120 KliRRERA. GEOLOGY. 



extend so far, and is not followed by the same conse- 

 quences which here arise from the existence of fragments 

 of trap in the conglomerate rocks. It has been shown 

 that the conglomerates of that island are of a local nature, 

 and peculiarly connected with a succession of partial de- 

 posits of trap rocks. But it has been proved, that in the 

 island now under review and on the adjoining continent, 

 the conglomerate is part of the extensive deposit which 

 separates the primary from the secondary rocks ; or that 

 at least it is analogous to this, and equally to be con- 

 sidered as the first in order of the general series of these 

 latter. It therefore follows that the fragments of trap 

 existing in this rock have, like the other substances, 

 been derived from previous masses of that rock of a 

 date prior to the deposition of the secondary strata, 

 and equally therefore to be ranked with the other pri- 

 mary strata, by the fragments of which it is accom- 

 panied. 



It may be deduced from many parts of this work, that 

 the trap rocks are usually associated with the secondary 

 strata, and are, in by far the predominant examples, 

 superior to them; while the instances in which they 

 occur in union with, or immediate superposition to, the 

 primary, are comparatively rare. Even hi Morven, where 

 extensive masses of these rocks are found in immediate 

 contact with gneiss in some places, the existence of inter- 

 mediate fragments of secondary strata prove that the 

 whole deposit is of a date posterior to these, and that it is 

 only incidentally and occasionally connected with the pri- 

 mary rocks. The remarks formerly made in the descrip- 

 tion of Sky, also show that such a state of immediate 

 superposition and contact is no proof of an original 

 connexion in point of time ; as the peculiar mode in 

 which this rock appears to have been deposited, enables 

 it to possess an intimate connexion with every rock 

 from the most ancient to the most recent. Under these 

 circumstances, of its predominant connexion with the 

 secondary strata, and of the absence of positive proof 



