II. Pelmatozoa. 11 



Just below the floor of the calycal cavity it gives off 5 interradial branches which 

 soon fork, and the adjacent secondary branches converge. They are all connected 

 by a ring-canal which contained intraradial as well as interradial commissures ; 

 and the latter also received a short lateral branch from each of the two secondary 

 cords which it joined. The position of the basals must be marked by that of the 

 primary interradial canals as in other Criuoids ; and, since no actual trace of them 

 is present, they would seem to have been overgrown or absorbed by the radials, 

 but not fused into a solid piece beneath them, as supposed by Carpenter. 



Wachsmuth & Springer (*) have discovered, as has also Beyrich, that the sup- 

 posed central plate in the vault of Haplocrinus has no real existence. But as in 

 Allagecrinus the vault consists exclusively of five large plates, the interradials of 

 W. & S., who now admit, however, that Carpenter & Etheridge were right in call- 

 ing them orals [see Bericht for 1887 Ech. p 6]. The posterior oral, which 

 is pierced by the anus, is larger than the other four, and has a forward pro- 

 jection which completely separates the two postero-lateral plates, and which some- 

 times appears to be cut off from its base, so as to look like an isolated central 

 plate. W. & S. also abandon their view that the central plate in the vault of 

 the Camerata represents five anchylosed orals, and that the proximals are 

 interradials. The four anterior proximals are orals, as maintained by Carpenter, 

 while the central plate, his orocentral, is the posterior oral, pushed forward and 

 inserted between the other four, in the same way as in Haplocrinus. There is 

 therefore no further need to suppose that the posterior oral is divided into two 

 parts by the anal system, the two small plates, hitherto regarded as representing 

 it, being really anal plates. - Tazocrinus, and probably therefore the Ichthyo- 

 crinidae generally, had an external mouth in a disk which was covered with small 

 and irregular perisomic plates, but had a well defined ambulacral skeleton. Around 

 the mouth are five oral plates, the posterior being much larger than its fellows, 

 and of these the antero-lateral is larger than the postero-lateral pair. The orals 

 are separated by the skeleton of the ambulacra which pass in between them to- 

 wards the peristome. The anal plate bears a short appendage like that of Thau- 

 mafocrinus, which is connected with the perisome by the small plates at its sides. 

 The authors conclude by discussing the relations between the earlier and the 

 later Crinoids. Their discovery of an open mouth in the Ichthyocrinidae shows 

 that it was not subtegminal in all Pakcocriuoids, as hitherto supposed ; and the 

 other characters which have been relied on as distinguishing this group from the 

 Neocrinoids are also subject to numerous exceptions. A natural classification of 

 Criuoids must therefore be entirely independent of geological age ; and W. & S. 

 propose the following as natural primary divisions:- - 1. Camerata. 2. Inadunata 

 (including Encrinus) . 3 . Articulata (= Ichthyocrinidae with Thaumatocrinus] . 

 4. Canaliculata (= remaining Neocrinoids). 



Wachsmuth & Springer ( 2 ), having examined several examples of the Crotalo- 

 crinidae, withdraw their former views respecting the subtegminal summit plates 

 and double ventral covering of this family, as they were based on a misinter- 

 pretation of the restored and incorrect figures of Angelin [see Bericht for 1886 

 Ech. p 7, 8].' The affinities of the Crotalocrinidae are with' the Camerata rather 

 than with the Ichthyocrinidae ; while the relations of the higher radials to the 

 primaries and to one another resemble those found in the Platycrinidae, to which 

 Crotalocrinus is also allied in the structure of its vault. This was composed of 

 five orals, large interradials separated by the ambulacral skeleton, and the anal 

 plates. In fact the calyx of C. is that of a dicyclic Marsupiocrimis. But the pre- 

 sence of an axial canal in the arms, the joints of which are articulated and con- 

 nected laterally into a network, distinguishes the family sharply from the other 



