Till' HYDR08PIBES WD BP1BACLES. 93 



But in Acentrotremites (PI. XIII. fig. L9) and Mesoblastus (PI. IV. fig. 4 ; PI. VI. 



fig. 10; PI. VIII. fig. G; PI. XI. fig. !•".) the lancet-plate is relatively narrow, and 

 does not fill the radial sinus, so that the side plates rest partly upon it and partly 

 upon the sloping hydrospire-plate, which is moulded to receive them; and this con- 

 dition also presents itself in Granutocriintx Xoriroodi (PI. XI. fig. 1 I). We have not 

 heen aide to evt any satisfactory sections showing this point, the best being that of 

 j\Tcsolt/a*fits angulatus (PI. XVII. fig. 9). 



The formation of the pores between the hydrospire-phvte and the radials is well 

 seen in the moulds of the interior of the calyx of Granatocrinus Norwood/, which are 

 figured on PI. VII. figs. 7-9. The last of these figures represents a specimen from 

 which the original limestone plates have not been entirely removed, and the relations 

 of the lancet-plate and deltoids to the surface of the mould beneath them are very 

 clearly shown. The hollow impression lodging the lancet-plate is bounded — outside 

 the limit of the deltoids — by a strong ridge, which is broken up transversely into a 

 number of smaller portions. This is the cast of the upper part of the hydrospire- 

 sac, the cleft-like opening of which is divided up into a series of pores in the manner 

 already described ; but this subdivision is not always so evident in the casts as is 

 seen in PL XI. figs. 19, 20. Many of these moulds of Granatocrinus Worwoodi are 

 hollow, and in their interior may be seen the duplex sac-like portions of the hydro- 

 spires, just as is shown in PL VII. figs. 1, 2. In many of the internal casts of 

 Schizoblastus Sayi, though not of Granatocrinus Norwoodi, the ridges representing 

 the hydrospire-sacs go but a very short distance beyond the radio-deltoid sutures, 

 although the external ambulacrum with its marginal pores reaches far down into the 

 radials, the limbs of which are twice as long as the body. A similar inequality in the 

 internal and external aspects of the ambulacra appears in many Penfremites, notably 

 in P. conoideus (PI. III. figs. 4-G) ; and its occurrence in Schizoblastus Sayi will be 

 better understood when the structure of the Pentremite type has been considered. 



Except in one essential particular the general relations of the hydrospire-sac are 

 the same in Pentremites, Pentremitidea, and Troostocrinus, as in Mesoblastus and 

 Granatocrinus. The point of difference is the absence of any hydrospire-plate in the 

 three first-named genera, so that when the side plates of the ambulacra are removed 

 the hydrospire-slits are directly exposed (PL I. figs. 6, 7 ; PL IV. figs. 12, 14 ; PL V. 

 fig. 3 ; PL XII. figs. 12, 13). The same is the case in Orophocriwus Orbignyanus and 

 in Phccnoschisma acutum, J'. Archiaci, and P. caryopkyllatum (PL XL fig. S ; PL XIV. 

 figs. 2, 3, 6, 7,11, 1G); but the hydrospire-slits of these two generic types are mostly 

 situated on the sloping sides of the radial sinus and not concentrated in its deepest 

 part, as in the Pentremitidse and Troostocrimts, and to some extent also in Oropho- 

 crinus verus and 0. pentangularis (PL I. figs. 6, 7 ; PL XV. figs. 4, 10). 



The complexity of the hydrospire-sac in Penfremites appears to vary very consider- 

 ably. Thus in P. elongatus there are only three lamellar tubes on each side, as in 



