90 



K. KISHIXOUYE. 



ventral plate are one and the same ; but later tlie ventral plate does 

 not keep pace in its growth with the ventrum. The ventrnni extends 

 very quickly, as is known from many karyokinetic figures in it, while 

 the dorsum seems to increase very little after its formati(^n (after tlie 

 differentiation from the underlying liypohlast), if it increases at 

 all. As the result of this increase of tlie ventrum, tlie egg is flattened 

 dorso-ventrally, and the ventrum is in the céphalothorax reflexed up- 

 wards and forms apparently a ])art (^f tlic dorsum, and in the abdomen 

 is elevated as in fig. IV . The retinal portion of the eye is always at the 

 maro-in of the ventrum, iust as tlie eves of the medusae are alwavs on 

 the umbrellar margin, between the exumbrclla and the subumbrella. 



The maro-in of the ventrum of the cephalic lobe develops not only 

 dorsally but also posteriorly. Thus, though the lateral eyes are found 

 at the level of a thoracic a]ipendage in later stages, they are not pro- 

 duced there, but are formed in the cephalic lobe like the eyes of other 

 bilateral animals. 



Table showing the chief early changes of the Kabutogani's eggs, 

 which were fertilized on July 29, 1890. 



