Introduction ot I believe that the remarkable results which folloy froma syste- matic daily study of the buds as described in Chapter 15 of the Vortex of Life, can be seen as a discovery of a hitherto unknown Fhenomenon in the natural world. In a realm of such subtie form where everything 18 s0 nev to us, we are bound to find ourselves surrounded by unanswered question, and 8 very great deal of further research {s needed. This book {s pibl tshed with three main purposes: in view:~ firstly in the hope that {t will stimulate sane others to take up this vork practically— secondly that other readers will be able better to Judge the value of the work— and lastly to make a permanent record of the results achieved so far, uhich vill include much that has been subsequent to the publication of the Vortex of Life, hence the Inclusion of te word Sequel in the prevent ti tie. Basically the method employed Le to gather a nunber of buds each day from the species belng studied (normally 20 {8 found to be ade- quate), to find their mean Nand to count thie aa the Aof the day for that species. Doing this dafly, this allows us to See hov this species 13 varying through the weeks and months, Obviously Lf this york 1s to be in ny vay valid, the most rigorous care mist be tak— en to ensure absolute comparability In all the condi tions of working: such a the gathering of the buds, thelr mounting for photography, the measuring and calculation~ from day to day. The following chap~ ters give some indication of the general precautions which need to be taken; but each species presents Its ovn problemy,and Wiese will be dealt with In the chapters devoted to such species. Gathering ‘The task of gathering the buds {9 the most sensitive and vital one which the vorker faces. for this provides the basic data on which the research is to be founded. The need for strict consistency in one's method of working, from day to day, and from week to week can— not be too strongly enphasized. Any carelessness or thought lessness can render the whole research’ (nval ld. ‘The buds should be picked at the mane hour each day. There te seme evidence that there may also be 9 circadian rhythm at work here. It has not yet been properly investigated. If one 1s looking for evid= ence of a rhythm stretching over a matter of veeks, any interference by a datly thythm must be carefully excluded. Before starting work on a tree one should study 1t careful ly. Bude growing on its north aspect may have \ slightly, but consistently, different from those on te southern. Dud size ord maturity may very from one aspect and one branch to another. In some species one must, distinguish between terminal buds, srowing at the tips of their stem and lateral ones growing out of the side of the stem. It mant seem beat to be very restricted in one's cholce,— to confine oneself, let us say,to only the largest terminal buds growing on one particularchosen branch. But this means imposing a severe restriction on the nunber of buds avaliable. However, by confining oneself to four or five chosen branches, picking 8 fixed nunber of buds from each per day, vith a fixed ratio of terminal to lateral, keeping as far as may be @ constant mean size of bud for each daily set, {t is feasible to maintain strict comparability, while using @ wide range of buds from different parts of the tree. ‘A large and mature tree will be able to yleld 1500 buds or sometimes even a fev more, In the course of a season, without showing efgno of subsequent Impover {shment and the graph which follows fron this Will be improsaive-looking. But often, If the tree ts smaller, or many of ita buds are Inaccassible, we have to be content with mich leoa. These shorter graphs should not be despised, Trey have been made vith the sane care and give valuable informaticn, specially for determination of Une phase-snirt, Atter I have used @ tree heavily for one season | do not return to it for several years. Ve mst be careful not to injure those organ iemg which ve are studying. When ve come to tne flower buds ve meet somewhat similar problens with regard to comparability, One of the difficulties here 3 the shortness of the budding seascn. Ve need to start work Just as early as possible, at a time when buds are few to find, and ve have to go far afield to find as many as we need. But buds growing in a sneit— ered situation cannot be compared directly vith those on an exposed Hillside. However If we are careful to take exactly the same pro- Portion each day from each 5! tuation, conparabi Ii ty can be retained. In Chapter 7 of The Vortex of Life a description 1s given of the considerable vartations in A which the flover buds undergo during the course of thelr develooment and opening. Once this is realised It becomes cbvious that such work with the flover buds will not be valid unless sone vay 1s found of gathering the buds always at the same exact stage of thelr development, week by week. With great many species this is a matter of considerable difficulty, but a few grow In such a way as to make It comparatively easy, These are the ones In which the buds start their life totally enclosed in their green sepals, but during the course of their develcoment at a certain stage thrust throush to the open air. The point at which the coloured tip of the bud first Just appears 19 an easily Identiflaole moment of de- velopment, and this is the moment at which they should be picked. Avoid picking any buds which are growing near to the hish tension cables, elther overhead or underground, or near to any other pos~ sible source of an electric or magnetic field. unless you are spe~ cifically aiming to investigate the effects of such a field.Measurement and Calculation a3 ‘Ag s0cn as they have been gathered the buds begin to lose moteture, very slowly to begin with, and at an increasing speed later, and in the course of this they gradually change their shape. It 13 there~ fore necessary to photograph them imediately after they have been picked, For measuring they are laid out, side by side, on a glass measuring plate. | place them Iishtly on a strip of transparent sticky tape. lying sticky side up, #0 they do not roll about while being carried to and from the camera. There Is probably no one single 'right" way: of laying them out, but whichever way 1s adopted {t must be achered to strictly throughout any series of observations, In order to ensure absolute comaradi lity. | divide my tventy buds into two sets of ten, the larger buds going into the first set, and the smaller ones into the second. The mean \ for each set is found, and the \ for the day {9 taken ag the mean of these tyo means. WorkIng this wey one {8 con~ stantly avare of how mich the size (le. the maturity) of the bud affects {ts \s species, and even individual trees, vary much in this respect. If the correlation between bud-size and Ais at all marked it 19 a warning to use extra special care to see that the mean size of each datly set of 20 remains absolutely constant throughout the Once @ set 1s photographed each bud 19 11fted off the plate, rotat~ ed through 90 dearees about. its long axis, replaced én the plate, and the whole set {s re-photographed. Thus a set of ten buds gives Us twenty pictures to analyae, and. the fimire for each day 18 the mean of no lesa thar forty mets of messurenets. Each bud 1s, a5 tt were, sem from tvo directions Of space. The protographs next: have Yo be messured. The methods for this and the subsequent. calcula Hons, are described in Chapter § and