1895.] MB. F. A. BATHEK ON UINTACBISUS. 995 



bifascial articulation ; but there is no defined ridge, and of course 

 no such skewing from right to left as in the brachials. The 

 pinnules gradually decrease in size towards the distal end of the 

 arm, and eventually become of exceeding tenuity. Since they are 

 rarely preserved or traceable for their whole length, measurements 

 are difficult to make. A brachial 3-0 mm. wide bears a pinnule 

 of which the proximal ossicle is 1-0 mm. wide. A brachial 

 2-5 mm. wide bears a pinnule of which the proximal ossicle is 

 •8 mm. wide, and the total length more than ll'o ram. A brachial 

 2-0 mm. wide bears a pinnule of which the proximal ossicle is 

 about -5 mm. wide, and the total length fully 16-5 mm. 



3. The Eelations op UiNTAominrs. 



As already stated, there are no forms with which direct com- 

 parison is obvious. The discovery of the ancestry of Vintacrinus 

 must therefore be a long process of induction. In prosecuting 

 such an inquiry, the first step is to clear away secondary and 

 accidental characters, so as, in any comparison, to utiHze only 

 those that are essential. It is, for instance, fiitile to lay any 

 stress on the fact that Uintaerinits is an unstalked crinoid, and 

 for that reason to compare it with its contemporary Marsupites. 

 Worse still to follow H. A. Nicholson and P. H. Carpenter (7), 

 and to place the two genera in a single family, though their 

 organization differs in almost every other respect. As well place 

 it with Saccocoma, or -with Ac/assizocrinus, or any other stalkless 

 crinoid. Surely the argument is absolutely the reverse. Features 

 in which unstalked and free-swimming crinoids agree -nath^ one 

 another are, it is probable, features due to similarity of environ- 

 ment rather than similarity of descent. The resemblance is 

 physiological, not morphological. In short, one infers that such 

 features are secondary, and not essential. They are the ones to 

 be cleared away. 



Let us consider the general and common characters of un- 

 stalked crinoids. All agree in the absence of a stem in the adult ; 

 but, \A'hen further compared, they are soon seen to fall into three 

 distinct groups. Tirst, the group in which a portion of the stem 

 remains, becoming modified into a cirrus-bearing centrodorsal, as 

 in Antedon, Eudiocrinus, and Thaumatocrinus. These forms anchor 

 themselves by their cirri, and though capable of crawling, climbing, 

 and swimming, do not often exercise their faculty of locomotion. 

 Secondly, the group in which either a portion of remaining stem, 

 or the lower part of the cup (i. e. basals or infrabasals), becomes 

 soHdified, usually by additional deposition of stereom, into a knob, 

 which, one may suppose, serves as ballast or as a sea-anchor ; such 

 forms are Agassizocrinus, Edriocrinus, and Millericrinus pratti. 

 Both of these groups have a small calycal canty with thick walls, 

 and there can be little doubt but that all are attached by a stem 

 in the earUer stages of ontogeny. The third group, comprising 

 Marsupites, Saccocoma, and Vintacrinus, has no trace of a stem or 



