178 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



mechanical or oecological conditions, and of the continued pressure of the ancestral 

 characters, should evince a decided tendency to revert to the latent ancestral type 

 of structure through which, and not directly from the early type, it finally reaches 

 its ultimate most highly specialized and perfected condition. 



The very simple structure of such types as the Larviformia and Boihriocidaris 

 does not indicate that they represent the true phylogenetic prototypes from which 

 all the later crinoids and echinoids have developed, but rather suggests that they 

 are new and aberrant types in which the sudden and remarkably perfect mechanical 

 readjustment has for the moment inhibited all inherited tendencies which, however, 

 will slowly reassert themselves just as soon as they can adapt themselves to the new 

 mechanical conditions. The Larviformia and Boihriocidaris form the structural 

 starting point for the crinoids and for the echinoids as we know them ; but I believe 

 that both types are very aberrant, abnormally simplified, if I may so express it, 

 and therefore give far less accurate a clue to the true affinities and ultimate origin 

 of their respective groups than do the Flexibilia or the Archseocidaridse of later 

 occurrence. 



Logical and connected proof of this hypothesis is not possible ; but many facts 

 may be found in any group of which we have an adequate knowledge which appear 

 to substantiate it. For instance, the first cetacean to appear is the Eocene genus 

 Zeuglodon, which in many ways presents fewer mammalian characters, and cer- 

 tainly is far more fish -like than any of the latter forms; again, the earliest holo- 

 thurians of which we have any record, Eldonia, Laggania, and LouiseTla, are, 

 superficially at least, much less close to what we commonly regard as the typical 

 members of the group than the great majority of the subsequent genera. 



It was this curious specialization of primitive types through the temporary 

 dominance of the effect of an entirely new cecological or physical environment which 

 led me at one time, by a rather natural misinterpretation, to make the statement 

 that among the crinoids the early forms are phylogeneticallyjio less advanced than 

 the later. 



The calyx plates of the crinoids respond immediately to any change in the 

 mechanical forces bearing upon the dorsal cup. A very flexible and yielding column 

 allows of the retention by the calyx plates of conditions which more or less closely 

 approximate their original relationships; with increasing rigidity of the column 

 comes increasing compactness and solidity of the calyx, necessarily accompanied 

 by increasing reduction of the calyx plates, which eventually culminates in their 

 almost complete degeneration, so that instead of forming the capsule within which 

 the visceral mass is situated, and by which it is protected (their original function), 

 they merely form a small, flat, closely knit platform, upon which the center of the 

 visceral mass is supported. (For the details of this process see p. 341.) 



This condition reaches its extreme development among the comatulids, hi many 

 of which the calyx is so reduced as to serve for little else than a central point for the 

 attachment of the arms, for the comatulids are attached to the sea bottom or to 

 objects upon the sea bottom by numerous cirri springing directly from their dorsal 

 pole, and are therefore the most firmly and immovably fixed of any crinoids. 