Appendix 3 of The Vortex of Life. ‘They Involve measuring the diameter of the bud at a series of equai ly-spaced level along the long axis of the bud. These calcula tions are tine consuming and tedious, and these days {tte better to uss'a comuter. Ona fol loving poe ve sive a simple basic program Ua tinien does “thin calculation for us. LAM Genands as input First the number of levels which we have decided to measure (7 18 usually found to be adequate), and mfter. that ve have to input ona By one the measured dianeters’ starting with the loveat one (Ive.nearest, the base ) upwards. The computer then dellvers to us, )\ ,a9 calculated by four different methoss described In Arpendix 3. [f‘the form which ve are measuring 19.8 perfect path curve these four answers will of course be identical. But vhen working with living nature there will be small di frecesces. Te relative virtues of these four methods are Giscussed tn thet appendix. Apart from the morphological method. tt Seems that there i not mich to choome between them, but of course whichever method 19 chosen must. be adhered to throughout any given Series of cbservat ions. in'my oun work | normal ly use the project ive method, more for aesthetic reasons than any others It goes In a very Girect ‘vay stralaht to the heart of the whole processes we are using.a4 At this point we are faced with two alternatives. Ve nay proceed by ‘the method known as ML (the mean of the lamdas). For this ve calcu- late \separately for each bud and find the mean of these 40 values, each bud having of course bean photoaraphed twice. Or ve may use that knovn as LH (the landa of the means). In this case ve find the mean value for the diameter at each level. giving us the dimensions Of what ve may call the mean bud of the whole set, and ve find the ‘of this. These tuo methods are both valid, and {t does not very much matter which one we use, but of course once a choice has been made 1t mist be adhered to for the whole of any one eet of observa tions. It 1s found by experience that the values given by LM are nor~ aily very alightly below thoes given by ML, and that from day to day and week to week they tend to be rather more consistent, {.e. less erratic; and it le for thie latter reason that | always use LM in my york. It 1s possible, simply and at very small expense, to construct a cam- era unicn makes the photography speedy and economical. dispensing altogether vith the use of negative material. it is illustrated at the right of this colum. L tsa light en~ Closed {na box with a translucent screen, G, below It. 5 isa slot into which the giass measuring plate on which buds are laid out, will fit. Hig a leney a normal camera lens works very well for this Job. P ts a plate on uhich the photographic pa— per vill be laid. An image of the bud 1s Ehrown cn to the paper and ts recorded on it. The lens,H, needs to be moveable, up and down, for ine purposes of focussing. ond the piate,P, similarly for the purpose of getting the right etze of image. The camera in my darkroom {g about seven feet tall, and vith thig ene can get, within reason, an image of any size one wishes, The image, unich 1s a unite sf lhoustte on a dark ground, Is crisp and clear and is Ferfectly adapted for precise measurement . Apart from the light within its llght box, none of the apparatus needs to be bowed In} one simply works in a darkened room, The taking of the photographs to easy and very quick. The time consuming part {9 when one comes to measure them afteruacde! Al} my work up to the summer of 1987 was cone on this machine which Is why up to that date most of 1t vas done with only ten buds per species (i.e. twenty pictures) per day. This proves just suffictent for sianiti- cant reeults,but twice that amount fe much to be desired. a L g s HaS tes the necessity for photography. ‘This renarkable machine not only does al one's ’calcurat tote Be, 2120 makes the measurements on which they are based, It cerelony essentialy of a computer with a wide range of memon'y an erent eet tor sercen and a T:V. camera. Unen the buds, munted on tate gees measuring plate, are put into the macnine, an enlarsed Inase off then gepeers on the monitor screen, ‘and from thers a massage ta sent te the computer giving, firstly, the x-y co-ordinates (as seen on ths monitor screen) of ine tip of the bud, then of the bose of the bude and thereatter of the left and right-hand ends of each diameter, yorking fron the base end upvards, This data wows Inte’ freer frie for each bud, numbered from 1 to 20. The prostem;Ut, printed heres after, then goes from t11@ to ttle picking up ail the cata, converte co-ordinate pairs into distances (1.0. levels uards fren tre base to the tip, and dleneters at each of these levels), finds from these the Glmensions of the meen bud of the whole set, and delivers 1s As calculated according to the original projective method sr aloo Bccording to the regression method describad in oppendin 3, When the nachine arrived it was found to give results satisfactorily consistent with those which were already being achieved by the moot careful hand measurement and of equal reliability and accuracy. But where the machine real ly scores 15 In ite speed of vorking. By using {t I vas able to do between four to five times as much work ine set time 9a previously. After this nearly all the work haa been done with tventy buds per day (1.e. forty pictures) instead of the previous ten, thus giving sreatly Increased rellability, whilst covering twice ao many species. Program, L, also printed here, works from the same data but delivers the A for each separate bud.’ | should Tike to record here ny very great thanks to the Mergaret Wilkinson Research Fund for the great help to the work tnat thie has meant, In the Autumn of 1981 the Fund bought another such machine, and this 1s now In use in the south of England, Just about the same time my Erlend. Graham Calderwood, of Campni I! Aberdeen, completed the ec struction of another similar machine, ‘with which he has since been Going most useful work. So since the beginning of 1992 we can hepe: to have results from at least three independent observers separated by five hundred miles of countryside.a6 The Graphs ‘As far as 1t has been found possible and convenient all the graphs have been drawn to the same scale in order to make comparison bet— ween different years and different species easier; but necessarily there have been a few exceptions to this. Tyo curves appear on each graph. One 19 made of thin lines joining those points which represent the actual daily observations as they they vere made. The other, thicker, curve joins those points vhich have been calculated by the method of three—day-means. We call thie the ‘trend curve’ because It Irons out the accidental variations which are alvays lable to appear in observations, in this way, but shows us the true general trend of the way that A ts varying: ‘The usual vay of calculating thtes-day-mans 1s to take the sum of yesterday's, to-day's and to-morrow's observations and to divide it by 3. But In order to be really fair one fecla thet extre stress should be given to to-day’s result, over and above those of yesterday and to-morrow. So the mathed enployed here 1s te take once yester— day's result, twice today’s and once tomorrow's and to divide the um by 4. This gives results which are very clove indeed to those of the more ordinary method but, one feels, a somevhat truer picture. It {9 Insteuctive to watch the difference between these two