2. The Bargate Beds of Surrey, and their Microscopic Contents.' By Frederic Chapman, Esq., F.R.M.S.

3. 'On Deposits from Snowdrifts, with Special Reference to the Origin of the Loess and the Preservation of Mammoth-remains.' By Charles Davison, Esq., M.A., F.G.S.

4. Additions to the Fauna of the Olenellus-zone of the Northwest Highlands.' By B. N. Peach, Esq., F.R.S., F.G.S.

5. Questions relating to the Formation of Coal-Seams, including a New Theory of them: suggested by Field and other Observations made during the past decade on both sides of the Atlantic.' By W. S. Gresley, Esq., F.G.S.

6. Observations regarding the Occurrence of Anthracite generally, with a New Theory as to its Origin. By W. S. Gresley, Esq., F.G.S.

After discussing Dr. J. J. Stevenson's theory of the origin of anthracite, the author describes the nature and mode of occurrence of the anthracites of Pennsylvania, and gives his reasons for concluding that the de-bituminization of coal was not produced by dynamic metamorphism during mountain-building, but rather by previously applied hydrothermal action. He further discusses the applicability of his theory to other cases of anthracite, including that of South Wales and Ireland.

6

7. The Igneous Rocks of the Neighbourhood of Builth.' By Henry Woods, Esq., M.A., F.G.S.

In south-west Radnorshire (just north of Builth) there is an area of Ordovician and associated igneous rocks, surrounded on all sides except the north-west by Silurian beds; this is shown on Sheets 56 S.W. and S.E. of the Geological Survey map, and was described by Murchison. In this paper the author gives a map of the southern half of this area, and a description of the igneous rocks-andesites, andesitic ash, rhyolites, diabase-porphyrite, and diabase. The diabase-porphyrite is intrusive in the andesite, and the diabase in the Llandeilo Shales. The andesitic ash rests on the andesite, and is overlain by the Llandeilo Shales. The author concludes that the andesites, andesitic ash, rhyolites, and diabase-porphyrite are of Lower Llandeilo age; and that the diabases are post-Llandeilo and pre-Llandovery.

8. 'On the Relations of some of the Older Fragmental Rocks in North-west Caernarvonshire.' By Prof. T. G. Bonney, D.Sc., LL.D., F.R.S., F.G.S., and Miss Catherine A. Raisin, B.Sc.

In a recent paper on 'The Felsites and Conglomerates between Bethesda and Llanllyfni, North Wales,' it is argued that, in the well-known sections on either side of Llyn Padarn, a great unconformity separates the rocks into two totally distinct groups.

The authors of the present communication discuss at the outset the great physical difficulties involved in this hypothesis; a subject

which, in their opinion, was passed over too lightly by the author

of that paper.

They further affirm, in the course of a description of the sections, which are most clear and afford the best evidence :

(1) That the strike in both the supposed rock-groups is generally similar.

(2) That the same is true of the dips.

(3) That very marked identity of lithological characters may be found in rocks on either side of the alleged unconformity, specimens occasionally being practically indistinguishable.

(4) That in no case, which has been examined, can any valid evidence be found in favour of the alleged unconformity, and that in the one which is supposed to be the most satisfactory proof of it the facts are wholly opposed to this notion.

LXIX. Intelligence and Miscellaneous Articles.

ON THE ENERGY OF THE AMPERIAN MOLECULE.

To the Editors of the Philosophical Magazine.
GENTLEMEN,

WITH reference to the paper by Mr. Fawcett and myself published in your last issue, Prof. Oliver Lodge has kindly sent me the following interesting considerations.

If temperature is due solely to translational and vibrational motions of molecules, it is possible that an appreciable time may be required before the sudden alteration of the rates of rotation of the molecules of a piece of iron can cause an alteration in the temperature of a neighbouring thermopile. It follows that, if this be so, an inseparable connexion may still exist between molecular spin and molecular magnetism, in spite of our negative result.

It is true that our experiments lasted on the average from two to three minutes, and in two cases for four minutes and a half, without any marked temperature effect appearing; but an examination of our results in the light of Dr. Lodge's suggestion does certainly show that in every case (except curve f, where the temperature was probably changing rather rapidly) there is a very slight upward tendency of the curve after the first sudden bend is over, which is in the direction of the effect sought; and this tendency is also most conspicuous in the earlier readings (p. 477), where the time-interval was greatest.

