374 Review of Dana's Mineralogy. 



and the vibration that takes place. Mr. Ames states instances in which 

 a large bar of iron, used as an axle through a heavy wheel of cast iron, 

 broke square off in the middle, after use for a few months ; and in one 

 instance, there were two other fractures on either side of the centre. 

 In these instances, the bar was rendered coarsely crystalline, and was 

 wholly unlike the original iron. The accident which took place in 1842, 

 on the Versailles railroad, was owing to the breaking of an axle, which 

 was rendered brittle by the same cause." 



The chapter on " blowpipe characters," contains in a tabular 

 form the most important reactions of the principle oxides and 

 earths with borax, salt of phosphorus and soda, being reduced 

 from the works of Berzelius, Plattner, and others. 



The much vexed question of classification, occupies the fourth 

 part of the second section of the volume, and we extract the fol- 

 lowing judicious remarks on the subject, (p.128.) 



" The arrangement of objects according to any assumed system, is 

 styled a classification. By using different classes of characters to mark 

 the grand divisions, various modes of arrangement may be made out. 

 Of these there is one natural system ; the rest are artificial classifica- 

 tions. 



" Artificial classifications may sometimes be used to advantage for the 

 convenience of comparison in identifying species ; but farther than this, 

 they only lead to error, by suggesting false affinities and unnatural as- 

 sociations of species. An arrangement of this kind is adopted in this 

 treatise, founded on the crystalline forms. Excepting the purpose for 

 which it is instituted — the determination of the names of minerals — it 

 subserves no important end to the mineralogist; on the contrary, it 

 brings together species the most unlike, and separates those most closely 

 allied. 



" The natural system is a transcript of nature, and consists of those 

 family groupings into which the species naturally fall. In making out 

 such a classification, instead of conforming the whole to certain assum- 

 ed principles, the various affinities of the species are first ascertained, 

 by studying out all their peculiarities and resemblances, and from these 

 the principles of the system are deduced. There should be no forced 

 unions to suit preconceived ideas, but only such associations as nature 

 herself suggests. 



" Unlike the other branches of natural science, mineralogy admits also 

 of a chemical classification, or one founded on the chemical constitution 

 of the species ; and as minerals proceed from chemical instead of vital 

 action, there is some reason for the adoption of chemical characters in- 

 to the natural system. When the chemical relations of the elements are 



