NO. 1846. ON CERTAIN ELEVTHEROZOIC PELMATOZOA—KIRK. 73 



probabilit}' we need not go back into the Paleozoic for such a form. 

 Subsequently among the descendants of such a type we may pos- 

 tulate a tendency toward the acquisition of a proximale. The essen- 

 tial factor in such an evolutionary process is naturally a decrease in 

 the formation of ossicles immediately beneath the theca and a 

 lengthening of the stem by the intercalation of new columnals distad 

 to the proximal columnal. It is doubtful if there are many forms 

 among which formation of columnals immediately beneath the cup 

 does not take place even in post-larval stages of development. The 

 differentiation of a persistent proximal columnal may no doubt be 

 accounted for very largely on a purely mechanical basis, thouo-h con- 

 cerning its exact nature one can not be entirely certain. It is suffi- 

 cient to note that at some stage in the development of the animal 

 the topmost columnal becomes intimately associated with the proxi- 

 mal elements of the theca. This comes as the direct result of a 

 gradual diminution in the size of the infrabasals. Such a tendency 

 toward the elimination of the infi'abasals brings them within and at 

 the apex of an mverted cup in the base of the theca. This results 

 in the introduction of the proximal colunuial %Wthin the basal circlet, 

 as regards horizontal position. 



Having acquired a proximal columnal that retains its position as 

 such for an extended period during the adult stages, there begins an 

 appreciable differentiation of this ossicle, as separating it from the 

 remainder of the columnals. These differences he cliieflv in a leno-th- 

 ening of the columnal, which may become very marked. (PI. 7, fig. 1.) 

 At the same time there is a more and more marked inclusion of the 

 ossicle within the basal circlet accompanied by an increasingly close 

 union between it and the infrabasals. Eventually in some of the 

 later stages of development there is a complete fusion between the 

 infrabasals and proximal columnal. At this stage we ma}' call the 

 columnal a ''proximale" in its strictest sense. 



Subsequent to the differentiation of the proximal columnal there 

 was no doubt a constant tendency on the part of the crinoid to break 

 loose from its column immediately beneath tliis ossicle. The process 

 may in a general way be held to have evolved somewhat in the man- 

 ner to be observed in the case of MiUericrinus prattii, certain speci- 

 mens of which as already described are essentially comatuhd, barring: 

 the lack of specialization relative to the centro-dorsal that is to be 

 noted in the case of the later forms. From tv^pes in wliich detach- 

 ment occurs infrequently and perhaps at different points in the 

 column we come to forms among which detachment becomes uni- 

 versally effective. Detachment at first is to be found only in adult 

 stages, but, as affected by acceleration, was pushed farther and 

 farther back in the ontogeny of the organisms. 



