is then slid under the gallipot, the specimen being thus raised to the positions adopted by Prof. Szabó, without any of the difficulties that often arise from the jarring or stiffness of motion in more elaborate supports.

For

The dimensions above given are those adapted to a Bunsenburner of ordinary height and ordinary diameter of base. packing, the erection can be taken down, and the 5-millimetre plate and the smaller corks can be kept inside the gallipot till required.

Observation of Fusibility.-Szabó's scale may be thus summarised. It is unfortunate that the degrees are numbered in the reverse order to those of von Kobell, but this difference is easily learned.

The numbers of the scale here given refer to the result obtained, whatever part of the flame is used; thus, a mineral may have a fusibility of 1 in the point b and of 3 in the point c. The product must be examined with a lens.

0. Infusible.

1. Edges and corners alone rounded.

2. General form unaltered, but edges, corners, and faces fused. 3. Form altered, but not to a globule.

4. Fuses to a globule.

The time of heating is in each observation one minute, the specimen being tried first in the position a (base of flame), then moved to b (5 mm. above the base), then to c (5 mm. above the iron cone), notes being made of its appearance on withdrawal from each portion of the flame.

On these observations Szabó founds 8 degrees of fusibility among minerals, depending on the positions in which the above described reactions take place.

Determination of Sodium and Potassium.-Held in position b (first row, fig. 10), the assay imparts a certain degree of colouration to the flame. Five degrees are recognised, No. 5 being the most intense. The observer, with a drawing of these degrees before him, notes down the figure corresponding to the flame given by his assay, and picks up an indigo prism or cobalt glass 5 mm. thick, through which he views the flame with the object of detecting potassium. Three degrees of the characteristic violet-red flame may be distinguished by a good eye (lowest row, fig. 10). All this can be done in the one minute assigned to the observation; at its expiry the wire is withdrawn and the degree of fusion also noted.

The cone is put on and the same assay brought to c, the fusion-place. In one minute similar observations of sodium

(second row, fig. 10) and potassium (lowest row, fig. 10) are made, and the fusibility is again observed. The colourations

should be slightly stronger than in b, the degree 3, for instance, now representing an intenser flame than 3 in the previous observation.

The assay is now dipped in distilled water and then into powdered gypsum, which thus adheres to and surrounds it. On reheating in position c, the gypsum assists decomposition, the sodium and potassium being converted into sulphates. The observation should be made when the assay has been two minutes in the flame.

It is unnecessary to observe the sodium reaction except as a check; but the potassium flame is intensified and four degrees are distinguishable (fig. 11), No. 1 representing a quantity too minute for previous detection.

Determination of Calcium.-The lime in felspars and allied minerals can generally be inferred from the diminution of the soda and potash; but its flame may be seen as follows::

Put fragments of the mineral in a glass tube with cold concentrated hydrochloric acid sufficient to completely cover them. Close with wax and leave for 24 hours. Then open and plunge in a fairly thick platinum wire, coiled at the end. The drop thus brought out is held in position b, and the first colour observable is due almost entirely to the calcium. A directvision spectroscope shows the red, orange, and green lines distinctly. The flames due to sodium, potassium, and lithium follow in order, and the degrees observed during this operation often serve to distinguish minerals in which the actual percentage of soda, &c., is the same, a higher degree being shown by those more easily decomposed by acid.

A table based on those prepared by Szabó is given above. The degree of fusion is stated according to the modified scale adopted on p. 81. The observer draws up a blank form on the same lines, notes into it each observation numerically as soon as made, and compares the whole result with the series given in the table. A good plan is to take each result separately and write down all the felspars or allied minerals to which it might possibly correspond. A comparison of the brief lists thus formed enables one to pick out the name that occurs most frequently, or the two

between which the mineral must lie. These reactions, performed upon minute grains, and occupying altogether about ten minutes, have a great value to the geologist, however simple they may at first appear to the chemist, mindful of his refined but lengthy methods of analysis.

The determination of potassium in the flame is rendered easier by the following process, which allows of a complete decomposition of the silicate, and which is unaffected by the presence of a bright sodium flame. On a loop of platinum wire, 2 mm. in diameter, form a bead of sodium carbonate. Moisten this with water, and pick up about 2 cubic mm. of the mineral— i.e., about twice as much as is used in Szabó's process. Place the wire on the support in position c, 5 mm. above the cone. The specimen does not tend to fall off, but is soon attacked and dissolved. Leave for two minutes by the watch.

Then observe the flame through 5 mm. of blue glass. The bright sodium flame is cut off, except for a marginal column, which comes through blue. If potassium is present, it is revealed by a pink-violet inner fringe to this blue column. Judging from its extent and also its intensity near the assay, three grades can be established.

Grade 1 = about 4 per cent. of potash.

The ease of manipulation in this method, and the completeness of decomposition, seem to recommend it. It is also of service in examining the glassy or fine-grained groundmasses of igneous rocks. All true orthoclases fall in grade 3; but each observer should establish the grades for his own eye upon specimens already known.

СНАРТER IX.

EXAMINATION OF THE OPTICAL PROPERTIES OF MINERALS.

In a book written for the geologist rather than for the mineralogist, a detailed account of the optical properties of minerals, as observed in slices cut from them in known directions, would be out of place. The appearance of the common rock-forming minerals in *G. Cole, "Potassium in Silicates," Geol. Mag., 1898, p. 103.