FORMATION OF THE GERMINAL LAYERS IN TELEOSTEI. 201 



be quite separated from the inner one, and to constitute the viscous substance 

 which serves for the attachment of the egg. This layer now frequently pre- 

 sents a laminated structure. From this description it will be seen that 

 Hoffmann regards the viscous layer as a metamorphosed part of the zona 

 radiata. In this case it must be regarded as a secretion of the vitellus. 



Egg Membrane. — Kupffer has described two egg membranes, an outer 

 zona radiata and an inner one, comparable with the true vitelline membrane of 

 other forms. Hoffmann has shown that in their origin these two portions of 

 the egg membrane are very similar in structure, and that they are both to be 

 regarded as portions of the zona radiata ; there is no true vitelline membrane in 

 the herring ovum. My investigations support Hoffmann's conclusions. 



Yolk. — According to Kupffer, the entire contents of the egg envelope are 

 divisible into three parts — 



1. A superficial layer, consisting of strongly refractive shining homogeneous 

 globules *008 to "02 mm. in diameter, which he terms yolk granules (Dotter- 

 korner). 



2. Immediately below the thin superficial layer the yolk consists of larger 

 and less refractive yolk spheres, which have a rounded or egg-shaped form. 

 They vary in size from "05 to "08 mm., and constitute the greater portion of the 

 yolk. These are the Dotterkugeln of Kupffer. 



3. There is also a scanty viscous mass of protoplasm, which is mixed up 

 with the other two. 



According to this observer, the fine granular layer of yellowish protoplasm, 

 which later forms the blastoderm, is not present in the unfertilised ovum, nor 

 is there a germinal disc. 



As Hoffmann has already pointed out, Kupffer was mistaken in supposing 

 that the germinal protoplasm does not already exist in the ripe unfertilised 

 ovum. The reason why it is not visible in the living egg is twofold. In the 

 first place, it does not form a distinct layer around the yolk, as is usually the 

 case in Teleostean fish ova; and in the second, there is so slight a difference 

 between the colour of the yolk spheres and that of the protoplasm, that in the 

 living egg it is impossible to distinguish between them. A section of the unfer- 

 tilised egg shows nevertheless that the germinal protoplasm exists in the herring 

 as well as in other fish eggs. There is, however, this difference, that whereas in 

 fishes generally the bulk of the germinal protoplasm is collected as a distinct 

 layer on the surface of the yolk at the time the egg is ready for fertilisation, in 

 the herring the protoplasm is distributed throughout the yolk. The egg con- 

 tents consist in fact of a mass of delicate protoplasm, in which the yolk spheres 

 are imbedded. This structure is well brought out in sections of ova stained in 

 Mayer's carmine. I have obtained the best results by leaving the eggs in the 

 staining solution from twenty-four to forty-eight hours. Subsequent treatment 



