298 HELEN DEAN KING. 



from the surface as I have found to be the case in the embryo of 

 Bnfo lentiginosus. At a later stage of development, vegetative 

 cells grow up from the sides of the archenteron, and gradually 

 cover up the invaginated animal cells which now form an 

 unbroken sheet of mesoderm across the dorsal wall of the arch- 

 enteron. A portion of this mesoderm in the mid-dorsal line is 

 subsequently cut off from the lateral mesoderm to form the 

 noto chord. 



In the posterior region of the embryo of Bnfo Icntiginosns a 

 portion of the dorsal plate of cells which forms the mid-dorsal 

 wall of the archenteron becomes cut off from the rest of the layer 

 to be added to the notochord. If we attempt to trace the origin 

 of this dorsal plate, we find that it is composed of cells invagi- 

 nated from the surface of the egg before there was any division 

 of the cells into ectoderm, mesoderm and endoderm. These 

 invaginated cells form apart of the upper wall of the archenteron 

 for a comparatively short period of development only, and those of 

 the cells that are subsequently added to the splanchnic mesoderm 

 soon lose their identity entirely, and cannot be distinguished 

 in any way from the other cells of the mesoderm. The later his- 

 tory of the chorda-endoderm cells I have not followed. 



As the endoderm cells that grow up from the sides of the 

 archenteron and meet under the notochord are unquestionably 

 derived from the yolk portion of the egg, the archenteron even- 

 tually becomes lined throughout its whole extent with yolk cells, 

 and, therefore, the result is the same as if the archenteron was 

 originally formed by a splitting between yolk cells as is believed 

 to be the case by Robinson and Assheton (17), Houssay (9) and 

 Moquin-Tandon (14). 



According to Morgan, Wilson (21), Eycleshymer and others, 

 there is an invagination of surface cells at the dorsal lip of the 

 blastopore during the gastrulation of the frog's egg, and these 

 invaginated cells come to form a part, if not all, of the dorsal 

 wall of the archenteron in the posterior region of the embryo. 

 In subsequent development, as the studies of Schwink and of 

 myself show, these invaginated cells are not covered over by an 

 upward growth of yolk cells from the lateral walls of the archen- 

 teron as is the case in the toad embryo. A few of these cells 



