560 DR. J. F. GEIVMILL ON THE 



change of shape without delay. This contractile protoplasm may 

 perhaps be compared with the niyonemic layer in the cortex of 

 many protozoa. 



(13) Skeleton. — My material has not allowed me to make a full 

 study of the development of the skeleton, but the following points 

 may be noted : — 



1 . The aboral skeleton ai-ises in the foi-m of scattered plates 

 which, as in the case of Solaster, do not exhibit a deSnite radial 

 and interradial ari'angement as do the jDvimaiy plates of most 

 other echinoderm lai^-vse. In particular, there is no single ter- 

 minal plate at the end of each ray, but a number of s-mall 

 ossicles, one or more of which may be in the middle line of the 

 ray, while the others are moi^ laterally placed. 



2. The adult has a large terminal plate at the end of each ray. 

 Each of these plates arises by fusion- of a number (not less than 

 five) of the small first-formed ossicles.. This appears to be a 

 point of very considerable interest. The symmetrical arrange- 

 ment of the primary plates in typical starfish development must 

 be an acquired feature,, a result,, not a pi-ecursor,, of general 

 radiate symmetry. We can hardly doubt but that in the first 

 echinoderms the skeletal ossicles were diff'usely distributed in the 

 dermis.. In Cribi-ella accordingly, whether through survival or 

 reversion,, the primitive mode of origin of the primary plates 

 by coalescence of scattered calcifications is still exhibited in the 

 development of the terminals. 



It will be shown later (p.. 562), with reference to the mouth- 

 angle' and firet ambulacral plates in Solaster, that the converse 

 ontogenetic process, namely, division of an originally continuous 

 calcification into two or luore movably articulated ossicles, can 

 occur.. 



. III.. SOLASTEB ENBECA. (PI. II. fig. 11-) 



Exantination o-f material recently obtained from the Millport 

 Marine Station enables me to supplement or correct my former 

 account of the development of this species in i-egard to the 

 following points : — (1) origin of pharyngeal or perioral ccelom ; 

 (2) orig-in af perihaemal pouch IX ./I.*; sequence in formation 

 of hydrocoele pouches and final position- of remains of sucker ; 

 (4) formation of the- terminal, first ambulacral, and mouth-angle 

 ossicles, 



1. Pharyngeal or perioral cmlotn. — This coelom originates by 

 interradial outgro-wths from the posterior cojlom, and normally 

 such outgrowths occux> in all the- interrad'ii, those- in Y'lII./IX. 

 and IX. /I. being the latest to form. Sometimes, however, blanks 

 a,ppear to- be left in one or- more interradii, these- blanks in the 

 end being filled up by extensions from adjacent pouches. 



Masterman, in Crihrellci (9, p.. 392), first described the forma- 

 tion of the pharyngeal ca4o-m by interradial pouches from the 



* See footnote on p. 555. 



