138 BULLETIN OF THE 



a subsequent part of his paper (pp. 29-31), where he gives an account of 

 the development of thb ovarian egg, his statements seem to be conflict- 

 ing as to the relation of the spiral canals to the granulosa cells, but at 

 the end he repeats that the canals are outgrowths of cells as stated by 

 Kblliker. The interstitial matter (Zwischensubstanz) is arranged in 

 lamellae which are parallel to the surface of the egg. By the swelling 

 of the lamellae fissures arise which have the appearance of processes 

 from the canals. 



I have studied the ovarian eggs of Perca killed in October, February, 

 and May. It is probable that the formation of the egg membranes is 

 less advanced in the American species of this latitude than in the Euro- 

 pean species at a corresponding season. 



Contrary to Hoffmann's statement that in October the capsular layer 

 and zona are of equal thickness, not a trace of the capsular layer, dis- 

 tinct from the granulosa, could be found at this time of the year. The 

 zona is well developed, and is differentiated into two layers of about 

 equal thickness. The outer layer is radially striate, while the inner 

 appears to be structureless. The granulosa cells lie immediately in 

 contact with the zona radiata (Fig. 23, Plate III.). I have not been 

 able to find the micropyle in October eggs. 



In February the zona remains practically as it was in October, but 

 vacuoles — which may be caused by the method of treatment — are to 

 be seen in the inner portion (Fig. 25, Plate III.). They are much flat- 

 tened radially, and thus suggest an approach to a stratified condition of 

 this portion of the zona. The radial striations of the outer half of the 

 zona are more strongly marked than at the earlier stage, and much 

 fainter striations may also be seen traversing the inner half. The lat- 

 ter, though less distinct, are just as numerous as, and continuous with, 

 those of the outer half. At this date the capsular layer is already well 

 developed, but it has attained only half the thickness which it has in 

 May. 



Up to the month of May the thickness of the zona radiata has not 

 changed, but the pore-canals can now be more readily traced passing 

 entirely through it. They still remain much more evident in the outer 

 than in the inner half of the zona. This is due to the greater calibre 

 of the canals, not to their being farther apart in the outer half. 



The different descriptions of the capsular layer are in part due to the 

 fact that it presents different conditions according to varying circum- 

 stances. The radially arranged spiral structures traversing this layer 

 arise as funnel-shaped tubules, one beneath each cell of the granulosa. 



