424 DESCRIPTIONS OF ROCKS. 
almost all kinds of igneous and metamorphic rocks, inelud- 
ing trap or doleryte. | 
For a particular account of the distinguishing character- 
istics of minerals studied by microscopic methods, reference 
must be made to treatises on the subject. 
IV. KINDS OF ROCKS. 
1. Rocks are generally mixtures of two, three, or four 
prominent mineral constituents, with also others, it may be, 
of less importance. Each mineral adds a distinctive fea- 
ture, and might be a reason for a new name. But it is 
usual with lithologists to base the distinction mto kinds of 
rocks on the ¢wo chief minerals, and make the others acces- 
sory species and the basis only of varieties. This method 
is prompted by convenience, and also by the fact that the 
more important characteristics are commonly contained in 
two of the constituent minerals. It has many exceptions, 
however, and particularly where a third mineral has special 
peculiarities and abundance. 
2. Difference in kind of rock is naturally based on dif- 
ference in chemical or mineral constitution, and identity, 
accordingly, on essential identity in this respect. Conse- 
quently when there is no essential difference in chemical or 
mineral constitution, there is no sufficient reason for a dis- 
tinction in kind or a difference in name, unless the wide 
distribution of a particular variety, and the permanence in 
its characters, make the distinction in name a geological 
necessity. 
In accordance with this statement, the distinctions among 
crystalline rocks of coarse or fine in texture; of being por- 
phyritic or not; of containing glassy grains among the 
stony or not; of being foliated or not in crystallization, are 
of little value compared with the real mineral constitution, 
and are a fit basis only, at the best, for varieties. But the two 
rocks of like composition, trachyte and felsyte, retain their 
characteristics so widely, that geology needs both names, 
and only demands that their essential identity should be 
held in mind. 
The same kind of rock is in many cases both of metamor- 
phic and eruptive origin; still the difference of origin is 
not a sufficient basis for a distinction of kind unless there 
is some marked difference between them, and an extended 
