Vol. 59.] MAGNETITE-MINES NEAK COGNE. 61 



have often been more or less fractured ; the rough and granular 

 condition of their outer edges suggests aggregation, and they include 

 or are pierced by well-formed prismatic flakes of serpentine ; the 

 latter mineral also frequently exhibits a microfoliation. This is 

 equally true of many of the Alpine serpentines in my collection, 

 some of which are actually slaty. But that structure occasionally 

 exhibits marked flexures, so that either the movements have lasted 

 long enough to allow of changes in direction, or there have been 

 two sets. Again, part of the magnetite occurs in minute granules ; 

 these may be powdered grains, but, as larger ones also occur, we 

 must not forget the possibility that the smaller may represent the 

 iron-oxide ejected from the olivine during serpentinization. The 

 frequent ophitic relation of the two minerals suggests that either 

 the magnetite was an incoherent powder when the serpentine-flakes 

 were forming, or that there was, after the crushing, a more or less 

 complete reconstitution of the serpentine, which had previously 

 exhibited, as replacing olivine, the ordinary ' network' structure. 

 Beyond this alternative we cannot, I think, proceed, until we obtain 

 further evidence. 



Next, how can we account for the presence of these great masses 

 of magnetite in the serpentine ? Has the latter been converted 

 into the former by the aid of steam or hot water (pneumatolysis, as 

 it is sometimes called) as a vein of tourmaline-rock is produced in 

 a granite ? One of these serpentines generally contains from about 

 40 to 42 per cent, of silica and 38 to 40 of magnesia, and not 

 more than 10 per cent, of iron-oxides — that is to say (neglecting 

 the water and small quantities of such constituents as alumina and 

 the oxides of titanium, chromium, and nickel, and lime) more than 

 four -fifths of an enormous mass of rock must have been replaced by 

 peroxide of iron. 



"We must therefore fall back upon differentiation. But how has 

 this acted ? Sometimes, as shown by Prof. Vogt, 1 the more basic con- 

 stituents in a molten mass work towards the exterior or cooler part. 

 There is, however, in the section now visible nothing favourable, 

 and much opposed, to the idea that the magnetite, once disseminated 

 in a molten intrusive mass, has congregated in certain parts of it. 

 So, if this mineral has been separated by differentiation from the 

 original peridotitic magma, that must have been done before the latter 

 reached its present position, or deeper down in the earth's crust. 

 Here the magnetite must have become sufficiently solid to be 

 brought up like huge included fragments in the other material. As 

 the former would still be near its melting temperature 2 some of its 

 outer part might again become liquid and mix with the adjacent 

 magma, both probably being rather viscous : thus producing the 

 intermediate rock. Something of this kind must have happened at 

 Ovifak, where, as is well known, metallic iron not only has been 



1 Zeitschr. fur prakt. Geol. vol. i (1893) pp. 4, 125, & 257. 



2 The melting-point of olivine is about 1370° 0. ; see E. Cusaek, Proc. Roy. 

 Irish Acad. ser. 3, vol. iv (1897) p. 412. That of magnetite (teste Prof. Jolv) 

 varies between 1250° and 1450° C. 



