/8,
Nuovo Cimento Letters, in press
17. F. Gfirsey and L. Marchildon, Journ. 1~th. Phys. 19,
942 /1978/; Phys. Rev. ]11, 20)8 /1978/ --
18. J. Lukierski, Lecture at Symposium on Math. Methods
in the Theory of Elementary Particles, Lib1ice
/Czechos1ovakia/, June 1978; Wroc~aw University pre-
print No 436, July 1978, Czech. Journ. Phys., in press
19. M.M. Ansourian, Phys. Lett. 70B, )01 /1977/
20. M. Gdnaydin and F. Gfirsey, Phys. Rev. ~, ))87 /1974/
21. G. Domokos and S. K6vesi-Domokos, Journ. Math. Phys.
12, 1477 /1978/
See [22] where the composite space-time coordinates
are expressed by two twistors with commuting compo-
nents. See also Rzewuski et all [2),24] who introdu-
ced in analogous way the spinor space as the funda-
mental geometric manifold.
QUARKS AND FERMIONIC GEOMETRY 199
I. Montvay
ABSTRACT
I. INTRODUCTION
201
202 I. MONTVAY
v = I Ekin
2~+Ekin
(1.1)
( 1.2)
= R!2~'.
Ek'1.n
(1.3)
(1.4)
If the tw.o nuclei w.ould just g.o en t.op .of each .other filling
the cemmon .overlap v.olume unif.ormly (a situati.on, .of c.ourse, c.om-
pletely unrealistic but still instructive) the density w.ould be
(1. 5)
where
(1. 6)
is the density .of the undisturbed nuclei f.oll.owing fr.om Eq. (1.4).
There is a c.onsiderable rise in the relativistic regime f.oll.owing
fr.om the ever increasing degree .of L.orentz-c.ontracti.on.
Due to the interacti.on .of the tw.o nuclei the kinetic energy
.of the nucleens in the c.m. system is certainly rand.omized t.o seme
extent. The degree .of randemizatien ("thermalizatien") is, .of
c.ourse, an interesting and impertant questi.on. Assuming cemplete
thermalizatien (i.e., all the energy transfermed t.o thermal energy)
the temperature .of the preduced het and dense piece .of hadrenic
matter ("fireball") can be calculated .once the thermedynamic
equatiens .of state (in particular, the specific heat) .of the
nuclear matter are kn.own. Taking again the simplest case .of
symmetric c.ollisiens (i.e., central cellisiens .of twe equal nuclei)
and assuming, fer a first guess, ideal relativistic gas equatien
.of state fer the nucleens, gives fer the temperature T the
fell.owing equatien:
(1.7)
3 Kl (x)
R(x) = - + - - - - - ; > + 1-
2x
+ (1.8)
x K~(x) ex + 00)
/1 + Ekin' =
2~
I(~
T' T
m'IT) (1. 9)
+ _3_ +
2~
a (1.10)
o v m 3
'IT 'IT
Here is the bootstrap parameter equal to a typical hadronic
volume:
...... 4'IT -3
=-m (1. 11)
3 'IT
NUCLEAR FIREBALLS IN HEAVY ION COLLISIONS 205
ao = ~ K (mTI) (1.12)
m
TI 1 T 0
Numerically, it follows from Eq. (I. II) that To ~ 217 MeV.
2Ap A t Ek'1.n
(I. 13)
(A +A )2m...
p t N
We can see from here that the symmetric case A At is the mos t
p
favourable one for producing high temperatures.
"freezing in" conserving its actual state later on. This "frozen
in" state is available for experimental observation. The hard
problem is, of course, that we have to figure out the whole story
from this "frozen in" picture.
'" In2
\)
=V (1.14)
* 'IT
In the fluid phase the nucleons have the strong tendency to form
big clusters consisting of many nucleons, whereas no such strong
tendency is present in the gas phase equivalent to a free gas of
clusters. Numerically, using the value in Eq. (1.11) for the
characteristic hadron volume v , the density in Eq. (1.14) is
'IT
about 1/3 of the standard nuclear density Po in Eq. (1.6).
n
dv (t) Kl (mpB) [ v'IT(t) Qp(B)]
'IT 00
nr
L V (t) -
dt n=2 n K2 (mpB) p Q (B)n
'IT
(2. I)
dVp(t) 00 Kl (mpB) n Qp (B) ]
L r V (t)
dt n K 2 (mp B) [-Vp(t) + 'IT Q (B)n
n=2
'IT
(2.2)
The reactions taken into account by Eq. (2. I) are the formation
and decay of the resonance:
(2.5)
are equal to 1.
00 1-1 2
L gl Bn Po 1 (M ) (2.6)
1= I '
(2.7)
m.v.
B. = _1._1._ (2.8)
1.
(2.9)
m for n =0
m
n { 1T
n~ for n > I
(2. II)
is
(2. 13)
. 8(b, k
L b. ) Pk (2.
1
m
n
-+ m2. ,
nl
... , m2. )
nk
i=1
for b = 0
(2. 14)
for b > 1
dVnb (t) co
snb(t) + L
dt
n'=n
v
A = nb (2.16)
nb d(n,b)Q (S)
n
for b = 0
(2. 17)
{(I A )b/b'
for b, b' of 0
n'b'
00 00
E(t) L L V b(t) m R[m S(t)] (2. 18)
n=o b=o n n n
(2.19)
(2.20)
214 I. MONTVAY
(2.21)
k
• IS (b, L
i=1
b.)
~
... , ~)
The continuous analogues of Eqs. (2.15), (2.19) are:
aV(M 2 ,b; t)
at
00
L
I Jkn dM~
-,
00
L
k=2 k. i=1 ~ b.=o
~
(2.22)
and
constant and one can consider the pion condensation at fixed pion
number density. By lowering the temperature T the condensation be-
gins at T=T*. As it was shown in detail In Ref. [16] by the use of
the grand-microcanonical ensemble, below the temperature T* the
critical pion gas coexists with a condensed phase consisting of in-
finitely large pion clusters.
PEFEPENCES
I. M. Gyulassy 1n Proc. Int. Symp. on Nuclear Collisions and their
Microscopic Description, Bled, Yugoslavia, 1977, Physica.
8. A. Mekjian, Phys. Rev. Lett. 38, 640 (1977); Phys. Rev. C17,
1051 (1978).
