MUSEUM OF COMPARATIVE ZOOLOGY. 137 



described by him. The tubules are independent of their jelly matrix, 

 and in chromic acid preparations they can be separated from the latter. 

 When the eggs are deposited, the granulosa cells probably fall off, leav- 

 ing shallow depressions having polygonal outlines, from the centres of 

 which " tubules " arose. 



Ransom ('68) described the canals passing through the outer portion 

 as having a double contour for each wall, and as tilled with material 

 containing vacuoles ; but they do not seem to him to convey anything 

 either fluid or solid into or out of the egg. This outer layer is separable 

 only by tearing it from the yolk-sac (zona), and does not leave a dis- 

 tinct outline. The tubes divide at their inner terminations into branch- 

 like roots, and adhere closely to the zona radiata. The internal ends are 

 not expanded as Miiller described, and it is rarely that filaments pass 

 from one to the other. He supposes that the granules seen by Miiller 

 were only vacuoles. The eggs when deposited are arranged in the form 

 of hollow tubes with the micropyles all turned to the inside. 



His ('73) mentions having seen the micropyle, but neither figures nor 

 describes it. 



Brock ('78) describes the zonoid layer, and finds its striations inter- 

 mediate in fineness between those of the villous layer and those of the 

 zona. Judging by his drawing of Alburnus lucidus there are about three 

 striations in the zonoid layer to four in the zona. The latter, he says, 

 makes its appearance before the villous layer. 



Hoffmann ('81) finds that in October the zona and the villous layer 

 are of equal thickness. The latter is said to be composed of numerous 

 small projections which correspond exactly to the villi of the Cyprinoids. 

 At the free ends of the villi lie the granulosa cells. In February the 

 zona is differentiated into two layers, of which the inner is four times as 

 thick as the outer. There arise from the outer layer long fibres with 

 triangular bases and with their distal ends expanded to form a continu- 

 ous layer on which the granulosa cells rest. Each filament corresponds 

 to, but is not a process of, a granulosa cell. 



Owsjannikow ('85) recognizes the usual divisions of the egg membrane. 

 The contents of the distal ends of the filaments ai*e granular, which has 

 given rise to the belief that they are nuclei. The filaments end ex- 

 ternally in funnel-shaped enlargements described by Miiller. He suc- 

 ceeded in forcing granular matter from the yolk into their deep ends. 

 The latter divide and enter the pores of the zona, through the whole 

 thickness of which they can be traced. He states (p. 7) that, contrary 

 to Hoffmann's belief, the filaments are derived from the granulosa. In 



