Mineralogy of Sky. 9 



arts. From the identity of rocks in different tracts a conclusion 

 may be drawn with regard to the identity of the resulting soils, and 

 the analogies discovered in them may perhaps with less trouble and 

 as great certainty as any chemical analyses, lead agriculturists to 

 hazard those experiments in improvement which seem justifiable 

 from similarity of composition no less than similarity of climate and 

 exposure. I shall therefore I trust be excused for dwelling a little 

 longer on a subject intimately connected with geological science, and 

 in itself among the most practically interesting that can occupy our 

 attention. 



It is already well known that many of the rocks of the trap 

 family, like some lavas, afford on decomposition one of the most 

 fertile soils with which we are acquainted. On former occasions 

 I have mentioned this fact, and noticed at the same time the great 

 differences which appeared In this respect in different situations 

 where there was nevertheless a considerable resemblance in the 

 rocky substratum. The different proportions of calcareous earth, of 

 alkali, and of carbon, which latter appears to be an ingredient in 

 some traps, are probably the circumstances which constitute the chief 

 differences in this case ; but with this we must also consider that the 

 various rocks of this class differ in composition in perhaps a greater 

 degree than any others with which we are acquainted, and that 

 together with this essential variation in their composition, they also 

 possess infinite varieties in hardness, and in the property of decom- 

 posing by the action of air and water. Many considerations must 

 therefore enter into our views before we can decide on the fertility 

 of the resulting soil, unless it be actually exhibited in those places 

 where decomposition has already taken place. 



In the district of Sky which I am now considering, there appears, 

 as far as an accurate mineraloglcal investigation can determine such 



Vol. hi. » 



