THE SPERMATOGENES1S OF AGALENA N.KVIA. 143 



then we might suppose that these large, faintly stained cells 

 represent degenerating spermatozoa and as they seem to be about 

 as numerous as the deeply stained spermatozoa their presence 

 might mean that only one half of the latter become functional. 

 It seems more probable, however, that a spermatozoon would 

 uncoil in the process of degeneration and it also seems likely 

 that the chromatin in the head would retain its staining capacity 

 for a long time. I am, therefore, at a loss to explain the presence 

 of these large, oval cells. 



SIGNIFICANCE OF THE ACCESSORY CHROMOSOMES. 



Until comparatively recent times the problem of sex-determina- 

 tion has been approached chiefly from the outside and much- 

 experimental work has been done in the attempt to prove that 

 external factors, such as nutrition, temperature, etc., do influence 

 sex. Within the last few years, however, minute cytological 

 research has directed attention to internal factors, namely, to the 

 nuclei of the germ-cells themselves, and there is now good reason 

 to believe that the problem can be put on a morphological basis. 

 McClung's ('02^) brilliant idea that dimorphism of the sperma- 

 tozoa caused by the presence or absence of the accessory chromo- 

 somes might have a direct bearing on the determination of sex has 

 been strongly supported by Wilson ('05^, b, c, '06) and Stevens 

 ('05^, 'o6) and, more recently, by Boring ('07) in their work upon 

 insects. These authors find cases in which the somatic cells of the 

 male contain one less chromosome than the somatic cells of the 

 female or, in cases where one half of the spermatozoa contain a 

 very small chromosome represented in the other half by a large 

 chromosome, the somatic cells of the male and female show 

 corresponding differences. 



In the spider Agalena mvvia I have shown that dimorphism 

 of the spermatozoa obtains, one half of them having two acces- 

 sory chromosomes and one half of them lacking these elements. 

 The comparative number of chromosomes in the somatic cells of 

 the two sexes could not be determined, but a comparison of the 

 developing germ-cells was made with reference to the presence 

 or absence of the accessory chromosomes. In the spireme stage 

 of the growth period of the primary spermatocytes, the two rod- 