W. NAHM
CERN - Geneva
Abstract
219
220 W.NAHM
1. - INTRODUCTION
T-
and for its mass one has
(2 )
S " f~ ~ ~ - r7fGM 2
.
222 w. NAHM
S ~ ~ (4)
To'
When the radius of a black hole has decreased to (2rrT )-1, the
o
entropy of a hadron of the same mass starts to get bigger than
that of the black hole. According to the original version of
Hagedorn's bootstrap, the volume should be comparable, too.
Thus it becomes very probable that the black hole will change
into a hadronic fireball.
this yields
yielding
~ Li h(E,)) (10)
that is
E~
h(€) ::
-~ID (11 )
Thus
E GE~
SeE) - t ~ rs (€ ) -::: -=;=: -I- i r. T:.
o • D
(13 )
-dE
dt
~
2'17;
-GE ( 14)
This is rather slow, and in the last second much less than 10 9 g
will be freed. In the conventional model the explosion energy
will be higher, as already the electromagnetic radiation yields
a contribution
- ( 15 )
T :=
(17)
T(M) ~ To . (19)
(20)
Now Hagedorn compared with the experiments and found that F(x,E)
is practically independent of E, if the collision energy is
high enough. Later this phenomenon was rediscovered by Benecke,
Chou, Yang and Yen and called limiting fragmentation.
x .. (23)
but
(24)
ASTROPHYSICAL APPLICATIONS OF HAGEDORN'S BOOTSTRAP 231
(25)
(26)
One might even argue that at these low masses the exponen-
tial increase is just an accident, and that it would be suffi-
cient to use just the resonances of the naive quark model. How-
ever, the fireball excitation function f(x,M,E) might well have
a tail reaching to much higher masses. In particular one should
note that the thermodynamical model allowed a successful calcu-
lation of the production rates of anti-deuterons and anti-He 3 18)
In QeD one may also expect the production of larger fireballs
in collisions where multigluon exchange yields flux tubes with
a colour flux in a higher SU(3) multiplet.
(28)
t ",...., ,
whereas evidently no hang-up in fireball decay is observed. On
the contrary, for black holes just such a phenomenon is to be
expected, as according to Eq. (3) log p(M) is proportional
to M2.
(31 )
c (AE) o-(E) I
For open surfaces one can add unit pieces of surface anywhere
along the boundary, which yields some multiple counting. If
~(E) is the length of the boundary, one finds
.1.
For smooth surfaces, should increase like E2, such
that
(35)
(36)
)(f~ '"
..l ...J ..
.( &:
~ r
'"
-->n th
where p is the momentum on the n constituent. Obviously,
238 W.NAHM
(I) (38)
;v,M (39)
J
(40)
and
(41 )
20)
Gorenstein, Miransky, Shelest and Zinoviev ,and
Hagedorn, Montvay and Rafelski at this workshop introduced
versions of the bootstrap in which both non-associativity and
a volume proportional to the fireball mass contribute to the
exponential rise of the mass spectrum. Such models are threa-
tened by Occam's razor, but there are more serious objections,
too. The volume effect alone yields one non-degenerate leading
ASTROPHYSICAL APPLICATIONS OF HAGEDORN'S BOOTSTRAP 239
5. - CONCLUSIONS
When people tried to apply the bootstrap to hadronic
matter at large energy densities the results have been wrong
(as for black hole decay) or inconsistent with the observations
(as for entropy production in the early universe). There is no
point in looking for improved applications, as an analysis of
the foundations of the bootstrap shows that its fireballs should
be objects with an energy density of order 1 GeV/fm 3 • We have
to accept that our laboratory experiments yield no direct obser-
vations of the states of high energy density relevant for cos-
mOlogy. Of course, states of high energy density are initially
created by colliding Lorentz contracted protons, but they are
far from thermodynamic equilibrium.
REFERENCES
Johannes Ranft
Sektion Physik, Karl-Marx-Universitat,
Leipzig, G.D.R.
ABSTRACT
After a short review of some properties of hadronio
multi-particle systems which show the presence of a
statistical mechanism, we describe particle production
in soft hadronic collisions using the quark recombi-
nation model. Finally particle production in hard
collisions, in particular large transverse momentum
processes and the occurence of gluon jets in deep
inelastic processes is discussed within the framework
of QCD perturbation theory.
1. THE STATISTICAL DESCRJPTION OF HIGH ENERGY MULTI-
HAVRON SYSTr;MS, SUCCESSES AND FA ILURES
I will discuss multihadron production in hadron-
hadron scattering in the framework of the Statistical
Bootstrap Model /1/, the Thermodynamic Model /2, 3/
and the Independent Cluster Emission Model /4/ with
clusters decaying according to the SBM. Only a few
characteristic sucoesses and problems of these models
are discussed.
Suocesses:
(i) Limited transverse momenta of produced particles
following from the finite temperature T ~'30 -
180 MeV belong to the firmest predictions of
these models and have been observed at least in
241
242 JOHANNES RANFT
_I 2 2 2'
due to the term exp(--Yp + Pit + m /T). For Pions,
Kaons and Antiprotons, the only particles where
data were available, the agreement with experi-
ment was good. Also the production of the J,~
(3.1 GeV) particles was consistent with this
rule /3/. Recently it was found by Satz /5/
that even uncorrelated multihadron systems and
~epton pairs produced in hadron collisions
behave like (1).
(v) It was first found from the analysis ot inclusive
and semiinclusive two particle correlations:
that particle production in the central rapidity
region is dominated by central clusters or fire-
balls with masses <m>",1 - 2 GeV /4/. In the
recent years more and more experimental evidence
has become available, that particle production
in the central region is largely via the pro-
duction of well known resonances.
(vi) The effeots ot. Bose statistics on multihadron
systems produoed in hadronio ool!isions has
been predioted since long /6/. Two particle
IS THERE EVIDENCE FOR HADRONIC CONSTITUENTS? 243
a.cr I
cJ..~ f r =
~ d..v..1 d...~2... clx:3
x-::-
X" 3 -x F3 (x,,/)(2. Xa) ~3
(7)
(><,.,)<2. X""x.\
-+ 10 -, I I .. 'J
do-I
d'l< =o(b ~"
(oi-X)
2.
t
(J
J(
~
jrAX-"
x' X1
PI'''' to X 0
o
The production of protons, neutrons and A hyperons
could well be described /14/ by this method.
