EVOLUTION OF THE ECHINODERMS — FELL 465 



Figure 5 shows some details which have recently been elucidated on 

 material of GMnianaster and Villebrunaster. In figure 5, B, the 

 ambulacral ossicles are seen to be arranged in opposite pairs; this is 

 the condition in the middle and base of the arm. In figure 5, H, how- 

 ever, they are seen to be placed alternately; this is the condition in 

 young specimens of Villebrunaster^ and at the tip of the arm in older 

 specimens. In figure 5, D, are seen rows of virgalia in Chinianaster; 

 they terminate in a spine-shaped virgalium. In figure 5, G {Ville- 

 hrunaster) , virgalia are seen overturned on the right side and erect in 

 the normal position on the left ; from this specimen, and others, it has 

 become clear that each fully developed virgalium was shaped like a 

 segment of a tram rail, while in the growing region they were nearly 

 cylindrical rods. 



Figure 5, E, shows a more detailed reconstruction on the basis of 

 the type material of GMnianaster^ and of other specimens since dis- 

 covered. It explains the relation of the ambulacral ossicles to the 

 adjacent virgalia. Tube-feet were evidently placed in cup-shaped 

 depressions on the lower side of the ambulacra, here sketched upside- 

 down, as in ventral aspect. The arm had a frondlike structure 

 to the very tip, as seen in figure 5, C. 



Figure 5, F, is a general i-econstruction of Chinianaster^ seen in 

 ventral aspect, to incorporate these data. The arrows show the direc- 

 tion of the inferred ciliary currents in the food grooves between the 

 rows of virgalia, and along the main radial groove leading to the 

 mouth. The material on which these details have been studied was 

 generously placed in my hands by Prof. Georges Ubaghs, of Liege. 



Referring again to Archegonaster^ figure 5, 1, the more starfishlike 

 of the somasteroid fossils: Note the buttresslike appearance of the 

 virgalia, especially near the ends of the arms, where they seem to 

 make a bridge between the marginal plates and the axial ambulacral 

 plates. This illustration, published by Spencer in 1951, remained 

 with me as a dimly remembered image after I read his paper. Mean- 

 time I had embarked on a long series of dissections of modern genera 

 of starfishes, searching for characters which might enable a better 

 correlation between the Paleozoic sea stars and the modern forms, 

 and thus lead to a more satisfactory classification. It happened that 

 one of the genera whose skeleton I had prepared some years prior 

 to 1951 was Persephonaster^ illustrated in figure 6, E. The dissection 

 was used for a nmnber of years in our practical classes, but although 

 students had several times questioned me about the rodlike ossicles 

 between the ambulacral ossicles and the marginal series (for no such 

 rods were mentioned in the textbooks) , I am afraid a good many years 

 went by before it dawned upon me one day that there was sometliing 

 oddly familiar about these so-called superambulacral ossicles. Of 

 course, as must now be obvious to you, they present a remarkable 



