564 EEPORT OF COMMISSIONER OF FISH AND FISHERIES. [1 lOj 



as a portiou, perbaps the most of it, is concerned in the development of 

 the subnotochordal vascniar tract of the tail. In Fig. I, which is a 

 section throngh the caudal knob cm of an embryo of the same age as 

 that shown partly in K, the muscle plates W22? on either side of the uoto- 

 chord are found to be continuous with the caudal mass and to embrace 

 between them the swollen, undifferentiated, caudal end of the chorda ch, 

 which is somewhat quadrate in section at this time. The hind gut when 

 finally formed as the end of the tail grows away from it, is perceptibly 

 swollen and ends blindly. Dorsally, between the muscle plates, the 

 caudal end of the neurula n is quite distinct in sections a little forward 

 of the tip of the outgrowing tail, as shown in Fig. I. It is solid, however, 

 like in the cross-section from the body of a somewhat younger embryo 

 shown in Fig. H. The caudal end of the neurula cannot, however, be 

 traced over the end of the tail at this stage, but ends in the same api- 

 cal mass of undifferentiated cells as the muscle plates, chorda and post- 

 anal strand of cells. The only differentiation of layers at the tip of the 

 caudal knob, in fact, is the skin or epiblastic stratum which covers it. 

 In order to understand in part the means by which this arrangement 

 of parts is established, we will be obliged to look closely into the man- 

 ner in which the blastoderm grows over and incloses the yelk. The 

 large size of the latter necessitates such a spreading of the blastoderm, 

 because up to the time that the tail buds out, the yelk diminishes but 

 slightlj^ in volume and does not exhibit any signs of segmentation like 

 the deutoplasmic pole of the frog's ovum. In Fig. D a blastoderm is 

 shown in outline at two consecutive stages of growth, in order to illus- 

 trate the fact that as the rim r is pushed out to the position r' the em- 

 bryonic shield es is lengthened towards hi by what would appear to be 

 a concrescence of the rim r in a line with the primitive groove. This 

 progressive fusion of the blastodermic rim along the neural axis lengthens 

 the embryo posteriorly. While this appears to be the fact, it is not to 

 be forgotten that, inasmuch as the embryonic cells have certain powers 

 of movement or translation conferred upon them in virtue of a power 

 of more rapid growth in one direction than another, which is again de- 

 pendent upon the operation of certain hereditary and fixed laws of 

 cleavage, all of which is to be considered, no less than the histological 

 forces which make the concrescence spoken of possible. It is at first 

 hard to understand in what manner the whole of the rim of the blasto- 

 derm is incorporated into the body of the embryo fish, but of this we 

 have such overwhelming proof in observed fact that it will be unneces- 

 sary to appeal to other evidence. His and Eauber, who were the first 

 to clearly describe this process of concrescence or j)recession of the rim 

 of the blastoderm, have been criticised by Balfour (Comp. Embryol. II, 

 p. 254). From what the latter remarks it is evident that he never wit- 

 nessed the closure of the blastodermic rim at the vegetative or caudal 

 pole of the living Teleostean ovum, in some of the rapidly developing 

 forms of which it may actually be seen in the process of transformation 



