132 MR HAIDINGER on the Natural-Historical 



only serve farther to increase the embarrassment, for it is not un- 

 der such circumstances that a complete h'st of quotations, or a 

 thorough knowledge of the authors consulted, can be of any use. 

 We must, therefore, recur to Nature herself, and conform our opi- 

 nions to her dictates. 



The principal properties in which Diallage, in its most dis- 

 tinct varieties, is said to differ from other minerals, more or less 

 resembling it, consist in the grass-green colour, and in the facili- 

 ty of being reduced by fracture into thin laminae, of a bright 

 pearly lustre. All the works hitherto published on the subject, 

 state these to be produced by cleavage, and yet they are not 

 owing to cleavage, but merely to composition. 



It has been reserved to the most recent period, to ascertain 

 the real and important difference between faces of cleavage and 

 faces of composition, which, indeed, are sometimes hardly distin- 

 guishable, if the latter possess a regular disposition in the inte- 

 rior of massive or crystallised varieties that seem to consist of a 

 single individual. As far as I know, Professor MOHS is the first 

 author who succeeded in giving an accurate definition of the phe- 

 nomenon of cleavage in crystallised bodies. He confined it to the 

 property of a simple mineral, of allowing its particles to be sepa- 

 rated in more or less regular faces, in various yet determined di- 

 rections. Only this property, and not the faces themselves, ex- 

 ist before the cleavage had been rendered visible by the assist- 

 ance of mechanical force *. It is quite another tiling with those 

 faces of composition, which preserve constant directions within re- 

 gularly formed crystals of certain minerals. They really exist be- 

 fore the mechanical division of the particles has been effected. 

 This can be proved by many examples, and, among others, with 

 a peculiar degree of evidence, by one of the most common in na- 



* MOHS' Grund-Riss. Th. i. p. 264., &c. 



