Experimental Studies on Germinal Localization. 261 



justified, I believe, that these factors differ in different types of 

 eggs from the beginning, and that they become steadily more 

 specialized and limited as the development progresses. With the 

 advance of development, accordingly, the response becomes cor- 

 respondingly more limited. 



This is shown by such a series as the following, including the 

 sea-urchin, nemertine and mollusk. In all these the J/2 or Ya- 

 blastomere produce an embryo that closes and gastrulates. In 

 the nemertine or sea-urchin any of these embryos may undergo 

 complete development, since the first two cleavages are symmetri- 

 cal and quantitative, distributing to each cell all the elements of 

 the original system. In the mollusk, however, the AB-half or 

 the A-, B- or C-quadrant, though undergoing certain characteris- 

 tic differentiations, is unable to produce a complete embryo, owing 

 to the absence of necessary specific material contained only in the 

 D-quadrant or the AB-half.^ Beyond the 4-cell stage all of the 

 forms exhibit limitations of potency, not primarily due to decrease 

 of size (as is proved by Zeleny's observations on the upper quar- 

 tet of Cerebratulns) , but to qualitative Internal factors. Cells of 

 the upper quartet, or the entire quartet, produce closed embryos 

 which in the nemertine or mollusk are unable to gastrulate (again 

 owing to the lack of specific material), but in the sea-urchin may 

 do so provided the third cleavage does not exclude a certain 

 amount of entoblast-stuff from the upper cells. The isolated pri- 

 mary trochoblast of the i6-cell stage completes its predestined 

 twQ divisions and differentiates typically except for slight changes 

 in the relative position of the resulting cells; but the remarkable 

 change of position, which in the complete embryo leads to the ac- 

 curate fitting together of the rows of cilia, at first disconnected, 

 to form continuous ciliated rings, fails to take place. Its sister- 

 cell (i^) likewise divides and differentiates typically; but owing 



ij pointed out in the preceding paper that the failure of the CD-half to 

 produce a perfect larva may not improbably be owing to the fact that owing 

 to its great susceptibility, the larva is unable to sustain itself long enough to 

 assume the normal conditon. It is theoretically possible that the same may 

 be true of the AB-half; but the actual facts are that the latter shows from the 

 first certain definite defects that do not exist in the former, the CD-larvce 

 showing merely a lack of the proper proportions. 



