The microscopical study of shales, slates, and phyllites is assisted by comminuting the rock specimen and examining the particles after they have been cleansed of chloritic matter and freed from attached carbon. The study of the trend of reconstitution processes is also assisted by analyses of fractional extracts of the groundnlass.

Hutchings ¹ has shown that (a) re-crystallization begins at a very early stage in the consolidation of clay material: secondary mica may be present in abundance in a shale which is soft enough to be cut with a knife; (b) this primitive mica is essentially sericitic but impure--in the sense that it contains magnesia, lime, and ferrous and ferric oxides, which are Mien to pure muscovite as represented by Tschermak's Kmolecule, and that as recenstitution proceeds, the sericite sheds its impurities; (c) the development of sericite is accompanied by that of a chlorite-like substance, which is at first obscure and indefinitely crystallized; but in slates and phyllitcs chlorite is sharply separated from the mica, and the latter is definitely muscovitic.

The further observations which this paper records indicate that the trend of reconstitution in shales, slates, and phyllites is towards the establishment of a metastable ternary system of white mica, chlorite (or chloritoid), and quartz : it is essentially a process of molecular differentiation— into (a) monad-oxides, type R_O, combined with alumina, silica, and water; (b) diad-oxides, type RO, combined with alumina, ferric oxide, silica, and water ; and (c) free silica ; accessory minerals such as rutile, ores, epidote, &c., being products which, iu large measure, are consequent upon this essential differentiation. In the early stages, differentiation is imperfect; for example, appreciable quantities of magnesia, lime, &c., are temporarily accommodated in the micaceous mineral, and alkalis are to some extent absorbed by the obscure chloritic matter; but as reconstitution proceeds the mica tends to lose its magnesia, lime, &c., which thus supply increment to the chloritic substance; similarly, the latter loses its alkalis and some alumina, which thus supply increment to the mica. The alkali-content of the rock-mass exercises an early and a preferential claim upon the available alumina; the identity of the mineral-species towards which reconsti. tution of the chloritic substance trends during any phase depends largely upon the ratios R₂O₃: Re and MgO: Fee. Both R₂O₃ and Re are in part dependent upon the reduction of ferric to ferrous oxide.

In the particular rocks discussed in this paper the lime content is of special interest. In the Skiddaw slate the finer-grade mica is relatively rich in lime; but as the mica is re-crystallized, it sheds an appreciable proportion of this lime. In the Calvario phyllite the mica has a negligible lime content, though abundant titanite is present as an alteration product of ihnenite and the mica is especially course and muscovitie about the margins of altered ilmenite crystals ; it is suggested that in this case also primitive mica on re-crystallization sheds its lime content and that this lime is fixed mainly in the titanite encrusting or partly replacing ilmenite.