The mineral here described was exhibited at the meeting of the Iron and Steel Institute held on May 6, 1920, in illustration of a paper by Mr. J. F. Wilson, who found that slags yielding these crystals had a composition approximating to the orthosilicate ratio. The slag was obtained from a metal-mixer ladle of about ten tons capacity, and consisted of a network of interlacing rhombie prisms, from which the mother-liquor had withdrawn, leaving material on which it has been possible to make fairly accurate measurements. Mr. Wilson very kindly arranged to place at the disposal of the present authors the specimen on which this account is based.

Rock-gypsum occurs as extensive deposits of considerable economic importance in the neighbourhood of Nottingham. The mode of occurrence and origin have been studied by several geologists, to whose work references were given in a recent communication to this Society. I Investigations have proceeded chiefly on field lines and microscopic examination has been somewhat neglected, except with respect to the contact relations of gypsum and anhydrite. Nevertheless, examination of thin slices from specimens collected in different parts of the district yields information of considerable interest and. it may be, of importance.

There are certain foreign minerals upon which the South African diamond has crystallized readily. The chief of these are garnet, ilmeuite, olivine, and iron-pyrites. In addition, zircon and mica are also probable inclusions. Cognate inclusions are graphite, bert of every kind, and diamond itself of unconformable orientation.

It is the author's good fortune to collaborate with Dr. J. W. Evans in describing the rock specimens which he collected on the expedition to Caupoliean in Bolivia during 1901-2. These specimens contain a very large proportion of shales, slates, and phyllites, and in their investigation certain petrographic methods devised for a detailed study of the slates of North Wales, Devon, and Skiddaw were followed. The present paper is intended primarily to describe these methods as applied particularly r the Bolivian rocks and to a Skiddaw slate facies which, for purposes of comparison and contrast, is of special interest. It is not suggested that between groups of rocks from places so widely separated as Bolivia and Skiddaw close correlation is possible; the products of the reconstitution processes which the two groups record differ considerably in detail; but as providing data for the study of the mechanism of these processes the two groups can be conveniently examined together.