zenith of the observer's station; in other words, by repeated determinations of geographical latitudes. But, notwithstanding very long and refined determinations of the geographical latitudes at some of the principal observatories, beginning shortly before the middle of the present century, only very uncertain and discordant traces of the phenomena in question were found.

The reason for this want of success is now very clear. Astronomers had limited their researches too narrowly to the last-mentioned typenamely, to the supposed regular ten-monthly periodical movement of the pole of the rotational axis around the pole of the principal axis. Too easily it had been admitted that all the existing variations of the distribution of terrestrial masses were far too small for altering sensibly the position of this principal axis itself.

It was Lord Kelvin, at the Glasgow meeting of the British Association (1876), who first drew the attention of the scientific world to the consideration of the great natural transports of masses of air and water, and of various masses by the water, going on continuously and periodically in the form of currents and circulations of different kinds, as well in the atmosphere as in oceans and rivers, and the effects of the enormous periodical deposits of snow and ice. He showed that these very considerable variations of the distribution of masses on the earth could not only produce sensible displacements of the principal axis of inertia, but that such displacements of this axis could have an amplifying effect on the total amount of displacements of the rotational axis.

For if the principal axis were itself not in a constant position, the theoretically required movement of the rotational axis around the principal axis would become a very complicated movement, differing entirely from the simple form which to that time had appeared in the researches of

astronomers.

This epicyclic character of the movement of the pole of the rotational axis could considerably modify not only the length of the period, but also the whole geometrical character and amplitude of the curve, in such a way that in longer periods epochs of very small variations of latitude could alternate with epochs of considerably increased variations of latitude. Possibly, as a further consequence of this complication of the displacements of the two axes, and as a consequence of the still existing plastic state of certain parts of the earth, as well as by the damping effects of the fluid parts, even progressive-though very slow and unsteady progressivedisplacements of the rotational axis in the earth could still result.

The field of this research was thus decisively cleared by Lord Kelvin ; and finally, about four years ago, by the co-operation of some Observatories with the International Geodetic Union, clear evidence was obtained, and in the last three years, with the aid of an expedition sent by the International Union to Honolulu, decisive proofs of such displacements have been found. I consider it a special honour and pleasure to be enabled to submit some of the newest results of this international co-operation to a meeting of the same Association which, twenty years ago, had been witness of the almost prophetic assertions of one of its most illustrious members regarding the real conditions of this important phenomenon.

I have prepared a diagram showing these newest results. In this diagram is given a representation of the wanderings of the pole of the rotational axis of the earth on its surface during the last twenty months, from October 20, 1892, to May 1, 1894.

This sketch is founded on nearly 6,000 single determinations of latitude made in the Observatory of Kasan (Eastern Russia), Strassburg, Elsass (41° 21' west of Kasan), and Bethlehem, Pennsylvania (124° 30' west of Kasan). The observations are condensed in twenty monthly mean results, numbered from zero to 19. Every one of these resulting monthly positions of the pole indicated by the centres of the small circles is thus the mean result of about 300 single determinations.

Figure showing movement of the North Pole of the rotational axis of the earth.
Derived from observations made at Bethlehem, Strassburg, and Kasan :-

The figure is drawn on the scale of two millimetres to one-hundredth of a second of arc, and the maximum amplitude of the curve is nearly fifty-hundredths, or half a second. The amplitude of these movements of the pole on the surface of the earth is between 40 and 50 feet.

The general character of the movement is quite in accordance with what has been mentioned concerning its complicated and somewhat spiral

character. The sense of the motion is from west to east. The velocity is apparently very variable, and it seems as if we are now approaching an epoch in which the amplitude is considerably diminishing. It is also evident that such a character of movement can very easily produce slow progressive motions, and for this reason the whole phenomenon wants to be watched incessantly and very carefully.

The astronomers and geodetists who are now associated in the International Geodetic Union have invited geologists to associate with them in this common research. Such an international organisation will be also useful and almost indispensable for a great part of the work of astronomical observatories.

It is to be hoped that Great Britain will now participate in this International Union, embracing all other civilised nations. Such organisations, with their clear and reasonably limited aims, involve not only real economies and refinements of mental work, combined with diminutions of material expenses, but it is hoped that they will also have great importance as slowly growing foundations of human and terrestrial solidarity.

A Lecture-room Experiment to illustrate Fresnel's Diffraction Theory and Babinet's Principle. By Professor A. CORNU, F.R.S.

[Ordered by the General Committee to be printed in extenso.]

THE diffraction fringes bordering the shadow of an object illumined by a point of light present us with one of the most striking phenomena in optics.

Dr. Thomas Young was the first to connect these fringes with the wave theory. According to him they were due to the interference of the direct wave with the wave tangentially reflected at the surface of the body screening the light.

Fresnel, following the way opened by Young, proved, however, by a very simple experiment, namely, by the identity of the fringes produced by the edge and by the back of a razor, that the reflected light has no appreciable influence on the production of these fringes. He proved that the phenomenon is exclusively due to the mutual interference of all the vibrations proceeding from the whole of the wave not intercepted by the object.

A mathematical investigation, and even a superficial analysis, shows that the resulting vibratory motion producing the fringes is equivalent to the vibration which would be caused by a permanent wave fixed at the edge of the body screening the light, provided that the acting part of this wave is reduced to a small breadth variable according to the obliquity of the rays. Consequently, everything happens as if the source of light were taken away, and the body were surrounded by a true luminous source forming a sort of border round its apparent contour. This result, proved for a single luminous point, is immediately extended to a circular source of light of any diameter considered as composed of points acting independently.

