M. Keilhau's Theory of Granite, and other Rocks. 9T 



out of consideration all these artificial speculations, and do not 

 require to think of emanations, whether with reference to dolo- 

 mite or to crystalline siliceous rocks. We should otherwise be 

 led into a dark region, which will not for a long time be illu- 

 minated by the light of chemistry. . It can be of very little 

 service to trouble ourselves with hypotheses wherein chemistry 

 teaches us to respect its operations, but of which it shews us 

 that not one half can be made available. It is, therefore, a 

 much more natural proceeding for us to confess candidly, that 

 the moment is not yet come when we can give a full and satis- 

 factory account of the phenomena observed, and that we should, 

 therefore, as yet limit ourselves to the examination of the facts, 

 such as they are, which are the first steps to an explanation, 

 without for the present entering on the more remote causes. 



We are undoubtedly in possession of the following two most 

 important facts : — First, that masses in a state of compact ag- 

 gregation can crystallize or undergo morphological changes, 

 without, so far as we see, previously acquiring, in any degree, a 

 fluid condition ;* and next, that substantially newly formed bo- 

 dies, crystallized or uncrystallized, can be developed in com- 

 pact masses likewise, without our remarking that the matter, 

 immediately before the formation of the new body, was melted 

 in any of the known methods. These two facts are all that we 

 require, but they are absolutely necessary to enable us to ex- 

 plain a very large number of geological phenomena, both small 

 and great, — so much so, that it really appears quite extraordi- 

 nary that proper attention has not been bestowed upon them. 

 For my own part, ever since I directed my attention to these, 

 though obscure, yet very important facts, I have constant- 

 ly endeavoured more and more to shew their value in the 

 science of geology. Even in 1828,f I endeavoured to advance 



* Gaj Lussac, after narrating the well known fact regarding barley-sugar, 

 expresses himself in the following manner : " Ceci prouve que, dans un corps 

 solide, les molecules peuvent changer de position et prendre la forme crystal- 

 line. Cet exemple a fait dire que les molecules des corps solides peuvent, dans 

 certaines circonstances, prendre de nouveaux arrangemens. C'est qui a lieu 



pour un grand nombre de sels Ainsi les molecules des corps solides 



ne sont pas fellemenl liees entre elles, qui'elles ne puissent changer de place 

 el former d'autres groupes."— (Cours de Chimie, II. 26, Lefi. 24.) 



t Poggendorff 's Arlnalen, xiv. 134. 

 VOL. XXV. NO. XLTX. JULY 1838. G 