246 JOHANNES RANFT
n d. ....
,~ ~ 5(11- ~ ><j) ,
l"''' "j J:.-1
We have ns kinds of sea part one (quarks or gluons).
The 'Cotal number of' partons is
111\. =. 3 + 1IIl... .j.. 11It..2,. 1- .•• of M..."'s (12)
The f' ,1(X) are valence quark matrix elements, the
fs i(~) are sea quark matrix elements. Different f'rom
Kut~ and Weieskopf /16/ we ohoose
d>C-1-)() =: (15)
= ~
1>11 -+ ~
2 ("Po) f
"=0
(~"Lt· I ex
"""" h I
"Po)
with
(V
t\- It
$:- 1(;) .
~-)(
S(~
'))
.f - }() ... - i) 0
S( ~)
~
.. i-. \A. d~
(18)
and
1-X
2'(~}-1-X):'~· S~)c) (t>'r(-i~)()d)C _ h.L-1~)(] (19)
Cl.
We insure the proper normalization of' the parton dis-
tributions
~ ~cLfM.lX.,) ... ) X/~J == -1 (20)
248 JOHANNES RANFT
~A ~ Jrf'{J
3
Jf~1
I
d.;,i
I
r ,(Xi) \f~(I\_£ Xi)
til',. I'=-'\
(21 )
r
l.t(x) .: tlA.
I
V CX) 1\ )(elx
~
, . "':..,
2. -
Xct.
r V(><Iot,)
"" d..t.d, t",
2., l:
t (KiA.)
1..,11
~ ('\ - x -)(\4. -
.a..
(22)
xJ
~
F(X k ,.) X;;:,,) = f ..,v (){ ...) fJ/".r,s) t\ Sd: . . ~.( ""2.. t(
(23)
j
~ (1\-'1( v..~v -X'",-2.. -}(J)
,tt.c.;,.., v (l(\.t;L) TcrdV(X.J V) '1' r.{, 11\.1'
3 valence quarks
( 24)
1 sea quark
S (x) ~ ts (x.) 1\ (d.K"'1\ lAx: ""a.. cil<'tA,. r (x·) ( eX) (25)
J X',.
-"
l(
t.\~
)(~ t'\ V
CII.'
14.,\ t\.t, v ,
IA.~.
f,,(x) =X
of - U(. to)
j
r
ts Cx) =~1.,
(26)
With this simple choiiCe all integrals can be evaluated
:in closed form
IS THERE EVIDENCE FOR HADRONIC CONSTITUENTS? 249
(27)
(29)
~;_lpp ... 7-
= ----------------------
¢~"'-(I\-)() ~j(tc.t(J()+~(x.-X")c!)(,,
where the phase space integrals th 0j (1 - x) are de-
fined as follows ~~
250 JOHANNES RANFT
Ph (t:A.'lCj tJ,.'lCj r .r ~
41ij{oi- lC ) =j lei 'lCj ti(~i) tj(~j) yr(I\-)C - X, - Xj) (30)
The J(" +/Jr- ratio (29) depends exp1icite1y on the sea quark
matrix elements. Therefore we can fit the experimental
data to fix Ksea/13/. In this way we obtain a good des-
~~il~!~nb~~a!~gd~l~~ :(~)~if1_!)a§~9s~~n~~~:n~i:l~!-
data on lepton pair productio~. After this fit also the
shape of the Jrfo';r:']r distribut ion is very well a.escribed,
see Fig. 2 and no free parameter is left in the model.
A second problem is the normalization of the sea
quark a.istributions to De used in the quark recombi-
nation model /13/. In hadronic collisions, a long time
is available to form the resulting hadrons. During
this time the gluons fluctuate continuously into quark
antiquark pairs and can contribute as such to the re-
combination into hadrons. Therefore we choose in the
model only to keep two parton recombinations into mesons
and three parton recombinations into baryons but with
enhanced sea quark distributions, enhanced so much that
they carry the momentum of the gluons as well.
pp-+l[tx PT .0.8
!.i=45
10
/
j
8
.. :/
~
!)~t
II
....... V'
.- ~~
t#
o
o 0.5
x
10' .'--
.:---r--.
1'--r-... ~
:-... r-.. ............ ~
~ ~~
~ "
~1\
IAi'
i\ .\
~
O.S
x
( 32)
IS THERE EVIDENCE FOR HADRONIC CONSTITUENTS? 253
1"<-
1"- .
•
If
____ iir
==~---===~------vp x-o..,
r
J[-
/ klif;.h,,,,,J;
10 PP-Jr!·x It..r~"
/""'" tr.~~'r
t..rf-o" rair r -M'
s-
..,.. ,"£..sO..,,,
/
G,
4 1"'''''0.0-1
2-
~-o.o.
--... J:.O.11
J .. o.1b
0
0 ·l .~ ·6 ·8 1.
X
r
are calculated in the quark-parton model from the e%-
pression
dO-A" = L~,r..
d.. a. dx" F"'A ix..) F.,.. t~.) d......... (3)
Order g 2 : qq
- .....,fl+
-
(35)
--"""---r
256 JOHANNES RANFT
d"A ... ~ ~ ~ cI. x.. G( Xb 1'.;.,,,(. ~ p') F.,,, ex., Q') a.cr-" b
(36)
I
5
5 10 15 20 p.. GlNlc
<k2.>
p~ ~ 1c.a.2.). .£,.
pp-JET' X
data: Bromberg et al.
• ~ 3GeVJc
-28
• >4 '
• untJjasE'd
::- -29 '.'.
~ - - - (k:~ -1 GrNlc
Q7 •
"e
&
-30 --' a
a. "31
~
.g
w -32
.§' •
-33
-3"0 1 2 3 4 5 6 7 8 R.GeVJc
r- s2-
1- o~·
IJ ~ - J;. z.
'J 3
260 JOHANNES RANFT
olo:;- (45)
de =
and for definite transverse polarizations of the two
quarks
cto-1'+ no<;
= s2.
~t
d~~ 1r 0<.$
l...
