Experimental Studies on Germinal Localization. 223 



grouping of these cells is essentially the same as in the upper half 

 of a normal 32-cell stage; though, owing to the flattening out 

 of the group, the normal position of the cells is somewhat modi- 

 fied and (as in the case figured) the cells are often rather loosely 

 connected. 



In later stages such groups invariably broke up more or less, 

 and no larvae were obtained which had not lost some of the cells. 

 Nevertheless these larvae close up more or less completelv, form- 

 ing irregularly pyriform structures with an apical organ at the 

 smaller end and a group of trochoblasts at the larger one. The 

 largest of these larvae obtained probably represents at least three- 

 fourths of the first quartet. This individual (24 hours) is shown 

 in Fig. 67, drawn from a preparation; in life it swam very ac- 

 tiv^ely about. This larva is clearly a purely ectoblastic structure, 

 and shows no trace of archenteron. It is of an irregular flattened 

 pyramidal form, with an irregular group of apical cells at the 

 narrow end, forming an unmistakable apical organ. The larger 

 end is occupied by a group of trochoblasts, which form a some- 

 what irregular series around the margin, but also extend some- 

 what over the base. The remainder of the embryo is formed of 

 small ectoblast cells that have on the lower side extended more 

 or less into the basal region. The exact number of trochoblasts 

 cannot be determined, but there are at least 14, and probably a 

 larger number. It is clear that this larva represents a partly 

 closed and distorted prae-trochal region, with that part of the 

 prototroch derived from the first quartet, minus a certain number 

 of cells that have separated. The other larvce were similar in 

 type, but evidently represent a smaller portion of the same 

 region. This case, taken In connection with the other facts de- 

 termined, renders it practically certain that the first quartet as a 

 whole is here incapable of producing a complete dwarf, but gives 

 rise to essentially the same ectoblastic structures as in a whole 

 embryo. This result is entirely in agreement with that after- 

 wards obtained on Cerehratiiliis by Zeleny ('04), who, at my sug- 

 gestion, undertook a comparison in this form of the upper and 

 lower quartets of the 8-cell stage — a question particularly inter- 

 esting in this case since the upper quartet is larger than the lower. 



