182 PROCEEDINGS OF THE ACADEMY OF [1881. 



ring, and this of only four pieces. Periechocrinus, like Mari- 

 acrinus an Upper Silurian genus and like it without underbasals, 

 has instead of four or five basals, only three. In all other 

 respects tliese genera agree so perfectly with Glyptocrinus that 

 they cannot be distinguished, proving again liow closely llhodo- 

 crinidne and Actinocrinidse are linked together, and that they 

 are in fact variations of one great group. 



Where underbasals are unrepresented, families have frequently 

 been created upon the number of the basals, and Angelin based 

 his entire classification upon the number of proximal plates, 

 whether basals or underbasals. Convenient as this scheme of 

 classification may seem, it is altogether artificial, and combines 

 forms which are widely different, while it separates others which 

 are clearly allied. 



We have in the introduction to this work, page 11, dwelt at some 

 length upon the basals or first ring of plates below the radials ; 

 and believe we have shown that the basal disk, whether composed 

 of one, two, three, four or five pieces, can almost invariably be 

 reduced to five elementary pieces, and that all deviations from 

 this number have been produced by anchylosis of two or more of 

 the original segments. This, of itself, is a strong argument 

 against a classification based upon the number of these plates. 



Among the Actinocrinidas, only a few genera with the original 

 five basal plates are known, and these are confined to the Silurian ; 

 indeed we have good reason to believe that only the very earliest 

 representatives of this group possess a base divided into five 

 pieces. Genera with four basals commence in the Silurian and 

 terminate in the Devonian ; while genera with three basals are 

 found from the Upper Silurian to the close of the Warsaw lime- 

 stone where the family becomes extinct. The genera, with four 

 basals have been referred by us partly to the Actinocrinidte, and 

 partly to the Calyptocrinidffi. The latter family has four basals 

 throughout, but even here this number cannot be considered a 

 family character, since Melocrinus and Mariacrinus^ which have 

 four basals, belong nndoubtedly to the Actinocrinida?. Species 

 with three basals are found among both Actinocrinid* and 

 Platycrinidae, and the latter are by no means restricted to this 

 number, as Dichocrinus, which has been by most systematists 

 placed in the same group with Plati/crinus and Rexacrinus, has 

 but two basals. 



