THE GASTR^A THEORY AND ITS SUCCESSORS. 9/ 



forming the micromeres, and the large spherules project 

 into and fill what would otherwise be the blastula 

 cavity. In consequence of this there can be no im- 

 migration, but . the same result is achieved by the 

 micromeres growing round and enclosing the large 

 micromeres. An indication of the possibility of such 

 a process can be seen in some of the Hydromedusae as 

 Laodice, where the migrating cells are much larger and 

 of a different structure than their fellows, and Met- 

 schnikoff has also described a more pertinent case in 

 Polyxeiiia, in which the segmentation is as a rule equal, 

 resulting in the formation of a morula by precocious 

 immigration, but occasionally the segmentation becomes 

 decidedly unequal, so much so that the process of 

 endoderm formation resembles closely epibole. I have 

 noticed in Cyanea, in which the segmentation is appar- 

 ently equal, an exceptional case in \Vliich one pole of 

 the rotating blastula consisted of a few large cells, while 

 the other was formed by a number of smaller ones. 



The invaginate gastrula I do not then consider to be 

 a phylogenetic form — the Gastraea never existed. The 

 ancestor which must take its place is the Parenchy- 

 mella. This ancestor is perhaps reproduced in the 

 ontogeny of the Metazoa in the structure which results 

 from the closure of the blastopore. This is a phe- 

 nomenon which frequently occurs, which is difficult to 

 explain under the Gastrasa theory. It is on the other 

 hand what might be expected from the Parenchymella 

 theory, as I have endeavored to extend it. Where 

 epibole occurs, owing to the nature of the process, the 

 breach made in the walls of the blastula by the immi- 

 gration is not closed at once as in typical immigration. 



