1898] WACHSMUTH AND SPBINGEB ON CBINOIDS 341 



and Springer, who call the first Order Inadunata ; the second, 

 Canierata ; and the thh-d, Articulata. But were these the only 

 characters to guide us, we should often err in referring families to 

 their respective Orders ; or, to put it another way, a collocation of 

 families by these structures alone would bring together forms that 

 other evidence forbids us to regard as related. We know for cer- 

 tain that we should place the Pentacrinidae and Antedonidae side by 

 side, because this is just what has always been done. Fortunately 

 Wachsmuth and Springer have discovered a fresh character, which 

 they believe is of considerable value. They maintain that in all 

 undoubted Inadunata and Canierata new columnals are developed next 

 the cup, so that the top one is always one of the latest formed, and 

 continually moves from its proximal position as new columnals 

 develop ; but that in all undoubted Articulata the top columnal is not 

 the latest formed, but a persistent element, for which I propose to 

 use the term ' proximale ' ; it often fuses with the infrabasals. 

 This discovery is" one of the chief novelties in the present monograph, 

 and it appears to be of great importance. There is, however, no 

 great attempt to prove its universal application ; in fact the whole 

 question is disposed of in the few following lines. " In the young 

 Comatula [i.e., Antedon], in which the top joint subsequently develops 

 into a centrodorsal, in the recent [sic] Mesozoic Millericrinus, and 

 probably in the recent Rhizocrinus and Calamocrinus, and in all 

 Ichthyocrinidae, so far as observed, the new nodal joints were formed 

 beneath the top joint, and the latter remained permanently attached 

 to the calyx. In Apiocrinus, in which for some distance the upper 

 end of the stem is greatly inflated, and the proximal joints extremely 

 long, it is possible that the nodal joints were introduced below the 

 inflated part, for there appear to be no immature segments between 

 the upper joints." This last sentence is not quite correct. One 

 often meets with a specimen of the Bradford pear-encrinite (A. elegans, 

 Defr.) with imperfect ossicles in the proximal cone ; indeed similar 

 incomplete portions of columnals may occur above the proximale itself, 

 as in the British Museum specimens registered 46,234 and 34,520. 

 Whether these imperfect portions are immature, in the sense that 

 they might have become more mature and complete, may be dis- 

 puted, but they certainly resemble the columnals which Mr Alexander 

 Agassiz has, no doubt rightly, regarded as immature in Calamocrinus. 

 In such genera as Ichthyocrinus and Taxocrinus it is by no means 

 easy to satisfy oneself that there is a persistent proximale. Of 

 course when the top columnal fused with the infrabasals, it must 

 have been persistent ; but it did not always so fuse. I do not mean 

 to say that there is any objection to this statement by Messrs 

 Wachsmuth and Springer ; but when it is introduced by them as 

 one of the three " most important characters for dividing the Crinoids 