curves: Sometimes the thin line l1es closely along the trend curve, indicat: ing that the tree or species was growing consistently tharmoniously one might say?) at that time; but at other times 1t may oscillate wildly ebove and below the trend curve giving us the feeling that some kind of disturbing factor has entered into the growth of the plants we are studying. On eech graph arrovs are planted showing us where dips (downward— pointing) and peaks (upyard-point ing) might be expected in the curve according Uo past expertence. The position of each arrow is calcul ated according to the moment of the lunar alignment with that planet which has proved to have connection with the species in question, Adjoined to each arrow 1s a number which shows the phase-shift at vnicn this arrow nas been placed. Tnis phase-snirt is taken from the phase-shift chart, printed at the end of this section, which shovs the average phase-shift of ail the species being studied, at this particular moment. The arrow 1s simaly there to auide one's eye. To determine the actual phase-shift of the specimen being studied one mist notice the position of the dip or peak in the trend curve in relation to the position of the arrow. If it co-incides with the arrow then of course the phase-shift 19 the same as the nunber with which that arrow 19 labelled.A Statistical Consideration a7 It Ls important. to bear in mind that nearly all the figures involy- ed In vork of tnle kind are mean numbers, and by the very nature: bf the case such mean numbers can never be considered to have very exact values. In the Course of tne days and weeks, due to the rendon collection of the buds there are bound to be accidental rises and falls in our curve, which are not sienificent for the life of the tree. If ve believe that we have found such a significant change, it {5 important to be able to know whether this 19 larger than any pro~ bable accidental change which may have been registered. To do this we must keep a check on the general level of the standard deviation of our reaults. This standard deviation gives us 8 measure of how far the numbers arrived at by measuring our forty pictures Of buds spread on either side of their mean value. Coviously if the Spread 1s very large then the resulting man value will be less re- Table and the chances of accidental change will be all the greater. it can be shoun statistically that {n any such set of numbers, we can expect to find 95.4% of then crowded within the space of 2 stan~ dard deviations plus and minus of their mean value. But ve are not concerned here, directly. vith the variation of the actual A-values of our individual buds, but rather with the varia~ tion of their mean values from day to day and week to week. Clearly ve need a way to find the effective standard deviation of our mean Values; and equally clearly, the greater the nunber of buds ve are dealing with each day, the more reliable these means are likely to be. Statistically 1t Con be shown that {f 9 1s the standard devis— tion of @ set of n buds then thelr mean has a standard deviation of 9 divided by the square root of n. We consider now the trend curve on our grapha. The accidental varta— tions on this curve have been smoothed out by Joining points which have been areived at by taking the meen of three days’ observations tie. 120 bud pictures. Their standard deviation will therefore be found by dividing the standard deviation of the buds tnemeelves by the square root of 120,- for practical purposes, by 11. If ve double this ve find the spread plus and minug of the ‘central mean value, within which 95.4% of our points ought to Ife by probability. Any point lying outside this zone does 50 vith 95.4% probability of oig~ nificance rather than accidentally. That ts to say that the points lying outside this zone probably number abcut 4.6% of tne total But In our work ve are normally only concerned with variation in one direction. If on any particular day we are concemed to know whether 3 Significant dip has cccured in the trend curve ve are interested Only in those points which fall below the limit set by the tvo-stand~ ard-deviation threshold, {.e. In 2.3% of the total. Therefore ve can say that any point lying below this threshold does so with a propabi— lity of at least 97.7% And this, as ve shall see later, gives us a good measure of the significance of our results.a8 The Phase-shift The strange and unexpected phenomencn of the changing phase-shift and the method of 1ts char ting have been aeseribed in chapter 15 Of The Vortex of Life, and full deta!ls will not be repeated here. We should renemcer that the rhythn of the plants has shown itself, over the years, to be getting slowly and consistently out of step with that of the planetary alignments, the dips in the \-curves coming earlier than the corresponding al janments. hen the bud is resoond— ing 2 days early we say that ft has a phase-shift of “21 clearly bY the tine that the phase-shift has increased to -14 the rhvthns are ‘on tine’ again and the phase-shift has returned to zero; 2 dava be fore that the buds vill have been responding 2 davs late, and the Bhase-shift will have been +2, while half way in between these two times there will have been a moment vhen the phase-shift vas #7. Oving to the fact that the periods between allgneents vary for diff erent planets from about thirteen and a half to neacly fi éteen dave the matter {5 more comp! {cated than thig and as the work progresses: more sophisticated methods of presentation will no doubt be needed: but by ensuring that the numerical value of the phase-shift never Increases beyond 7 any inconsistencies are reduced to a minimum and the present simple method of notation can be a valuable tool for mapping the general form of the phenomenon. On page a10 we print the phase-shift chart made to the latest date. Each point shows the phase-shift for a whole species in the parti— cular season where It fo marked. Owing to the fact that a phase shift of -2 1s of very nearly the sare significance as one of +12, etc. each point, and of course the whole curve, appsara in a series of positions along the vertical axis of the chart, out only the cne. in the middie needs to be conaldered. It will be @ long time before ail the complications of this phenomenon are understcod but a few things can be vald Immediately: 1) The curve appears to be repetitive,t.e. cyclic. 2) The gradient is never positive, and almost all tne time nega tive,1.e. the earthly part of the phenomenon is nearly all the time running a I1Utle ahead of tne celestial part. 5) It appears to have a periodicity of about seven years. 4) Tre phenomenon seem to cut rignt across the differene= between one species and another, and between one planet and another, e.g. Une point representing Mars and the mianty dak I1es on the same curve as that for the Sun and the delicate Primrose. 