106 



further duriiit? Ihc next days, I suspected that it was due to the lacking 

 incilanient ol a natural bottom. Some of the larvae were then trans- 

 ferred to a jar with sand on liie bollom, and after a few days these larvae 

 had completed their metamorphosis, while those left in the jar without 

 sand on the bollom showed hardly any further advance in the process 

 of metamorjjhosis. 



The shape of the fully formed larva is that typical of Clypeastroid 

 larvtv (PI. IV. Fig. 2). There arc small vibratile lobes, as usual; that those 

 of the dorsal side appear lo be much larger than the ventral ones in the 

 figure is due to the fact that the posterodorsal arms are directed nearly 

 horizontally on account of the contraction of liie muscle uniting the bases 

 of their rods. The frontal area is remarkably small, the oral area being 

 very large. The nervous system is not distinct in the preserved specimens, 

 neither does the eso])hngus distinctly show llie gizzard-like structure found 

 f. i. in the Aravhiwidcs-larva. There is a consj)icuous coloration of red 

 and yellow in the point of the arms, scattered red pigment cells occurring 

 otherwise irregularly in the body, often in larger number along the preoral 

 and postoral band and in the posterior end of the body. The stomach 

 has a faint yellowish-green tint. 



The skeleton (Fig. 12) is of the typical Clypeastroid-structure. the body 

 skeleton forming a basket structure. The body rod and recurrent rod are 

 thorny, often even fenestrated, and form in the posterior end a large, 

 fenestrated, thorny plate. Even the transverse rods may develop into 

 narrow, fenestrated plates. The upper ventral transverse rod is entirely 

 absent or quite rudimentary. Postoral and posterodorsal rods of the usual 

 fenestrated type; anterolateral rods thorny in the outer j)art. The small 

 posterior prolongation from the dorsal arch appears to be constantly 

 curved and more or less thorny or fenestrated. 



The larva of the closely allied species Mellila 5-pcrforala (or tcstiidi- 

 nala Klein), which was reared by Caswell Grave, likewise through 

 metamorphosis, has not been adequately described. From the outline 

 figures given by Tennent in his paper on Echinoderm Hybridization 

 PI. 1, Figs. 10 — 12, it appears to dilTer quite notably from the present 

 species, especially through the much less developed fenestration of the 

 posterior part of the body skeleton and through the presence of the upper 

 and the absence of the lower ventral transverse lod. A more detailed 

 description of this larva is, however, desirable 



