472 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [18] 



large and few in number. What is meant here by the term yelk spheres 

 must also be explained. We denominate as yelk spheres those large, 

 finely granular masses of protoplasm which are involved in a delicate 

 matrix of germinal protoplasm. The ultimate granules or spherules 

 which compose these spheres or yelk masses are altogether different, 

 vastly more minute, and are the same as what we have called yelk 

 spherules and granules. The large yelk masses are very clearly seen 

 in the shad's egg, that of the herring, and in Elacatc, but not so clearly 

 in any other forms known to me. In many types the yelk is almost 

 homogeneous under a low power. In Cyprinoids it is finely granular, 

 and again wholly different from what is seen in the eggs of Clupeoids 

 and Gadoids. These distinctions are important, and not less so is the 

 distinction drawn between the mode of formation of the germinal disk 

 of the Gadoids and Clupeoids. The germ disk of Gadus is wholly devel- 

 oi)ed from the external germinal layer of the vitellus, that of the Clu- 

 peoids is apparently not entirely so formed, but derives part of its sub- 

 stance from the protoplasmic matrix involving the large yelk spheres 

 or masses below it. 



The germinal pellicle, cortical layer, has certain common characters in 

 all the forms, however ; these are its superficial position and the vesic- 

 ular or refringent bodies it incloses prior to the development of the disk. 

 In the eggs of the shad and of the herring the cortical or germinal layer 

 contains large refringent bodies, different from the vesicles which occur 

 in the same layer in the ova of Oadits, Tylosurus, and Cyhium, in the last 

 of which I have watched their slow disappearance and apparent absorp- 

 tion into the surrounding germinal plasma. In this process they become 

 graduallj' smaller and finally disappear under the iidluence of impreg- 

 nation. It would seem as if these refringent structures were somewhat 

 similar to those seen in the germ of Acipenser ruthenus, and described as 

 islands by Salensky. In Chq)ea and Alosa, as the peri])heral germinal 

 matter is gradually gathered into a depressed, somewhat conical germi- 

 nal disk, the refringent globules or spherules which were originally dis- 

 tributed over the whole vitellus become less numerous and mostly dis- 

 appear, except over a portion of the vitelline surface and in spots usually 

 near the margin of the disk. After the disk has segmented into a large 

 nnmber of cells and advanced somewhat beyond the morula stage, these 

 refringent spherules have disappeared entirely. The space figured un- 

 derneath the germinal disk in the yelk of the egg of Glupea harengus by 

 Kupffer is probably a product of the hardening reagents used, and is not 

 to be regarded as a latehra in the sense of the structure of that name 

 found in the ovum of birds. Actual sections of several species of fishes of 

 very early stages of development have served to convince me of the cor- 

 rectness of this interpretation. Germinal matter which does not take 

 part in segmentation lies below the disk, as shown by Gi^llacher, and ex- 

 tends over the vitellus as an almost structureless membrane, Dotterhant 

 of CEllacher. It is this membrane which protects the vitellus and forms 



