Professor Forbes on iJte Geology of Auvergne. 15 



larger the fissures ; and the greater the diameter of the crater the 

 smaller the sum of the interrupted spaces of its circumference, 

 are obvious upon mere statement ; and since the comparisons 

 with nature do not admit of more precision, any formal calcula- 

 .tion goes for very little. 



Yet there are very strong grounds for adopting the conclu- 

 sions of MM. Ehe de Beaumont and Dufrenoy, nor do I find 

 it easy to conceive on what grounds the earnest hostility mani- 

 fested to their views has been maintained. Passing over most of 

 the arguments which have been urged by themselves and others, 

 I will point out one or two which have struck myself with pecu- 

 liar force, as regards the Cantal and Mont Dor, and especially 

 the former. 



1. The Cantal presents a crater-hke basin, in which rise 

 phonolitic cones, and from which radiate numerous valleys, like 

 jdeep scars, so similar to the divergent cracks of a piece of ice or 

 of a starred piece of glass {etoile), as to lead us to imagine that 

 the excellent map given by MM. Elie de Beaumont and Du- 

 frenoy had been drawn under the inspiration of theoretical views, 

 did we not find it fully confirmed by the unbiassed authority of 

 the accurate Cassini. An examination of the valleys, too, would 

 lead us to the conclusion that, in general, they are valleys of dis- 

 ruption, not of erosion; but admitting, as I do, after paying 

 long continued and earnest attention to this important and diffi- 

 cult question in geology, (the origin of valleys), that there seems 

 no universal and unimpeachable criterion between the two, I 

 would not now press that assertion. There is, however, this 

 most striking in these valleys, (I speak with most confidence of 

 the valley of the Cer in the Cantal, which I examined with 

 most care, and the same may be added of some valleys of the 

 Mont Dor), that they often rather contract than increase in 

 width, as we retire from the centre of the group, at the same 

 time that they are less profoundly excavated, till at last they 

 merge in the slightly swelling surface, exactly as the theory of 

 elevation would prescribe. They are frequently bounded on 

 either side by mural precipices, and of lateral valleys there are 

 almost noiie. 



2. M. Elie de Beaumont has justly pointed out the regularity 

 of the succession of layers or coats which occur over the conical 