5) The latest observations seem to Indicate that the symmetry 13 ot exact, the curve running dovnhill a little less steeply between the beginning of 1983 and the end of 1984, than between the beginning of 1990 and the end of 1991. The curve appears to Nave been horizontal about December 1982 ‘and again in December 1989 indicating 3 period of seven years, but whereas the phas=— shift nad become about ~3 by January 1985 {t was fully —4 oy January 1992 giving a period of only six and a half years. We must vatcn for the possibility that althougn this 1s basically Seven year periodicity it may vary a Little over and under that figure in the course of long periods of time.ad Consideration of no 4 above which shows that planetary bodies as different from one another as Saturn, Mars and the Sun all seem to be treated alike by the phase-shift phenomenon has led to a doubt being expressed as to whether it will be possible to find a true celestial correlation for it, and to the suagestion that it might possibly be an effect of terrestial origin vnich is suerimoosed on the astronomical one. Gne of the next important steps to take is to test this possibility. If this should prove to be true it vould mean making a fresh phase-shift chart for each locality on the surface of the earth. For this reason, the phase-shift chart printed here con~ tains only points arising from observations mace at Strontian, and, until further evidence accrues, must be judged aa valid only for that locality, Except those label led otherwise, every graph in this book was made near Strontian, and in all these the arrows mark the current phase~ shift of the generality of buds at that time, as taken from this phase-shift chart, They are there in order to assist the reader's eye in judging quickly and easily just how accurately the species under consideration was conforming to the general behaviour. Clearly this could not be so for the graphs mace at other places since no phase-shift charts have yet been made for them. Therefore in these other graphs the arrows have been put in, and labelled, at places which seen most appropriate to the behaviour of that graph. One dav, If {t {8 found to be needed, these arrows may assist in making 3 phase-shift chart for that locality: ‘The chart printed on page a10 Is central to everything in this book and all the graphs need to be carefully studied in relation to it.PHASE-SHIFT CHART 14 ae of Tag 3 oho Faq é ®. hoo , om Pe Bag Fon woez | wes | soe | 1985 | r9e6 | ser | teee | tae | 1990 | s90te! 1902Beech a The Beech proviciem buds Uhich are some of the most frultful for this kind OF vork. A'mature tree wil bear. Titsraliye Milt tons of bude."and many” of thase are Usually easily. seceseibie., Trey are Nell ‘sdepted for accirale nescurerent ard they turn eut tebe good ath curve. forme, wlth $0e Unusl ly, Getceer 14 ond 3 enc thet A-values have a satisfactorily low standard deviation to give accu- Gola tant Deller lisse alansidoe east eecincane cumin Cait ae Selachteved uniesa very great care Ta observed In their col lection ond herding A cursory examination of the tree reveals uhat, at first sight, seem to be both terminal and lateral buds. However If one returns to the sane tree a feu monthe later one finds that many of the bude vhich one had supposed to be lateral are now clearly terminal, growing out of their oun little side stems. It ts probably correct to say that tn essence al beech buds are terminal but that at any particular mo- ment some of then are not mature enough to show themselves {n this form, Never theless these 1e9s mature buds are growing in the lateral situation and are subject ta the restraints in form vhich are due to their closeness to the main stem, and it is not fair to compare them directly with the ordinary terminal buds. Therefore in this text ve shall continue to refer to then ag either ‘terminal’ or ‘lateral’. It te usual to find that the lateral bude give rather higher) than the terminal ones. It Is therefore essential, after an initial exam- ination of the tree. to decide on a fixed ratio of terminal to lat~ eral which one will pick each day, and to stick to this rigidly all through any one series of measurements ‘The next point at uhich great care must be exercised 1s in the cut- ting ef the buds fron their aten. Nost terminal buds viIl be found to have a small asymnetrt— cal ly placed notch near their base,and this 1s a useful place at which each bud can be cut from its stem. If on an occasional tree this notch 1s not @as! ly seen, then one can usually find 8 faint horizontal ring round the base of the bud where the cut can he mada. Tha lateral buds usually do not show these things, and tn their case the cut. should be made just’ at that point where the bud emerges from the main stem. ‘The buds are formed of counties exceedingly thin scales. placed one over the other. In spt val formation. They are so thin that {f one of them is carefully removed it does not make any sianificant difference to the form of the bud. However in the course of cutting and mounting it occasional ly happens that one of these scales becomes loosened forming a bulge in the side of the bud, and {t can then serfously deform the shape of the bud. In the course of the work one must alvays watch out for this, and renove the offend-b2 ing scale before proceading further with the process of measure- ment Another feature of these buds vhich has to be paid careful attent- ton to is thelr curvature, Close examination of these buds 1s, 1 think, enough to convince one that their essential nature is to be straight. Nevertheless, oving to certain asymmetries in the process Of thelr growth {for instance the light falls more strongly on them from above than fron below! they do often tend to exhibit a silant degree of curvature along their long axis. On most trees thig cur— Vature is very slight Indeed, but it 1s there, and since it affects Yeele h, ft le important that all buds should be mounted in exactly the same vay with respect to this feature. The surest way to do this te to lay each bud down on the glass measuring plate with {ts curve facing directly away from the camera. They will thus be seen on the photo or the noni tor screen as though they are completely straight — {n wnat | call their strateht aspect. Then when they are lifted and turned through S0° they are geen in ther maximum curved aspect and the mean between these two 1s the figure accepted. This ts a more Consistent way to deal vith them than to have them {n various inter— mediate espects. In general a curved aspect. tends to give a slightly Righer A than the straight. | have the definite impression that the curvature of the buds 1g Increased when the tree ls growing under the high tension cables but thls 1s something which haz not been subject to exact research yet. Avery important feature to deal with is the size. i.e. the length of the buds. On the ame bough one will find larger end smo! ler buds growing side by side. By choosing with the greatest care the sane Bind of buds from the same parts of each branch, each day, one can enoure a good degree of uniform! ty, but for really consistent results one needs to do more than this. My method 1s to choose my 20 buds 92 Garefully as possible, and then to take two extra ones, one larger than average and the other smaller. When the 20 buds are laid cut on the measuring plate, they are first measured for length, and the mean Jength of the Set calculated, If this mean length comes greater than the standard lenath which | have already adopted for this particular tree then the smaller of the extra buds 1s substituted for one of the larger ones {n the set, but if the value cones out too small’. then the larger extra one {3 substituted for one of the smaiter ones. In this way one 1g able to keep the mean length of each dally set cons— tant to within one fifth part of amillimetre. Only then is the meas- Urenent for A proceaded with. This extra precaution definitely adds Yo the consistency of one’s results. In general cne usually finds tne smaller buds ylelding rather higher figures than the larger. If all these precautions are