ON THE SPERMATOGENESIS OF THE SILK-WORM. 



149 



In the first division of the sperm-mother-cells (" Sperma- 

 tocyten 1. Ordn."), the twenty-eight chromosomes in the equa- 

 torial plane of the spindle divide transversely, as stated above, 

 and produce two rows. Both Henking and vom Rath affirm 

 also the existence of two rows of chromosomes in the equator 

 of the spindle of this stage. But this, according to vom Rath 

 (28), is due to the doubling of the chromosomes in the growing 

 stage of the sperm-mother-cells, while Henking (15) believes it 

 to be formed by the double arrangement of the chromosomes, 

 already existing, so that according to this latter author, the 

 reduction of the number of chromosomes takes place by 

 this division, which according to all other authors occurs by 

 the second < r) . 



The reduction of the number of chromosomes in the genital 

 cells of plants has been worked over, so far as I know, only by 

 Guignard (10). The mode of the reduction which he observed of 

 Lilium martagon, however, differs entirely from that of animals. 

 According to this investigator the reduction occurs without a 

 nuclear division in a stage of the "cellules mere definisives" (pro- 

 bably corresponding to the sperm-mother-cells of animals). My 

 investigation on the formation of the pollens of Lilium tigrinum 

 and Allium fistulosum is not in accordance with this, but shows 

 much similarity with that of animals. This I trust will be 

 soon published. 



Summary. 



1. A large cell in the blind end of the follicle which is de- 

 signated by the name of " Keimzelle " by Verson, is not a genital 

 cell but a supporting cell which connects the genital cells with 

 the follicular wall. It always has a large nucleus with finely 

 granular chromosomes collected here and there and staining 

 very deeply with hasmatoxylin, safranin, carmine, anc' anilin 

 colours. 



(1) According to the investigations of Moore (24), the sperm mother-cells of the 

 rat contain sixteen chromosomes, which are reduced into eight, and " the reduction 

 would appear more comparable to the type described by Guignard during the forma- 

 tion of vegerable pollen, than, to that pointed out by Hcrtwig in the spermatocytes 

 of Ascaris." This however appears to me very improvable. 



