24 PE. p. H. CARPE]Sia.'EIl ON CEBTAIN POINTS 



the two remaining orals of Caryocrinus are represented in tlie 

 latter type (fig. III.), which seems to occupy a curiously 

 intermediate position between Caryocrinus and Semico smites. 

 Alternating with the three plates m are three others, situated 

 interradially or nearly so, which are marked n in Yon Koenen's 

 figure ; and outside these again, in the direction of the ambulacra, 

 are three larger plates, two of which, Gr, Gr, are interradial, while 

 the third, H, rests directly on the top of radial 17. These six 

 plates are well shown in the summit of Hemicosmites, as figured 

 by Miiller *, and I have lettered them accordingly in my copy of 

 his figure (fig. I.). Two points, however, are noteworthy. Semi- 

 cosmites has an anterior interradial. A, between radials 11 and 12, 

 which is not represented in Caryocrinus or Juglandocrinus ; and 

 the anterior summit-plate, Gr, thus rests directly upon it instead of 

 on the suture between the two radials. In the second place, the 

 ambulacra of Hemicosmites are external. There is a triradiate 

 peristome in the centre of the summit which extends outwards as 

 grooves on to the surface of plates Gr, Gr, H, the whole structure 

 being roofed in by smaller plates, which are probably ambulacral 

 covering-j)lates, very much as in Gyathocrinus. 



The close agreement between the varying conditions of the oral 

 plates in different Cystidean genera and those of the later Crinoids 

 is very remarkable. We have already traced the resemblance 

 between the summit of Caryocrinus and that of an Actinocrinoid. 

 The five orals of CyatTiocystis, with their distal angles abutting on 

 the ambulacral skeleton of each ray, reajjpear under similar con- 

 ditions in the young Flatycrinus symmetricus, recently figured by 

 Wachsmuth and Springer t. So far as can be judged from the 

 condition of the fossils, the orals were movable in Cryptocrinus, 

 SphcBronis, and Glyptosphcsra (PL I. fig. 15), which were the fore- 

 runners of the vQcent Hyocrinus, HJiizo crimes, and Thaumatocrinus. 

 Lastly, in Ascocystis the orals, if ever developed, must have dis- 

 appeared as completely as in a recent Pe7itacrinus or Comatula. 

 Barrande seems to have been somewhat puzzled by the con- 

 dition of the ventral surface with its five reniform compartments |, 

 and believed " que I'ouverture etait rameuse et coraposee d'arcs 

 entourant les compartiments renif ormes, d'unemaniere comparable 

 a celle que nous voyous dans divers autres genres de Cystidees, 



* Abhancll. d. k. Akad. d. Wiss. Berlin, 1863, Taf. vi. figs. 4, 5. 

 t Proc. Acad. Nat. Sci. Philad. 1888, pi. xviii. fig. 15. 

 % Op. cit. p. 117, pi. xxxiii. fig. 13. 



