BLACKMAN. — SPERMATOGENESIS OF THE MYRIAPODS. 497 



plasm (Figures 5, G, and 7). The remaining portion of the nucleus is 

 occupied by an irregular reticulum of linin fibres. 



As the cell continues to grow this amorphous deposit increases in 

 amount (Figures 8, 9, 10), until finally, in the vesicle stage, it forms 

 quite a conspicuous portion of the nucleus (Figures 9, 10). Just 

 what the significance of this material may be I am at a loss to say. 

 From its straining reactions, however, it is seen that it is chemically 

 similar to the nucleolar substance. In Lithobias mordax it is much 

 more abundant than in either of the other species studied, while in 

 Lithobius sp. 1 (Figures 14, 17) very little if any occurs. Thus it varies 

 inversely with the amount of nucleolar material contained in definite 

 nucleoli. In Lithobius multidentatus less is present than in L. mor- 

 dax, but in the former species (Figures 11, 13) it occurs in much 

 more definite bodies, and these often approach nucleoli in form and 

 structure. So the conclusion would seem justified that these deposits 

 represent nucleolar material which has never become arranged in the 

 form of distinct nucleoli. 



We will now return to the early spermatocyte and consider the evo- 

 lution of the extra-nuclear structures. First, as regards the centro- 

 some. In Scolopendra heros (Blackman, :05) I was so fortunate as to 

 be able to find the centrosome during all the stages from the spermato- 

 gonium to the mature spermatozoon. Here it was readily found and 

 recognized, owing to the fact that during the growth period it is en- 

 closed in a specialized portion of the archoplasm. In Lithobius, how- 

 ever, such aggregations of archoplasm do not occur, and the centrosome 

 has not been observed during the growth period. In Figure 7 two 

 small, densely stained bodies enclosed in a denser portion of the 

 cytoplasm are shown. It is, however, impossible to state with any 

 certainty that these are centrosomes, owing to the fact that such 

 structures have not been observed at other stages of the growth 

 period. However, it does not necessarily follow that no centrosome 

 exists during the growth period, as no determined effort has been 

 made to find it and study it, as was done in Scolopendra. 



The structure of the cytoplasm in Lithobius has been studied by P. 

 et M. Bouin (99), and by Meves und von Korff (:0l). The brothers 

 Bouin report the presence within the cytoplasm of Lithobius forfica- 

 tus of certain irregularly formed bodies, which they have designated 

 "formations ergastoplasmiques." These are believed to arise by the 

 disintegration of the astral rays. Just before the prophase these de- 

 posits undergo what is described as a gelatinous metamoi'phosis and 

 are not seen later than the early proj^hase. Meves und von Korff 

 (:0l), working on the same species, find these bodies in all phases of the 



VOL. XLII. — 32 



