412 Mr. E. Billings on the Structure of 



have been exclusively ovarian in its function. The liydro- 

 spircs have no connexion Avhatevcr with the arms, and are, 

 moreover, all of them entirely separated from each other. If, 

 then, they represent the anibulacral system of the recent Eehi- 

 noderms, it is (juite certain that that system was at first (or in 

 the undeveloped stage in which it existed in the Cystidea) 

 destitute of the cesophageal ring. 



In Codaster a further concentration of the res])Iratory organs 

 is exhibited. There are here only five hydrospires, and they 

 are all confined to the circle around the apex. Two of them 

 are incomplete, in order to make room for the large mouth and 

 vent {m v, fig. 2). They are each divided into two halves by 

 an ami al o2, &c. They are only connected with the arms to 

 this extent, that these latter lie back upon them. The amis 

 are provided with pinnula? ; but it is not at all certain that the 

 pinnuUc were in any direct communication with the hydro- 

 spires. It is evident that in all the Cystidea (and in none 

 is it more obvious than in Caryocrinus) there w\as no con- 

 nexion between the hydrospires and the pinnule. The main 

 difference (so far as regards the evidence of the presence or 

 absence of such a connexion) between Caryocrinus and Co- 

 daster consists in this, that in the former the arms are erect 

 and do not touch the hydrospires, wdiereas in the latter they 

 are recumbent and lie back ujion them. Each of the arms of 

 Codaster has a fine anibulacral groove; and all of the grooves 

 terminate in a single central aperture. But, as this aperture 

 ■was covered over by a thin plated integument, as in the Blas- 

 toidea, I have not shown it in the diagram, but only the five 

 pores, J)- 



No one wdio compares a Codaster with a Pentremites (the 

 internal structure of the latter being visible) can doubt that the 

 liycU'Ospires of the two genera are perfectly homologous organs. 

 If we grind off the test of a species of the latter genus, select- 

 ing one for the pur])ose which has broad petaloid ambulacra, 

 such as those of P. Schidtzii, the structure exposed will be 

 that represented in the diagram fig. 3. In Pentremites, as in 

 Codaster, the iive hydrospires are divided into ten equal parts 

 by the five rays, a 1 , a 2, &c. In Codaster these ten parts re- 

 main entirely separate from each other ; but in Pentremites 

 they are reunited in pairs, the two in each interradial space 

 being so connected at their inner angles that their internal 

 cavities open out to the exterior through a single orifice or 

 spiracle {s, figs. 3 Sc4). This is best shown in fig. 4, intended 

 to represent the structure of P. ellipticus (Sowerby), as de- 

 scribed by Mr. Kofe, Geol. Mag. vol. ii. p. 249. In this spe- 

 cies the hydrospires, instead of being formed of broad sacs 



