XV111 UROCOPTID^E. 



In each of the subfamily groups certain members of the 

 several phyla have been similarly modified. Thus in the 

 Eucalodiinae, Ccelocentrum, C&lostemma and Epirobia are 

 alike in having narrow vertical riblets on the tubular axis, 

 although not otherwise closely related. Anisospira, Elasmo- 

 centrum and Bostricho cent rum, though rather widely separ- 

 ated in a phylogenetic classification, have axial armature of 

 almost identical pattern. In the Urocoptinae such conver- 

 gence is common. Thus Mychostoma and Idiostemma, ter- 

 minal twigs of two divergent branches, have developed an 

 almost identical armature of hooks upon the axis, in both 

 cases by modification of vertical axial ribs ; and further, these 

 ribs were undoubtedly absent, or rather, latent, in the com- 

 mon ancestor of the two groups. The convergence has gone 

 so far in some cases that the position of a group cannot be 

 predicted by shell characters, even when there has been great 

 specialization. Thus, I was formerly deceived in the rela- 

 tionships of Amphicosmia, of the series of slender west 

 Cuban Tetrentodons, of Spirostemma, etc., etc. 



This evolution of identical complex structures, de novo, 

 in two or more widely divergent phyla is what Lankester and 

 Osborn have called homoplasy " independent similar devel- 

 opment of homologous organs or regions giving rise to sim- 

 ilar new parts. " (Osborn, American Naturalist, 1902, p. 

 261.) The structures in question are not homologous, because 

 that term implies community of origin, whereas in the in- 

 stances now under consideration only the tendency toward 

 certain modes of modification is common heritage. It is as 

 if only a strictly limited number of possible lines of shell- 

 specialization had been assigned to the primitive Urocoptid. 



In general, the modes of internal specialization in Urocop- 

 tidce are very different from those of the Clausiliida, the 

 only family of long slender shells of comparable extent. In 

 the Megaspiridcz and Achatinidcz (subfamily CoeUaxina), 

 there is more similarity to some Urocoptid structures : com- 

 pare Holospira s. str. and Sectilumen with Cceliaxis, Thomea 

 and Distcechia, and with Perrieria. Whether the similar 

 structures in this case are homoplastic, or whether they are 



