Ch. XIV.] OF THE STRATIFIED ROCKS. 319 



The objections already advanced appear to us to have no 

 little weight, but our dissent from the prevailing theory does 

 not solely rest on these grounds ; for we are prepared to 

 show that the slates, at their junction with granite, not only 

 exhibit a difference of composition from that of the adjacent 

 mass, but that the latter also contains an element which is not 

 to be found in sedimentary clays, marls, and sands, from 

 which they are supposed to have been derived. And, in 

 order to establish this point, chemical analyses must be again 

 referred to, and though these are certainly not so numerous 

 or accurate as might be desired, yet they, perhaps, approxi- 

 mate near enough to the truth, to enable us to estimate 

 the value of this argument. The compositions of several 

 substances, which bear on this question, are given in the 

 subjoined table. 



From this table we learn that the proximate principles of 

 all the compound crystalline slates, with the exception of 

 quartz, contain no inconsiderable proportion of alkali, either 

 of potash or of soda : but this substance does not form a com- 

 ponent part of the secondary slates. This, indeed, is what 

 might be expected a priori ; for felspathic rocks, during their 

 decomposition, lose their alkali, as in the formation of porce- 

 lain clay from granitic protogine and some kinds of eurite : 

 indeed, the solution and abstraction of alkali from decom- 

 posing felspar appears to facilitate the rapid disintegration of 

 the rocks which contain this mineral. 



Those who support the conjecture concerning the dolomiza- 

 tion of limestone will, of course, attribute the presence of 

 alkali to a similar cause, to the transfusion of this substance 

 through the primary slates ; and in this case the opinion 

 seems more tenable, magnesia being exceedingly refractory, 

 whilst the alkalies may be fused and even dissipated in vapour, 

 by an intense heat. Here, however, the same difficulty occurs 

 as in the case of dolomite ; the layer of rock immediately 

 adjoining the granite does not always contain alkali, but is 

 sometimes almost entirely quartzose. 



