106 BRITISH PALEOZOIC ASTEROZOA. 



more and more distinctly stellate plates " (Schucliert, 85, p. 144). They are 

 situate well within the margin, being separated from the infero-marginalia by 

 several rows of inter-marginalia. In P. sjjeciosiis, as judged by the figures, one 

 can recognise differentiated supero-marginalia throughout the length of the ray, 

 although even in this species there is considerable similarity in the appearance of 

 all the apical ossicles at the extremity of the ray. In U. hiiialiani it is very 

 difficult to decide which are the true supero-marginalia of the distal half of the 

 arm. An examination of PI. V, fig. 4, shows that there are three rows of 

 precisely similar ossicles interior to the infero-marginalia, and that it is not the 

 row adjacent to the infero-marginalia which is in series with the ossicles described 

 above as supero-marginalia, but the second row. We can suppose either that the 

 supero-marginalia are not as yet completely differentiated throughout the arm, or 



60 



61 



Text-fig. 60 (on left).— Outline drawing of ossicles on the apical surface of a iwrtion of the disc of 



Uranasler kinahaiii (based on E. 194). x 6. 



Tbxt-kig. 61 (on right). — Outline drawing of the ossicles on the apical surface of the median portion of 



an arm of Uranasler kinahani (taken from E. 13110a). x 8. 



that perhaps a differentiation already acquired is now in process of being lost. It 

 is impossible to decide which of the two theories is correct until we know more of 

 the ancestry of the forms. 



The middle of the arm is occupied by small irregularly shaped plates which 

 probably represent radialia, for the greater part broken down. Only at the 

 extremity of one arm of one specimen (E. 13110^) is there any suggestion of 

 radialia in series such as are preseut in P. speciof-'us. 



Oral Surface (PI. IV, figs. 1, 5; Text-figs. 62, 63).— A noteworthy feature 

 of the oral surface is the presence of numerous ventro-lateralia in the interradial 

 angles of the disc, best seen in Text-fig. 63. They are numerous and almost 

 granular in appearance. At the proximal apex of the ventro-lateralia may be seen 

 a ridge to which the mouth-angle plates were attached, 'i'hese latter plates are 

 much longer than the succeeding adambulacralia and are much more Ophiuroid- 

 like in character than Asteroid (compare the text-figure with PI. I, fig. 10). The 



