MORPHOLOGICAL PART. 89 



While thus among recent Crinoids the orals are readily recognized, their 

 identification among Palaeozoic forms has been the subject of much contro- 

 versy. Allman* in 1863 expressed the oj^inion that the group of plates in 

 the centre of the vault of many Palseocrinoids is a representative of the sim- 

 ple oral system of the young Comakila ; but as those plates often consist of 

 more than ^lyq, it was not made clear to which particular ones his homology 

 applied. In most Palaeozoic Crinoids there is no oral opening, and the 

 arrangement of the plates at the summit is irregular and quite variable. 

 This is the case particularly among the Camerata, in which the median 

 portions of the disk are generally occupied by a large, centrally located 

 plate, surrounded by eight or nine others, of which four are large and 

 similar in form and size. These four larger plates are directed toward 

 the anterior side of the disk, forming at their outer edges re-entering 

 angles, which are filled by three rather large plates, radially disposed ; 

 while the four or five smaller plates of the proximal ring are directed 

 posteriorly, and are followed by numerous more or less irregular pieces, 

 directly or indirectly connected with the anus (Plate III., Figs. 17, 18, 20, 

 21, 22, and 23). Occasionally the larger plates are separated from one 

 another by small, supplementary pieces (Plate III., Fig. 23). This is the 

 case in some of the larger species, in which the small pieces were intro- 

 duced in the growing Crinoid. There are also species in which the larger 

 plates are not represented at all, and the whole ventral disk is composed 

 of minute pieces without definite arrangement, leaving only an opening 

 for the anus (Plate III., Fig. 24). 



The interpretation of these plates has proved the more difficult because 

 in other groups, notably the Larviformia, the tegmen consists of but few 

 pieces, which have a diffi3rent arrangement. In Allagecriniis and Myrtilo- 

 criniis (Plate III., Fig. 13), the whole ventral surface is covered by five large 

 interradial plates, resting upon the superior edges of the radials, exactly as 

 the orals in the Pentacrinoid larva of Antedon. Haplocrinus (Plate III., Fig. 

 14) has ^YQ similar plates, which were at one time supposed to surround a 

 small central plate. f SymhathocrinuB (Plate III., Fig. 25) has a pyramid of 

 five large plates, four of them resting upon the edges of the muscle plates of 

 the radials, and partially surrounding a larger one, wedged in from the pos- 



* Trans. Roy. Soc. Edinb. Vol. XXIII., pp. 245-251. 



f Our supposed discovery of this plate in a specimen of Haplocrinus mespiliformis proved afterwards to 

 be a mistake, due to tbe peculiar fractures in the specimen. Carpenter, to whom we submitted the specimen, 



12 



