MUSEUM OF COMPARATIVE ZOOLOGY. 67 



the latter case the author maintains that it is homologous with the 

 vitelline membrane ! 



The fact that the layer called by Beddard vitelline membrane becomes 

 radially and coarsely striate, suggests comparison with the villous layer 

 of Lepidosteus. As in the ganoid, so in Lepidosiren this is the layer 

 which is produced first. 



e. Teleostei. 



The numerous descriptions which have been given of the egg mem- 

 branes in different osseous fishes show that there is not uniformity either 

 in their number or structure. Besides the wall of the follicle with its 

 epithelium, the granulosa, there is perhaps only one investment of the 

 egg which is universally present, the zona radiata, and even this may 

 be wanting, or at least may wholly disappear in the case of certain 

 viviparous fishes. I believe it is certain that the homologue of the zona 

 radiata is invariably present in oviparous fishes, and I am likewise 

 of opinion, notwithstanding the inability of some observers to discover 

 the presence of pore-canals in the eggs of certain fishes, that the canals 

 are also always present. 



In many cases there is a membrane intervening between the zona and 

 the granulosa, sometimes thin and homogeneous, at other times of a 

 more complicated nature. It may even (villous layer) resemble some- 

 what the zona itself, although in reality very different from it in struc- 

 ture. Whether the thin homogeneous layer is homologous with the 

 more complicated villous layer cannot as yet be definitely answered, but 

 Balfour's account of the origin of a similar structureless membrane in 

 Elasmobranchs makes me incline to the belief that it is. 



In addition, the cells of the granulosa undergo in some instances a 

 remarkable metamorphosis, accompanied probably by a process of secre- 

 tion, and thus furnish still another primary envelope to the ovum. 



As to the existence of membranous structures inside the zona radiata, 

 there is much less certainty. The presence of a structureless vitelline 

 membrane has been maintained with more or less confidence by Vogt, 

 Aubert, Thomson, Lereboullet, Kolliker, Eimer, and others, and more 

 recently by Owsjannikow and Scharff. Its existence has also been 

 denied by eminent authority. 



The question as to the presence and nature of a so called zonoid layer 

 seems, especially in the light of Brock's recent contributions, to demand 

 a more extended and thorough investigation. In the case of some of 

 the eggs studied by Eigenmann ('90), structural conditions have been 



