found in t?ie Trap-Rocks of Scotland. 89 



classes, and to give them very expressive names, if the minerals 

 which they contain were well known. The unicoloured crystal- 

 line traps occasionally contain labradorite, and the presence of 

 this mineral may account for the soda that has generally been 

 found in them, as well as in basalt, to which they are closely al- 

 lied ; but it does not, in some cases, account for the whole of the 

 soda, so that they probably contain other minerals, of which this 

 alkali forms a constituent part. I have analyzed two of these 

 rocks, that contained labradorite, and the analysis differed but 

 little from those of some basalts. The neighbourhood of Glas- 

 gow affords some very fine specimens of columnar trap, which 

 all contain labradorite. The range of large and regular columns 

 that stands on the side of a wooded hill, a little to the south of 

 Strathblane, is well known. It is composed of a dark green trap 

 or greenstone, containing crystals of that mineral. The same 

 may be said of the black basalt, that forms the remarkable hori- 

 zontal columns, near Altmarry turnpike, on the road from Glas- 

 gow to Drymen ; and also of the columns that are found in more 

 than one place on the Kilpatrick and Cathkin Hills. So that 

 labradorite is probably one of the minerals of which basalt is 

 often composed. 



I have made these remarks upon the situations in which I 

 have found labradorite, because I beheve that we cannot at- 

 tempt to arrange the trap-rocks, or give them expressive names, 

 until we are better acquainted with their composition. The pre- 

 sent nomenclature of these rocks is very defective, and, not be- 

 ing formed upon any fixed principle, it cannot be improved. 

 The names must either be multiplied to such an extent as to 

 render them very burdensome to the memory, or be applied so 

 loosely, as always to require a definition. The short and simple 

 terms now in use have not the flexibility which names, applied 

 to objects that are constantly varying in their characters, should 

 possess. It would be better to employ in geology, as often as 

 possible, compound descriptive names, that would indicate the 

 structure and composition of rocks. Such names would, in many 

 cases, be shorter than those now employed, as they never would 

 require a definition ; and, when formed upon a fixed principle, 

 they might be multiplied without loading the memory. As far 



