THE DKVKLOPMENT OF LEPmOSIREN PARADOXA. 13 



the use of the vague phrase " differeutial growth " — nothing 

 can be said. The absorption of the fluid in the segmentation 

 cavity which is associated by Samassa with the invaginatory 

 process of Amphioxus is excluded as an explanation of the 

 phenomenon here^ as the first obvious result of such absorp- 

 tion would be the collapse of the very thin and delicate roof 

 of the segmentation cavity, and such collapse is conspicuously 

 absent. 



In transverse section the archenteron is seen to be, in this 

 stage, a tube rounded in section — in other words, showing no 

 signs of splitting laterally, and about '2 mm. in diameter, 

 strikingly narrow in proportion to the diameter of the egg as 

 compared with most holoblastic forms. 



Towards the end of period b the archenteron approaches 

 the margin of the segmentation cavity, and now we have very 

 distinct evidence that the growth of the archenteron is not 

 due to splitting, for the cells round its tip become pushed 

 definitely into the segmentation cavity forming a rounded 

 bulging into it (PI. 2, fig. 8). As the process goes on the 

 large-yolk cells become laid up against the original roof of the 

 segmentation cavity, which, already two-layered, alters little 

 in character and will later become definitive epiblast. The 

 further stages in the obliteration of the segmentation cavity 

 I will deal with later. 



c. In the later stages of gastrulation we have certainly to 

 do with a process of true invagination, the end of the archen- 

 teron being always quite smooth and rounded, with cuticular 

 lining, and there being never any trace whatever of splitting 

 (of. PL 3, figs. 11 and 12). The precise character of this inva- 

 gination could only be settled definitely by experiment upon 

 the living egg, and such experiments, though attempted, 

 proved absolutely fruitless on account of the tough egg 

 capsule and the soft nature of the egg contents. From the 

 study of sections ^ of the eggs I am disposed to believe that 



' 111 my account of the external features I pointed out that against the 

 probability of such a backgrowth taking place, was the fact of the blastoporic 

 lip not assuming tlie form of an arc of gradually diminishing radius with its 



