388 HAROLD H. PLOUGH. 



B, compared with a normal control larva of the same age in C. 

 All the defective larvae had a complete tripartite gut like the 

 normal. In the A pair one shows a skeleton like the control but 

 with the large left postoral arm lacking, while the other has only 

 a rudiment of a spicule on the left side the right side completely 

 lacking. These larvae were followed for two days longer and the 

 latter never developed any farther, while the former never re- 

 placed the defect. In B much the same result is shown, perhaps 

 more distinctly, but rather more skeleton appeared in the smaller 

 larva. As indicated above Runnstrom (1914) described such 

 larvae in earlier stages as occurring regularly in Paracentrotus, 

 and suggested that the undivided egg showed a bilateral organiza- 

 tion. The finding of these plutei in Arbacia and the sand dollar 

 constitutes very clear proof that a localization of skeleton forming 

 material has taken place in the undivided egg, and that this may 

 be divided unequally by the first cleavage plane. It also indicates 

 a bilateral organization of this material. It seems quite probable 

 that the reason that this bilaterality is not more often evidenced 

 is that blastomeres with a marked deficiency of skeleton-forming 

 material often fail to develop the skeleton rudiment at all. 



It may be asked in view of this interpretation how it happens 

 that two complete plutei ever develop from both blastomeres of 

 one egg. To this it may be answered that my data indicate that 

 such cases are extremely rare in Arbacia and the sand dollar. 

 When the stratified material is about equally divided apparently 

 the initial bilaterality may be reorganized from the start. That 

 such reorganization may occur infrequently even in later stages 

 seems to be proved by the history of one Arbacia blastomere, in 

 which the steps were noted with great clearness. 



In the case illustrated in Fig. 5 one isolated blastomere began 

 its development as a partial one, but later a complete but small 

 skeleton was regenerated or postgenerated. In this case both 

 blastomeres of the egg went through the normal one half cleavage, 

 becoming rounded up at the blastula stage. Two one-half-size 

 gastrulae were formed, but one never became motile and soon 

 died. The other apparently formed a single triradiate spicule and 

 developed at the end of two days a half skeleton with a single 

 postoral arm, and the crossbars as indicated in the sketch. On 



