MUSEUM OF COMPAEATIVE ZOOLOGY. 131 



impossible to follow a single filament throughout its whole length. It 

 often happens (Fig. 10) that several filaments are parallel to each other 

 for a considerable distance. In sections the filaments are found to lie 

 in between the bases of the granulosa cells, and also to rise between 

 these cells nearly to their outer sui'faces. 



In the ripe ovarian eggs the basal ends of the filaments pass directly 

 through the granulosa layer, and the greater part of the filament thus 

 comes to lie between the outer portions of the granulosa cells, or even 

 quite outside of them (Fig. 9). The regularity of their windings can- 

 not be seen as well as in eggs of the third size. The filaments are of 

 varying lengths, but most of them are several times as long as the 

 diameter of the egg. The distances between filaments are not materially 

 altered during the growth of the egg ; but since the surface of the egg 

 increases during its development to many times the size which it had 

 when the filaments first appeared, the total number of the latter must 

 also be greatly increased. The earliest stages in the formation of new 

 filaments would be difficult to find after the egg has reached its second 

 stage, because they would be hidden by the larger filaments. 



In ripe eggs forced from the ovary, the filaments extend out from the 

 egg for some distance, and then form a network, several filaments deep, 

 over the whole surface (Fig. 11). 



Concerning the origin of the filaments it may be said that they do 

 not have any connection with the granulosa cells at any stage of their 

 growth (Figs. 3, 4, 6, 8). In tangential sections it is seen that they 

 arise at places corresponding to the boundaries between two or three 

 cells. In a ripe egg examined in the fresh state under pressure (Fig. 12) 

 indistinct processes are seen to radiate from the base of each filament, 

 forming a stellate figure. In no case, either in fresh specimens or sec- 

 tions, could the filaments be traced into the substance of the zona radi- 

 ata. They are outgrowths from a thin membrane which lies outside of 

 the zona and is formed before the latter, not processes of the zona itself. 



When the filaments first make their appearance, the egg membrane, 

 as stated above, is much thinner than the diameter of a filament, and 

 the granulosa cells are lens-shaped, barely touching by their mar- 

 gins (Fig. 8). In the largest eggs found in the ovary of April 2d, the 

 granulosa was about 8 //. thick, but the egg membrane had only reached 

 the thickness of 2 fx. That it is radially striate is rather to be in- 

 ferred than directly seen. In places the outer surface shows slight 

 elevations at regular distances, which I believe to be prolongations of 

 granulosa cells sunk into the radial canals (Fig. 13). While the largest 



