1890.] NATURAL SCrEXCKS OF PHrLADEH'IItA. 365 



are lociitcd at some distance from the orals, being placed in the sim- 

 pler forms, with but two arms to the ray, close to the outer margin 

 of the vault {Agaricof.rinus), directly over the point at which the 

 bifurcation of the aml)ulacra takes place. AVhen there are four arms 

 to the ray, they are removed relatively further inward, and are fol- 

 lowed by two similar but smaller i)lates; but when there are three 

 arms to the ray by one plate only, which is directed to the side of 

 the bifurcation. 



It is now very interesting to find also that the axillary plates of 

 the covering pieces upon the ambulacra are frequently protuberant. 

 In Encladocriniis millebrachiatus all along the free appendages they 

 are strongly nodose, and their tops, if the ambulacra were subtegmi- 

 nal, would naturally extend into the vault, and be exposed to view. 

 That all calyx plates, and especially the " radial dome plates," were 

 capable of secreting an enormous amount of limestone matter, is well 

 shown by the fact that in some species of Dorycrinus the plates are 

 extended to the length of three or four inches, and their ends, if acci- 

 dentally broken during the life of the Crinoid, were at once replaced. 

 Our explanation that the radial dome plates were developed from 

 the covering plates, seems to us exceedingly probable, and has been 

 favorably received by Dr. Carpenter. If the plates were special 

 structures covering the ambulacra, as heretofore supposed, they would 

 have to be regarded as true vault plates. 



In many of the Palaeozoic Crinoids we find upon the tegmen ele- 

 vations, which sometimes take the form of ridges and pass out from 

 near the center to the arm bases. These ridges which are best pre- 

 served and most frequently found in Silurian Camerata, are formed 

 either by the covering plates, or by the so-called smaller vault pieces 

 which pass up from between the rays. Similar ridges occur upon 

 the disk of recent Comatulidae, but these are always formed by the 

 covering pieces, and the plates are movable ; Avhile in Palaeozoic 

 forms, in which covering plates enter the surface, these are united 

 by a suture. Ridges of this kind are found in Actinocrinus qainquau- 

 gularis Angl. (Iconogr. Crin. Suec, PI. XVI, fig. 28), Habrocrinus 

 ornatus (Ibid., PI. XXVII, fig. 5), Marsvpiocrinus depressus and 31. 

 radiatiis Angl. (Ibid., PI. X, figs. 16 & 21), and Pktfijcrinus syvi- 

 metrieus W. & Sp. (Proc. Acad. Xat. Sci. Phila., 1888, PI. 18, fig. 

 15). The mouth in these species is closed either by the orals, or, 

 when these are absent, by the uppermost covering })ieces, which inter- 

 lock with those of adjoining rays. Very prominent ridges occur al- 



