200 AN ATTEMPT TO DESCRIBE THE ANIMALS THAT MADE 



2.3 inches ; between the middle and hind toes, 3.3 to 4.7 inches. 

 Distance between the roots of the front toes and the root of the 

 hind toe, 0.8 to 1.3 inch. Length of the middle toe bejond the 

 rest, 1 to 1.6 inch. Toes slightly curved; the two front inner ones 

 inward, and the outer one outward. Axis of the foot correspond- 

 ing nearly with the line of direction. The whole length of the 

 tarsal bone reaches the ground usually in walking. Track shown, 

 of the natural size, on Plate 8, fig. 4, and Plate 10, figs. 1, 2, 3. 



Localities. — Wethersfield, at the Cove, on gray shale, or mi- 

 caceous sandstone, at Turner's Falls, and Cabotville. 



Remarks. — Nearly all the facts within my reach would indicate 

 that this animal was a bijied. Yet the long heel and side toe, so 

 like a lacertilian, have long led me to suspect it might be a quadru- 

 ped. I have sometimes found two tracks almost in the same spot, 

 as is common with quadrupeds. But still the most instructive case 

 of this kind, already referred to under the third general character, 

 does not confirm this supposition. By a careful dissection of No. 

 171 in my cabinet, I found, on three successive layers of the rock, 

 three impressions so unlike as to perplex the most practised eye ; 

 but I think I now understand them. The uppermost layer presents 

 a track as exhibited on Plate 10, fig. 1, having five toes in front and 

 one articulated to the tarsus, or tarso-metatarsus. The lowest 

 layer, represented on Plate 10, fig. 3, shows five toes most sym- 

 metrically arranged, and scarcely exciting a suspicion that there 

 could be two tracks. But I felt quite confident that existing an- 

 imals would not allow us to give six toes to the foot of any biped 

 or quadruped ; and therefore I ventured, at the risk of spoiling 

 the specimen, to cleave it asunder once more ; when I was pre- 

 sented with the outline shown on Plate 10, fig. 2, which seems to 

 me to solve the enigma to a considerable extent. It shows, in my 



