1862.] DAWKINS HY^NA-DEN. 121 



on the floor, was not so coated. This deposit may, perhaps, explain 

 the absence of round balls of Album grcecum, which, assuming that 

 the cave at the time was more damp than that at Kirkdale, would 

 be trodden down on the floor by the hysenas, instead of presenting 

 a rounded form. The stone also itself exhibits tooth-marks, and 

 probably was gnawed by the hyaenas, like the necrosed antlers, for 

 amusement. Dogs are very fond of exercising their teeth in this 

 way. This discovery also proves that violent watery action had but 

 small share, if any, in tilling the cave ; for in that case the soft 

 Album grcecuyn would have been removed from the stone. 



The section made in cutting this passage presented irregular layers 

 of peroxide of manganese, full of bony splinters, and in general 

 covered by a layer of bones in various stages of decay. These layers 

 disappeared in the upper portion of the passage. There were masses 

 of prismatized stalactites scattered confusedly through the matrix. 

 After excavating the vertical branch as far as we dared (for the 

 large stones in it made the task dangerous), we were compelled to 

 leave off", having penetrated altogether only 34 feet from the cave's 

 mouth. In this vertical branch, the bones, stones, and red earth are 

 cemented together by carbonate of lime, — a circumstance which 

 added materially to the diiflculty of the excavation. 



A short distance from the entrance the cave gives off a lateral 

 branch to the left, which tends obliquely upwards, and is abruptly 

 closed by stalagmite. This forms a marked contrast to the rest of 

 the cave, being covered with stalactite and stalagmite, and free from 

 debris ; while the other parts are full of debris, and at the same time 

 free from any but the merest traces of carbonate of lime, except in 

 the case of the vertical branch above mentioned, where, however, it 

 does not assume a stalagmitic form. 



There are numerous caverns in the vicinity which, in all proba- 

 bility, are connected with the one under notice, and which, to say 

 the very least, are parts of the same great system*, and all open 

 upon the same ravine. And even this probably is but a cavern 

 unroofed by the chemical action of the carbonic acid in the air, 

 by which the insoluble carbonate of the stone is changed into the 



* By a system of caverns I mean all those which open upon a common ravine. 

 Through this a stream often flows, supplied in many cases by feeders out of one 

 or more caverns. On close examination of a number of the caves in the Mendips, 

 I find them, in the main, ranged round their ravines as branches are arranged 

 on a tree. Burrington-Comb, Cheddar Pass, and Wookey-Hole Ravine, each 

 siu*rounded by its system, are eminently typical. I do not see the reason why 

 the change of insoluble carbonate into soluble bicarbonate of lime, by which 

 swallow -holes and parts of caverns are perpetually and gradually being enlarged, 

 should be limited in its effects, if infinite time be granted, and why it should not 

 have been the chief agent in forming the ravines so common in all limestone dis- 

 tricts. By this process one of the caverns at the top of Cheddar Pass is gradually 

 being unroofed, and is becoming a miniature ravine. On tliis view, the great 

 majority of limestone ravines are but ruined caverns. The loose stones on the 

 summit of the Mendips in many cases present a ground-plan of a system of 

 caverns on the upper surface, by the chemical action of the carbonic acid, the 

 main channel being surrounded by niiraerous accessory ones, which collect all 

 the moisture on the surface in their ramifications. 



