228 



TRANSACTIONS OF THE 



ascertain, it continues in connection with the scapula 

 downwards to the pelvis. (See pi. 10, fig. 3.) 



They are constructed of several pieces, the sutures of 

 their junctions are visible in some, while in others they 

 are entirely obliterated, but I had the good fortune not 

 only to find specimens in which these sutures were very 

 plain and could be separated, but I found also separate 

 parts imbedded in the rock. Mr Say, in concurrence with 

 Mr Miller, says, (loco citato) " the pelvis is composed 

 of three pieces, tw^o pentagonal and one quadragonal.'' 

 Neither the form of the scapula nor that of the intersca- 

 pula is given by these authors ; it is only said ^' they are 

 large, and deeply emarginated for the reception of the 

 ambulacra, and truncated for the junction of the sub- 

 rhomboidal interscapular.'' Nor is the description pub- 

 lished by Dr Goldfuss and M. de Blainville correct, ac- 

 cording to our specimens. The result of our observations 

 shows that the whole surface of the animal, not including 

 the ambulacra, is composed of thirteen pieces. The 

 base, or as it is called the pelvis, contains three pieces, 

 one pentagonal, and two heptagonal, having a re-enter- 

 ing angle. We have endeavoured to show these several 

 pieces, which we found isolated, imbedded in the rocks, 

 in pi, 10, fig. 1 and 2 ; the pieces alluded to are marked 

 a a a m the two figures. Five pieces which receive 

 the tips of the ambulacra, having more or less the form 

 of a horse shoe, fig. 1 and 2, b 6; and five pieces 

 having four sides of a subtriangular shape, fig. 1 and 

 2, c c c. 



Seven species seem to be known of this genus. 



1. Pentremites globosa, Say, supposed to have been 

 found at Bath in England. We have found it in Ala- 

 bama, at Mount Sano and vicinity j Tennessee, Crab 

 Orchard mountain ; and Illinois. 



2. P. pyriformis, Say. Tennessee near Sparta, 