careful ly fol loved the Beech buds will give us excel lently consistent results, usual ly with a standard do~ Ginech Se meet vedos, Nene oe should consider the individual graphs. Graph cl. Although in the tuo previous years I had acquired some evi~ Genes linking Satuen with the Beech this was the first time that, | had been able to do a really long series on this species, and thisbs marked something of a turning point in the work, as this vas ficst time that the phase-shitt was 50 strongly’ shaun as tenrsree itself on my attention. The arrove on this graph are shoving the actual dates of the alignments, i.e. at phase-snift zero, because at that time | uss not expecting the dips in the curve to cone a ary other tines. Hovever 8 glance at the graph shows that the dip which I had been expecting on October 12th had come about one and a half days early, and the following dip fully two days before ex- pected. In fact during the whole of that Autum the phase—shift Veried between minus one and a half, and minus tvo, days. With hind~ Sight now ve can see that this graph vas already demonstrating the Varying speed of change of the phase-shift, al thoush | did not be- come avare of It at the time. Gy the end of February and in early March we see that the dips are coming tvo and a half and finally fully three days ahead of time thus showing a rate af change in the phase-shift of about a day in six months, a rate vhicn has since proved to be fairly general in the middié part of the phase-shift chart, Another point of interest that winter vas the general form of the graph. We see a strong variation during the Autunn getting gradually weaker as midwinter approaches. From the middle of December unt! the first week of February, apart from one dip about January 13th, the fortnightly variation has almost completely cessed. Then in tne Spring months the variation resumes, quite strongly, although pec— haps not quite as strongly as in the previous Autumn. Now if one con— sults the graphs vhich | had arrived at two years earlier for the Oak ‘and the Cherry one sees almost exactly the same general form portray— ed. These sinilarities were so striking that this led me to believe that this is a quite general form for the mid-vinter months,~ so mich so that I was led to give the phenomenon a name,-The Midwinter Sleep. Since then. as vill appear in the following pages, a number ©f Beech trees have been found which have contrived to retain their variation claht through the vinter, with considerable strength, al— though even with those I think the’ampl {tude of the variation tends to get smaller round about the turn of the year. So | think that ve must say that although the general form of Diagram cl nay be fairly general !t {s by no means universal ly so. ‘Any doubts as to vhat the phase-shi ft vas doing vere resolved sone seven months later with the making of Graph cz. Here ve see a firm fortnightly rhythm with a clear phase-shift of minus four days: in fact the dip of October 24th comes at nearer minus five. This means change In phase-shift of fully a day, or maybe even a ii ttle more, in the course of the preceding six months. One year following, graph c4, proved to be one of the most unsatis~ factory ones that | have made. Starting very early in the season the buds were inmature and still groving towards their final size and form; and I had no means of knowing vhether work with then in this state would be at all valid. My circumstances during that period Prevented me from working 38 regularly as I would have wished. The record 1s broken and rather thin, At the end of six weeks work allba | could honestly say was that the results could be considered con— sistent with a tortnightly rhythm of dips but could certainly not. be brought forward as strong evidence to prove It. Howaver, exper— lence of several further years Of work has shovn that nia graph was in fact foreshadowing future tendencies with remarkable truth. Compare this grapn with graphs cB, cll, cl4, cl6 and Gl? amd one finds five independent trees In four different years all behaving in almost exactly the same way as the tree of c4. Wnen one starts so early in the season, \ starts high, and then undergoes a pro- gressive period of decrease, punctuated by fortnightly dips corre— lated with the Moon/Saturn alignments. Examining these graphs with care one Tings that they donot significantly differ from one anoth— er except in the phase-shift uhich, of course, varies from year to year. Of particular interest 1s a comparison of cl6 and cl7; these tuo trees, groving 150 niles apart, yielded graphs of almost ideni— ical torm, the chief difference being that the Aberdesn tree gave amore lively performance, varying with greater amplitude. ‘A somewhat. unusual phenomenon 1s shown by graph c7. This was at the time when the electronic scanner had Just arrived and for the first time | felt able to observe two trees of the same species side by Side. The first of these yielded graph c6, an absolutely typical result for the very early weeks of the Beech season. The other, C7. growing @ fev hundred yards avay, in, as far as one could see, id~ entical conditions, started identically, showing a vell-marked dip on Septenber 16th.’ However, from the last week of Septenber this variation went into abeyance. The next two alignments are not ac: knowledged at all; but just beyond the middie of Cctober, after 8 short period of som: agitation, the next dip, of October 29th, is strongly shown. A somewhat similar thing had happened the previous Spring,~ graph cS, It started with a well-marked dip on January 18/19th, and ano— tuner. not so wel I~marked, on January 3ist. The folloving rise failed to take place satisfactor! ly and after a period of some agitation in the first week of February, the next dip failed to materialise con pletely. | had to be avay from home after this, but on my return was able to resume work on this tree. Doing so | found, vith relief that the next dip, of March 12th, vas clearly shown. Since this rhythm had thus been called in question I selected yet another tree for cb— servation at this time, just to make sure, The results trom this are are shown in the top part of the righthand section of cS. It had a sonevhat higher A, but otherwise was behaving identical ly. An even more striking example happened the fol loving year, shown in graph 10. Here 1 was fol loving two Beech trees, one down by the val— ley and the other on the hill, over a considerable period of time. The first 19 shown in the upper part of clO. Notice that the first dip, of Cctober 2nd, ts distinctly higher than any of the others. The usual period of decrease at the beginning of the season was just. coming to an end, Thereafter the next three dips in the curve cone impeceably, but then, again after a moment of sone agitation, about November 20th, the usual rhythm ceases. The second tree shows the ehythm complete and unbroken right through to the end of the series.bs This phenomenon of 3 tree which follows the planetary rhythms vith quite strict accuracy for a considerable period and then. for no ascertainable reason, ceases to do so is one that I have only found With the Beech, and even there but rarely.- in fact for three times in nearly ten years work, The fact that in two of these cases | was fol loving a second tree !n paral lel. and that this further tree con ‘inued vith the planetary rhythm unbroken, gives ne to bel {eve that the cause of this phenomenon i@ not celestial. 