MUSEUM OF COMPARATIVE ZOOLOGY. 97 



remnants of the follicular cells." He further adds, that occasionally 

 it appeared as though there were between these two membranes a thin 

 flat expanse of granular protoplasm in which the filaments terminated. 



His (73, pp. 14, 15), who examined perch eggs in April, also confirms 

 the opinion of Kolliker, and gives a figure to show the relation of the 

 radial processes to follicle cells. The radial streaks consist, he says, of 

 a turbid substance, which stains in osmic acid, and is continuous with 

 conical nucleated bodies which form a continuous layer between the fol- 

 licle wall and the outer surface of the capsule. Kolliker is therefore 

 right in considering the layer as "granulosa," and the capsule as its 

 product. 



Brock ('78, p. 556) gives the following clear, and I believe correct, 

 account of the capsular membrane of the perch. The follicular cells, 

 which at first are in close contact with the young egg, are raised up from 

 the zona radiata by the developing gelatinous layer, and with the advan- 

 cing growth of that layer are drawn out on the side toward the egg into 

 long processes which can be followed up to the zona. In older eggs 

 these follicular cells, separated by considerable intervals, (an indication 

 that their multiplication soon stops,) lie in shallow depressions of the 

 gelatinous capsule, and with their lower pointed ends continuous with 

 the processes. These appear to end at the zona with conical enlarge- 

 ments, but the author will not affirm that this is a constant feature. 

 Brock also maintains that Serranus hepatus has a very similar gelatinous 

 capsule. The follicular cells, however, are very peculiar. They form a 

 network of thin flat cells, which are in contact with each other only by 

 means of lateral processes, while perpendicular processes, which are 

 sometimes branched and exceedingly fine, can be traced through the 

 jelly to the zona radiata. Concerning the development of the capsule, 

 nothing is known. 



So far as regards the capsular membrane of the perch, Hoffmann ('81, 

 pp. 19, 20, 27-29) comes to totally different conclusions from Brock, and 

 expresses views which seem to me untenable. In October, ovarian eggs 

 from 600 to 700 /x in diameter possess a membrane 5 /x thick, which is 

 composed of two layers of nearly equal thickness. "The inner is the 

 true zona radiata ; the outer is composed of very numerous, small knob- 

 like projections, which stand very close together and correspond exactly 

 to the villi of the cyprinoids. On the free surface of the conical villi lie 

 the granulosa cells." 



It cannot be denied that the figure cited (I. c, Taf. I. Fig. 9) corrob- 

 orates the description given. But there is one fact which I should im- 



VOL. XIX. — NO. 1. 7 



