Till: HYDROSPIRES AND 8PIEACLES. 95 



the hydrospire-sacs. These extended right down to the end of the radial sinus as it 

 appears in an external view id' the plate. But the inner face of the same plate 



(PI. XVIII. fig. 7) seems to show a much larger bodj and correspondingly smaller 

 sinus and limbs. This is due to the closure of the lower part of the sinus by an 

 extension of the calcareous tissue forming the body and the lower ends of the limbs 



of the radial, which is seen deep down in the view of the plate from the outer side 

 (PI. XYI1I. fig. 8). The consequence is that in a radial winch retains both its 

 ambulacrum and its hydrospires the former sometimes seems much longer on the outer 

 face than the latter appear to be in an internal view ' (PI. I II. figs. 5, 6). In reality 

 however, the hydrospires extend to the full length of the ambulacrum, but their 

 distal ends are contained (so to speak) within the substance of the radials, and so do 

 not depend into the body-cavity at all. The impression of an ambulacrum on the 

 upper surface of an internal cast is therefore much shorter than the ambulacrum 

 itself, just as has been noted in Schizoblastus Sayi. This is excellently shown in 

 PI. III. tig. 4, which represents a partially exposed internal cast of I', nl n mites 

 conoideus; while tigs. 10-12 are different views of a cast of its variety P. KonincM- 

 anus with quite short ambulacral impressions, though in reality the ambulacra are 

 much longer, as shown in PL II. figs. 16-23. 



The enclosure of the ends of the hydrospire-sacs within the substance of the radials 

 so as to shut them off from the body-cavity is not always so complete as in P. conoi- 

 deus, but we believe it to be of very general occurrence in the genus. In PI. II. 

 fig. 13 is represented a section of a radial of P. Godoui just above the point at which 

 it begins to enclose the hydrospires ; while fig. 31 shows a similar section near the 

 tip of an ambulacrum of Pentremites sulcatus. The number of folds in the hydro- 

 spire-sac is much reduced, only those nearest the middle line being continued to the 

 extreme end of the ambulacrum, and a great thickness of radial intervening between 

 them and the body-cavity. In the genus Tricoelocrinus, so far as we have been able 

 to make out, the greater part of the length of the hydrospires is thus enclosed within 

 the substance of the radials. Figs. 10, 11 on PL XVIII. are inner and outer views 

 of the same radial ; while fig. 13 represents a sectional view of the end of the plate 

 seen in fig. 12, and another section is shown in fig. 14. The amount of surface which 

 the hydrospire-sac would present to the water contained in the body-cavity must 

 therefore have been extremely limited. 



In Troostocrinus Reinwardti the hydrospires appear to reach the end of the radial 

 sinus in the usual way. But in T. (?) lineatus, to judge from a fragment in Mr. 

 Wachsmuth's collection, they lie within the substance of the radials for some distance 

 from the end of the ambulacrum, as in Tricoelocrinus. This point may be of use 

 if it should become necessary to limit the genus Troostocrinus to T. Reinwardti, 



1 Figs. 7, 8 on PI. III. show that this is nut always the case. 



