238 Messrs. R. EtlicrRlge, Jan., and P. II. Cavpentei- on 



margin of these pieces {G. Sa>ji) — or, in other instances, 

 piercing directly through them, so that each pair appears 

 externally as a single opening ((r. melo and G. Norwoodi), 

 though they divide into two distinct canals before passing 

 entirely through the plates. The typical forms of this genus 

 also have the interraduxl pieces proportionately much larger 

 than in the true Pentremites, though this is not a constant 

 character." 



On a comparison of the species thus separated from Pen- 

 tremites and united under Granatocrinus^ we find that they 

 belong to two well-detined morphological groups. To the 

 first of these, having for its type Pentremites N^orwoodi, 0. & 

 S., we pro})Ose to restrict the name Granatocrinus ; and to the 

 other, typified by Pentremites Sai/i, Shum., we apply the 

 name Sc/iisohlastus. 



The lancet-plate of an OYd'ma,Yj Pentremites does not fill the 

 radial sinus, but only occupies its central portion. Between 

 it and the walls of the sinus therefore a groove is left, at the 

 bottom of which are the slits of the hydrospires. This groove 

 is continued from the peristome on either side of the apical 

 end of each oral piece and down to the end of each ambula- 

 crum. It is not, however, left open, but is converted into the 

 " hydrospire-canal " by the side plates, which are wedged in 

 between the lancet-])late and the walls of the sinus. In the 

 proximal portions of the ambulacra these walls are formed by 

 the oral plates, the central ridge of which is sometimes com- 

 paratively prominent, so as to separate the proximal side 

 plates of adjacent ambulacra, but sometimes so reduced that 

 these plates meet their fellows over the top of the oral ridge. 

 In either case, however, the passage from the peristome be- 

 tween the lancet-plate and the calycular plates becomes con- 

 verted into a canal, which is roofed over by the side plates 

 and opens at the summit by the so-called spiracle. This is 

 the structure of the summit in Pentremites proper. 



In Granatocrinus Norivoodt, and in all the species met 

 with in British rocks, with one exception, the lancet-plate 

 almost entirely fills up the radial sinus, and the narrow 

 hydrospiral canals are continued upwards through the suhstance 

 of the oral jy^iftes^, opening externally usually by five, but in 

 one case (^G. llojii) by ten, apertures f. Most of the other 



* This structure was oi-igiiially described by Dr. C. A. White la G. 

 Norwoodi (Boston Journ. IN at. Hist. 181)3, yH. no. 4, p. 483). 



t We have a specinieu of G. Noncuodi iu ^vhich cue of the oral }>lates 

 is pierced by two spiracular openings instead of by one only. This is the 

 natural condition of four of the oi'al plates of the Ijritish G. llojii, the 

 fifth one having a large anal spiracle as in all the species of this genus. 



