458 CANNA. GEOLOGY. 



The similarity between the trap rocks and those of 

 volcanic origin has been too often noticed to require 

 a repetition of that remark. It is equally known to geolo- 

 gists that volcanoes are situated in this class of rocks ; 

 a position which appears to countenance the opinion that 

 there is a connexion in the origin of the two ; and that the 

 recurrence of trap in any given place, such as it is 

 observed in the island under review, is a phenomenon 

 closely linked with the actual existence of volcanoes 

 in districts of trap rock. 



It is still necessary to point out a circumstance of 

 no uncommon occurrence in the trap of this island, as 

 well indeed as in many others which I have examined ; 

 partly because it confirms this conjecture and is an 

 additional proof of the resemblance between the traps 

 and the lavas, and partly because it has been doubted 

 or denied by many geologists who have treated of these 

 rocks. This is the existence of cavities resembling 

 in every respect those which are contained in the 

 scoria of volcanoes, or in cellular lavas. They occur 

 also abundantly in Mull and in the trap near Oban; 

 and are particularly conspicuous in some parts of the 

 little Cumbray, where I shall have occasion to describe 

 them more fully. 



The results which follow from the existence of such 

 cavities are evident, since they remove the necessity 

 of supposing that the trap rocks, admitting them to 

 be of igneous origin, have been formed by submarine 

 eruptions and under great pressure. The distinction 

 must cease with the loss of the proofs on which this 

 supposition is conceived to rest ; and these rocks might 

 therefore be assigned, as far at least as any proofs 

 to the contrary have yet been produced, to causes of 

 the same nature with those which now operate in 

 producing lavas ; to ancient volcanoes, of which the 

 other and perhaps more characteristic traces have dis- 



