BULLETIN OF THE UNITED STATES FISH COMMISSION, 149 



It is this structure which Cuvier and Valencienues * alhide to in the 

 following- words : " Le vitellus a deux tuniques, completes I'uue et I'autre, 

 quoique tr^s-fines. L'exterue se continue par sa lame exterieure avec 

 le peau et par I'interieure avec le peritoine; I'iuterne, tres-vasculeuse, se 

 continue avec les membranes des iutestins et leur tunique peritoneale ; 

 la cavite donue directement et visiblement dans celle do I'intestin, et la 

 matiere du Jaune y afflue." This quotation shows that Cuvier, to whom 

 it is in all probability to be ascribed, was aware of the existence of a 

 double envelope over the yelk, but in no instance have I found what 

 could be considered a communication between the cavity of the intestine 

 and the vitellus. \^on Baer is stated by Balfour t to speak of two types 

 of yelk-sack, one inclosed within the body wall, and the other forming 

 a distinct naked ventral a])pendage of the embryo, from which it is clear 

 that the great German embryologist never clearly understood the man- 

 ner in which the vitelline globe or yelk is inclosed by the blastoderm. 

 Nor can I confirm Lereboullet's view that a connection of the vitellus 

 and cavity of the intestine exists between the stomach and liver, because 

 the stomach is not ordinarily differentiated in young fishes while the 

 yelk persists. In the young California salmon {Oncorhynchus) the mus- 

 cle plates grow down on either side between the epiblast externally and 

 the splanchnopleure internally. The hypoblast covering the remains of 

 the yelk is traversed externally by a network of blood vessels, as may be 

 learned from an examination of transverse sections through advanced 

 embry^os. In this way it results that the segmentation cavity is oblit- 

 erated or filled up during the latter part of embryonic life by the down 

 growth of the somatopleure and splanchnopleure between the epiblast 

 and hypoblast, both the splanchnopleure and somatopleure having 

 originated from the mesoblast. This, I believe, is essentially the mode 

 of development of the blastoderm in teleostean embryos and the way in 

 which the segmentation cavity disapjiears. Any view, however, accord- 

 ing to which the yelk is looked upon as a mere nutritive vesicle, and not 

 at all times in intimate organic union with the embryo, betrays a want 

 of comprehension of the way in which the teleostean ovum is developed 

 in the ovary, the manner of the formation of the germinal disk, the de- 

 velopment of the blastoderm and inclusion by it of the vitellus, and, 

 finally, of the relation of the heart and blood system to the vitellus. 



STRUCTURES DEVELOPED IN ' THE EMBRYO MACKEREL FROM THE 

 ELEVENTH HOUR TO THE TIME OF HATCHINO. 



Starting with the stage represented in Fig. 9, when the medullary 

 canal m, in section, the notochord or cartilaginous axial rod c/i, and the 

 somites or muscle plates on either side of the medullary canal are de- 

 veloped, it is apparent that the embryo by the fourteenth hour, repre- 



•Histoire Naturelle des poissons, tome i, p. 399. Paris, 1828. 

 t Comparative Embryology, vol. ii, p. 65. 



