130 BULLETIN OF THE 



Kblliker ('58) corrected the mistake made by Haeckel as to the posi- 

 tion of the filaments. 



Hoffmann ('81) found filaments on the eggs of Heliasis, Gobius, and 

 Blennius. 



Ryder ('82 c ) described the filaments of the eggs of Belone longirostris,' 

 and, in passing, mentioned the probability of their existence in the eggs 

 of Mugil albula. He afterwards ('86*) found them on the eggs of Fun- 

 dulus heteroclitus, and has also ('83) shown that the eggs of Menidia 

 (Chirostoma) notata are provided with four of them. 



I have examined eggs of Fundulus heteroclitus at intervals of about 

 two weeks from October, 1887, till June, 1888. The eggs undergo 

 scarcely any change between October and April. A series taken between 

 April 1st and June 1st shows all the stages covered by the longer period. 

 The filaments can best be studied in fresh material. They appear in 

 the form of hyaline threads, which are more highly refractive than any 

 other part of the egg membrane. In an ovary of October 27 there were 

 filament-bearing eggs in three stages of development. 



In the smallest eggs — about 0.16 mm. in diameter — in which 

 filaments can be seen (Plate I. Fig. 1) they appear as hyaline dots, or as 

 conical bodies with rounded bases, uniformly distributed over the entire 

 surface. They either lie wholly below the granulosa, or the tips of the 

 longer ones may lie in between the granulosa cells (Fig. 8). In this 

 stage the diameter of the threads is much greater than the thickness of 

 the membrane, which can scarcely be distinguished in sections. I was 

 not able to discover sheaths enveloping the filaments such as Haeckel 

 describes for the Scomberesocidee. In other slightly larger eggs belong- 

 ing to this same stage of development (Figs. 2-4, 6) the filaments are no 

 longer conical, but appear in the fresh condition as short, curved threads 

 equally blunt at both ends. 



In the second stage, the eggs being intermediate in size between those 

 just mentioned and the largest, the threads (Fig. 5) are of about the 

 same thickness as those of the earlier stage, but they are much longer, 

 and taper near the free end. They do not seem to be closer together 

 than in the smaller eggs. The filaments are bent in a more or less 

 regular manner, first to one side and then to another. On stained sec- 

 tions it was to be seen that the threads usually follow the margins of the 

 granulosa cells, and that they are correspondingly curved (Fig. 6). 



On the largest eggs — about 0.4 mm. in diameter — the filaments are 

 much longer, and cover about as much of the surface of the egg as they 

 leave exposed. They are so long and so tortuous that it is almost 