This optical equivalence is not only a symbolic or geometrical result, but a real physical fact, and the experiment, which I propose to show, gives the most striking evidence of the reality of this source of light.

To show the acting part of the wave surrounding the screening body three conditions are required:

1. The plane of distinct vision must be made to coincide with the edge of the diffracting body.

2. The greatest portion of the diffracted rays must be preserved for producing the phenomenon.

3. The rays not employed in producing the phenomenon must be cut off.

It is not difficult to realise experimentally these three conditions, and the result obtained is extremely brilliant with sunlight. Though much less intense with the electric arc, it is nevertheless sufficiently distinct to be projected. With a feeble source of light, such as an oil-lamp, the phenomenon is still quite visible, but only by direct vision.

Let us take a luminous source of circular form (the disc of the sun or a circular hole in a diaphragm conveniently illuminated). Let the rays diverging from this source fall upon an achromatic lens (0.5 to 1 metre focal length) and produce at the conjugate focus a bright circular image of the source. Exactly at this focus let us place a black circular disc having precisely the same diameter as this image.

The eye placed behind this opaque disc will not see any point of the luminous source, because all the rays are cut off by the disc; it will perceive, however, a luminous line round the edge of the lens. This is in itself an illustration of the acting part of the Fresnel's wave, for the edge of the lens is really a screen which intercepts the incident wave. But the experiment becomes extremely striking if any object is put on the surface of the lens: this object appears on a dark field as if surrounded by a luminous line following even the smallest details of its edge. [In the experiment projected before the Section the object shown was a branch of fern.]

Moreover, each particle of the unavoidable dust lying on the lens produces a luminous point, and for this reason the field is never quite dark. Lycopodium thrown in front of the lens exaggerates this effect and produces a sort of luminous nebula.

The intensity of the bright bordering line depends upon the accuracy with which the circular screen masks the focal image of the source; if the diameter of the screen is too small the field becomes luminous, if it is too large the brilliancy of the line diminishes. If another shape be given to the screen the luminous line is interrupted normally to the directions in which the screen extends much outside the image of the source.

With the circular screen it is easy to observe the important particular case to which Babinet called attention, viz., that a very thin opaque line diffracts light exactly in the same manner as a transparent slit of the same form and size.

It suffices for this purpose to observe the shadows of successive wires, straight or curved, of decreasing diameters. The dark space lying between the two luminous edges in the case of a thick wire diminishes as the wire becomes finer, and vanishes when its apparent angle becomes sufficiently small; then the wire appears on the dark field like the incandescent filament of a glow-lamp.

The slits corresponding to the wires are made with lines of decreasing breadths traced on smoked glass. The appearances are exactly the same as before, a double bright line being produced by a broad slit, a single bright line by a narrow slit, so that a fine transparent slit and a fine

1894.

11

opaque wire give exactly the same phenomenon. This is the simple and useful result called sometimes Babinet's Principle, and, in any case, the experiment just described gives immediately the means of verifying whether the conditions required for the correct application of this principle are sufficiently fulfilled.

All these results can be collected under a pretty form which illustrates one of the finest natural phenomena. It is well known that when the sun is hidden behind the top of a mountain, but near the crest, all objects in its neighbourhood are surrounded by luminous borders, and minute objects, branches, leaves of trees, flying birds, &c., appear on the sky as if incandescent.

To imitate this appearance it suffices to cut out in cardboard an irregular edge representing the crest of the mountain, and to border it with some blades of grass or moss representing trees. This object placed on the lens reproduces an image of this beautiful phenomenon, which seems to have been observed many centuries ago in deep valleys when the sun rises; for Shakespeare says (Richard II., act iii. scene 2): 'He fires the proud tops of the eastern pines.'

The Connection between Chemical Combination and the Discharge of Electricity through Gases. By J. J. THOMSON, Professor of Experimental Physics, Cambridge.

[Ordered by the General Committee to be printed in extenso.]

THE intimate connection between chemical change and the passage of electricity through liquids has been universally recognised ever since Faraday discovered the laws of electrolysis which bear his name. These laws state that, whenever electricity passes through an electrolyte, chemical changes take place in the electrolyte, and that the quantity of electricity which passes through the electrolyte is connected in the most intimate and simple manner with the amount of chemical change which has taken place during its passage. For each unit of electricity which passes through the electrolyte a definite amount of chemical change takes place, and the chemical changes which take place when equal quantities of electricity pass through different electrolytes are chemically equivalent. But although chemists have largely availed themselves of the light thrown by electrolysis. on chemical phenomena, the subject of the passage of electricity through gases does not seem to have attracted their attention. We have strong evidence, however, that the connection between the discharge of electricity through gases and chemical change is not less intimate than that between electrolysis and chemical change. Thus, for example, when the electric discharge passes through steam, the steam is decomposed, an excess of hydrogen appearing at one electrode and an excess of oxygen at the other; these excesses of hydrogen and oxygen are proportional to the quantity of electricity which has passed through the steam, and are equal to the amounts of hydrogen and oxygen which would be liberated if the same quantity of electricity passed through a water voltameter. We have here evidence for connecting chemical action with the discharge of electricity through gases of precisely the same kind as that which has connected it with electrolysis. Again, as I hope to show later in this paper, the passage of