1 think we mist real~ {se that ve are dealing with subtle and very delicate reactions here and there may be many Interfering factors, both terrestial and celes- tial vhich have yet to be explored. Another fact of considerable importance 19 shown by graph c9. In some previous yeers (see for inetance the Stitchvort and Geranium? evi- ence had been found to show that when Hars and Saturn are aligned with one nother they have a mutually inhibiting effect, but {t was fot until the Spring of 1988 that there was a chance to test this With the Geach. Consulting the phase-shift chart ue find that the Shift for the generality of species in that spring was plus three and a half daye. In the normal way therefore ue might have expected 3 dip in the curve abcut February Ist to 2nd. Hovever Mars was al~ Teady approaching Saturn; there was In fact a dip but It came two days late and showed only feebly. 4 fortnight later, when the two planets were only three ond 9 helf degrees apart no significant dip was recorded at all, Much stronger evidence vas to come later. Graph ci2 shows the result of 8 lone series of observations almost throughout 3 whole winter. By calculating 7-day means one can smooth out all the local varia~ tion, and this hao been done to produce the central continuous curve: showing a very slov rise and fall throughout the whole period. above and belov this, ond parallel with 11, are dotted curves showing the tvo-standaré-deviation limits for these buds. On any particular day, adip coming below the lower curve has a probability greater than 97.7% of statistical significance. These limits are calculated for tne heavy, trend, curve, not for the lighter one shoving the actual daily observations. We see that between October 19th and the first Week of February there were, during the periods when observations: Were being made, eight occasions when dips might have been expected, nd on every one of tnese a dip was actual ly recorded, no fever than Seven of them reaching down to or belov the two-standard-deviation limit. It 1s interesting to note that the two dips which fel) to B well below thts limit both occur during the time which we nave de— Signated as that of "winter sleep". BUL on Februery 20th Saturn and Mars came into conjuncticn. The nufbers on the graph show the dist— ance Separating these two bodies; and we see that from the moment when they are less than about five degrees apart no more dips are to found, Indeed almost all statistically significant vectation ceases. This f9 not to say that such variation as still seems to show is not real, but that 1t 1s so small that, with the frequency of obverve- {ons which were being made, 1t could not be statistical ly verified. Had this been all that I nad to go upon | would nave been incl inedbe to consider the apparentiy residual variation in the graph as just Tnoise? without any real significance. Hovever | vae at that time running another Beech tree sin{larly, in parallel with this one. The result of this is sean in graph cl3, and it te very instructive to compare this with graph cl2, Again ve see that of the eight ai— fanments unich fall within the perieds of cbservation between the middie of October and mid-February every single one 1s marked with 3 dip in the curve and again ne less than seven of these dips fal! weil belov the lover two-standard-deviation limit, the only excep- tion being the cip of Nov ard. And again, as with graph ci2, from the middie of February cnwards there are no further dips co-inciding with the ai fanmente of Moon ang Saturn, Furthermore it i@ noticeable that what variation remains stays for the most part well within the tuo-standard-deviation limita; little of 1t can real ly be seid to ce statistical ly stsnificant. Nevertheless the sureriging thing 1s that these tuo cases of residual variation ere quite remarkably sim lac, so much 30 that one 1s driven to ask unether there 1s perhaps sone sianificance here. We see dipe,conmon to the tye grapha, round about February 23rd and March 15th, ‘and peaks on March Sth and 18th. Ine dips are separated by twenty cays and even if they are significant they cannot possibly have any lunar correlation. But the peaks are just thirteen to fourteen days apart and our attention therefore is bound to fall upon them. It happens that they coincide closely with @lianments of Moon and Jupiter. One has to ask whether, ence the in- Eluence of Saturn hss been annulled, it would de possible to find another influences, from Jupiter, working softly in the pack@round. Well, of course {t would! But just two cases like this are certainly not enough to establish o fact. It 1 a possibility that has to be heid in mind for future consideration only. Just a year later we have a set of observations vhich gives a grach, C15, which le at first glance very similar to the two we nave Just been studying. The curve for the 7-day-neans shows 3 gradual rise and fall. The first three altgrmants are strongly marked by dips and the fourth rather less markedly. Thereafter all significant varia~ tion again ceases. This tine the interfering Influence seams to oe coming from Jupiter. In the last week of February Jupiter moved to within five degrees of alignment with Saturn and thereafter. for the remainder of the graph, stayed closely in oppositicn. it would seem that _we have to enunciate the rule that saturn work in thia vay on the Beech only when unencunbered by alignments with other heavenly bodies, In the winter and spring of 1692 ue had three cbservers working on this, Ada Ruth Dogger near Stroud In the south west of England, Gra— han Calderwood In Aberdeen cone 500 miles to the north, and mrscif in Strontian, some 140 miles west of him, and also S00 miles north of Stroud. The preliminary results of dual working con be clearly geen in grapes ci6 and cl7. As has been already described on page b4 the very early covelopmont of the Beech buds had been followed in @ vhole number of previous years and had been found to run in quite 2 definite pattern. It te pictured in graphs c4, 6, cll and cid. The earliest measurements shov comparatively high values for A,b7 particularly at the beginning of Septesber, and then, during the Succesding veeks this value undergoes a progressive decrease unt! | about the middie of October vhen 1t reaches a more steady value, this decrease being punctuated by the usual fortnightly dips in the curve correlating with the alignments of Moon and Saturn. The Autumn of 1991 vas no exception In this respect and examination of Graph C16 reveals a result absolutely typical of past years. It is unfor— tunste that 11 Inese prevented me from getting actual measurements for the last dip (Cctober 13th), but the results for October lath, 15th and 16th were enough to assure mo that the dip had really sc~ cured. Having had results similar to this in a number of past years it wos a matter of very great interest to see what would eventuate from a different environment and a different observer. Grach cl7 shows the result. Although at first sight 1t appears to be rather different, careful examination shows that this Aberdeen tree was in foct behaving very similarly to iis sister tree growing 14) miles to the wast,— 3 long progressive decrease punctuated by fortnightly dips which were almpst simu! taneous vith those at Strontian, AN UA expected difference was that the Aberdeen tree vas mre ‘lively’, the ampli tuce of the variation being much greater. The phase-shitt for both places seemed to be the sane, at about -3 days Later the same year ve