Experiments on the Early Development of the Amphibian Embryo etc. 623 



This retardation renders it possible to follow more easily the 

 exact method of development, and we can determine the following 

 facts : 



1. There is no closing of the blastopore previous to the formation 

 of the (temporary) anus as in normal development, but the yolk 

 plug remains exposed until after the external gills and the heart 

 have been formed. By this time the primitive gut is well developed 

 and it's different parts are performing their individual functions. 

 Hence, in these embryos the formation of an anus capable of function- 

 ing as such is not at all dependent upon a previous closure of the 

 blastopore. 



2. The anus certainly occupies the anterior portion of the blasto- 

 pore so long as the yolk plug remains exposed. This seems at 

 first sight a direct contradiction of the results obtained by Morgan, 

 Clarke and others, but it will be found upon further examination 

 that it really supports their statements. Let it be remembered that 

 the retardation is more effective upon the passive yolk cells than 

 upon the active pigment cells. Hence we may expect to find any 

 process in normal development intimately connected with these yolk 

 cells very much delayed. It might even be possible for one stage 

 in the process to occupy a considerable period of time, and that is 

 just what occurs in the present instance. 



In the normal development of Amblystoma, the archenteron forms 

 a cavity which is large and deep anteriorly, but posteriorly narrows 

 from before backward as it passes down to the anus, while it re- 

 mains wide from side to side. This was first pointed out by Morgan 

 and has been amply confirmed during the present investigations. It 

 is well shown in fig. 21, which represents a longitudinal median 

 section of a normal embryo just after the closing of the blastopore. 

 We see that in front the cavity extends down almost to the ventral 

 surface of the embryo, while posteriorly it is pushed upward to the 

 dorsal surface. The whole ventral portion in this posterior region 

 is occupied by the mass of yolk cells which gradually withdraw 

 into the blastopore as the latter closes. This closure of the blasto- 

 pore, as recently determined by Eycleshymer and Morgan for 

 Amblystoma, and by Jordan for the Newt takes place as follows: 

 There is first a convergence of cells from all parts of the margin 

 at a uniform rate. This process is gradually modified by the more 

 rapid movement of cells on either side, which press in toward the 

 mid-line until the blastopore becomes a longitudinal slit. Soon they 



