178 L. B. WALLACE. 



McClung has suggested that the heterochromosomes might be 

 sex-determinants and he is supported in this view by Sutton, 

 Blackman and others. Paulmier regarded them as degenerating 

 chromatin. Montgomery agreed with him in this and considers 

 that they are " chromosomes that are in the process of disappear- 

 ance, in the evolution of a higher to a lower chromosomal num- 

 ber." If only one fourth of the spermatids contain the hetero- 

 chromosomes, as in the spider, they can scarcely be sex-deter- 

 minants. Neither do they here show the characteristics of degen- 

 erating chromatin. That they represent some form of specialized 

 chromatin, I cannot doubt, but that they are specialized for some 

 metabolic function and comparable to nucleoli, as has been sug- 

 gested by Montgomery, does not seem to me probable in view 

 of the facts. I, therefore, venture to offer a fourth theory in re- 

 gard to the function of the accessory chromosomes in the spider. 

 The breeding season of Agalcna ncevia comes late in the sum- 

 mer, the eggs being laid in August and September. While 

 examining preparations made late in September I noticed a great 

 many degenerating cells side by side with the ripe spermatozoa 

 found in abundance in the lumen of the testis and in the ducts 

 (PI. II., Fig. 49). These cells are without nuclei and many of 

 them are fragmented. My first conclusion was that the presence 

 of so many fragmenting cells was due to the lateness of the 

 season but on examination of sections. prepared early in August 

 a few cells undergoing degenerative processes were found even 

 though but few spermatozoa had yet reached maturity. Sections 

 made from the testis of PJwlcus plialangiodes at the very height 

 of the breeding season, in July, were most convincing. In PJwl- 

 ciis the spermatozoa are much more elongated than in any other 

 species of spider examined (PI. II., Fig. 51), but when mature 

 they curl up as the others do. In the lumen and ducts were 

 found hundreds of spermatozoa thus coiled up but a far greater 

 number of round cells with rounded nuclei in which the linin 

 and chromatin are no longer distinguishable. These cells vary 

 in size --some as large as the germ cells, some very minute and 

 in every way they resemble cells in the process of degeneration 

 (PI. II., Fig. 50, d). Further study showed that as a rule the 

 degeneration does not begin until after the spermatozoa have 



