[71] EMBRYOGRAPHY OF OSSEOUS FISHES. 525 



appropriation of its cells to form the caudal end of the embryonic axis. 

 When on the eve of closure, the limbs of the rim of the blastoderm in 

 Tylosurus form an acute angle with each other, and the yelk blastopore 

 has the form of a wide oval with the narrow end next to the caudal 

 swelling. In Elacate the evidence in relation to the concrescence of 

 the rim of the blastoderm in the middle line is very striking. Here, the 

 limbs of the rim of the blastoderm on the eve of closure, where their 

 substance is continuous with that of the muscle plates anteriorly, 

 form an acute angle with each other, and there is no caudal swelling 

 intervening between them as in Tylosurus. Not only is it evident in 

 this case that an actual concrescence of the limbs of the rim of the 

 blastoderm occurs, but it is also plainly evident that a transverse seg- 

 mentation into segments has occurred in the lower layer of the limbs 

 before their concrescence. The segmentation affects only the lower or 

 somatic layer of the blastodermic rim and extends some distance behind 

 the caudal end of the embryonic axis already formed. This is the only 

 instance in which I have found evidence of a normal process of concres- 

 cence of the rim of the blastoderm along the median line in embryos of 

 osseous fishes before the formation of the caudal plate. The concres- 

 cence, therefore, takes i)lace also in the plane of the nervous axis as 

 well as in the enteric. It would appear as if the yelk blastopore in 

 such cases migL t be the true blastopore of the gastrula stage of devel- 

 opment. It is remarkable, however, that I should meet with such a 

 state of affairs only in Elacate and not in other forms, as regards the 

 fate of the inner edge of the blastodermic rim. Only in Elacate have 

 I ever met with any evidence of direct marginal apposition, concrescence, 

 and convergence of the blastodermic rim on the eve of the closure of 

 the blastoderm ; in other forms it closes as a round pore, as in Gadus^ 

 Cyhmm, and Alosa, and segmentation into muscular somites of its lower 

 layer never occurs during its closure to form the caudal plate. 



In these ways the rim of the blastoderm is completely nsed up in the 

 different species to form the caudal end of the embryo, the most of its 

 substance being finally converted into the muscle-segments of the tail. 

 But the growth of the tail outward is a most remarkable phenomenon, 

 in that there is as yet in some forms no vascular system whatev. r for 

 the conveyance of nutrient matter, in spite of which the tail continues 

 to elongate, evidently gaining bulk mainly to build np the lateral mus- 

 cular masses the material for which must of necessity be transported 

 outwards and backwards somehow from the yelk or other pre-existing 

 tissue. The way in which this is accomplished is not clear to me ex- 

 cept upon the theory of growth proposed by De Bary and Eauber. 

 They regard cell-division as a consequence of growth, not growth a 

 consequence of cell-division. Then, if we suppose with Eauber that 

 cellular protoplasm has a structure consisting of vacuoles or lines ra- 

 diating from a center, which favor intussusception of plasma from inter- 

 cellular spaces, we may i)erhai)s have an approximate explanation of 



