the Crinoidca, Cy.stidca, and Blastoidea. 259 



111 a large hydrospirc there are about twenty of those tubes ; 

 their greatest breadth is at their mid-lengtli, where they are 

 crossed by the suture (c c) ; and as they ])eeome narrower 

 accordingly as their k^ngth decreases, the one in the middle 

 projects the deepest into the perivisceral cavity. In con- 

 s('f[uence of this arrangement, when a section is made across 

 the hydrospire at tlie suture, c c, fig. 4 a, the form h is obtained, 

 w here c c is the surface of the shell, while the comb-like struc- 

 ture below represents the tubes. 



Specimens of C. ornatus almost entirely empty are often 

 found ; and in some of these the internal foiTn of the hydro- 

 s[)ires is sometimes preserved. Those that I have seen have 

 tiie form of small rhomboidal pyramids, with four slightly 

 convex sloping faces, and composed of a number of vertical 

 [)arallel ])lates (the casts of the interior of the tubes), the sub- 

 stance of the tube itself not being preserved. I have, how- 

 ever, several polished transverse sections in which I think tlic 

 thin walls can be seen. 



The structure of the hydrospires is such that there can 

 scarcely be any doubt that they are respiratory organs. I'he 

 sea-water entered tlirough the pores, and aerated the chyla- 

 queous fluid contained in the perivisceral cavity by trans- 

 fusion through the exceedingly thin membranous shell that 

 composed the walls of the tubes. The number of pores varies 

 with the size of the individual. In large specimens there are 

 from 800 to 1000. 



It has been stated by some authors that the pores were pas- 

 sages for the protrusion of internal organs connected witli the 

 vitality of the animal. The fact, however, that the pores do 

 not penetrate into the general cavity of the body disproves 

 this theory ; and, moreover, through many of the tubercles 

 (those with a vesicular and spongy summit) such protrusion 

 would be utterly impossible. 



In Caryocrinus ornatus there are thirty hydrosj)ires, arranged 

 as follows : — 



1. Ten at the base — half of each on a basal plate and the 

 other half on one of the subradials, their longer diagonal 

 vertical. 



2. A zone of six around the fossil at the mid-height, their 

 longer diagonals horizontal. These seem to be imperfectly 

 tlcveloped ; for, on the inside, the tubes occupy only a small 

 space in the centre. 



3. A third band, of fourteen — two of them with their longer 

 diagonals vertical, and the others arranged in six pairs, the 

 diagonals of each pair inclining toward each other upward 

 at an angle of about 30^. There are only three interradii in 



