522 



NA TURE 



[September 21, 1905 



thread. Whichever of these explanations is the correct 

 one, the doubling gradually disappears and the thread 

 becomes distributed through' the nuclear cavity and again 

 appears single ; it becomes shorter and thiclcer and once 

 more becomes aggregated around the nucleolus. This 

 may be, as Miss Sargant suggests, a second sinapsis. At 

 this stage the chromosomes appear, but reduced to half 

 the number of those which appeared in the previous 

 divisions, so that they may be regarded as bivalent or 

 double chromosomes. They become shorter and thicker, 

 and gradually become grouped in the equatorial plane of 

 the nucleus, where they become attached to the spindle 

 fibres. Each chromosome now divides into two halves, 

 which pass to the respective poles of the spindle, to form, 

 without the intervention of a complete resting stage, the 

 division figures of the daughter-nuclei. 



The exact mode in which the division of the chromo- 

 somes into two halves takes place is the subject of much 

 controversy. The studies of Weissman on the pheno- 

 mena of heredity led him to the conclusion that the 

 chromosomes consist of more than one complete ancestral 

 germ-plasm, and that consequently these must be reduced 

 in number in the sexual cells to escape the extraordinary 

 complexity which would arise if the ancestral germ-plasms 

 were doubled at each sexual fusion. As the longitudinal 

 division of the chromosomes divides them into two equal 

 halves it is obvious that this will not reduce the number 

 of ancestral germ-plasms, and therefore Weissman pre- 

 dicted that a transverse division of the chromosomes would 

 be found to take place by which the reduction would be 

 brought about. This was soon discovered to be the case 

 for many animal cells, the reducing division taking place 

 during the formation of the sexual cells, but in plants 

 this was not so easily determined. Belajeff, Dixon, 

 Atkinson, and others maintained that a true reduction 

 division took place in the cases examined by them ; but 

 the majority of observers. Miss Ethel Sargant, Strasburger, 

 Farmer, Mottier, and many others, maintained that there 

 was no transverse division, but that all the divisions were 

 longitudinal. Recently, however. Farmer and Moore have 

 re-investigated the whole sequence of events in both 

 animals and plants, with the result that a true reduction 

 division is found to occur in the heterotype stage. In 

 many investigations which have recently appeared this 

 transverse division is confirmed, but the exact details of 

 the process are not yet agreed upon. Farmer and Moore 

 state that the spireme thread first becomes longitudinally 

 split, the two longitudinal halves then fuse again, and 

 subsequently bivalent chromosome loops are formed which 

 divide transversely in the middle, and so produce two 

 monovalent chromosomes which pass to opposite poles of 

 the spindle, as already described. Gregoire, on the other 

 hand, states that the threads at the first sinapsis become 

 approximated together and then fuse ; the double thread 

 thus produced breaks up into chromosomes, which are 

 thus bivalent in a different sense from those of Farmer and 

 Moore, the monovalent chromosomes being produced by a 

 longitudinal splitting of the thread, which divides it into 

 the two original halves which fused together. 



Which of these two methods will ultimately be found 

 to be the correct one remains to be seen, but Allen has 

 recently published an account of the process as it occurs 

 in Lilium canadense, in which he agrees substantially 

 with Gregoire, and states definitely that the first appear- 

 ance of the double nature of the thread is not due to 

 a longitudinal splitting of a single thread, but to an 

 approximation of two threads, which ultimately fuse 

 together to form a single continuous thread in the nuclear 

 cavity. This thread at a later stage undergoes a longi- 

 tudinal splitting, possibly into those which formerly united ; 

 but this is not certain. The double thread then divides 

 up into segments, the chromosomes, and in the subsequent 

 series of events the longitudinal halves of these chromo- 

 somes become distributed to the opposite poles of the 

 spindle. Each chromosome is thus seen to be bivalent ; 

 but whether each half of the chromosome is to be re- 

 garded as a monovalent chromosome is doubtful, as the 

 fusion of the original threads was complete, and there is 

 no means of deciding as to how far the subsequent longi- 

 tudinal division of the completely fused thread separated 

 it into its two original parts. 



