In the classification proposed by the author in a previous paper meteorites are divided into five groups according to the ratio of iron to nickel in the nickeliferous iron they contain. In the case of the meteoric irons such a classification, as shown by Farrington's list of analyses, is a natural one, since the structure of the irons, as revealed by etching, varies with the amount of nickel they contain. Thus Group 1, in which the metal is poor ill nickel, with a ratio of Fe to Ni of 13 upwards, includes nickel-poor ataxites, hexahedrites, and Coarsest and coarse octahedrites ; Group 2, in which Fe : Ni =8-13, the rest of the octahedrites from medium to finest ; Group 8, in which Fe : Ni = 5-8, nickel-rich ataxites; Group 4, in which Fe: Ni = 2-5, ataxites still richer in nickel; while Group 5 may include the doubtful Oktibbeha meteorite in which the ratio of iron to nickel is even less than one.

The substance was met with in the course of work on steel-furnace slags, a general account of which has been communicated to the Iron and Steel Institute. These slags were for the most part melts consisting principally of the three oxides FeO, MnO, and SiO2, and yielded the silicates fayalite and rhodonite. The small amount of calcium silicate usually present is held in isomorphous solid solution ill these silicates, but when the amount of lime exceeds about eight per cent. a new silicate appears, which does not seem to belong to any of the known mineral groups.

This form of goniometer was designed in order to provide for the use of students of elementary crystallography, at as moderate a cost as possible, an efficient instrument for the measurement of the interracial angles of crystals. It is of the customary type, in which the direction of reference is given by the reflection of a distant object in a small plane mirror without optical aid, and the axis of the grHduated circle is horizontal. The mirror measures about 45 by 25 ram. and its frame is rigidly attached to the ball of a ball and socket joint, so that it is adjustable in any direction; the socket is split, and a screw linking the halves enables it to grip the ball firmly.

In August 1904 I discovered in the Lengeubach quarry a specimen on which were three small, round-shaped crystals, about the size of small shot, grown upon a prismatic crystal of sartorite. I measured them, and owing to their small size experienced some difficulty in determining the position of some of the faces with the goniometer at my disposal, which was of the ordinary single-circle pattern.

Considered from the crystallographical point of view the outstanding feature of the gnomonic projection is the fact that all great circles on the sphere, and consequently all zones on the crystal, appear on the diagram as straight lines. In order to secure its full advantages, it is necessary to take the plane of the paper at right angles to the axis of a zone, and preferably one of the principal zones of symmetry on the crystal.

In the Geological Magazine for 1918 there appeared three papers discussing the origin of tile system of cracks ill septarian nodules. Dr. A. Morley Davies opened the discussion in a paper where, after a summary of earlier literature, he again put forward an older idea. Without adducing evidence, he considered :

1. (a) that interposition of particles between those of the original matrix caused expansion which set up tensile strain in the growing body ;

2. (b) that the cracking of the nodule was due to the relief of those tensile stresses.

The following is a simple general proof that, on the hypothesis that crystals have a homogeneous cellular structure, the only possible axes of symmetry are those with cyclic numbers 2, 3, 4, or 6.

According to this hypothesis, every crystal is entirely made up of minute space-partitions or ceils, which (apart from those of purely surface layers) have all the same form, size, internal configuration, orientation, and external relations. Accordingly, every point in one cell has an equivalent point exactly corresponding to it in every other cell.

The material examined forms part of the late Dr. Henry S. Peele's collection, which shortly before his death he decided to present to the British Museum. Dr. Poole was at one time Government Inspector of Mines in Nova Scotia, and his collection is rich in Nova Section ores and minerals from many well-defined localities.

To mineralogists probably few spots in the world surpass in interest the little quarry hollowed out of the dolomite, just where the latter is cut by the Lengenbaeh, and about half a kilometre above the point where this stream ends its hurried course by joining the Binna at Imfeld, a hamlet lying in the main valley three kilometres above the village of Binn.

In the following notes I have endeavoured to give in a convenient and simple form a complete account of the relations connecting different twinning-laws or operations, which, when applied to the same crystal-structure, produce indistinguishable results.

A rigorous analysis of the possible relations between the compound structures of twin-crystals, based on the principles of the equality of the structural distances in the plane of composition, will be found in the author's papers, ' The geometry of twin-crystals' (Proc. Roy. Soc. Edinburgh, 1912-1918, vol. xxxii, pp. 416-432, 433-457) and 'Die Geometrie der Zwillingskrystalle' (Zeits. Kryst. Min., 1918, vol. lii, pp. 827-371). See also the paper 'Twin-planes and cross-planes' in this Magazine (1910, vol. xv, pp. 390-397).

Suppose a crystal has two good cleavages, and that a thin section is cut in a random direction. Cleavage-cracks are seen in the section which consist of two sets of parallel lines inclined at an acute angle ϕ, which may have any value whatever, depending on the direction in which the section is cut. The question suggests itself: what is the chance that ϕ may lie between assigned limits? This question proves to be incapable of a direct mathematical solution, but it can be solved graphically. This is worked out here for two typical cases, and the general method of performing the graphical solution in every case is indicated.

In the Rose-Tschermak-Brezina classification of meteorites the intermediate class of stony-irons, in which iron and stony matter occur in approximately equal amounts and to which Maskelyne gave the name siderolites, is divided into (1) siderolites proper, including the groups of the pallasites and the siderophyres, and (2) lithosiderites, comprising the groups of tile mesosiderites, glahamites, and ]odranites. Mesosiderites are defined as consisting of iron and crystalline olivine and bronzite, and grahamites of the same constituents with the addition of plagioclasefelspar.

The Amana meteorite, known also as Iowa County, Homestead, and West Liberty, is fully described in O. C. Farrington's 'Catalogue of the Meteorites of North America'. After the appearance of a brilliant meteor which on the night of February 12, 1875, passed with loud detonations from S.W. to N.E. over northern Missouri and southern Iowa, about 100 meteoric stones, weighing altogether some 500 lb., were found scattered over an elliptic area of about eighteen square miles from Amana in Amahs township to Boltonville in Iowa township, both in Iowa Co., Iowa.

The pale celestite-blue calcite which occurs on the eastern half of Lot 13, Range V, of North Burgess township, Lanark county, Ontario, contains very remarkable lattice-like inclusions of a hydrous magnesium silicate.

The calcite forms crystalline masses belonging to the Grenville series, which has a very extensive areal development in certain parts of Ontario. The texture varies from medium to very coarse, in which the individual cleavage-rhombs of calcite measure one inch or more in diameter. Repeated twinning on e (0112) or (101) is almost invariably exhibited, and is especially pronounced in the more coarsely crystallized varietiea Both the twinning and the crystalline character are results of the very intense metamorphism to which this Archaean limestone has been subjected.