=
Cltt- &2.
c{fY+- i- d...~_
i 1-j
'" =
cU:. d..t
=
{ 0.' (47)
A£..J... c10+i-
+- d.-<11--
o.LtS I=j
~ oU:
J b
b
J
b
b
b
b
O. .2 .4 f> .8 1.
x.
=L
the process quark + quark~quark + quark + gluon
according to
~ ~~ =t(49)
The azimuthal angle ~ dependence of the production of
two large p~ particles for a fixed value of the total
transverse energy EL~ p~ + PL2-+ PoL (p~ is the
transverse momentum balaJcing ~1 + 3p;) ~as measured
by Cobb et a1. /31/. If the jets are produced at
9 = 90 0 the variable E~ plays for the large p~jets
a similar role as rs in e+e- -.hadrons or the hadronic
energy W in deep inelastic collisions. The large 'cross
section for large E..L. around e~ 1800 is evidence for
the jet structure of large P..L. events. In this experi-
ment the usual trigger bias effects, which occur if
one or two particles are selected according to their
transverse momentum, are not present.
The cross section measured by Cobb et a1. /31/
was howevgr also found to be rather large at e values
around 90 where only rather small contributions are
262 JOHANNES RANFT
o
III
X - all final state jets
~ 40 -- qucuk final state jets
VI
~
~20
;
~ O.j....-L.~...;u.=.,....~
0- 90-
Fig. 8.
duction of two *'
The az imuthal aneile <l> de-pendence of' the pro-
at large p. t"or fiJeed E~
= P~1 + p~? + PL1· The data \s due to Cobb et
ale The carves are calculated by Kripfganz
and Schiller according to QCD three~jet
contribut ions qq - qqg.
IS THERE EVIDENCE FOR HADRONIC CONSTITUENTS? 263
(51)
-"""-'1
if
p------ p
The two processes with emlssion or a gluon dominate at
presently available energies. The rise of hed.ron trans-
verse momenta relative to the direction of q due to the
processes (51) was studied by Altarell~ and Martinelli
133/.
Azimuthal asymmetries due to these processes were
predicted by Georgi and Politzer /34/, Cleymans /35/
and Mend~z /36/. Georgi and Sheiman /37/ studied the
cross section do/dT~, where ~iS defined as
( 52)
They found that the H.L distribution, which at large
is less dependent on nonperturbative effects, can be
u:.
predicted more safely than average values like <P,L'> •
Ranft and Ranft /38/ studied the 3 jet cross sections
(51) in terms of the variables spherocity S and thrust
T. In the following we will discuss these cross
sections. The variables S /39/ and T /40/ are defined
as follows .
S=(~t~i'" (z.lrlIV- (53)
tr L.1p..I)
(54)
T==
For an ideal jet at ra~1l:I one expects S .. 0 and T .. 1.
For an ideal spherical symmetric t'irebaLl S .... 1 is ex-
pected. At finite energies nonperturbative effects,
the fragmentation of quark jets into hadrons with
limited PL ' leads to Sand T distributions of forms
like
ctN
d.S
<'V S
(AS)~
e..xrJ(-.L).
\ AS ) (55)
264 JOHANNES RANFT
dN
dT
N (" - T)
(AT)t. e.X)' -
. - T)) j
(-1AT (55)
References
/1/ R. Hagedorn, Supple Nuovo Cim. 3, 147 (1965).
/2/ R. Hagedorn and J. Ranft, Supple Nuovo Cim. 6,
169 (1968);
Nucl. yhys. B48, 157 (1972)
/3/ J. Ranft, Nucl. Phys. B105, 139 (1976); this
paper describes the last version of the T. M.,
see also J. Ranft, contribution to the Work-
ShOp on Theoretical £hysics: Hadron Thermo-
dynamics, Erice 1975.
/4/ reviewed in J. Hanft, Fortsch. Phys. 23, 467 (1975).
/5/ H. Satz, preprint CERN-TH 2532 (1978).
/6/ u. uo~dhaber et al., ~hyS. Rev. 120, 300 (1960)
/7/ J. Ranrt and G. Ranft, Phys.Lett. 57B, 373 (1975);
Nucl.Phys. B92, 207 (1975).
/8/ Data by V. ~. Kenney et ale pre~ented by R. E.
Diebold at the 19. Int. Conf'. on High Energy
Physics, Tokyo 1978.
266 JOHANNES RANFT
H. Satz
CERN - Geneva
University of Bielefeld*)
ABSTRACT
*)
Permanent address
269
270 H. SATZ
I. INTRODUCTION
(a) (b)
(0 ) (b)
Fig. 2: Metal in (a) gaseous and (b) solid phase
*)
This problem can be avoided by formulating the bootstrap model
at constant particle density [see Ref. 8)]
FROM HADRON TO QUARK MATTER 273
(2)
(3)
(5)
with
(6)
(7)
Z(S) • 1 + B~
o (8) + (l+a)
(8)
(10)
( 13)
276 H. SATZ
(15)
with
(16)
( 17)
(18)
2(S)
(20)
10)
is, however, found to yield the same qualitative behaviour as
Eqs. (11)-(19), so that we here concentrate on the latter.
~0(8) 2 In 2 - I (21 )
(22)
and thus
(23)
N-2
Q ::
ex>
L (B~o) IN! Q' :: aQ/az (26)
2
vanishes when
FROM HADRON TO QUARK MAnER 279
Z - 2( 1+a.)Q In Z. (27)
Z(Jj) E(Jj)
FG
To To
(a) (b)
In Z = I, (28)
z (~)
~-I
(31 )
For our partition function (19), this regime ends with a=a •
c'
for a < ac' d~O/dZ never vanishes and hence we always have a
finite energy density. In Fig. 5, we show s(S), as obtained
from Eq. (19) and (24), for various a < a
- c
E(/3)
/3-1
Fig. 5: The hadron-quark energy density for different
resonance couplings ac > a 3 > a2 > a l •
(dS/dT)B = O. (32)
conditions
(33)
iii) for a < ac , the specific heat is once more always positive,
the temperature is unbounded, and the energy density in-
creases again monotonically with T.
(34)
KT = (-I/V)(av/ap)T (36)
between the gas and the liquid states of the system (cf., Fig. 6).
