146 BULLETIN OF THE 



Belone lougirostris or ('84, p. 457) the cod, and states ('85, p. 145) that 

 the eggs of Gambusia patruelis do not possess any membrane. I have 

 found striations in the membrane of the fresh cod egg. It may be stated 

 here that the striations of the zona sometimes show plainest in fresh 

 eggs, sometimes not until reagents have been applied. Haeckel says, for 

 forms related to Belone longirostris, that the membrane is structureless, 

 but that it is covered with minute black dots. These dots were doubt- 

 less pore-canals seen from the surface. The zona radiata of Osmerus 

 eperlanus was found by Buchholz ('63) to consist of an inner and outer 

 portion, joined together in the micropylar region only. On deposition 

 of the eggs the outer membrane is turned wrong side out, and serves 

 to attach the eggs to foreign substances. These conditions have been 

 redescribed by Cunningham ('86). Hoffmann ('81) found that the zona 

 is differentiated into two layers in all adhesive eggs, the outer portion 

 being ultimately transformed into a viscid mass. 



Ryder ('86) describes a peculiar arrangement of the egg membranes 

 of Ictalurus albidus. He says : " The egg-membrane is double, that is, 

 there is a thin inner membrane representing the zona radiata, external to 

 the latter and supported on columnar processes of itself, which rest upon 

 the inner membrane j there is a second one composed entirely of a highly 

 elastic adhesive substance. The columns supporting the outer elastic 

 layer rest on the zona, and cause the outer layer to separate very dis- 

 tinctly from the inner one." I have found similar conditions in Amiurus 

 catus (Plate II. Fig. 45), but am inclined to think that the two mem- 

 branes represent the outer and inner portions of the zona radiata ; for 

 the outer shows the striations peculiar to the zona, and the columnar 

 layer is of varying thickness. The inner membrane, being closely asso- 

 ciated with the yolk, would cling to it when the yolk contracts; the 

 protoplasm in the pore-canals being partially withdrawn would give rise 

 to these columnar processes. Where the two membranes were separated 

 for a considerable distance, the columnar structure was destroyed. Simi- 

 lar conditions obtain in the eggs of Clupea vernalis, but in this case the 

 columnar structures lie between the zona radiata and a thin outer homo- 

 geneous layer which is in contact with the granulosa. There cannot be 

 the least doubt concerning the meaning of the columnar layer in Clupea 

 vernalis, for the two membranes lie directly in contact in some parts of 

 the egg. The peculiar structures in Ictalurus and Amiurus doubtless 

 have an origin and meaning similar to that of Clupea. 



The eggs of all the species of fishes examined by me possess a perfo- 

 rate zona radiata. The radial striae could never be made out on the 