I confess that it had not occurred to me as possible that several minutes might elapse before an alteration in the rotations of a set of molecules could make itself felt in their translational velocities, and for molecules consisting of more than one atom this still seems to me very unlikely; but if the rotations dealt with are those of single atoms spinning about their own centres, between which it is perhaps not unreasonable to assume the absence of any marked tangential forces during collision [and I take this to be the case contemplated by Dr. Lodge], there seems to be no reason why Phil. Mag. S. 5. Vol. 38. No. 235. Dec. 1894. 2 R

the time of adjustment or distribution of energy among the degrees of freedom should not extend to long periods.

The difficulties connected with any considerable increase in the duration of the experiments will undoubtedly be great, as it will be necessary to maintain the temperature so very uniform; but we hope shortly to attack the problem again from this point of view. I am, Gentlemen,

University College, Bristol,
November 12, 1894.

Yours obediently,

A. P. CHATTOCK.

ON THE PROPAGATION OF ELECTROMAGNETIC WAVES IN ICE, AND ON THE DIELECTRIC POWER OF THIS SUBSTANCE. BY M. BLONDLOT.

In a previous note (Comptes Rendus, July 25, 1892) I enunciated the following proposition:-The length of the waves which an electromagnetic oscillation can emit is the same whatever be the insulating medium in which the experiment is made; in other words, the wave-length depends on the oscillator alone, just as in acoustics the wave-length of a pipe depends only on the length of the pipe.

The confirmatory experiments described in the Note cited referred to oil of turpentine and to castor-oil; the law holds perfectly for both these substances, and everything leads to the belief that this will be the same for other dielectrics.

There is, however, a doubt about ice, in consequence of the exceptional properties ascribed to it. The experiments of M. Bouty (Comptes Rendus, March 7, 1892) show in fact that ice has a dielectric power of 27, that is to say incomparably greater than that of all other substances. Suspecting that the law relative to the propagation of waves might not apply to a dielectric so different from the others, I resolved to subinit the question to experiment.

For these investigations I availed myself of the intense and prolonged frosts of the winter of 1892-93. M. M. Dufour has helped me in carrying them out, which the rigour of the cold rendered difficult and even painful. I thank him for his extreme kindness on this occasion.

The method which I adopted was the following, which, with slight modifications necessitated by the solid character of the dielectric, is the same as that I used in the case of turpentine and castor oil.

Electromagnetic waves were transmitted along two tinned copper wires 2.5 millim. in diameter, stretched horizontally and parallel to each other at a distance of 0.8 metre. A resonator of gilt copper is placed in a fixed position between the wires; the portion of the transmitting wires beyond the resonator is contained in a wooden trough 4 metres in length. The trough being filled with liquid, the position is sought at which a movable bridge must be placed joining the wires beyond the resonator to cause the spark to disappear; the distance from the bridge to the resonator is then

a quarter of the specific wave-length of the resonator; the position of the bridge is accurately noted.

That done, I surround the part of the resonator forming the condenser with a watertight bag of parchment-paper which I fill with distilled water, and then freeze this water; the layer of air is thus replaced by one of ice. Measuring the wave-length afresh, it is found to be considerably greater than in the first

141

experiment, having become of what it was.

100

The trough is then filled with water which is frozen, and then the position of the bridge for disappearance of the spark is again sought. For this purpose the ice at the distant end of the trough is broken and progressively removed. I ascertained that this position is exactly the same as in the first case, when the dielectric was air.

The experiment four times repeated, varying each time the capacity of the condenser, always gave the same result. The proposition relative to the wave-length is therefore true for ice as well as for other dielectrics. Hence, as shown in my previous Note, Maxwell's relation that the dielectric power is also equal to the square of the refractive index also holds for electromagnetic waves in the case given.

The preceding results, partly unforeseen, led me to determine the dielectric constant of ice, using electromagnetic undulations. The experiment cited above gave all the data necessary for this determination.

For if A and A, are the wave-lengths corresponding to a given resonator, working respectively in air and in a substance of dielectric power K, we have

The experiment repeated a dozen times always gave the same result. I consider the relative error does not exceed, for the plate of ice was almost entirely free from air-bubbles. According to this, ice does not present exceptional dielectric properties.

It remains to be explained how MM. Bouty and A. Perot obtained values of a totally different order for the dielectric power of ice. In the first place, in M. Bouty's method the charge and discharge were enormously slower than in my experiments. Is it not probable, then, that the physical magnitudes measured by M. Bouty and myself were in themselves very different. In any case we know at present too little about the dielectric properties of bodies to be surprised at the divergences of numbers obtained by two methods so dissimilar, however great they are.-Comptes Rendus, October 8, 1894,

INDEX TO VOL. XXXVIII.