In 1879 the late Dr. C. O. Trechmann described in the pages of this Magazine some orthorhombic crystals from Cornish tin furnaces, which from the analyses of the late Mr. J. H. Collins were stated to be metallic tin. Shortly afterwards Baron Foullon described similar crystals from tin furnaces in Bohemia, arriving at the same results, and thus apparently confirming the observations of the previous authors. At a later date, in 1898, Dr. W. P. Headden analysed crystals of similar appearance, also from Cornish tin furnaces, which he found to be stannous sulphide (SnS), hut he stated that these crystals were monoclinic.

The apparatus here described was designed in the first instance to provide a means for the rapid preparation and accurate adjustment of the liquid mixtures now commonly used for mineral determinatiom For this purpose refractometers of the Pulfrich type are inconvenient, not only on account of the risk of error due to the complicated centring of the prism or hemisphere and scales, but also because the transfer of a drop of liquid to the refracting surface renders the determination especially liable to errors arising from temperatm'e variation, from contamination of the drop, and from evaporation. Ful%her, in the preparation of liquid mixtures the frequent transference of small quantities to a refractometer is the source of much inconvenience.

Dr. Harker has familiarized us with the use of variation diagrams, in which the Pereentage amounts of a constituent are plotted against the amount of silica present. He applied these diagrams chiefly to study selected rock groups, but also gave certain generalized eurves fo illustrate the chemical properties of his 'Atlantic' and 'Pacific' facies. Such variation diagrams based on an adequate body of analytical data form a useful means of constructing diagrams for the comparison of rock analyses. Some means of generalizing analyses becomes a necessity as their number accumulates and the labour of comparing individual analyses increases.