139 



ground enriched by numerous undulating striae, some of which unite 

 with the rays of which the stars are themselves composed. At Planche 

 XLI. Fig. 6, a. and 6, b. is a fossil which also bears a general resem- 

 blance to our fossil, but differs from it in having, instead of undulating 

 lines, superadded rays, which diverge in straight lines from the central 

 star. The fossil, Planche XLI. Fig. 5, which he describes, Astroite ma- 

 mellone, avec des rayons en tete de Meduse qui partent de chaque 

 mamelon, would, if reduced to a smaller size, approach much nearer 

 to our fossil than do either of the others. 



The specimen, Plate XII. Fig. 4, like the former, is highly inte- 

 resting, from the difficulties which accompany an inquiry into its ori- 

 ginal nature; rendered still greater, by the want of any analogous 

 bodies, with which any instructive comparison might be made. This 

 fossil, also obtained from St. Peter's mountain, is formed of the same 

 substance as the former. The organic part of this fossil is confined 

 to the superior part, and to a small cleft on one side, and is composed 

 of a surface so finely granulated, as to give an appearance like the 

 pile of velvet. On this general surface are disposed, in small depres- 

 sions, oblong bodies, in an oblique but almost horizontal direction, 

 each body being composed of six similarly formed tubes or fibres, 

 from an eighth to a quarter of an inch in length. Most of these bo- 

 dies taper a little, at their more detached ends, and thereby obtain 

 somewhat of a conical shape. 



A slender fibre branches from about the centre of some of these 

 bodies, and is inserted in the side of the adjoining depression. In. 

 some there are two of these fibrous connecting processes, and in others, 

 no traces of them are observable. But from considering the delicacy 

 of their structure, and their liability to be broken, and observing in 

 that lower stratum, at the one end of the specimen, which has been 

 protected in a considerable degree from injury by the superincumbent 

 layer, that this slender process is attached to every one of these cy- 

 lindrical bodies, I think there is very little reason to doubt, that this 



