34 



and has its corners produced into a pair of narrow, forward directed 

 processes. Further there is a large fold across the dorsal side, so that the 

 vibratile band along the dorsal side of the body is produced into a pair 

 of processes in the middle of the body. - - All the arms have a small ac- 

 cumulation of red pigment at the point; also in the vibratile lobes there 

 is a series of red pigment cells under the vibratile band; few scattered 

 pigment cells are seen in the body. The stomach is of a faint yellow colour. 

 The first pedicellaria has been formed; it is situated in the miclline in 

 the posterior end, deeply sunk in a groove. Also the amnion has begun 

 to form. There is thus no doubt that we have here the fully formed larva, 

 which is about to begin its metamorphosis. The metamorphosis of the 

 larva could not be followed, as the sojourn at Tobago ended by this linn 1 , 

 and there were only very few larvae left, so that it would have been 

 useless to try to bring them along with me on the voyage. 



Fig. 7. Outer end of postoral rod of the Tripneusles esciilentus-larva. 400 / l . 

 The unfenestrated part is shortened. 



The skeleton. The postoral rods are fenestrated only in the lower part; 

 they continue beyond as three, farther out only two, separate rods, 

 one of them ending a little way beyond the fenestrated part (Fig. 7). 

 The postero-dorsal rod is simple. The posterior transverse rod is a simple, 

 thin, straight rod, ending in two simple, diverging processes, the upper 

 quite short, the lower somewhat longer, but not nearly as long as the 

 postero-lateral processes. The other rods, including the dorsal arch, 

 are thin and smooth. 



Tripneustes gratilla (Linn.). 



PI. VIII, Figs. :>, 6. 



Fertilization of this species was undertaken in the end of March and 

 the beginning of April 1915 at Honolulu and at Ililo, Hawaii. At the 

 age of three clays the first larval stage was reached. In the shape of the 

 larva and the structure of the skeleton it very closely resembles Tr. 

 esculentus. The skeleton having been dissolved in the larva 1 preserved, I 

 cannot give a figure of it, but judging from my notes there would appear 

 to be no noteworthy difference in this regard between the larvae of these 

 two species. The skeletal rods of the "basket" are quite smooth and the 

 postoral rod fenestrated, as in Tr. csculeniiis. A small group of red pig- 

 ment is found already at this stage in the end of the postoral arms, and 

 small red pigment cells are found scattered in the body. 



