500 Observations on the 



their usual order of operation, as the mouth was doubtless 

 at the side of the neck, near the base of the arms. This 

 aperture, however, has not been observed, although a sepa- 

 rated sack of C. dilatatus has been carefully examined, 

 together with more than half its neck, without the dis- 

 covery of any aperture whatever, and four plates of the 

 anal series seen, with no better result. 



Other peculiarities of all the species at present known, 

 are, — the primary radials are only two in number to 

 each ray ; each of the second radials supports two arms, 

 which are composed of from four to ten joints to the first 

 bifurcation, which are subcylindrical, and lie folded against 

 the neck of the ventral sack. 



The joints of the upper parts of the arms in one spe- 

 cies do not reach quite across, but alternate in such a 

 manner as to give somewhat the appearance of a double 

 series. 



The column of all the known species is round, and com- 

 posed of thin joints of nearly equal size ; and the surfaces 

 of all the plates, except those of the sack, are without or- 

 namentation, and the sutures are not impressed. The anal 

 space is rather wide, and the series of plates, so far as 

 seen, consists of a double, alternating series, the first plate 

 resting in the retreating angle between two of the sub- 

 radials, and the second, on the upper truncated side of one 

 of these subradials. 



The plates of the ventral sack are rather small, tumid at 

 the upper part of the sack ; and in two species they are 

 radiately ridged. Some separated parts of these sacks 

 show the plates arranged in somewhat regular rows, which 

 pass over the top, from side to side, their arrangement 

 being more confused on the two opposite sides ; yet the 

 symmetry of the sack does not seem to be affected by the 

 disposition of the plates. 



Should these peculiarities prove to be of generic im- 