have, on erach ci8, results for two Beeches: in Strontian. Of the five dips shown, two, (on November 6th) come at a phase-shift of 4 days and those of Uscenber 20th vould seem to be at about minus three and = half days, and this final result should probably be taken as correct for the end of that year, These results can now be compared with both Aberdeen and Stroud, on graph C19, The dip of December 6th does not show on the Aberdeen araph Dut is strongly marked on the Stroud one, although one day later than at Strontian, The dip of December 20th 1s clearly shown by all four trees. Coming now to graph c20 we have a long series of observations on a Strontian beech. This graph Is remarkably similar to those of some Previous years; conpare it with those of rages cl2, cl3 and cl5. There 18 a long slow increase in}, followed by the beginning of 8 decline tovards the end of the season. | have marked, as before, the mean value obtained by taking 11-day means, and above and below {t, In dotted curves the two-standard-deviation limits of 97% can— fidence. Ve see the curve coming well below the lover limit at each of the first three alignments and at no other time. On February 27th Mars approached to within 5 of Saturn, and the conjunction was com plete on March 6th. We see the curve apparently approaching another dip round about February 25th, falling to do 96 and thereafter cess— {ng to show any further variation outside the confidence limits. It 1s of great interest to compare this vith the results for 8 Beech taken In Aberdeen at the same time, shown in graph c21. Again we see. 8 gradual rise from the beginning of the year. culminating about the middle of February. Again ve see the fortnightly dips, al though here the mean phase-shift would appear to be nearer -5 daye than the —4 of araph c20. Again, although this vas a different tree from thatbs of grach cl7 it also shows the mark of greater liveliness, the am piituce of variation being nearly twice as large as at Strontian. This should alsc be conpared vith graph c22, of tko further trees taken at the same time In Aberdeen. It vas not possible to cbtain such an unbroken record of these tuo trees but apart from the dip of January 4th/Sth, which for some reason got displaced from agree— ment with all the others, this graph confirms in all respects the qualities seen in graph c21. The fact that thie wider amplitude of variation appears in al! of four separate cases promts the question whether this quality ought not to be associated vith the indivicual trees thenselves, but rather vith the locality in which they were srowings and this point mist be born in mind with further research, At this point we should remind ourselves that vhereas in all the Strontian graphs the arrows mark the phase-shift for the general ity Of buds at that time, taken from the phase-shitt chart, with these Aberdeen graphs it {s different, The arrows are marked in at the points which seem most appropriate to the coservations unger con— sideration, and these may one day form part of a basis for making @ phase-shift chart appropriate for the neignbdournood of Aberdeen. Such a chart, !t 1s clear, vould bear points at minus three to minus three and a half days for the Autum of 1991 and -5 days for the Spring of 1992, ‘Thus ve see that the correspondence between the measurenents made in Aberdeen and in Strontian has so far been remarkably good, and the difference in phase~shift lalmost negligible in Autumn 1991 and at the very most one day in Spring 1992) 18 not great enouah to be cer— tainly of significance. Hovever if this small difference continues consistentiy in future years 1t will have to be taken Into account. With the observations made at Stroud (the lower part of graph ci9, and graphs c23, c24 and c25) the situation 1s not quite so clear. During the pericd from the end of November 1991 to the end of Feb~ ruary 1992 when Hars moved into conjunction ui th Saturn. we have ten allanments to consider for these three trees and {f we adoot a fig- Ure of ~2 dave as thelr phase-shift at this time, no less than elant of these alignments are represented by dips in the curve with a mod- erate degree of regularity: arrows have therefore been marked into the graphs at these points. Hovever the curves show many other dips at points which cannot possibly correspond te any allannents. Indeed the overall impression one gets from these graphs is not of @ ld~day but rather of = 7-dsy rhythm, the internediate dips corresponding fairly regularly,with the square aspects of Moon and Saturn. i.e. when they are 96° apart. Against thie we have to eet the fact that in ten years of work at Strontian, and also recently at Aberdeen, such a phenonench has not bess observable. It sesns that at this early stage ve can only return an open verdict~ not proven: but that If It eventual ly proves significant to Interpret these resulte in this way then we must attribute to Stroud, for the Spring of 1992 2 phase-shift of ~2 to -3 days. We can only come to a Final decision as to what these results mean, on the basis of much more work in the south of England.as| 23 2a boo2 BEECH 7 Nov 17 18 Oct 28 28COPPER seecrl 1986 7 y 34 |} — AHN 13 48 Jan 23 26] 215 it Sept 16 2i 26 2BEECH 1987 ail 13 dn 28 7 Fob 12 mat 16 goBEECH 1987 18 Septze 3 i@ ig Oct aa 29 a 8 90BEECH 1987 14 i9Septaa ag @ 1419 Oct 24129 a 28 BEECH 1987BEECH 1988 1 6 Feb 11 16TWO,BEECHES 1988 oFen 1988 BEECH 16 Sept2t 26 nzo1989 BEECH3.0} 28 26 17 Sept voBEECH 1991 c15gp28, 2 1 6 ii Septi6 2726 © Oct 1 76 2bo18 Two BEECH Strontian 1991 l 7 22 Nov 27 2 7 12 Dec 17 22 2 =f 25] 7 22 Nov 27 12 Dec 17 22Aberdeen Two BEECH 1991 Stroud 2 7 12 Dec 17 22 c19oz BEECH 1992 Strontian © Wunz 26 3 5 1 IS Feb20 2 1 6» BEECH 1992 Aberdoon, lzoa Aberdeen Two BEECH 1992 eo 16 Jan 21 26 10 15 Fep 20BEECH 1992 Stoud a ; 2 Yr , a tI a NY f aa PN J Ny a 8 ive 0 2 2 <7 Webi 22 7 @ zoStroud BEECH 1992 PA MJ VAM V7 | 19 Jan 18 3 28 2 7 32 Feb 7 22 ar Py veoaire BEECH 1992 | | | 26 Jan 31 5 10 18 Feb 20 25 1 6 geoOak dt The leaf bude of the Oak provide excellent examples for this kind of work, They are, for the most part, very good path curves and are foeaiiy shaped for precise and easy measurement. At first sight it vould seen that they are hardly well adapted for our purposes, be— ing usually rather email, and always very hard and voody. It is not easy to imagine then undergoing these subtle rhythmic variations Nhich we are finding in other species; and each Autumn Iam aston— {shed afresh when | find them indeed responding again so strongly ard regularly. But there are difficulties. There are many different types of oak and several of these produce buds which are too small for accurate measurement. When one finds a tree producing good-sized buds all too frequently one finds that most of then are high up and quite out of reach. After only a fev weeke of working the supply of acressibie buds runs out, and one has to start work on another tree. This 1s why we ave so many short graphs charting the behaviour of the oak In fact. when we cone to actually vork with them, ve find these buds to be more, not less, strongly variable than most other species: and the ronae of hon any one cay 19 often rather igh. This introduces: another problem which has to be coped with. In all the other species Geseribed in this book | come avay from my gathering not naving. for one reason or another, any idea whatever whether the average |) of the cay'e set io aianificantly high or low. This 1s partiv due to