and a portion of the glacier where they did not exist was presented to us simultaneously with the bands upon another portion. Their proximate origin and final completion were thus before us. The explanation offered in the paper is, that the dirt scattered by avalanches and winds over the upper regions of a glacier is redistributed by the passage of the glacier down a cascade where the ice is shattered, and the dirt broken up into detached patches. On reaching the bottom, where the force becomes one of compression, the patches of dirt are squeezed longitudinally and drawn out laterally, being thus converted into stripes of discoloration, which, owing to the speedier motion of the centre of the glacier, are convex towards its lower extremity. On consulting the map of Prof. Forbes, it will be seen that the dirt-bands commence at the base of the ice-fall of the Talèfre, while none exist above the fall. Those shown on the Glacier du Géant, we are led to infer, commence at the base of the Cascade of la Noire, which, however, is not sketched on the map. The theory of Prof. Forbes is, that a glacier, throughout its entire length, is composed of alternate segments of hard and porous ice; that the dirt is washed from the former, but finds "a lodgement" in the latter, thus giving rise to the phenomenon of the bands. We are unacquainted with the experiments on which this theory is founded, and have only to state, that the bands which we have observed seem accounted for by the simple explanation given above. In the paper, experiments are described illustrative of this view, and in which the bands are accurately exhibited on a small scale.

In the course of the inquiry much assistance was derived from the use of troughs of various shapes and sizes, through which mud, formed by the mixture of finely-sifted pipeclay and water, was permitted to flow. Coloured circles being stamped upon the mud, from the distortion of these the character of the forces, whether compression or tension, acting upon the mass was inferred. It is needless to say, that this mode of experiment does not pledge the authors to the view that ice is similarly constituted to mud. Where the mud stretches, the ice breaks; and the experiments were mainly instituted for the purpose of examining the relation of the forces acting upon a glacier to its directions of fracture.

It may, however, be urged, that, after all, the foregoing experiments do not prove the viscous theory to be wrong. The mere fact of

bending a prism of ice by fracture and regelation, does not prove that ice is non-viscous. This is perfectly true; nor is it conceived that the onus rests on us to prove the negative here. All that is claimed for the foregoing experiments is the referring of certain observed phenomena to a true cause, instead of to an imaginary one. An illustration may perhaps serve to place this question in its true light. By Newton's calculation, the velocity of sound through air was one-sixth less than what observation made it; and to account for this discrepancy he supposed that the sound passed instantaneously through the particles of air, time being required only to accomplish the passage from particle to particle. He supposed the diameter of each air-particle to be 13ths of the distance between two particles, and nobody ever proved him wrong. Still, when Laplace assigned a vera causa for the discrepancy, the hypothesis of Newton, and other ingenious suppositions, were discarded. The proof indeed in such cases consists in the substitution of a fact for a conjecture; and whether this has been done in the case now before us, the intelligent reader must himself decide.

916

January 22, 1857.

Dr. W. A. MILLER, V.P., in the Chair.

The following communications were read :

I. "On some of the Products of the Destructive Distillation of Boghead Coal."-Part I. By C. GREVILLE WILLIAMS, Esq., Assistant to Dr. ANDERSON, Professor of Chemistry in the University of Glasgow. Communicated by Dr. SHARPEY, Sec. R.S. Received November 25, 1856.

(Abstract.)

The paper, of which the following is a brief abstract, constitutes the first part of the author's examination of the hydrocarbons con

tained in boghead naphtha. In it he gives the results of his experiments on that portion of the fluid which resists the action of monohydrated nitric and sulphuric acids. He had previously stated the fact of his having obtained a substance possessing the composition and vapour-density of butyl*, and had expressed a belief that he should succeed in isolating not only that radical, but also propyl, amyl, and caproyl. The composition of the radicals varies so little, that to determine the boiling-points it was necessary to take the density of the vapour of all those fractions which distilled anywhere near their known boiling-points; and in each case he regarded that fraction which gave the nearest result as representing the boilingpoint of the radical as obtained from the source mentioned.