The problem is resolved by introducing two phases in coexistence
and consequently replacing P(V,T) by the Maxwell construction
peT) for V' > V > V' with VA' and V' determined by the equal
A - - B' B
area rule. For V > VA' we then have a homogenous gas, from VA
to V~ gas and liquid in coexistence, and for V< V~ a homogenous
liquid. The coexistence region becomes smaller with increasing T,
until it vanishes at the critical point T • T , where the Van der
c
Waals equation provides a diverging compressibility and hence in-
finite volume or density fluctuations: the system at the transition
point oscillates between the dense liquid and the less dense gas phase.
€(Jl)
P <--> E: T <--> a
V <--> T
FROM HADRON TO QUARK MATTER 285
between a Van der Waals system and our Eq. (19). The phase
transition in the former is associated to volume or density
fluctuations, in the latter case to temperature fluctuations. In
Fig. 8 finally, we show the comparison between the P-T diagram
of the Van der Waals system and the corresponding E - a diagram
in our case.
p E
RG QG
G
__------------a c
a >a c
€
Fig. 9: Temperature variation in the hadron-quark system.
********
REFERENCES
13) L.D. Landau, Izvest. Akad. Nauk , Ser. Fiz. 12, 51 (1953).
John H. Schwarz%
Laboratoire de Physique Theorique de l'Ecole Normale
Superieure, 24 rue Lhomond 75231 Paris Cedex OS, France
ABSTRACT
Various trends and problems in supersymmetry theory are discussed.
The requirement that gravitation be consistently unified with the
other interactions in a finite theory is very restrictive with far-
reaching consequences for the other interactions. A specific example
based on the spinning-string theory is described in some detail. In
this model the elementary particles lie on linear Regge trajectories
with a common slope determined by Newton's constant. As a result the
spectrum suggests a limiting temperature corresponding to the Planck
mass, rather than to the pion mass as in hadronic string theories. The
superspace approach to supersymmetry is also presented. While this
formalism has so far only been applied to the reformulation of known
theories (in a more geometrical way) it is hoped that it will provide
a framework appropriate for obtaining insights and extensions that would
otherwise be elusive.
where P r and Jr~ are the translation and Lorentz generators respect-
ively. e is a dimensionless constant. characterizing a de Sitter
radius, that vanishes when one contracts to obtain Poincare invariance.
The only other super Lie algebras that have been seriously studied
are the superconformal ones SU(2,2;N) which contain the conformal
group SU(2.2) and SU(N) as subgroups. While algebraically elegant
theories have been constructed [7J based on this algebra, they
involve higher-derivative interaction, so that their interpretation
as particle theories appears problematical.
The SO(N) theories with e = 0 have irreducible representations
whose particle content is easily inferred from (I). For the massless
case (which is the one of greatest interest) one finds that the I.R.
contains 2N physical helicity states with the multiplicities of the
various helicities given by
N!
(2)
where Jmax denotes the maximum helicity in the I.R. Evidently the
z
294 JOHN H. SCHWARZ
Jmax _ N
(3)
z 2
When Jmax F N/4, the TCP conjugate multiplet containing the opposite
z
helicities must be added in. Supergravity multiplets are, by defini-
tion, ones for which Jmax = 2. The multiplet is evidently self
z
conjugate (with respect to TCP symmetry) when N = S and contains
spins exceeding 2 for N ~ S.
Interactions can be introduced among the fields of a supergravity
multiplet in such a way that the spin-two (vierbein) field interacts
according to Einstein's theory and supersymmetry is preserved. When
such theories are constructed, they turn out to have surprisingly
good ultraviolet behavior. Namely, on the mass shell the divergences
appear to cancel through the two-loop order in perturbation theory
(S] . This is in striking contrast with what happens when gravity lS
contain all of known physics --namely SU(2) x U(I) for the electro-
weak interactions and SU(3) for QCD. One can embed SU(3) x U(I) x U(I),
but then the ~ bosons are missing [9] . This is a very serious
problem in this approach. One might imagine that other states arise
dynamically and perhaps even fill out SU(8) multiplets, for example.
This is wild speculation, however, and I know of no reason to
seriously expect such good fortune.
Even if the problem of the group is overcome for extended super-
gravity, there are other difficult problems that must be confronted.
In particular, when Yang-Mills self interaction is introduced so as
to gauge the SO(N) group with coupling constant e (10] one obtains
a cosmological constant at the same time that is some 60 orders of
magnitude larger than the experimental limit (de Sitter radius of
10- 30 em instead of 10 30 em) for reasonable choices of e.One may
hope, following a suggestion of Hawking (II] that in a quantum theory
such a term would only play an important role on the scale of a Planck
length and would not imply a large effective de Sitter radius on a
macroscopic scale. Alternatively one may hope that some cancellation
can be rigged up, preferably in a "natural" way. Another serious
problem concerns the breaking of supersymmetry. For theories with
N > 2 the usual Higgs type mechanisms (12) appear inapplicable, at
least based on fields occurring explicitly in the Lagrangian. The
theory must find some clever way of breaking its own symmetry. There
are simply no knobs to turn from the outside. In summary this is a
very grandiose program that is running up against some equally
imposing obstacles.
of this type are known to have a spectrum free from ghosts. The first
is the Veneziano model which exists in 26 dimensions (I time and 25
space), but contains a tachyon and no fermions. The second is the
spinning-string theory (14], which requires 10 dimensions (9 space
and I time). This is the theory we will discuss in greater detail
below. The third theory is a two-dimensional string theory [IS] with
extended (N = 2) supersymmetry in its gauge algebra. It turns out to
be equivalent to a supersymmetric nonlinear (J"" model [16] in two
dimensions x .
When first invented, the spinning-string theory was proposed as
a phenomenological theory of the hadrons. In the Born approximation
the state·s all lie on linear parallel Regge trajectories. The
universal slope~~ was naturally taken to be about I GeV- 2 ~n the
context of a hadronic model. Since string theories have an exponent-
ially increasing particle number density as a function of mass, this
implied the existence of a limiting temperature close to the value
suggested by Hagedorn and Fral1tschi based on statistical bootstrap
ideas [17] . What Scherk and I have suggested [18]in previous works,
and which I wish to review here, is an alternative interpretation in
which the states of the string theory are taken to be elementary part-
icles. In particular, a massless spin-two state in the closed string
sector is identified as the graviton. Then the requirement that its
interaction be of the usual Newtonian strength implies a universal
, -34 -2
slope controlled by the Planck mass, 0( ~ 10 GeV. When this
assignment is made one is left with a well-defined finite spectrum
of massless states and a plethora of string excitations having masses
of the order of 10 17 GeV, and hence quite unimportant for most
purposes. Of course, in this scheme the limiting temperature is also
taken to be of this magnitude. This suggests that the Hagedorn
XThe field theory version of this model was discovered some time
after the string theory. The identity of the two theories was not
noted despite some overlap in the authorship.
