MUSEUM OF COMPARATIVE ZOOLOGY. 117 



pore-canals, which rarely (Myxine?) branch. J. Miiller is wrongly cred- 

 ited with their discovery. 



25. An egg membrane genetically, but not always structurally, com- 

 parable with the villous layer of Lepidosteus, is found in several other 

 cases : possibly in Petromyzon, probably in selachians and Lepidosiren, 

 and certainly in several teleosts. This membrane is also produced by the 

 ovum, and earlier than the zona. 



There is some reason to believe that it exists in the herring and the 

 smelt (Osmerus). Hoffmann is probably in error in attributing the 

 presence of villi to the eggs of Perca in October, and Owsjannikow cer- 

 tainly is in asserting that the mushroom-shaped villi in Gasterosteus are 

 individual cells. 



26. The capsidar membrane is produced, as originally denned by 

 J. Miiller, by the follicle or follicular cells. It has often been confounded 

 with the zona, and also with the villous layer. 



27. Although the " tubules " in Perca have been described as possess- 

 ing root-like prolongations which penetrate the pore-canals of the zona, 

 they are genetically unlike the villi found on other eggs, being produced 

 by the granulosa cells alone. Hoffmann's statement to the contrary 

 rests on insufficient evidence. 



28. A comparison of the condition of the granulosa in Blennius pholis 

 and Esox with that in Perca, is believed to shed some light on the prob- 

 able steps by which the capsular membrane is produced. 



29. The most important paper on the origin of the zona and the 

 villous layer I believe to be that of Kolliker, whose conclusions I have 

 confirmed in the case of Lepidosteus. All authors who have maintained 

 that either zona radiata or villous layer is the product of the granulosa 

 I believe to be in error ; and in particular I maintain that the reasons 

 assigned by Cunningham to prove that the zona in Myxine is produced 

 by follicular epithelium are not adequate to establish his proposition. 



30. Tt is doubtful whether the micropylar canal has been seen in 

 Myxine. What W. Miiller called the micropyle was the micropylar 

 funnel, and possibly that which Cunningham describes as the micropyle 

 does not embrace the canal. 



31. What I have called the micropylar plug of granulosa cells was 

 first seen by W. Miiller in Myxine ; later it was figured, but apparently 

 not appreciated, by Hoffmann in Leuciscus ; it was described as hollow 

 by Owsjannikow in the case of Osmerus, who recognized its relation to 