Propyl presented itself under the form of a colourless, very mobile fluid, having a pleasant odour, and boiling at 68° C. At 18° it had the very low density of 0.6745. Combustion gave results agreeing closely with the formula

C H44 vols.,

confirmed by a determination of the density of its vapour by GayLussac's method, which gave 2.96, theory requiring 2.97. Propyl had not been previously obtained.

Butyl from the Torbane-hill mineral distils at 119°, and has a density of 0.6945 at 18°; its analysis coincided with the formula

C16 H18.

The vapour-density was found to be 3.88, theory requiring 3.94. Amyl distilled at 159°, and had a density of 0.7365 at 18°. On analysis, numbers were obtained agreeing perfectly with the formula

C20 H22.

The vapour-density was found to be 4.93; theory requires 4.91. Caproyl boiled at 202°; its density at 18° was 0.7568. The results of analysis accorded with the expression

C24 H26,

which indicates a vapour-density of 5.87, while experiment yielded

5.83.

The experiments detailed in the paper appear to demonstrate the

* Proceedings of the Royal Society, vol. viii. p. 119.

VOL. VIII.

2 D

radical nature of the hydrocarbons, and to negative the assumption of their being homologues of marsh-gas.

The paper concludes with a description of a method, by which, where numerous vapour-density determinations are to be made, the necessity is avoided of refilling the balloon with water or mercury in order to determine the residual air.

II. "On the Optical Characters of certain Alkaloids associated with Quinine, and of the Sulphates of their Iodo-compounds." By WILLIAM BIRD HERAPATH, M.D., in a Letter to Professor STOKES, Sec. R.S. Communicated by Professor STOKES. Received January 8, 1857.

You will probably recollect that I sent you some time since a small portion of an alkaloid, which at that time was called quinidin in Germany, but it has since been distinguished from it and named cinchonidin. You then examined it for epipolism or fluorescence, and you pronounced the opinion that it possessed this property only in a minor degree, and you imagined that this arose from the presence of a small per-centage of a-quinine.

I have since obtained, through the kindness of Mr. J. E. Howard, specimens of the perfectly pure alkaloids quinidin and cinchonidin, and find that quinidin, which I can now identify as the B-quinine of Von Heijningen, possesses the phenomenon of fluorescence or epipolic dispersion as powerfully as a-quinine; whilst cinchonidin, if perfectly pure, is devoid of it altogether; and recent experiments have shown me that a small per-centage of quinidin was the cause of the epipolic dispersion found by you in the specimen of cinchonidin sent by me.

It

may be as well to state that the cinchonidin tested by water of chlorine and ammonia gave no evidence of green tint, which it would have done if onlydth part of either a-quinine or quinidin had been present, according to some recent experiments of my own.

I have also found that 1 gr. of pure quinine or quinidin in 35,000 of water will give an evident “ epipolic" appearance; whilst when diluted with 70,000 grs. of water we have still very evident appear

ances of "fluorescence" upon the perpendicular wall of the glass vessel exposed to the incident light; whilst a bluish milkiness of "internal dispersion" may be seen when I gr. of either alkaloid is diluted with 700,000 grs. or 10 gallons of distilled water, well acidulated in all these cases with sulphuric acid.

Some other interesting results have followed from these investigations. When quinidin is dissolved in an excess of diluted sulphuric acid, and the solution mixed with about twice its bulk of spirit, and warmed to 130° F., and tincture of iodine then added in sufficient quantity, and subsequently set in repose, beautiful red acicular crystals are deposited; these, upon recrystallization from rectified spirit, acquire an increased size, become beautiful quadrilateral prisms, leaving a deep garnet-red by transmitted light, and possessing a clear bluish-purple reflected tint; they are optically doubly absorbent in a slight degree, and transmit a brownish-orange body-colour when polarized perpendicular to axis. The primary form appears to be a rhombic prism, and as far as my present analyses go, appears to possess centesimally the following composition :

The excess arises from the hydrogen twice calculated in water of crystallization.

These numbers agree very closely with the formula

(C35 H19 N2O4+12) SO3 HO+5HO,

and prove it to be the sulphate of iodo-quinidin, very analogous to the optical quinine compound, yet differing materially in its optical properties.

There is another alkaloid frequently associated with quinidin, which also crystallizes from spirit in the prismatic form like quinidin and cinchonidin, but is another example of epipolism or fluorescence.