159 



characters by which rocks may be distinguished, 

 it is necessary to bestow a few words on that of 

 colour, since it has often been made a ground of 

 distinction. It is, however, generally afi em- 

 pirical, and, frequently, a fallacious criterion of 

 the nature of a rock. In the case of mixed 

 rocks, it sometimes facilitates the investigation of 

 the different substances which enter into the 

 composition ; but, being subject to vary, it fre- 

 quently deceives, by its changes or its absence, 

 those who have been accustomed to rely on it. 

 In some instances, it appears to have misled 

 many observers respecting the true nature of a 

 rock, so as to have produced much confusion ; 

 its obvious nature encouraging a careless mode 

 of examination, and taking off the attention 

 from the more obscure, but more essential cir- 

 cumstances, in which the distinction consists. 

 The different substances ranged under the term 

 basalt, and the occasional confounding of the 

 primary and secondary sandstones, in conse- 

 quence of the red colour of both, will illustrate 

 this latter remark. Gneiss, will illustrate the 

 former ; being often confounded with micaceous 