NO, 1873, VOL. 72] 



Sinapsis. 



The term " sinapsis " was first given by Moore to 

 that stage in the prophases of the nuclear division of the 

 sexual cells in which the contraction of the nuclear thread 

 around the nucleolus at one side of the cavity of the 

 nucleus takes place. If this phenomenon is not a result 

 of the action of the fixing reagents, then it indicates some 

 striking change in the metabolic activity of the nucleus. 

 This activity is seen in the increased staining capacity of 

 the chromatin thread and in the changes which take place 

 in the nucleolus, by which it becomes very irregular in 

 shape and closely connected by threads to the chromatin 

 network. In many cases the nucleolar substance appears 

 as if being drawn out into the threadwork, and the 

 nucleolus appears as if some active change were taking 

 place in it. 



It is very difficult to escape the conclusion that we are 

 here dealing with a series of changes in the chromatin 

 thread which are intimately bound up with the activity 

 of the nucleolus, and it is probable that the increased 

 stainability of the chromatin is due to an actual trans- 

 ference of a portion of the nucleolar substance into the 

 thread. 



Experimoilal Observations on the Activitii 

 Nucleus. 



of the 



So far as I know no experimental investigations into 

 the causes which bring about the changes in the prophases 

 of nuclear division have been made, but it is not difficult 

 to imitate artificially some of the phenomena observed. 

 Olive oil is shaken up in a mixture of methylated spirit 

 and water of such a strength as will allow the oil globules 

 to float. A shallow Petrie dish, three or four inches in 

 diameter, is then taken ; the mixture of oil and dilute 

 methylated spirit is well shaken until the oil is broken 

 up into very fine globules, and the mixture is at once 

 poured gently into it. The appearance of the mixture is 

 that of a homogeneous mass of small oil globules dis- 

 tributed through the solution, and can be compared to 

 the granular appearance of a nucleus in a resting stage. 

 The spirit at once begins to evaporate, and currents are 

 at once set up in the solution in such a way that the 

 globules of oil gradually become restricted to certain areas 

 only, and a coarse granular network is formed somewhat 

 like the early stages in the aggregation of the chromatin 

 granules into a spireme in the nucleus. The network 

 gradually becomes more and more clearly defined, and 

 then, just as is the case in the nuclear network, it begins 

 to show a double row of granules, which finally becomes 

 very clear and distinct. The threads become shorter and 

 thicker and break up into irregular lengths, which 

 gradually mass themselves together into an irregular heap 

 or heaps of fusing-oil globules either in the middle or at 

 the periphery of the petrie dish. We have, in fact, a good 

 imitation of the earlier stages in the prophases of division 

 of the nucleus, and it seems not unlikely that the aggre- 

 gation of oil globules in our petrie dish may afford some 

 clue as to the possible means by which the aggregation 

 is brought about in the nucleus. 



I do not suggest that the complex phenomena which 

 take place in nuclear division are to be explained as due 

 simply to such phenomena as diffusion, surface tension, 

 and the like, or any other physico-chemical processes. 

 We must be very careful not to attempt to force merely 

 physico-chemical explanations upon such phenomena as 

 these. Without admitting the necessity of anything akin 

 to a special vital force, we are compelled to admit that 

 vital phenomena do not at present admit of a merely 

 mechanical explanation. But it does seem to me possible 

 that the metabolic activity of the nuclear material at 

 this stage may be accompanied by phenomena referable 

 in part to these agencies. If, for example, active metabolic 

 activities are set up between the nucleus and cytoplasm 

 through the nuclear membrane, as seems probable, it is 

 quite conceivable that this would bring about diffusion 

 currents which might be taken advantage of in producing 

 the aggregation of the more solid parts of the nuclear 

 substance into a more or less definite thread-like structure 

 and its aggregation into chromosomes. In any case such 

 possibilities must be taken into account in considering 



Jl 



