Some Points in the Spermatogenesis of Mammalia. 147 



The enquiry however need not end here. I have personally more 

 than once, when examining the general characteristics of the centro- 

 somes in comparison with those of the suiT<Junding cyto-microsomes, 

 experienced a sensation of panic, lest these apparently important 

 structures should lose their individual identity; and it is very probable 

 that the Japanese enquirer had some such experience in mind when 

 he put forward the ingenious hypothesis that the microsomes, cyto- 

 microsomes and centrosomes are one and the same thing. It appeai-s 

 to me that this whole question may tuin on one point, viz. whether 

 the microsomes and the cyto-microsomes are really similar. If it is 

 true that the microsomes in their totality represent the hereditary 

 substance of the cell, one would hardly think it probable. 



In November of last year, through the kindness of my friend 

 Mr. H. M. Bernard, I came into possession of some male Brancliipus, 

 in which the divisional phenomena of spermatogenesis, interesting in 

 themselves, bear so directly on this very question of the homology of 

 the centrosomes, that I need no excuse for briefly reverting to them 

 here. The cells in question have a fine reticulate appearance, both 

 within and without the nucleus. The meshes of this reticulum are much 

 the same size in both localities, but the stain was almost entirely 

 selected by the fibres which form the network, so that the nucleus 

 must either contain more staining material, or the microsomes and 

 cytosomes must be of a different natm-e. It is very probable that 

 both these causes operate in producing the effect. Just exterior to 

 the nucleus there are usually to be found from foiu' to eight dusky 

 bodies, anyone of which might be taken for a centrosome. When 

 carefully examined they appear to be neither more nor less than 

 larger angular spaces left between the constituent glolniles of a 

 "schaumplasm", and or the sake of clearness I termed these bodies 

 pseudosomes. ^) 



As metamorphosis proceeds the fine chromatic reticulum of the 

 nucleus becomes collected into ten dumb-bell shaped chromosomes, 

 while the clear nucleoplasm between seems to fuse with the corre- 



>) Cf. Qu. Jour. Micr. Sci. Vol. XXXY. p. 263. 



10^ 



