HISTORY OF CHROMOSOMAL VESICLES IN FUNDULUS. 263 



granules distributed fairly uniformly throughout it. Succeeding 

 this stage of general chromatic dispersion the vesicles appear to 

 shrink, expressing part of their water content; this phenomenon 

 is manifested doubly, first by the clear non-fibrillar area about 

 the nucleus, the extent of which before shrinkage is thus marked, 

 and second by the concentration of the contents of the vesicle, 

 which in fixed material appears denser and takes more orange G 

 than before. Coincident with this change the amount of chroma- 

 tin appears to increase and the aggregation of the granules takes 

 place, the larger masses of chromatin appearing along the peri- 

 phery and perhaps extending into the center along a single 

 thread. The aggregation continues until masses the size of the 

 chromomeres of the anaphase are formed and have become 

 arranged in a linear series. The linin sheath about this linear 

 aggregation of chromomeres forms early. This structure is the 

 new chromosome and must be thought of as chromosomal in 

 nature quite as much as the more familiar metaphase stages. 

 When it is completely formed the walls of the vesicle may be 

 seen to be gradually dissolved and the denser contents to diffuse 

 into the cytoplasm. About this time or a little later the split 

 of the new, or filial, chromosome takes place, each chromosome 

 and chromomere showing it about equally. 



Figs. 17 and 18 show the first sign of the next division. The 

 chromatin granules are well dispersed and the vesicles definitely 

 oriented, while the pre-mitotic shrinkage has taken place. In 

 Fig. 19 the aggregation of the chromatin granules has begun 

 and the vesicles in this state can always be recognized as about 

 to divide. In Fig. 20 is seen the linear arrangement of the 

 chromomeres, a process which is well advanced in Fig. 21. In 

 this latter case indeed the chromosomes are well formed and 

 the linin sheaths can be made out. In these figures the walls 

 of the vesicles are not as clear as in some cases, but the denser 

 gel-like portion still shows the form and relationships of the 

 vesicles. There would seem to be some variation as to the 

 exact time when the disappearance of the walls takes place, in 

 relation to the development of the spindle structures at least. 



Figs. 21 and 22 are prophases in the spireme stage, early and 

 later. Fig. 22 is an especially clear case showing the newly 



