490 RUM. GEOLOGY. 



same mass of rock may thus be disposed in both these 

 forms, it will follow that if a stratified and an unstratified 

 rock of the same mineral composition should occur in 

 different places, their difference of disposition is no proof 

 that the periods or causes of their formation were dif- 

 ferent. 



In the next place it is plain that the same rock (geo- 

 logically considered) may in its different parts possess 

 a different mineral composition ; since on the one extreme 

 of this rock, which forms a single deposit, is found augit 

 alone, and on the other felspar. Similar changes occur 

 in some of the trap rocks ; basalt or greenstone being in 

 many cases portions of a common deposit, of which 

 syenite, porphyry, or compact felspar is the other. The 

 observations made in Mull and Arran, hereafter to be 

 described, confirm the truth of this statement. The change 

 is not greater in the one case than in the other ; and it is 

 as easy to imagine the hornblende to be excluded from 

 a common mass of hornblende and felspar, as that the 

 augit should, as in this instance, disappear. It is true 

 that geologists have admitted these several rocks to 

 belong to one collection of overlying substances, but it 

 has still been often imagined that there were essential 

 differences either in their relative ages with regard to each 

 other or in the nature of the stratified rocks which they 

 accompanied. As far as any evidence to be derived from 

 the phenomena visible in the Western islands is valid, 

 there is no ground for this supposition; the several 

 varieties, or species of the whole tribe appearing in 

 most, if not in all, the instances examined, to be asso- 

 ciated by a common set of geological relations. 



The last conclusion to be deduced from these facts 

 relates to basalt. Notwithstanding the trouble which 

 mineralogists have taken to describe and limit this sub- 

 stance, it is as yet incapable of being defined by any 

 number of characteristic properties. The difficulty is in- 

 surmountable, since it consists in the very essence of this 



