The first division is again parted into subdivisions depending 

 upon the presence or not of metalhc lustre. Some innovation 

 having been here made upon the estabhshed practice, as regards 

 the application of these characters for purposes of distinction, I 

 am bound to give a reason for it. Certain minerals in the me- 

 tallic class possess a lustre which is intermediate between metal- 

 lic and non-metallic ; these have hitherto, generally speaking, 

 been ranked with metalhc minerals, which want the metallic 

 lustre, but there exists quite as much reason to arrange them 

 with those which possess it, for the fact is that they are often 

 not to be distinguished from them by this character alone. 

 Moreover, one eye often considers that to be metallic lustre 

 which another declares is not, and that in spite of the test of 

 the streak, which, however good, is not infallible. Influenced 

 by this reasoning, those minerals which are described as pos- 

 sessing the imperfect or semi-metallic lustre have been intro- 

 duced, throughout these tables, under both heads, "lustre metal- 

 lic," "lustre not metallic." This will, undoubtedly, occasion 

 repetition, but the student will be benefited by it, as it will save 

 him the trouble of referring, backwards and forwards, from one 

 table to another. It has always appeared to us, that a tabular 

 view should embrace all probable and even possible cases, as 

 well as positive, and we have endeavored throughout this, to con- 

 form to that opinion. 



The subdivisions in this grand division, where necessary, are 

 divided into sections, depending upon hardness ; these require 

 no comment farther than to remind the reader that they blend 

 one into the other, by which observation it is intended to convey 

 to him the hint, that if he fails to find the mineral he is in search 

 of, under one section, he will probably meet with it in the next; 

 and this remark applies to the character of hardness throughout 

 these tables. 



In those sections in which the minerals are numerous, sub- 

 sections are introduced: thus, in tables 9, 10, 11, 12, there are 

 sub-sections; 9 and 10 are made to depend upon the development 

 or not of the odor of selenium (like putrid horse-radish) before 

 the blowpipe; 11 and 12 upon specific gravity. As in the in- 

 stance of arsenical odor, Berzelius is again my authority for giving 



