of Mineral Species. 105 



is no similarity between the two in the mode of combination of 

 their ingredients, analcime being considered as a compound of 

 bisilicates of soda and alumina with water, while prehnite is con- 

 sidered as a compound of simple silicates of lime and alumina, 

 with a hydrate of silica. 



On another occasion *, I have described a very curious in- 

 stance of pyramidal forms, agreeing as near as possible with those 

 of the pyramidal scheelium-baryte, which consisted in their in- 

 terior of multitudes of columnar crystals of the prismatic scheel- 

 ium-ore. They were found at Wheal Maudlin in Cornwall, and 

 are partly implanted on quartz, arsenical pyrites, chlorite, &c. and 

 partly imbedded in cleavable blende. The chemical composition 

 of the two species is almost identically the same, at least not 

 more different than in the varieties of pyroxene, or other similar 

 substances. The chemical formula of the first is Ca W 2 ; that 

 of the second Mn W 2 + 3 Fe W 2 , different only in the isomor- 

 phous bases of calcium in the one, and manganese and iron 

 in the other, one atom of the protoxide of each of them be- 

 ing united with two atoms of tungstic acid. This curious re- 

 semblance of the chemical mixture was then pointed out to 

 me by Professor MITSCHERLICH, who supposed, that, from the 

 isomorphism of the bases, the varieties observed might be ge- 

 nuine crystals, of the same ingredients as wolfram, but with the 

 form of the scheelium-baryte : this was disproved, however, by 

 the observation of the mechanical composition of the masses. Of 

 itself, the hypothesis is plausible enough that such was origi- 

 nally the case, and that the cohesion among the particles was 

 so slight, as to be afterwards overpowered by the greater crys- 

 talline attraction of the same particles in hemi-prismatic crystals, 

 subsequently formed, and as they now appear ; in a manner ana- 

 logous to the decomposition of the common hydrous sulphates 



* Edinburgh Journal of Science, vol. i. p. 380. 

 VOL. XI. PART I. O 



