CRANGON VULGARIS. 127 



figured to siipport this view but they are to me far from 

 conclusive. Indeed they do not in the least appear to 

 Support hira ; but on the other band seem to be in füll ac- 

 cord ^vith the explanation advocated below. In a second 

 portion of bis paper (Ital. edit., pp. 191-194; French 

 edit., 267-268) he says that this mode of origin is in füll 

 accord with the theory of tbe gastrula, that the line of in- 

 vagination is to be regarded as an elongate blastopore, 

 the invaginated tissue is to be regarded as a meso-ento- 

 dermal layer and that gastrulation is here rudimentary 

 rather than falsified, the nutrition rendering a perfect gas- 

 trula unnecessary. Concerning the yolk cells, which are 

 clearly like those of other arthropods, our author says 

 that he has never seen them take any part in the forma- 

 tion of the endoderm. The lack of method in the arrange- 

 ment of bis ligures renders it a difficult task to follow 

 the sequence of bis sections ; but a study of figs. 3, 13, 

 14, 29 on pl. YII, figs. 6, 7 on pl. VIII, as well as many 

 others, would seem to show beyond a doubt that the 

 endoderm in the bee was formed by a migration of the 

 amoeboid yolk cells to the surface of the yolk and their 

 arrangement there into an epithelial layer inside the mes- 

 oderm and resting directly upon the yolk. The appeal to 

 figures like pl. YII, fig. 4, to prove that the yolk cells take 

 no part in the formation of the endoderm is far from con- 

 clusive. Further studies on the development of the bee 

 will be necessary before this question can be regarded as 

 settled. Grassi thinks that the results of the Hertwigs 

 and TlchomirofF ('82) can be explained to agree with bis 

 views, but this does not readily appear. 



From this review (which is not exhaustive) we see that 

 there have been almost as many theories as writers con- 

 cerning the origin of the endoderm in the hexapods and 

 its relation to the gastrula. It has not been our purpose 