SUPERSYMMETRY APPROACH TO UNIFICATION OF INTERACTIONS 297
(4)
The action formed from this Lagrangian is invariant under the global
infinitesimal supersymmetry transformations
SUPERSYMMETRY APPROACH TO UNIFICATION OF INTERACTIONS 299
i € r; "AQ. (5a)
L += f""o. E
r ll
(5b)
• (6)
mJ.. (7)
10 •
( ll.
t)b 0
ddt:. £.10
(Note that I take ~ 4 and 0<...
.p to have the same dimens ions
10
--namely the usual one of 4 dimensions).
At the same time as y dependence is dropped, the fields that
transform by a nontrivial representation of 0(9,1) have to be
decomposed into 0(3,1) representations. Thus the 10 vector A~ is
decomposed by the rules
t-"= 0,1,2,3
} m 1,2,3
r (0oc. O{W\)
}
m+ 3 : (8b)
'IS' ~
TYI 0 m: '.I J..) 3
r ll\+b =
"S- ® (~~ O~) (8c)
o -~
where the ()( I sand r'S are 4x4 real antisymmetric matrices
satisfying
pmp'n. ==
_ E"'lIr
r P 8"'''' (9b)
t..o(~.1 rtl.] = 0 ~
(9c)
I (10)
~
where °l:i are four 4-component Majorana spinors} and we have taken for
the charge conjugation matrix
(11 )
( 13)
302 JOHN H. SCHWARZ
i 1,2,3,4 (14)
( 15a)
(15b)
( 15c)
(16)
where
v = d-et V;
I5r 4J1' = ~t li'r •
The field Vr is the vierbein and ~~s is the connection field, which
is varied independently in the first-order formalism. R V'" Vl> Rrr~
= is
r s
the curvature scalar. The action formed from this Lagrangian is
invariant under the local infinitesimal supersymmetry transformations%
sV'"r = - ~ K (X r
'tho \V ( 17a)
S~ = ~l}o(.1 (17b)
%There is also a formula for S tU~rs, which is not needed when ~~s
is expressed in terms of the other fields using its equations of'
motion ("one-and-a-half-order formalism").
SUPERSYMMETRY APPROACH TO UNIFICATION OF INTERACTIONS 305
+ )
(18)
(19a)
-I
e·~ et + 0(
Y I
·r ~ (
rv )""
~ "''''tS'': r e~ ) ( 19b)
rs
parameter € (x, e) one has
+ • (20)
J.. I "5
+ .;L n"
+ X"-5 ) (21)
}
(22)
Z
where CXy are the structure constants of OSp(I,4) and the [ •• ~ 1
symbol means connnutation or anticonnnutation "as appropriate". In the
general case of curved supers pace one has
(23)
Z
where RXY are the superspace curvatures and torsions. One other
convention we have found convenient is to raise and lower spinorial
. ab -.I ab
tangent-space indices with the charge conjugation matrlx C (= 0 0
in the Hajorana representation).
In previous work [22,23J we showed that equations of motion and
transformation laws implied by eqs. (16) and (17) would follow from
making a special gauge choice in superspace provided the covariant
superspace equations (including a de Sitter charge e and axial-vector
external source J~ describing vector-spinor matter) were the following
308 JOHN H. SCHWARZ
-2e( ()rs)
ab
o
(24)
2 t JU
E rsu 5
Now these 420 equations describe the dynamics of the 112 fields ~
and hAs . Therefore they are highly redundant. It has been shown by
MacDowell [29] as well as by the Caltech group [30] that a subset of
112 of these equations can be selected from which the others follow
as a consequence of the Bianchi identities :
(26)
(27)
where
REFERENCES
F. Gliozzi, J. Scherk, and D. Olive, Nucl. Phys. BI22 (1977) 253.
2 L. Brink, J.H. Schwarz, and J. Scherk, Nucl. Phys. BI21 (1977)77.
3 N = I supergravity was formulated by
D.Z. Freedman, P. van Nieuwenhuizen, and S. Ferrara, Phys. Rev.
DI3 (1976) 3214
S. Deser and B. Zumino, Phys. Lett. 62B (1976) 335.
4 For a review of extended supergravity theories and a complete
list of references to them see :
J. Scherk, LPTENS 78/21, Invited talk at the July 1978 Cargese
Institute.
5 A. Salam and J. Strathdee, Nucl. Phys. B80 (1974) 499 ;
P.H. Dondi and M. Sohnius, Nucl. Phys. B81 (1974) 317.
6 R. Haag, J.T. iopuszanski, and M. Sohnius, Nucl. Phys. B88 (1975)
257.
7 M. Kaku, P.K. Townsend, and P. van Nieuwenhuizen, Phys. Lett.
69B (1977) 303 ; Phys. Rev. Lett. 39 (1977) 1109.
8 For a review see P. van Nieuwenhuizen, CERN preprint TH.2473,
Orbis Scientiae (Coral Gables) 1978.
9 M. Gell-Mann, to be published.
10 D.Z. Freedman and A. Das, Nucl. Phys. BI20 (1977) 221.
II S. Hawking, lecture given at San Francisco EST meeting, Jan. 1978.
12 P. Fayet, Nucl. Phys. BI13 (1976) 135.
13 "Dual Theory", edited by M. Jacob, North-Holland Publ. Co. (1974);
J. Scherk, Rev. Mod. Phys. 47 (1975) 123.
14 A. Neveu and J.H. Schwarz, Nucl. Phys. B31 (1971) 86.
15 M. Ademollo et al., Nucl. Phys. BIll (1976) 77.
16 P. Di Vecchia and S. Ferrara, Nucl. Phys. BI30 (1977) 93
E. Witten, Phys. Rev. DI6 (1977) 2991.
17 R. Hagedorn, Nuovo Cimento Suppl. 3 (1965) 147
S. Frautschi, Phys. Rev. D3 (1971) 2821.
18 J. Scherk and J.H. Schwarz, Nucl. Phys. B81 (1974) 118
Phys. Lett. 57B (1975) 463.
