CONCRESCENCE IN EMBRYO OF CRYPTOBRANCHUS. 259 



the anterior end forming the mouth. The separation of the two 

 sides of the embryo, shown in the vertebrates most clearly in the 

 early history of the mesoderm and the nervous system, was 

 primitively caused by the blastopore itself. In the great majority 

 of cases the original mode of closure of the blastopore has been 

 secondarily modified, but the mesoderm and the neural cords or 

 folds still follow the original mode of development, being laid 

 clown separately on either side of the region of the primitive 

 blastopore, and growing together along its line of closure (con- 

 crescence by apposition). In cases where the region of the 

 blastopore is occupied by a large mass of yolk, the process of 

 concrescence is much modified and exaggerated; to this extent 

 only, concrescence is to be regarded as a ccenogenetic character. 



Taking up now, in the light of this theory, our soecial problem 

 of concrescence in the vertebrates, we may note that there is no 

 valid objection to the extension of the theory to cover the 

 vertebrates on the ground that forms like Peripatns and the 

 earthworm are not ancestral to the vertebrates; granting that 

 these forms are probably far removed from the ancestral line 

 of the vertebrates, we may still rely on the fact of common descent 

 from the ccelenterates. Further, we may note that the ancestral 

 mouth of the vertebrates has been lost, and a new one acquired ; 

 the anus represents typically the posterior remnant of the primi- 

 tive blastopore. Wipe out the upper half of the figure showing 

 blastopore closure in the embryo of Peripatns (Fig. 44), and we 

 have left substantially the conditions found in the late gastrula 

 of Cryptobranchiis. If we give a phylogenetic interpretation to 

 the confluence of materials in the dorsal lip of the amphiban 

 blastopore, then this confluence is but a vestige of the more pro- 

 nounced shifting of tissues that caused the constriction of the 

 primitive blastopore. The observed confluence of materials in 

 the dorsal lip of the blastopore is such as this theory of con- 

 crescence would lead us to expect. I have suggested a possible 

 mechanical explanation, w r hich seems not wholly adequate. The 

 facts are not inconsistent with the palingenetic theory of con- 

 crescence. 



The process which I have called convergence is necessitated by 

 the large mass of yolk which greatly delays the completion of 

 the process of gastrulation. 



