THE TRAP ISLANDS. 65 



respect similar to the pale coloured and uppermost strata 

 of this series. 



In concluding these observations on the geographical 

 relation of the fragments of these strata, it is proper to 

 remark, that no traces of them are found on the continent 

 immediately adjoining Sky. But in Loch Greinord, far to 

 the northward, two minute and solitary portions occur; 

 highly interesting from their insulated position, and no 

 less so from their possible connexion, in times far distant, 

 with the remainder of these now disjointed fragments.* 



The last subject of inquiry is, whether these several 

 portions are, as I have here hinted, merely the separated 

 parts of a mass of strata once continuous, or whether 

 they have been originally deposited in distinct cavities. 

 It is possible that the original deposits, of some of these 

 strata at least, have been partial ; and that the present 

 resemblances are rather the result of analogy than identity. 

 This question is important, as it involves others of con- 

 siderable moment respecting the general changes of the 

 earth's surface and the theory of the secondary strata; 

 but the want of satisfactory evidence is such that I can 

 scarcely venture on it. It must probably remain for deci- 

 sion to some future period, unless the following argument 

 in favour of an extended deposit should be deemed sa- 

 tisfactory. 



It has been seen, that notwithstanding some partial 

 disturbances, there is a general conformity of position, 

 as well as x>f character, throughout the whole, wherever 

 the body of strata present is sufficiently considerable to 

 permit these circumstances to be examined. It may 

 therefore be presumed, that this identity bespeaks a con- 

 tinuity once more perfect among the several detached 

 portions ; as it is difficult to conceive that such a commu- 

 nity of character should have been affected by separate 



* This place unfortunately lies out of the limits of the accompanying 

 map. The reader may consult Arrow-smith's Map of Scotland. 

 VOL. II. F 



