b8 



K. KISHIXOUYË. 



Uli AugLi.st 7tli (10 clciy.s after .spawniiig), I found a t<iiia]l i-ouikI 

 whitish disc (û^. o). This UTadiiallv erdai-fi-ed and on Auo-yst 10th 

 (13 diiys after spawning) it had assumed an o^'aJ shape. The Linger 

 axis of the disc corresponds to the future median hue of the embryo, 

 and the disc itself is the rudiment of the ventral [)late. The disc had 

 divided into two unequal parts by a transverse line (fig. 4). The 

 smaller portion of the disc corresponds to the cephalic lobe and the 

 first appendage-])earing segment. Xew segments are successively cut 

 off from the anterior end of the larger, posterior portion. 



On the morning of August 11th (14 days after spawning), three 

 transverse lines and a median longitudinal line were visible (fig. 5). 

 The median line did not extend posteriorly beyond the third trans- 

 verse line. Thus at this stage the central plate is divided into four 

 segments, three anteri<jr ones of which are divided into two lateral 

 halves by the median longitudinal line. About noon of the same day 

 one more segment had been added (fig. 6), and the median line had 

 l)een produced to the posterior end of the new segment. 



On August 12th (15 days after spawning), five transverse lines 

 were seen, and the median line was produced beyond the penult seg- 

 ment (fig. 7). Although the change is \ery gradual, yet when we 

 compare the two ventral plates represented in fig. 5 and fig. 7, we 

 find tliat the ventral plate is much larger in fig. 7. Moreover while 

 the angle made by tlie lateral hahes of a segment with the median 

 line is almost a ri^'ht an<jle in fii!'. 5 and in the anterior region of fio-, 

 6, the ano'le has become acute in fiüf. 7 — the acuteness increasing' more 

 and more towards the posterior end. 



On August 13th (16 ditys after spawning), many imporfant 

 changes htid occurred. The external ends of the segments had united 

 with each other forming a white maroiiuil band on eacli side of the 



