140 THE CRINOIDEA 



It must always have been obvious that the Neocrinoidea were a 

 polyphyletic group derived from the Palaeocrinoidea ; the difficulty has 

 been to trace the relationship. This task, however, was forced upon us 

 when those orders were finally rejected. Since that rejection did not 

 carry with it the overthrow of Wachsmuth and Springer's sub-orders, the 

 practical result was the raising of them to the rank of orders, in which 

 Neozoic crinoids had to be appropriately placed. 



P. H. Carpenter (1889) referred all Neozoic crinoids to the Articu- 

 lata (W. and Sp.) as a sub-order, PINNATA, with pinnules ; while the 

 Ichthyocrinidae constituted a sub-order, IMPINNATA, without pinnules. 

 This had the advantage of making Articulata, W. & Sp., very nearly 

 the same as Articulata, Miiller. But there is reason to believe that 

 the Pentacrinidae are descended directly from the dicyclic Inadunata ; 

 and since Pentacrinus (i.e. Isocrinus) was Miiller's type, one can hardly 

 escape confusion in using the term Articulata for a group that 

 excludes the Pentacrinidae. FLEXIBILIA, Zittel (1895), is superior 

 and prior to Wachsmuth & Springer's proposed substitute ARTICULOSA 

 (1897), which, moreover, was used in a different sense by Jaekel (1894). 



One reason for the above reference of the Pentacrinidae is the 

 discovery by Wachsmuth & Springer (1897) that in the Flexibilia the 

 top columnal is not the latest formed, but a persistent proximale, usually 

 fused with the infrabasals ; whereas in Pentacrinidae, as in Camerata and 

 Inadunata, the top columnal is merely the latest formed, and continually 

 moves from its proximal position as new columnals develop. This, along 

 with the other characters, seems to confirm the independent nature of the 

 order Flexibilia ; at the same time, the resemblance of early genera 

 to contemporaneous Inadunata is so striking, that one must suppose the 

 Flexibilia Irnpinnata to be derived from non-pinnulate dicyclic Inadunata. 

 Then the want of links between Impinnata and Pinnata suggests that 

 the process may have been repeated, and that Pinnata were derived 

 from Triassic pinnulate Inadunata (Fistulata, W. & Sp.). Whether the 

 mode of stem-growth indicates affinity or parallel modification is a point 

 that demands investigation. 



The Larviformia of Wachsmuth & Springer are generally accepted as 

 the most primitive Criiioidea, and as representing the ancestral type of 

 all the orders. On this ground the group might be retained, not 

 as a sub-order of Inadunata, but as an Urgruppe, or an independent 

 order (Zittel, 1895). But their geological age forbids us to regard the 

 known genera and species as themselves ancestral to far older forms. If 

 then they be placed with other Inadunata, the question arises whether 

 the distinction of the dicyclic and rnonocyclic base is not more funda- 

 mental than the varying development of the tegmen. A complex tegmen 

 is a development from a simple one, but between dicyclic and monocyclic 

 the barrier runs back to archaean ignorance (Bather, 1893). 



The Camerata, like the Flexibilia, form an order fairly well defined 

 on a morphological basis. But here too there are grounds for suggesting 

 that the modifications may have occurred more than once. The evolution 

 of pinnules, of biserial arms, of fixed brachials, interbrachials, and the 

 like, even of a solid tegmen, may all be traced among Inadunata (e.g. 



