ASYMMETRY IN THE STARFISH. 121 



united with the posterior coelom; the hydrocoele has begun to 

 undergo marked thickening of its walls; the median anterior part 

 of the fused anterior coeloms has sent forward a long median 

 process into the preoral lobe of the larva which has now a trilobed 

 shape. 



At twenty-six days (Fig. 12) the median diverticulum of the 

 anterior coeloms has branched terminally into three parts, a large 

 median one and two small lateral ones. The preoral lobe is now 

 decidedly a trilobed affair. 



At the end of a month the larva takes on the characteristic 

 features of a simplified Brachiolaria (Fig. 13). The changes that 

 signal the arrival of the larva at the Brachiolaria stage have to do 

 with the differentiation of the median and the two lateral brachia 

 of the preoral lobe, organs that were already foreshadowed in the 

 trilobed form of the preoral lobe and in the three-branched 

 condition of the median anterior coelomic process. The middle 

 brachium develops a well-defined attachment disk (Fig. 16) on 

 which are present three pairs of papillae. The lateral brachia 

 remain simple knobs, slightly thickened and clubbed at the ends. 

 One branch of the anterior extension of the coslom runs into each 

 brachium, but the lateral branches remain relatively small. A 

 ventral sucker appears between the paired lateral brachia. The 

 hydroccele of the Brachiolaria has become lobed in some indi- 

 viduals, and in others it has given off some or all of the radial 

 branches, destined to form the radial water canals (Figs. 13, 14, 



15). 



Further than this it was not possible to go with certainty. For 



one month the larvae grew normally on a diet of the diatom, 

 Nitschia, but almost over night the best cultures underwent 

 retrogressive changes and died. One individual was found which 

 had rounded up into a thin-walled vesicle. At first it was hoped 

 that this individual had undergone or was undergoing meta- 

 morphosis, but in less than a day it also died. The same fate 

 attended three other carefully reared lots of larvae, all dying after 

 about a month, and none, so far as I was able to' determine with 

 certainty, completing the process of metamorphosis. In spite of 

 this ill fortune it should be emphasized that larvae with right- 

 handed (reversed) asymmetry were found among the most 



