6 E. W. MacBEIDE. 



When we review tlie account that is here given of these post-larval Holothurians, 

 we are, I think, confronted with a certain number of Echinoid features. 



The almost complete dorsal cuirass of overlapping plates certainly suggests the 

 corona of an Echinoid, especially if we recollect that, in the earliest Echinoids, the 

 number of rows of plates in the corona was indefinite in number and the plates over- 

 lapped. 



The closed axial sinus is another Echinoid feature. 



Some investigators have made an attempt to show that Holothurioidea are 

 widely separated from the rest of the Echinoderma. From the development of Synapta 

 digitata, the only Holothurian in which the life history has been thoroughly worked 

 out, it has been argued that since in this form the primary evaginations of the water- 

 vascular rudiment give rise to the buccal tentacles, whilst the radial canals owe their 

 origin to secondary evaginations alternating with these, the radial canals of 

 Holothurioidea are not homologous with those of other Echinoderms. But in our 

 specimens the canals occupying the interior of the tentacles can be traced into contin- 

 uity with the radial canals. The fact is that Synapta is about the worst form that 

 could have been chosen to represent the Holothurioidea. Its radial water-vascular 

 canals are only transitory larval structures, and its buccal tentacles not only spring 

 directly from the water- vascular ring, but, in contravention of the rule which prevails in 

 all the other groups of Holothurioidea, their number is no longer a multiple of five. 

 The functional importance of the buccal tentacles leads to their early appearance in 

 ontogeny, before the transitory radial canals appear. Professor Ludwig, who worked out 

 the early development of Cucumaria planci (4), points out that in this form the first 

 tentacles spring from the radial canals. 



If we compare the youngest post-larval form with the Auricularia which was 

 described in the beginning of this paper, we see that the alimentary canal in both 

 is very similar. The long stomach obviously corresponds in each case, as does the 

 short intestine. The short thick oesophagus of the post-larval form corresponds to the 

 inner part of the oesophagus of the Auricularia, whilst the shallow outer part becomes 

 (as we know from the life-history of other Auricularia) the oral vestibule. 



The chief difiiculty in the way of deriving Holothurioidea from a primitive form 

 of Echinoidea has lain in the difference between the Echinopluteus and Auricularia 

 larvse, and the apparent retention by the latter of more primitive features than 

 those shown by the Echinopluteus. I have shown elsewhere (5) that the stomach 

 and intestine of the Echinopluteus are directly converted into the stomach and 

 intestine of the young Echinoid. Further, in our specimens the stomach remains 

 unchanged in shape as growth proceeds, and the looping of the alimentary canal 

 is due solely to the growth in length of the intestine. This is also true of 

 the Echinopluteus. The " amniotic space " in the EchinopAuteus inside which the 

 oral disc of the young Echinoid is formed corresponds to the oral vestibule of 

 the Auricularia — the difference between the two cases being that in the latter 



