Some Points in the Spermotogenesis of Marnnialîa. 139 



the archoplasm (when at restj in Amphibia, in Helix, and jiiesumably 

 in l-Cchinodenns ^); but within the Nebenkern of the rat's spennatocytes, 

 no centrosomes are to be found. Successfully stained preparations 

 however show, during tlie resting condition and the prophasis of the 

 ensuing mitosis, two small and individually duplicated structures, 

 looking very like a pair of micrococeus dumb-bells, and lying quite 

 outside the archoplasm (Nebenkern) (Figs. 2, 3 c), generally between it 

 and the nucleus. 



The appearance of these bodies is so peculiar that 1 have repro- 

 duced the relations observed in two photographs (Figs. 35, 36). The 

 elongated spermatocyte (Fig. 36) shows in the upper portion of its 

 cytoplasm (to the right in figure) the conspicuous chromatic body ot 

 Hermann; a little lower (to the left in figure) is seen the dusky 

 archoplasm (Nebenkern), a while to the left (below), and between it 

 and the nucleus, are seen the bodies in question c. In the photograph 

 they appear single, the lower being eclipsed by the higher one. In 

 Fig. 35 the chromatic body is not included in the optical section, but 

 the dusky Nebenkern is seen in the tapering mass of cytoplasm below 

 the nucleus (to left in figure); while above it, and to the right there 

 lies the same structure as in Fig. 35 c; it appears V-shaped, but is 

 seen to be in reality made up of two rather long halves, applied 

 together at a wide angle. In the early phases of development there 

 seems to be only one of these V's, but as the process advances the 

 two halves separate from one another, and each becomes in like 

 manner broken-backed. By focussing up and down we can see that 

 there are two. During the differentiation of the chromosomes, and 

 towards the disappearance of the nuclear membrane, these V-shaped 

 bodies become further and further separated (Figs. 4, 5 c) until, simul- 

 taneously with an apparent break-down in the nuclear periphery, they 

 acquire a radial connection with the liberated chromosomes in the 

 surrounding cytoplasm, assuming the usual characters of centrosomes 

 in relation to the spindle figure (Fig. 6). 



From these appearances we must conclude that the V-shaped 



') Cf. F. Field, Anat. Anz. Bd. VIII. 1893. p. 487—49.3. 



