60 



lance to the Mespilia-larva, this larva must in all probability belong to 

 some Temnopleurid, possibly a Salmacis. 



Echinopluteus of Temnopleurid (?), species c. 



PI. XI, Fig. 3. 



Together with the specimen of the species a were found, in the same 

 plankton sample from off Jolo, three specimens of another Echinopluteus 

 species, which must be assumed to belong likewise to a Temnopleurid, 

 and which I shall designate as species c. One of these specimens has not 

 yet begun to metamorphose, the two others are in the beginning of meta- 

 morphosis. Although not in a very good state of preservation the younger 

 of these specimens shows several very characteristic features well; it was 

 therefore thought desirable to give a whole figure of it, although it had 

 to be partly restored ; thus the vibratile band is preserved only in the lower 

 part of the arms, so that the true shape of the arms cannot be seen; like- 

 wise it is impossible to ascertain the exact shape and size of the epaulets. 



As seen in PI. XI, Fig. 3 the absorption of the body rods has not yet 

 begun, in spite of the fact that the posterior transverse rod has already 

 been formed. The body is thus seen to be somewhat elongated as in those 

 Temnopleurid-larvse of which the young larval stage is known. A very 

 characteristic feature of this larva is the shape of the anal lobe, which is 

 very high, being supported by the very long ventral transverse rods; it 

 is concave in the middle, the corners being rather prominent. On the dor- 

 sal side the vibratile band is produced into a small lobe, supported by a 

 process from the dorsal arch. The postero-lateral process is very short and 

 small, but it is quite possible that it is considerably larger in the living larva. 



The skeleton (Fig. 22) affords several very characteristic features. 

 The body rods are branched at the end in the same way as appears to be 

 the rule in Temnopleurids. The postoral rod is fenestrated, but the holes, 

 which are very small, disappear completely a little above the middle of 

 the rod, and the outer part is simple, the three component rods coalescing 

 so completely that there is no trace of a compound structure. The postero- 

 dorsal rod is simple, shghtly curved. Both these rods are very sparsely 

 set with fine thorns. There are two sets of ventral transverse rods, one 

 very long, upwards directed, supporting the high anal lobe, the other 

 transversely directed, short; the latter may perhaps really represent the 

 recurrent rod. The posterior transverse rod is very pecuUar, formed like 

 a bow, with no posterior branch at the ends; on the other hand it has a 

 long posterior process in the middle, somewhat curved, and a small an- 

 terior process. In PI. XI, Fig. 3 it is seen to have developed before the 

 absorption of the body rods has begun; a similar case was found in the 