19 Chan Hong-Mo and J.E. Paton, Nucl. Phys. BIO (1969) 519.
312 JOHN H. SCHWARZ
F. Zachariasen
CERN
Geneva, Switzerland
1. INTRODUCTION
(1.1)
:: f 2. 1'-1 6
L
~ e 61. J" ~/( I) (It) s ( ,- l! )
. (z.,-l~ ( f..., - f¥.) 1; (" 1- ki I.). (1.2)
RUNNING COUPLING CONSTANT IN YANG·MILLSTHEORIES 315
f) (1) ;
/"
is the photon propagator in a gauge specified by the parameter ~ ,
and rv is the proper vertex function in which a photon of momen-
tum p-k connects to one of momentum k.
r (f. f-~' 1)
/ ~ J
::. (zr-I.)
7
r,
z.
(2. 1- I.)
r (k -
7
z:.
(ZI-I.)·/c
It
) r z. (1.3)
(1.5)
..
The equation 1S linear and may be solved by a Laplace transform.
In the IR limit, the solution is
t'r(l'! ,.,.) +
j
~(s)
S(,) ~ ds e
6
where
(,. /.) L)
b t.r J 'Co
4~1
I~ ~~
(i Z
f ~
I e
t'l ( ! ~ ,-/,1. J
It. J_ z.,' t.
I/. /==""
(1.6)
(1. 7)
1+
-)
) (1.8)
(1. 9)
Formally, this is also the leading log IR result, but since in QeD
b < 0, there is a singularity as q2 + 0 and log q2/A2 + -00.
Hence the result is not valid in the IR: we must go beyond a lead-
ing log calculation. Fortunately, the Dyson equation/Ward identity
approach will permit us to do this.
318 F. ZACHAR IASEN
What basis do we have for this hope? In QED the "moment form
factor" (essentially f2) is known to be free of IR singularities,
and, furthermore, the spino logical coefficient of it in the vertex
vanishes with the photon momentum. Hence it can be neglected. In
pure Yang-Mills theory we have not been able to find a proof that
the undetermined parts of the vertex are irrelevant in the IR. As
we shall argue in the following section, we can show that they are
irrelevant in the regime where anyone of the three momenta in the
three glue vertex is small relative to the other two. But, unfor-
tunately, we also need to know the vertex in the regime where all
three are comparable. So, at present, our use of the Ward identi-
ties to eliminate the vertex functions in the Dyson equation must
be looked on as an assumption.
~--
Fig. 1
320 F. ZACHAR IASEN
:: ()
(2.1)
(2.2)
- A
-
... (2.3)
\<]e define
(2.4)
so that n'N O.
ab
\fuen B= 1 and A = y, ~ becomes the free gluon propa-
( ~v
gator ~ O)ab. We shall henceforth suppress explicit colour in-
~v
dices, in the interest of simplicity.
*)
The extra term 1n (2.5) results from the constraint (2.1) •
~v - ( ~I - 'J-f. ) ,
+ (I
/'''
- Y,. ~ f" - (/- y)
~,.~
nil'
Yf- d - ~ ~" I
and
(2.6b)
8 : A :. (2.7)
+ 1 fa;.
(2.8)
. L\~'I1' (,') " 7;,~/1I (~~,,'-,.)
(2.9)
"r
('J
)., >'a..As
(1,.I 9a7 '.1) :.
.,.. ~ye.l"c. ,e r- ....d -,. ~.~j
(2.10)
RUNNING COUPLING CONSTANT IN YANG-MILLS THEORIES 323
(2.12)
with ·1
Z
A ~ Y8
so that
L3
7"
I, J :
ll'~") (I _ ~ If.,
,'\. 7~
",.
1- ~~)
",'"
(2.13)
-- i'/'~" J ~~/,)
/ .
'"
RUNNING COUPLING CONSTANT IN YANG-MILLS THEORIES 325
With the above form for IT , we have shown that the most
]JV
general expression for r consistent with the Ward identity, Bose
symmetry, and free of kinematic singularities is
- I
(2.14)
We note that the first line in Eq. (2.14) reduces to the bare
vertex if Z is replaced by unity. The second line has a struc-
ture similar to that of the vertex function in QED. The undeter-
mined parts involving F and G are all proportional to a pro-
duct (ql)Al(q2)A2(q3)A3 and hence vanish when anyone of the
three momenta vanishes. If indeed this is the relevant regime for
the infra-red, then we are led to drop the F and G terms in
(2.14). Thus (2.8) and (2.14) together yield a closed non-linear
integral equation for the function Z.
Given the ansatz (2.12), and the resulting vertex (2.14), the
Dyson equation (2.8) becomes an integral equation for Z(q2/A 2 ).
+
L
z
j f'l..'"
Lit. ~/I..) 1(1,')
(3.1)
(3.2)
To guarantee II
~v
(0) = 0, then, we must require the integral to
vanish. This means we must have
(3.4)
RUNNING COUPLING CONSTANT IN YANG-MILLS THEORIES 329
-
Z
I
OZ(J,) - ~('j.)
".{/,.-J"j
I
h L
(3.5
cont'd)
J It"-
-:;..
(3.6)
~(1- )
I -f -Y
y-' Ie L j ,~
~/4 J ~(I.' I L
RUNNING COUPLING CONSTANT IN YANG-MILLS THEORIES 331
- (3.7)
F(I(~ y/
/ ~
~(~ y)
(3.8)
for all y.
+ ...
332 F. ZACHARIASEN
and
C{y)
+
1- «.
with the same coefficient C('i). This again suggests that the
power a = I characterizes the infra-red singularity of Z, so
that as q2. -+ 0,
(3.9)
J-(Ic.}- r(J.'/
Ie. L_ 1:. "
In any event, we feel that the results obtained here are suf-
ficiently encouraging that the Dyson equation-Hard identity approach
is worth pursuing further. In particular, we need to go beyond the
small momentum expansion, to study the existence and uniqueness of
solutions. We also need to fully include the quadruple glue term,
to assure ourselves that it does not alter anything. And we need
to incorporate quarks and to clarify the connection between a
(1/q2)2 singularity and confinement.
