34 ART. 3. — S. IKEDA : CONTRIBUTIONS 



a considerable regularity. By continuous observation for one hour 

 or more, I finally made out that the egg was rotating regularly 

 with the axis of the embryo for its rotation-axis. Fig. 50 is the 

 anterior view of the egg. The direction of rotation as indicated 

 by the arrows, was from right to left i.e. contrary to the motion 

 of the clock-hands. The thick lines on the figure show the 

 different positions of the neural groove during rotation. The time 

 occupied by one turn of the egg was not quite constant, but the 

 differences were small, as the following measurements show : 2'.17", 

 2\21", 2'.20", 2'.30", 2'.20", 2'.15" etc. Thus the mean time 

 required by one turn was 2'.20". The position of the anterior and 

 posterior ends of the neural groove also varied slightly at every 

 turn. Such regular rotating motion of course could not have 

 been accidental, and the somewhat irregular movements which 

 I had noticed in the first two eggs must have been the beginning 

 of this regular motion. I can not understand the purpose of such 

 motion. Nor have I been able to ascertain the means by which 

 it is performed. If minute cilia are present, I could not bring 

 a sufficiently high power to bear on the egg to make them out. 

 So far as I know, such motion has not been noticed by any pre- 

 vious observer. 



Although my observations on the eggs of both Rana and 

 Bufo are somewhat meagre, the results as regards the most 

 important points, such as the mode of the blastopore closure and 

 the location of the embryo body, are exactly similar to what have 

 been obtained from the study of the Rhacophorus eggs. The 

 principal points in these results in the three genera may be 

 summed up as follows : — 



1). The dorsal lip of the blastopore always makes its first 



