REMAINS OF REPTILES AND FISHES. 109 



would take place by their coming in contact with the ground. 

 And again, the largest or oldest spines are uniformly the most 

 worn ; some, indeed, are reduced to mere stumps. In one such 

 specimen now before us, which is seven inches in circumference, 

 and which must have been one of the largest, only 10^ inches 

 are left. Another example, six inches in circumference, is only 

 seven inches long, including the portion buried beneath the skin. 



All this seems to demonstrate, beyond doubt, that these are 

 really paired spines, most probably pectoral. And from this 

 wearing we may fairly assume that Gyracanthus was a ground 

 fish, and that the spines assisted its motions at the bottom of the 

 water. 



The straight spines, or those which are not laterally bent, are 

 all regularly arched from before backwards ; and their distal or 

 pointed extremities are all perfect, not being in the least degree 

 worn. These are apparently dorsal spines ; and that there is 

 only one of such in each fish seems probable from the fact that 

 they occur in the ratio of one to two of the paired spines, as 

 shown by our previous division of the seventy-one specimens. 



The dorsal spines are considerably smaller than the paired 

 ones ; they are more compressed, and the posterior denticulated 

 keel is more strongly developed ; the extreme point is smooth, 

 compressed, and rounded in front. The largest are about eleven 

 inches long, and 3^ inches in circumference at the thickest part. 

 The paired spines are fifteen or sixteen inches in length, and up- 

 wards of 6^ inches in circumference. 



One or two specimens of the species denominated G.formosus 

 have likewise occurred ; and as the same spine of G. tuberculatus 

 is occasionally found with both tuberculated and smooth ridges, 

 the former can scarcely be considered a good species. M. Agas- 

 siz's figure of G.formosus,^''' like G. tuberculatus, is laterally bent. 



Large flat triangular bones are frequently found associated 

 with the spines, measuring sometimes 8^ inches long, and 6|- 

 inches broad at the widest part. Their structure is very open ; 

 and as they are seldom well preserved, they are probably only 

 imperfectly ossified, the bone fibre radiates from the apex to the 



* "Poissons Fossiles," vol. iii., tab. V., figs. 4, 5, 6. 



