VALUE OF CHEMICAL ANALYSIS IN LITHOLOGY. 31 



Section IV. — Chemical Analz/sis of RocJcs. 



At the present time the most that chemical analysis seems to be able to 

 do for the lithologist is to give the composition of the rock as a whole. 

 The many attempts that have been made to determine the mineralogical 

 composition of rocks by unaided chemical analysis appear to have been in 

 almost every case a fiiilure. This is natural, for this method alone is unable 

 to take into account the three different classes of minerals in rocks, and 

 in its statements has to proceed as if all the minerals were the pro- 

 ducts of free crystallization in the rock. But while the chemical composi- 

 tion remains about the same, every gradation in structure and mineral 

 composition is known to exist, — from a pure glass to mixed glass and 

 crystals, to a purely crystalline rock, and to one in Avhich all the mineral 

 constituents are secondary or alteration products. Since the chemical 

 composition of all these forms is essentially the same, the results of any 

 calculation of the percentage and kind of minerals inclosed, would be 

 nearly the same ; but how different from the reality are the results of the 

 calculation, except when the rock is composed of crystals of the second 

 class alone. Even here the correctness of the result would be a matter 

 of doubt. 



Chemical analyses of rocks, showing their ultimate constitution, if made 

 from specimens carefully selected and studied in the field, and further 

 studied microscopically, would aid greatly in lithological research. Typical 

 unaltered specimens are needed to establish rock species ; and for such Avork 

 the average specimens of collectors are too much affected by surfoce altera- 

 tion, or weathering, to be used. But a large proportion of rock analyses 

 appear to have been made from such unsuitable specimens, of whose struc- 

 ture, mineral composition, and field relations we know nothing, or next to 

 nothing; this, too, when the chief value of such anal3'ses is to enable us not 

 only to institute comparisons between the chemical composition of the rock 

 analyzed and that of other rocks, but also between that composition and 

 its origin, structure, mineral composition, and physical relations. 



Chemical analyses could be made of great service in lithology by taking 

 a graded series of rocks, beginning with the unaltered form, and passing 

 gradually into the extremely altered form, comparing step by step the chem- 

 ical comiMsition with the changes in structure and mineral composition. 



