BULLETIN OF THE UNITED STATES FISH COMMISSION. 199 



of the hind-gut below, so that the gastrula is practically but obscurely 

 realized by means of a strand of cells which connects the neurula and 

 hind-gut. From Balfour's account I infer that the development ofLepi- 

 dosteus is essentially similar. The hind-gut develops from behind for- 

 wards, and almost immediately after the tail commences to bud out the 

 vent appears beneath the latter. The gut has absolutely no connection 

 in any part with the yelk-sac, and the liver appears as an enteric thicken- 

 ing at first, and afterwards as a ventral diverticulum of the intestine, 

 lying upon the upper and hinder aspect of the yelk. The air-bladder is 

 a dorsal outgrowth of the intestine, which originates a little way behind 

 the origin of the liver, from what may be termed the anterior duodenal 

 region. The heart develops as a simple tube in the pericardiac region, 

 and almost from the first communicates with the segmentation cavity. 

 The body up to the time of the closure of the blastoderm grows from 

 behind forwards from the edge of the latter, adding somite to somite 

 behind with the progress of development. In some cases the rim of the 

 blastoderm commences to segment into muscular somites even before 

 the closure of the blastoderm. 



Comparing this form of development with that of the Amphibian and 

 JMarsipobranch, we find that we are forced to look upon the blastoderm 

 of the Teleost, exclusive to the yelk, as their complete morphological 

 equivalent, and that the Marsipob ranch justifies the comparison in the 

 peculiar way in which the body of the latter grows from behind for- 

 wards out of the caudal mass, just as the embryo Teleost grows from 

 behind forwards from the edge of the blastoderm ; so that, although in 

 other respects there are some essential differences, such as in the rela- 

 tion and homologies of the neurula, hind-gut, and the blastopore, other- 

 wise there is an evident similarity, which may give us a key to the com- 

 prehension of why it is that the embryo Teleost develops at and from 

 the edge of the blastoderm. 



It was not my intention, however, to enter into an embryological dis- 

 cussion in this place so much as to show that the development of the 

 teleost was peculiarly conditioned mechanically in consequence of the 

 peculiar organization of the egg, and that that had much to do in de- 

 termining its mode of development. The real cause of some of its singu- 

 larities appears to be the presence of the relatively enormous yelk which 

 must be included by the blastoderm in order to be absorbed. The pro- 

 cess of yelk-absorption itself in the teleost offers some strong contrasts 

 with that described by writers on the absorption of the yelk of birds 

 and Elasmobranchs, as we shall see in another place. 



A few words more on the mechanical conditions presented by the 

 ova of several genera of fishes, and I have done with this part of the 

 subject. I am familiar with the ova of four genera of salmonoids; 

 Salmo, Salvelinns, Coregonus, and Osmerus; all are characterized by 

 an abundance of oil drops embedded in the vitellus, but most abund- 

 antly just immediately under the germinal disk, where indeed most of 



