depending on their Internal Changes. 87 



successive alterations in the chemical composition of minerals, 

 by observing their jiatural repositories, veins and beds, and 

 mountain masses, exposed to the action of the atmosphere, 

 and of water, and to the mutual reaction of the mineral spe- 

 cies of which they are constituted. 



One of these examples, where the cause of a change in ap- 

 pearance is not so palpable, is the well-known one of the sub- 

 • stance usually named the Gray Andalusite. Its specific gra- 

 vity alone being above 3.5, while that of the real andalusite 

 never exceeds 3.2, would be sufficient to prove them to belong 

 to different species. But Prof. Mohs has found the gray crys* 

 tals actually to consist of a great number of small individuals 

 of disthene, with an easy cleavage, whenever they are large 

 enough to be distinguished from others, and Ijing in different 

 directions throughout the mass. Both minerals are found in 

 nodules of quartz engaged in mica-slate. From the analysis 

 by Arfvedson, it appears that disthene is a compound of one 

 atom of silica and two of alumina, or AP Si. Andalusite con- 

 tains about 83 per cent, of the same mixture, the rest being a tri- 

 silicate of potassa. — Beudanf s Mineralogy^ p; 333 and 363. 

 The loss of this ingredient sufficiently accounts for the chemi- 

 cal difference between the two bodies ; but we are at a loss to 

 conjecture in what manner such a change may have taken place. 



Mr Allan has in his cabinet several specimens from the 

 trap district near Dumbarton, exhibiting the shape of analcime, 

 but entirely composed of aggregated crystals of prelinite. Mr 

 William Gibson Thomson is likewise in the possession of se- 

 veral exceedingly distinct and instructive specimens of the same 

 description. There is one, among the former, where preh- 

 nite, aggregated in globular shapes, is implanted on icositetra- 

 hedral masses, once of analcime, but now likewise converted 

 into prehnite. The implanted varieties are green and translu- 

 cent ; I found their specific gravity e([ual to 2.885 : the por- 

 tions within the faces of the icositetrahedrons are white and 

 opaque, and give 2.842, both of them rather lower than the 

 usual results obtained, which are a little above 2.9, at least in 

 simple crystals. But the arrangement of the divergent indi- 

 viduals in the reniform shapes is highly remarkable, and throws 



