212 PROCEEDINGS OF THE \ \TI<>\ | /. MUSEUM. vol.38. 



interambulacral areas instead of their more primitive bicolumnar 

 areas, such an advance would not he accompanied by the dwindling 

 and disappearance of the sur-anal. Because the clavicles are small 

 or entirely absent in the mostly extinct Rat itse, while in all cases well- 

 developed in the mainly recent Carina! a' we can not pronounce them 

 unessential features of vertebrate morphology. 



The common ancestor of the Heteroradiate Echinoderms was mi 

 unattached organism with a central dorsal plate surrounded by two, 

 or possibly three, alternating circlets of plates. The Echinoidea, 

 which have remained unattached, retain to-day the original arrange- 

 ment in a slightly modified form; the Crinoidea, however, instead of 

 maintaining a, position in which the dorsal side is up and the mouth 

 down, became inverted, so that the mouth and peristome is up and 

 the dorsal side down. This brought the central plate into permanent 

 contact with the sea floor, and, the central plate being a physiologi- 

 cally inert skeletal structure, a, calcareous element whose shape and 

 thickness are in no way confined within narrow limits by physiolog- 

 ical or mechanical limitations, it became attached to the sea, lloor 

 by a simple increase in thickness. The facility with which organisms 

 with calcareous skeletons become attached, even though belonging 

 to normally free groups, is graphically illustrated by JEiheria, 

 Mulleria, Spondylus, Ostrea, Balanus, and various other genera, the 

 developmental stages of which show that the sessile mode of existence 

 is of comparatively late phylo'genetic inception. 



MacBride has observed that the larva' of Asterina gibbosa in the 

 early stages of development attach themselves for a short time by 

 the prseoral lobe, and from this circumstance it has been argued that 

 the ancestors of all echinoderms were attached. I quite fail to see 

 the force of this reasoning; the larva of Asto-'nia at the time of its 

 short attachment is at a very young stage. The larva' of echino- 

 derms are creatures of a high state of specialization, a specialization 

 along entirely different lines from that of the adults, fitting them for 

 a radically different mode of existence; and it seems to me that the 

 only logical course is to treat the larva' and the adults as different 

 classes of animals, modified for an entirely dilferent environment, 

 each highly specialized in its own way. Thus 1 consider that the 

 action of the Asterina larva is of interest only in comparison with 

 other echinoderm larva' of a corresponding stage of development . and 

 is and can he of not the slightest significance as regards the adult 

 life either of Asterina or of any other echinoderm: in other words, 

 t hat . in general, echinoderm larva 1 are only interesting or significant as 

 echinoderm larva', and not as elucidating the phy logenet ic path 

 which has been traversed by the adults. For instance, the larvae of 

 Echinus are highly specialized pelagic plutei, those of Antedon almost 

 annelidan in character and with a greatly reduced duration of free 



