direction was reversed according as lead or bismuth was present, viz., downwards to the left in the first case, and to the right in the second.

II. "On the Photographic Arc Spectrum of Iron Meteorites." By Professor J. NORMAN LOCKYER, C.B., F.R.S. Received December 22, 1893.

(Abstract.)

This communication consists of a discussion of the photographic arc spectra of the Nejed and Obernkirchen meteorites, the wavelengths of the lines being based upon those determined by means of the photographic arc spectrum of electrolytic iron, a paper concerning which has been recently communicated to the Society. The portion of the spectrum covered by the photographs extends from K to D.

The wave-lengths, intensities, and origins of the lines in the spectra are given in tabular form.

The following general conclusions have been arrived at:

1. The spectra of the two meteorites closely agree, both as regards the number and intensities of the lines, the slight difference in number being probably due to a difference of exposure.

2. The meteoritic spectra and the solar spectrum show considerable similarity. The iron lines in each have about the same relative intensity, thus indicating that the temperature of the iron vapour in the sun which produces the majority of the iron lines is about the same as that of the electric arc.

3. The results of the inquiry into the origin of the lines, in addition to those of iron, may be thus summarised:

Substances certainly present.

Manganese.

Cobalt.

Nickel.

Chromium.

Titanium.

Copper.

Barium.

Calcium.

:

Substances probably present.
Strontium.

Lead.

Lithium.

Cerium.

Molybdenum.

Vanadium.

Didymium.

Uranium.

4. Of the few faint lines in the tables, for which no origins have been found from the Kensington maps of metallic arc spectra, the majority are apparently coincident with lines mapped by Messrs. Kayser and Runge in the iron spectrum. These do not appear in the Kensington photographs, probably on account of insufficient exposure.

5. By noting the difference in intensity of identical lines in the two spectra, a rough approximation can be made to the relative quantities of the different substances present in the meteorites. Thus it is found that the chief chemical difference between the two meteorites is that there is a preponderance of calcium in the Nejed, and of nickel, barium, and strontium in the Obernkirchen, meteorite.

III. "On the Straining of the Earth resulting from Secular Cooling." By CHARLES DAVISON, M.A., Mathematical Master at King Edward's High School, Birmingham. Communicated by Professor POYNTING, F.R.S. Received January 10, 1894.

(Abstract.)

If the coefficient of dilatation (e) and the conductivity (k) are constant at every point within the earth, and if the temperature (V) was initially the same throughout, the depth of the surface of zerostrain after 100 million years is 2:17 miles, the total volume of the crust folded and crushed above that surface is about 184,500 cubic miles, and the mean thickness of the crushed rock spread over the whole surface of the earth is 4.95 ft. (taking e=0'0000057, к = 400, V = 7000° F.). The smallness of these figures has been claimed by some geologists as a new and strong argument against the contraction theory of mountain evolution.

In the present paper the problem is reconsidered on the supposition that the coefficient of dilatation is not constant, but increases with the temperature, the change in the former varying as the corresponding change in the latter. It follows, from experiments made by Fizeau, that, for a rise of one degree in temperature, the coefficient of dilatation increases on an average by about 1/888. Adopting this value, and taking the other constants as above, it is found that, after 100 million years, the depth of the surface of zero strain is 7.79 miles, the total volume of crust-folding about 6,145,000 cubic miles, and the mean thickness of the layer formed by spreading it over the whole earth 164.7 ft.

If the conductivity increases with the temperature, or if the material which composes the earth's interior be such that the conductivity and coefficient of dilatation are greater in it than in the surface rocks, or if initially the temperature increased with the depth, the figures given in the preceding paragraph must be still further increased. It follows, therefore, that calculations as to the alleged insufficiency of the contraction theory to produce mountain ranges are at present inadmissible.

IV. "Chemical Analysis of the Meteoric Stone found at Makariwa, near Invercargill, New Zealand, in the year 1886." By L. FLETCHER, M.A., F.R.S., Keeper of Minerals in the British Museum. Received December 13, 1893.

(Abstract.)

As the preliminary investigation of the Makariwa stone had already indicated to Professor Ulrich* the presence of mineral constituents having the physical characters of olivine, enstatite, glass, a substance resembling glass, nickel-iron, troilite, magnetite, hydrous oxide of iron, and possibly chromite, the quantitative chemical analysis presented difficulties, among which may be specially mentioned the fact that one chemical element (iron) enters into the composition of each of the above fine-grained and closely intermingled constituents. This chemical examination I was invited to undertake by Professor J. W. Judd, F.R.S., through whom the two fragments of the stone sent to this country by Professor Ulrich have been generously presented to the British Museum.

The composite method adopted for the analysis was as follows:

It was found advisable to make a preliminary separation of the powder by means of a magnet into attracted and unattracted material. Treatment of a portion of the attracted material with a solution of mercuric ammonium chloride, as recommended by Dr. Friedheim,† revealed the fact that the proportion of rust was too large to be negligible, and indicated the necessity of a preliminary reduction of the rust of the unattracted material, and of a subsequent extraction of the reduced metal with mercuric solution; this prevents the unattracted oxide of iron due to rusting of the alloy from passing into the hydrochloric acid extract with the oxide of iron of the silicate decomposed by the acid. Accordingly, after the sulphur and phosphorus of the unattracted material had been determined, mercuric solution was employed both before and after heating the material to low redness in a current of hydrogen (as recommended by von Baumhauer), the residual unrusted and rusted nickel-iron being thus separately removed. Analysis of the postreduction mercuric extract showed that there had been a small, but appreciable, action on the silicate portion of the meteorite during the reduction. Further, it became obvious that the troilite was largely affected by the heating in hydrogen, and in such a way that the greater part of the iron of that constituent, and practically all the iron and nickel of the schreibersite, had gone into the mercuric

*Roy. Soc. Proc.,' 1893, vol. 53, p. 54.

'Sitz. Ak. Berlin,' 1888, p. 345.