No. 2.] CLEAVAGE OF ARENICOLA CRISTATA. 



75 



cal sections show that they too are slowly elongating inward, 

 and that their nuclei are sinking farther and farther from the 

 surface of the egg. Finally, these cells extend as a column 

 through the egg to the lower surface of the ectoderm of the 

 upper pole (Fig. 9, En}. As is seen by this figure, the 

 amount of true overgrowth by the ectoderm is really slight, 

 only the latest stages of the closure of the blastopore being 

 effected in this way. The blastopore is at first nearly circular, 

 but during the later stages of gastrulation it becomes dorso- 

 ventrally elongated. To render this change in shape clear, the 

 growth of the somatic plate must be explained. 



The cells of the somatic plate, the derivatives of X, are at 



/v 



FIG. 7. 



FIG. 8. 



this time increasing rapidly in number, and the plate is extend- 

 ing itself posteriorly and laterally, the lateral growth being 

 much more rapid than the posterior. It is this growth of the 

 somatic plate that forces the sides of the blastopore together 

 and finally causes them to meet. Fig. 10 shows the rela- 

 tions of blastopore and somatic plate at a stage shortly before 

 closure. The lip of the blastopore, except the very narrow 

 space dorsally where it is formed by the derivatives of X, con- 

 sists of twelve cells (St}, four along each side and four ven- 

 trally. The cells are all derivatives of the third quartet of 

 ectomeres. I have followed their lineage exactly. At first they 

 alternate with cells of the second quartet, but by a series of divi- 

 sions they finally pass below these and then, dividing laterally, 

 shut them off entirely from the actual lip of the blastopore. All 

 the other cells shown in Fig. 10 are derivatives of X. 



