322 Analysis of Scientific Books. 



metal in its native state, when it so occurs ; then its combination with 

 other metals, when in the state of a natural alloy ; then combined with 

 sulphur ; with oxygen ; and, finally, as an oxide combined with an acid. 



The combustibles ibllow, beginning with sulphur, to which succeeds 

 carbon in its purest form, and afterwards its several combinations with 

 other bodies, as the base of the greater part of all the substances belonging 

 to this class. 

 The order of arrangement is therefore as follows : — 



Earthy minerals. 



Alkalino-earthy minerals. 



Acids. 



Acidiferovs earthy minerals. 



Acidiferous alkaline minerals. 



Acidiferous ulkalino-eaiihy minerals. 



Minerals (chiefly earthy J which have Twt been analyzed, or of which 

 but little is known. 



Native metals, arid metalliferous minerals. 



Combustibles. 



In our opinion, this is at once the clearest and best arrangement 

 Jiitherto suggested. Founded on the only rational basis, compo- 

 sition, it is so skilfully subdivided, that no confusion exists in any 

 part of it. In point of convenience, too, in cases of hasty reference, 

 the essential elements of any mineral are seen instantly by casting 

 the eye on the running title at the top of the page, where its 

 description occurs. It is hardly necessary to insist on the supe- 

 riority of the preceding arrangement to that abominable violation 

 of all chemical truth, which has placed crystallized carbon at the 

 head of the earthy minerals ! 



With respect to the figures of the crystalline forms which accom- 

 pany the descriptions, and which are neatly and accurately exe- 

 cuted in outline on wood, Mr. Phillips informs us, that the 

 measurements annexed to them are to be considered only as 

 approximations to their true value, especially of the secondary 

 planes ; " for in no instance has it been attempted to correct the 

 geometry of nature by a resort to the more rigid laws of calcu- 

 lation. It has been ascertained, by a comparison of the measure- 

 ments taken from similar and brilliant planes of different crystals, 

 that, owmg to some natural inequality of surface, the same precise 

 angle is rarely obtained, and hence those given in the succeeding 

 pages cannot be expected to be absolutely exact." The error, 

 however, rarely exceeds forty minutes, and is frequently not more 

 than one or two minutes ; and when the measurements of the 

 primary form have been obtained from cleavage planes, (which is 

 noted in the descriptions,) " they may be considered as approxi- 

 mating the truth much more nearly than when taken by means of 

 the natural planes." 



Where the regular solids, as the cube, regular octohedron, &c., 

 are the primary forms, our author has adopted the measurements 

 given by Haiiy, and denotes them by the letter H annexed ; but, 

 where the primary form is not one of the regular geometrical solids, 

 as the oblique, and doubly-oblique prisms, and the very numerous 

 class of rhombic prisms, he has determined their true measure- 

 ments by " subjecting the planes obtained by cleavage to the re- 



