549 



While the seminiferous elements are still quite young, i. e. several 

 generations before the final changes of the spermatogenetic process, 

 the foot and semen cells are arranged in a single row round the inside 

 of small globular cavities. From this period onward akinesis goes 

 on in the one, karyokinesis in the other, but their relative arrange- 

 ment changes in a striking way. Instead of remaining evenly dis- 

 tributed, all the foot-cells become aggregated round the central cavity, 

 and remain there till the semen cells by their repeated division are 

 three or four ranks deep; they then migrate, with a singular amoeboid 

 movement, back to the periphery, arranging themselves on the bound- 

 ing membrane, and assuming the normal-foot cell character, just before 

 the closing divisions of the spermatogenesis set in. 



The two last divisions are quite distinct from all those which go 

 before, but the first does not occur among small, nor the latter among 

 large growing cells, as in Mammals, the conditions in this respect 

 being quite reversed. 



There are twelve chromosomes in the penultimate division whether 

 we count them when emerging as chromatic rings from the reticulum 

 of the previous resting nucleus, in the "monaster", or as the divided 

 loops when nearing the opposite poles of the spindle figure, as sug- 

 gested by Boveei. 



The daughter nuclei produced, pass into a condition of complete 

 reposepriortothelastdivision of the spermatogenetic 

 series and the initial phases of the last mitosis as well as its whole 

 course, proceed in exactly the same way as in the penultimate division, 

 except that only six chromosomes emerge from the resting reticulum 

 instead of twelve! 



Now, a close examination of these twelve chromosomes of the pen- 

 ultimate division reveals the fact that they are each built up of four 

 small condensations or primary elements, united together in the form 

 of a ring; while those constituting the individual loops or rings of the 

 final division, are far more numerous. The small size and crowding 

 of the parts in the latter phase renders counting very difficult, but 

 numbers of readings satisfied me that in this division they are eight; 

 i. e. the twelve fourfold chromosomes of the penultimate division are 

 rearranged in the form of six eight -fold chromosomes in the last 

 division, while both divisions are normal mitoses in which there is n o 

 passage of unsplit chromatic masses to the poles. 



In the spermatogenesis of this fish then, there is reduction only 

 n the number of the chromosomes when they reappear from 

 rest in the final as compared with the penultimate division. But this 

 change is not brought about by either division, it occurs during 



