360 p. HERBERT CARPENTER. 



that it was possibly the top stem-segment, so that the first 

 ring of plates (fig. iii, 2, 2) would represent the elements 

 of the pelvis of the other Crinoids. This view was adopted 

 with a slight modification by Pictet/ who referred to the 

 absence of a facet on the central disc as indicating the 

 absence of a stem, a view in which, as will be seen later, 

 I entirely concur. I do not, however, agree with Miller 

 and Pictet in regarding the proximal ring of plates (fig. 

 Ill, 2, 2) in the calyx of 3Iarsupites as basals, and the 

 second ring (fig. iii, o, o) as subradials (= parabasals), 

 for, as shown above, the former are radial in position, while 

 the basals of Pentacr'mus and the other Articulata are 

 situated interradially. But I have referred to their works 

 in order to show that the view here advanced of the essen- 

 tially simple nature of the central disc of Marsiqntes is 

 not altogether a new one. There are no sutures upon it 

 which would indicate its composition out of five separate 

 elements. Yet this is the true nature of the basis of the 

 other Crinoids, which Loven himself admits,- though some- 

 times entire, to be sometimes composed of more or fewer 

 separate pieces, the hasalia, I have endeavoured to show 

 elsewhere^ that there is every ground for believing that the 

 basis is primitively a composite structure, and that its occa- 

 sional apparent simplicity as in Apioc7'i?ius, Rhizocri?ms , 

 Engeniacrimis is only the result of a very close anchylosis 

 of its component elements, and the disappearance of the 

 external markings indicating their faces of lateral union. 

 One important reason why the basis of the Crinoids should 

 be considered as typically consisting of five originally sepa- 

 rate plates is its mode of development in Comatula, the only 

 Crinoid of which the younger stages are known to us. The 

 ring of five plates (fig. i, 3, 3) which rest upon the top stem- 

 segment [= future centrodorsal piece (fig. i, cd)'\ have 

 been almost universally regarded as representing the basals 

 of the Crinoids generally. This was Allman's intepretation 

 of their nature, and it has been accepted by Wyville Thom- 

 son, Dr. Carpenter, Sars, and Gotte. They are, however, 

 termed parabasals by Loven, who compares them to the so- 

 called first parabasals of Marstipites (fig. iii, 2, 2) and to 

 the genital plates of the Echini (fig. ii, 3, 3). In this 

 latter comparison I entirely agree, but I cannot accept the 

 former, for the simple reason that the genital plates of the 



\ ' Traite de Palcontologie,' tome iv, p. 291. 

 2 Loc. cit., p. 72. 



' " Ou Some Points in the xVnatomy of Peniacriiivs and Bhizocrinus" 

 ' Journal of AiiJitoiny and Plij^siology,' vol, xii, pp. -IS, 49. 



