ON THE CiASTRULATION IN TETROMYZON. 2^> 



111 ;iii ovum willi ventnilly shifted segmentation cavity {Figs. Or/ 

 and i)d), the median sagittal section of which is represented 

 in T^Vy. 20, (lie micromeric layer assnmes epithelial character 

 throughout and is at its ventral limit (iQ sharply marked off 

 against the macromeric part, just like its dorsal limit at the 

 stage when the invagination was about to take place {Mffs. 14 

 and 23). 



Laterally the differentiation proceeds in a similar way as in 

 the median zone just stated, as will very well be seen in the 

 three sections represented in Figs. 25, 26 and 27. Of these 

 three figures, the last (Fig. 27) is a transverse section through 

 an ovum a little younger than that shown in Fig. 8a or Fig. 

 19 ; the lower depressed part (bjj.) shows a part of the blastoporic 

 funnel ; on either side of the macromeric cell-mass, we see very 

 well its connection with the micromeric layer which presents 

 the typical character of a columnar ejiithelium. The two other 

 figures {Figs. 25 and 2G) w^ere drawai from a series of horizontal 

 longitudinal sections through an ovum of about the same age 

 as that shown in Figs. 7a and 7d or in Fig. 18. In the section 

 taken from the dorsal part {Mg. 25), the differentiation of the 

 animal layer already reaches the junction-line, while in the other 

 section {Fig. 26) from the ventral part, the layer is still many 

 cells in thickness. 



Great care was taken to make out in every possible way the 

 actual mode by which the aggregated elements of the micromeric 

 layer are brought into the epithelial structure ; but all my efforts 

 have been fruitless after all. Although wanting positive evidence, 

 we can make some surmises from another point of view. There 

 can be only two possible modes : either the outermost cell-row of 

 the indifferent animal layer comes to be converted in situ into 



