184 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



as ought to have exhibited it. In fact, the tract from which the intes- 

 tine originates is independent of this outer yelk wall from the first. The 

 rudiment of the intestine, before the development of its internal cavity, 

 is merely a flat band of cells somewhat thicker in the middle line than 

 at its edges, and lies just below the tract in which the aorta and cardi- 

 nal veins are afterwards excavated. Upon referring to some of my 

 notes, bearing the date of February 27, 1882, in regard to tbe struct- 

 ure of the yelk sac of the land-locked salmon, I find the following re- 

 corded: "As to the structure of the yelk sac, in making a dissection 

 of a lively embryo, in a neutral salt solution, the epithelial (epiblastic) 

 layer was found to be quite free from the yelk, so that it could be stripped 

 entirely off from the surface of the latter." It evidently was not 

 continuous with the subjacent layer traversed by the complex blood-ves- 

 sels of the yelk, but between the two there was an exceedingly thin ser- 

 ous space. "On the ninth day after hatching, large numbers of red 

 blood corpuscles were still found in the pericardiac space. Later, in 

 diseased, or rather in what were probably injured specimens, numbers 

 of which were kindly brought me by Mr. Fred Mather for study, I found 

 large quantities of blood-cells in the serous space between the external 

 or epiblastic and somatopleural covering of the sac and tbe vascular 

 layer. In some cases the posterior portion of the former was abnor- 

 mally much distended, so that a large cavity was developed." To con- 

 tinue the reproduction of my notes, however, I further stated : " Beneath 

 the outer layer and forming the inner wall of the serous space around 

 the yelk, came the vascular hypo-blastie stratum in which the vitelline 

 network of blood-vessels was developed. This, like the outermost 

 layer, could be removed entire from the contained yelk. The seg- 

 mentation cavity, with which I identify one of the serous spaces so 

 resulting, may be either between the epiblast and vascular splanchno- 

 pleural layer, or between the latter and the yelk." But the homologue 

 of the segmentation cavity is probably the latter. Inside of the vas- 

 cular layer I encountered the yelk vesicle proper, comparable with the 

 palish amber layer of the shad. In its superficial portion I find the oil 

 spheres immersed. This stratum in fact is the " couche hwmatogme" 

 of Vogt, which is as well developed in the embryos of CoregonuH albus 

 of our lakes as in the European species, studied by the versatile natu- 

 ralist of Geneva. Here as in many other species there is a tendency of 

 the blood channels to present the appearance of irregular wide passages 

 over the yelk, somewhat lacunar in nature. This feature is observed, 

 however, only in such as have a vitelline circulation, as, for example, 

 in embryos of Apeltes, Tylositrm, Carassius, Itlus, Fundulus, Esox, Go- 

 arces, Sabno, etc., and not in Alosa, Cybinm, Parephippus, Pomolobus, 

 Gadus, which arc without a vitelline vascular system. But these two 

 types run into each other, for in some the intestinal or portal or else the 

 median subintestinal system of vessels may hereafter be found to take 

 a share in the process of blood developmei t. 