the very subtle nature of the forns we are dealing vith and. in the case of the flovere, to the fact that the buds are still hidden in their sepals. Once the buds are mounted on the measuring plate, and thelr imege {2 Blown up large on tne monitoring screen. | nave. with the experience of years, a fairly good idea of the approximate value which tne machine will deliver te me; but by then the die is cast: B11 subsequent operations are in the hands of the computer: nothing Icon do will affect them. And this is how 1 like it. It is easy 10 have confidence that the complete objectivity which ve must strive for is being maintained, But with the oak the differences beturen one bud and another are sufficiently great for me, at least some~ times to gain aome Kind of estimate, aporoximately. of their hat a glance while they are still on the tree. How can | be sure that | fn not, guite unconscicualy, affecting the leeus by selective choos~ {ng of the buds? The best way to cope with this {s to take the buds from the brench, net singly, one by ene, but by the tyigful. It is simply not possible to assess, at a glance, the mean ) of a set of fairly tightiy-packed buds on’a twig: and thus one comes avay again without any knowledge of what one has gathered. | take each day a twig from each of four chosen branches and use from each twig the four largest lateral, and the single terminal, buds which it con~ taine, thus preserving a conetent ratio, lateral to tecminal. of four to one. Normally the terminal buds have a rather higher \ than the lateral, e¢ the preservation of such a ratio unchanged during a complete set of observations {s important. Graph e1 vao mode differently from the others in this book, being a record Of measurements of @ single bud photograpned while it wasa2 actually growing on the tree. It was a method which was very cun- bersome! tO follow and wnich proved to have some serious oissdvan— tages: and | have not followed it since. It did, however, enable me to make @ long graon stretching rignt through tne winter, tne only such that I have been able to make for the oak, and it proved 1m @ vay historic in this branch of the work being the first one to show a clear planetary relationship. Especially during the Autumn one finds the dips in the curve co-inciding exactly with the days of the aifgnments of Noon and Mars, marked with arrows. Notice how the variation almost cissopears during the end of Decemeber and the first half of January, to be resumed, rather less certainly, there— after. | nad to be away from home during March, but returned in tine to record the beginning of the final dip which'marked the opening of the bud in April. Notice that all the dips co-incide punctual ly with the arrovs exceot the last one (Feoruary 28th) which came one day early. Was this the first manifestation of the phase-shift? If so It went unnoticed by me at the time. After this all arrows shown cn the graphs of the Gak mark Moon/Mars relationships. The top grach on page 22 shovs a typical Oak development during the autum of 1965. The generality of buds at that time were showing a phase-shi ft of abut ~4 days but this particular tree was running Silghtly ahead of schedule. Notice the tendency for a steadily in- creasing A during the period of cbservation, something that seams to be fairly commen with this species. Some four months Jeter we have the lower graph of page e2. At that time Hars had just passed conjuncticn with Saturn and the numbers on the graph show the degrees which then separated the two planets. The general phase-shift had then increased to —5 days but we,se= that the 'expected' dip about February 26th completely failed to haope. and the following one cane, rather uncertainly, nearly three day3 late. Just how close together these two bodies need to be te inhibit one another seriously 13 yet to be determined. The figure of S which | have been vorking with {s an arbitrary one. and this graph gces to confirm the impression given by graphs clz and cl3 that at any rave the after-effects of such a conjunction last considerably longer. ‘The graph of €3 should be compared with that of c4, made at’ the sane time. We see the Cak and the Beech behaving in opposite ways, the one increasing while the other decreases, which seems to be quite common behaviour for then early in the séascn. The right hand end of this graph 1s repeated in graph e4 showing the two- and tnree— standard-deviation limits for the probability of significance, and we see that the two dips. as vell as the intervening peak. reach Both {n graphs eS, e6. the lover right-hand graph of e7, and e8 ve Notice the tendency for a general Increase in} during the period of observation. This vas something which was not expected, and, | think. hardly noticed at the time. but which later on. when one com pares’ the results together, becomes very clear. Only in the upper43 graph of page e9 do ve see a case in vhich this phenomenon does Rot show itself. When one finds such mutual consistencies which, although unnoticed at the time, show themselves regularly year by year. it nelps to give one confidence in the reliability of one’s measurements. If we look back to eraphs el and e3 we see a very big difference in their quality. In the former, where the phase-shift was about zero, the dips In the curve are short and sharp, lasting barely more than a day each. In the latter. vbere the phase-shift has become as large funerically as possible, +7, the form of the dips has become open and gradual. the curve coming somewhat to resenble a sine curve, but hen we com to graph eS, taken at a time when the phase-shift has again become numerical ly very small, +1 and approaching zero, the form of the dips has again become, seven years later, very shert and sharp. This is so much so that the dips are almost obscured In the heavy. trend, curve, and one has to look to the lighter daily lines in order to see that the dips are really there. There 1s not enough evidence yet to show that this 1s a general rule. but 1t may be. The graphs of the Primrose, taxen over the same period. seem to shov a sinilar trend. They are well worth comparing with these of the Uak. A somevhat similar, though perhaps less well-marked trend. can be Seen in the case of the Stitchwort:- see the upper graph of page kd. More research is needed here. Graphs @7, 69 and e10 are all short, due to the sparsity of access— ible buds'on those trees, and perhaps no single one of them, taken by itself, can be considered highly signtficant. But taken all to- gether thay add up to a considerable body of evidence; in no single case did a dip fall to materialise at the time appropriate to the current pnase-shift. Graph e11 shows an Gak for the spring of 1992 and is interesting a5 It shows yet another case uhere Mars cones into alfgnnent with Sa- turn, The al tanment was exact on March 6th, and the nunbers printed above the graph show the number of degrees of longt! tude separating these two bodies. The phase~shift has now grown to ~4 and ve see the alisnments of January and February strongly marked by dips in the curve, but the March alignment, coinciding closely vith the conjunc~ tion of the two planets. cones punctual ly but with very much reduced strength. This confirms well what has been previously observed: ine Alignment of these tvo planets Inhibits their activity but does not alsays completely eradicate {t. Ye compare this with the graph of 20 uhere we see the effect of this same conjunction on the working of Saturn.