OLDEE PEEIDOTITES OF SCOTLAND. 385 



of the negative crystals, and occupy irregular spaces within them. In 

 some cases, like that of the olivines, the distribution of these 

 products of decomposition within the negative crystals is partially 

 determined by crystalline forces, and curious dendritic forms, of 

 microscopic dimensions, are the result. That this is really the ex- 

 planation of the dendritic forms seen in olivine, is shown by the 

 fact that the ends of the star-like masses are bounded by straight lines, 

 the sides of the negative crystals (see PL XII. figs, oa, ob, 5c). 



It is a very noteworthy circumstance that these negative crystals 

 are formed along certain definite planes within the original crystal. 

 The studies of the so-called " Aetzfiguren " by Exner, Baumhauer, 

 and others have shown that the surfaces of crystals, and of sections 

 of crystals, are very unequally afii'ected when submitted to the action 

 of appropriate solvents. The peculiar disposition of molecules in a 

 crystal which causes it to yield most readily along certain planes to 

 the mechanical forces applied in scratching and fracture, and which 

 permits the waves of light and heat to traverse it at diff'erent rates 

 in diff'erent directions, is equally manifested when the crystal is 

 operated upon by solvent agents. We can understand how, under 

 these circumstances, solution is set up along certain planes within 

 the crystal, innumerable negative crystals being formed, while the 

 products of decomposition are deposited within them. The action 

 may be compared to what takes place when a beam from the sun 

 or an electric lamp is passed through a block of ice. The beautiful 

 ice-stars partially filled with water appear to be exactly analogous 

 to the negative crystals formed by solvent action in augite, for 

 example, and occupied by the hydrated oxides which result from 

 its decomposition. Tschermak and others have pointed out that the 

 lines along which solution takes place most easily are not necessarily 

 the edges of cleavage-planes ; and it has also been shown that the 

 twinning of crystals modifies the " Aetzfiguren," as it certainly does 

 the position of the plane of most easy chemical action, as revealed 

 by the phenomena of Schillerization (see Ante, p. 379). 



There is one point in connexion with this subject which appears 

 to me to be specially worthy of attention, though, as far as I am 

 aware, its importance has hitherto been very generally overlooked. 

 The " Schillerization " of dark-coloured ferriferous minerals, like 

 proto-hypersthene and augite, results in the discharge of colour from 

 the substance of the crystal ; and instead of a dark green or brown 

 substance, we get a nearly colourless one, within which the innumer- 

 able deeply coloured enclosures are distributed. IN'ow, inasmuch as the 

 chemical composition of the whole crystal is scarcely, if at all, altered 

 by this molecular change, it is fair to conclude that the iron and 

 other oxides which gave the colour to the crystal have been dissolved 

 out and deposited in the substance of the enclosures. 



That this is really the case, we have, I think, a singularly beautiful 

 proof. The researches of many mineralogists have demonstrated that 

 in the pyroxenes, the micas, and many other groups of minerals, a 

 remarkable relation can be discovered between the proportions of 

 certain ingredients, notably the iron, in diff'erent varieties, and 



