INTRODUCTION TO THE TRAf ISLANDS. 237 



there is no apparent reason why they should not be 

 equally found overlying the rocks of primary formation ; 

 nor indeed, on any view of the causes of their deposition, 

 is this circumstance to be explained ; since, being the 

 uppermost of all rocks and the last in point of time, 

 they might be expected equally to cover all those whose 

 limits in elevation do not exceed the utmost height at 

 which they are found to exist. Although this connexion 

 should even prove very partial it is not the less worthy 

 of the attention of geologists ; but the solution, like that 

 of a thousand other difficulties which attend this infant 

 science, must be expected from future acquisitions of 

 knowledge. 



Whatever connexions more intimate than those at 

 present visible, may exist among the several detached 

 portions of trap, as they occupy either the islands or the 

 shores of the neighbouring continent, are for ever con- 

 cealed from us by the insuperable obstacle which the 

 ocean throws in our way. Hence it is impossible to 

 judge whether the Maddies and the Shiant isles are now 

 connected beneath the sea with Sky, or whether they 

 have ever been so united. The occurrence of St. Kilda, 

 so far remote from the whole, might perhaps rather induce 

 us to adopt the contrary belief, and to consider them as 

 independent formations. 



Contemplating the Trap islands in this general view, 

 the occurrence of the veins of that substance is not the 

 least interesting part of the subject. These are notoriously 

 common in the Western islands. I have pointed out many 

 of them already in the Gneiss islands, and they will here- 

 after be shown to be even more frequent in the other 

 islands of this sea. Wherever they are found, they seem 

 to abound most in the vicinity of the great masses of 

 trap. As proofs of this, I may quote their greater rarity 

 in the remote parts of the Lewis than in the Schistose 

 islands, which lie in the vicinity of Mull ; and their fre- 

 quency in Arran and the stratified parts of Sky, compared 