REFERENCES
Pietro Fre
f
The corresponding action is in explicit covariant notation that
of 5 :
= ~ ~B Reo Xe.
R/\ e
j6k.'I""
-.It
ASCDe
which yields the equation of motion:
335
336 PIETRO FRE
oc.P D
E"SC.DE I'.. A XE = 0
A
In the soldered gauge where )( = (0,0,0,0,1), equation (2)
contains both the Torsion and the Curvature equation of Einstein_
-Cartan-Trautman Gravi ty6, 7. In the second theory EAt=.1. J)X
A
and the action is 1 ,2: ~
~ AS ~ D E
(3) A.z= -d/~6ktJR I\EI\E J, ~ABcDE
In the soldered gauge this theory gives back usual gravitation
without cosmological term but with a modified energy-momentum A
tensor. Topological solutions of theory 1 can be found where/(
realizes a non trivial mapping of two S sPheres 2 • As pointed
out the starting point in the construction of manifestly cova-
riant Osp(4/N) supersymmetric actions is the symplectic spinor
transcription of theories (1) and (3). It is the following:
REFERENCES
Laszlo Gutay
Purdue University
ABSTRACT
339
340 LASZLO GUTAY
8000r-------------------------------~
95 fLm thick 6E detector
B p+ Xc ..,. fragment
7000 EF >15 MeV
6000
II)
~ 5000 c
w
>
~ 4000
o N
ffiCD 3000
:E o
:J
Z 2000
1000
5 6 8 9
Fig. 1
Ep AE Ef
Ap - Af
Thus Ef Ef
dn Ap
exp (- Ep ) = exp (_
dE p kTp kTp Ap - Af ) =ex( kTf)
EVIDENCE FOR TWO BODY BREAK-UP 341
20 Q Xe TARGET
+ Kr TARGET
18 -
16
14
>(lJ
~
12
t-
10 -
o 10 20 30
Fig. 2
Jerzy Lukierski
Institute for Theoretical Physics,
Universi ty of Wroclaw,
Cybulskiego 36, Poland
343
344 JERZY LUKIERSKI
References I
E. RECAMI
Istituto di Fisica Teorica,
Universita di Catania,
Catania, Italy.
REFERENCES
1. P.Caldirola,M.Pavsic & E.Recami: Nuovo Cimento B48:205(1978);
Phys.Letters A66:9(1978); Lett.N.Cim. (in press).
2. P.Caldirola: Lett.N.Cim.23:83(l.978); Nuovo Cim.A45:549(1978);
E.Recami: Rep.INFN/AE-78/5 & 6, to appear in.: "Gentenario
Einsteiniano" vol., M.Pantaleo,ed.,Giunti-Barbera,Florence,
1979); P.Caldirola: ibidem (to appear); D.D.lvanenko: ibi-
dem; P.Castorina & E.Recami: Lett.N.Cim.15:347tl976); 22:
195(1978); R.Mignani: Lett.N.Cim.l6:6(1976).
3. E.Recami, ed.: "Tachyons, monopoles, and Related Topics",
North-Holland, Amsterdam, 1978.
4. L.Fantappie: "Opere Scelte", UMI, Bologna, 1973, p.873.
5. A.Einstein: Preface in: "50 anni di Relativita",Firenze,1955.
ENERGY -MOMENTUM DISTRIBUTION IN e+e- ANNIHILATION
T(l+8) ~- 1
1 2 2
<. KI\ ") ZL K fL:J:. (ZL) (1)
~
2 2 2
2 .''IT (M2)
dP(K.l) 1
'(iiT(l;~)
zP/2 K (ZT) , (2)
dK,!. <Kl.) .ft _1 T f!._l
2
22 r 2 (p/2)
347
348 G. PANCHERI·SRIVASTAVA AND Y. SRIVASTAVA
v3i r( 1f.!.)
zT = r(#/2) ( 3)
The above formulae are applied to obtain the K" and K di-
stributions of single hadrons under the assumption that the hadro-
nic 4-momentum distribution follows that of the radiation. Figu-
res 1 and 2 show the theoretical prediction and the experimental
result from SPEAR at 7.4 GeV 4
10,.....------------, 10.------------...,
I do"
ddi<. I do"
(GeV·') 4 dX"
0.1 0.1
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
K. (GoV) XII
Fig. 1 Fig. 2
The present approach may well be a viable alternative to the
quark fragmentation models. For gluon jets, we obtain that ~.Kl.)
from quarks and gluons is roughly the same, then (n > glue »
>} ~ n '>quark' a result in qualitative agreement with results from
DORIS on Y-resonance.
References.
1. G. Pancheri-Srivastava and Y. Srivastava, Frascati Preprint
LNF-78/46 (1978); (To be published).
2. E. Etim, G. Pancheri and B. Touschek, Nuovo Cimento 51,276
(1967); G. Pancheri, Nuovo Cimento 60, 321 (1969); G. Panche
ri-Srivastava, Phys. Letters 44B, 109 (1973). -
3. G. Pancheri-Srivastava and Y. Srivastava, Phys. Rev. D15,
2915 (1977).
4. G.Ranson, SLAC-PUB-2118 (1978).
EXPONENTIAL HASS SPECTRUM: AND PION CONDENSATION
A. Tounsi
Laboratoire de Physique Theorique et Hautes Energies,
Th e POSS1. b 1. 1 1ty
. 0 f condensat10n
. ( 1 ) at h1g
. h energy 1n
. a
relativistic quantum gas of pions with fugacity A > 1 gives
the gas a structure similar to that obtained in the statisti-
cal bootstrap model. We consider(2) a more general system of
bosons with mass spectrum of the statistical bootstrap type:
( 1)
REFERENCES
Ludwik Turko
Institute of Theoretical Physics,
University of Wroclaw.
Cybulskiego 36, 50-205 Wroclaw, Poland
351
352 LUDWIK TURKO
REFERENCES
Re£erences:
1. B. Alper et al., Nucl. Phys. 1£, 395 (1972).
2. J.V'I. Cronin et al., Phys. Rev. D11, 3105 (1975).
3. A.G. Clark et al., Phys. Lett. 74B, 267 (1978);
A.L.S. Angelis et al., CERN-report 1978.
353
CONTRIBUTORS
357
358 INDEX