ioo BOTANICAL GAZETTE [august 



weak in development, so that the relation it bears to the blepharo- 



determined at this time. There is no 



question , however, that the blepharoplasts continue to occupy the 



lmme- 

 many 



diately preceding stages. In the later stages of division 

 extremely fine strands are present between the daughter nuclei. 

 Whether these are the remains of fibers passing from one blepharo- 

 plast to the other or represent the visible effect of the separation of 



was not determined. The cell 



asm 



sperm 



During the anaphases of karyokinesis a peculiar change occurs 

 in the blepharoplasts. For a time they lose their affinity for iron- 



many 



may 



more 



stained cells they appear as translucent bodies considerably larger 

 than during the earlier phases of division (fig. 16). They are no 

 longer solid but contain one or two large vacuoles, which give them 

 in section the appearance of small rings. It is probable that the 

 decrease in staining capacity is due to swelling through the absorp- 



erial. This 



condition exists only through the remainder of the division; when 

 the sister spermatids are well rounded away from each other the 

 blepharoplasts as a general rule stain deeply again. The vacuole 

 or vacuoles form an irregular cavity, and the whole structure soon 



ma 



form 



i 



7, 18). 



hich 



arranged in a row, usually at once (fig. 19). These pieces multiply 

 rapidly by further fragmentation and form a beaded chain extending 

 about halfway around the nucleus (fig. 21). Fig. 20 shows a mass 



■ 



of these granules just beginning to draw out into a row. 



It is at this beaded stage that the cilia begin to develop. From 

 the blepharoplast granules there are seen very fine strands extend- 

 ing toward the periphery of the cell (fig. 21). Whether more than 

 one of the strands, or rudimentary cilia, ever grow out from a single 

 granule was not definitely determined, but since the cilia of the 

 mature spermatozoid and the granules are approximately equiva- 

 lent in number, it is Drobable that a* a mlp parh crrarmle gives rise 



