108 Reduction Phenomena of Animals and Plants. [May 29, 



while the rings are opening out on the spindle and before the united 

 chromosomes of which it is composed have separated, and although in 

 some cases it is more or less concealed, this early fission rapidly 

 reappears, and the remarkable figures seen in the anaphases of 

 Tradescantia or Salamander, thus find their explanation. 



The deceptive appearances presented by lilies at this period also 

 find an intelligible explanation when the mode in which they are 

 attached to the spindle fibres is borne in mind, and a remarkable con- 

 firmation of this is forthcoming in the fact, which seems to have 

 hitherto escaped observation, that in some cases the daughter- 

 chromosomes, like those of Tradescantia, clearly show their original 

 longitudinal split (Fig. 6). This original longitudinal fission of the 

 spireme thread which is usually visible in the retreating daughter 

 elements is retained in some, probably in all, cases as a preparation 

 for a succeeding homotype division. 



It would thus appear from the preceding observations, that the 

 synaptic rest, culminating in the heterotype mitosis, is a phase specially 

 intercalated in the reproductive cycle. In it a reduction of the 

 number of the chromosomes is brought about by their adhesion in pairs, 

 but with the inception of karyokinetic activity the spireme thread 

 undergoes the longitudinal fission characteristic of ordinary somatic- 

 division, although the actual separation of these longitudinal halves is 

 deferred until the next mitosis. Thus the heterotype division not only 

 seems to be different in kind from an ordinary mitosis, but in this very 

 fact probably lies the reason of the extraordinary diversity of form so 

 often obvious among the chromosomes in this, as compared with any 

 other, type of nuclear division. 



We have purposely refrained in this communication from criticising 

 the results of others, and from discussing the general bearings of our 

 own interpretation of the phenomena upon heredity and other matters. 

 We have done so because we think this part of the subject will find a 

 more appropriate place in a memoir now in preparation, wherein the 

 evidence for our conclusions will be fully set forth. 



The illustrations that accompany this paper are intended merely 

 to serve as diagrams explanatory of the more important stages of the 

 heterotype division. 



EXPLANATION OF THE FIGURES. 



1. Spireme thread shoeing longitudinal fission. 



2. Looping together of the split spireme. 



3. Same at a later stage. 



4. The loops have separated, and the longitudinal fission cannot 



always be recognised. 



5 and 6. Stages in the separation of the daughter-chromosomes at the 

 equatorial plate. 



