778 TRANSACTIONS OP* SECTION K. 



approximated, the spireme being double from its first formation till it divides 

 into the longitudinally split chromosomes. On the spindle the longitudinal 

 fission closes to reopen permanently as the daughter chromosomes travel to the 

 poles. 



The chromosomes therefore which separate one from another in any given 

 division are the result of a longitudinal fission which is initiated in the pre- 

 ceding telophase, and the nucleus, alike of the sporophyte and gametophyte, 

 shows a double structure throughout its resting stages. 



2. On the Somatic and Heterotype Mitoses in Galtonia candicans. 

 By Professor J. B. Farmer, F.B.S., and Miss L. Digby. 



The question as to how the process of the reduction of chromosomes is 

 effected is still unsettled. Two principal views are current : — 



(1) That the reduction occurs during 'rest,' and that each meiotic division 

 is associated with longitudinal fission as in a somatic division. This view implies 

 the negation of chromosomal individuality. 



(2) That meiosis is associated with a temporary pairing followed by the 

 dissociation of somatic chromosomes, and their distribution to the daughter 

 nuclei at either the first or second meiotic division. Further, that the tem- 

 porary union is probably between homologous chromosomes derived from the 

 male and female parent respectively. Amongst those who adhere to the latter 

 interpretation of the process there is a difference of opinion as to the method 

 by which the result of distribution of the chromosomes during the first meiotic 

 division is achieved. 



One school, represented prominently by the investigators at Louvain, 

 believe that a side-to-side pairing at or before synapsis takes place, or that 

 the chromosomes first issue from rest as paired structures (pro-chromosomes) 

 and that there is nothing representing the longitudinal fission of the earlier 

 and subsequent mitoses, but that the appearances which have been thus inter- 

 preted are illusory and depend on the approximation and subsequent divarication 

 of distinct pairs of somatic chromosomes. 



Another set of observers, including ourselves, consider that the appearance 

 of early longitudinal fission is not illusory, but really corresponds to this process 

 as met with in ordinary dividing nuclei. The union of entire pairs of chromo- 

 somes, whereby the reduced number is effected, is regarded as being due to 

 an end-to-end union or non-separation, or else to be brought about by a pairing 

 at 'second contraction.' We do not propose to go into the whole matter fully 

 here; this has been dona in a paper in course of publication. We wish, however, 

 to draw special attention to certain facts, the general bearing of which seems to 

 have attracted less attention than they deserve. 



We refer to the results of a study of the complete series of events which 

 intervene between telophase and prophase in the last archesporial, and the 

 heterotype mitosis, together with the comparison between them and premeiotic 

 divisions, whether in the archesporial or somatic tissues. 



For purposes of illustration we have selected Galtonia candicans, mainly 

 because it is specially favourable for this investigation, and also because it 

 has been studied in part by others. We exhibit a series of drawings, a.s accurate 

 as wo could make them, to illustrate these most important stages, and we have 

 also added one of a tapetal prophase. 



It will be seen that in all of them, owing to the absence of ' resting ' stages 

 especially in the archesporial mitoses, there is a remarkable similarity, extending 

 to minute details, between the prophase of the heterotype, the division in dispute, 

 and the other and earlier mitoses in this plant. 



As the telophase of an early archesporial division, for example, comes on, 

 there is seen to be a condensation of the chromatin on the two edges of each 

 of the chromosomes, and even when these have lost their early distinctness the 

 duplicate character can still, and thus early, be detected. As the next division 

 supervenes, again the dots or lines aggregate in pairs — the new chromosomal 

 structures are longitudinally divided ab initio. The fission can be detected easily 

 up to the spireme, and only becomes for the most part temporarily invisible ae 



