peat under different headings the same mineral, in order to suit 

 such uncertain descriptions, as it is manifestly out of the power 

 of the compiler to distinguish that which is most correct; besides, 

 it is more probable that any similar discordance arises in a real 

 difference between two or more given specimens of the same 

 mineral, than that any error exists in description. 



Analogous reasoning has induced him, as before said, contrary 

 to custom, to introduce those metallic minerals which have an 

 imperfect or semi-metallic lustre, under both divisions, " lustre 

 metallic," " lustre not metallic," as the distinction between those 

 minerals in the intermediate stage and those positively of metal- 

 lic or not metallic lustre, is, as regards those characters, often 

 evanescent and inappreciable, a fact which is in harmony with 

 the present received opinion that earths are metallic oxides, and 

 metals deoxygenated earths ; the distinction which arises from 

 the presence of the true or semi metallic lustre has not, however, 

 been lost sight of 



In the construction of tables of this description there is more 

 risk of failure by attempting to be too definitive or exclusive,* 

 than the contrary. Assume in mineralogy what arrangement 

 you please, a neutral ground will always exist more or less on the 

 divisional frontier, which is moreover, better represented by a 

 broad band, whose colors imperceptibly blend, than by a line. 

 Agreeably to these views, the characters of hardness and specific 

 gravity have been expanded, as it were, so as to embrace all the 

 respectable authorities consulted. 



When the doubt originates in the compiler's inexperience, the 

 same plan has been followed, as it will probably happen that 



* To be usefully exclusive, a characteristic should be made, to include all respect- 

 able statements, as the attempt to be too definite by being too partial, may occasion 

 a given mineral to be omitted altogether, or occasion its introduction only under an 

 inappropriate heading. Also the attempt to contract too much by decimation those 

 two important characters, specific gravity, and hardness, narrows the sphere of their 

 usefulness, by introducing into descriptions a precision which probably does not ex- 

 ist in nature, or if it does, cannot be taken advantage of by the generality of students 

 owing partly to their own want of skill, but more to the varying nature of the evidence 

 they consult. The crystallographic character is perhaps the only one which will 

 admit of such nice distinctions, but its abstruseness and comparative infrequency of 

 occurrence are unfavorable to its being employed as a characteristic. 



2 



