mechanical action of light would manifest itself in pushing away the surface from which its rays are reflected, and that the surface into which they are absorbed would move towards the source from which the rays emanate.

In subsequent communications to the Royal Society, Mr. Crookes committed himself explicitly to the doctrine that the Radiometer (to which he also gave the name of 'light-mill') is driven by light, the mechanical effect of which he assumed to be proportional to its illuminating power; so that an exact measurement of the former would furnish an equally exact measurement of the latter. And thus, as a Thermometer, which measures heat by its physical action in producing expansion, is a far more trustworthy instrument than the human hand, the action of heat upon which gives rise to a sensation that is not capable of precise measurement and may be altogether deceptive, so, he contended, the Radiometer, which measures light by its physical power of repulsion, is a far more exact photometer than any which depends upon the physiological action of light upon the retina.

6

For the absolute measurement of the repulsive force exerted by radiation, Mr. Crookes employed the delicate torsion-balance already described; and found the mechanical effect of the light of a candle at twelve inches' distance, acting on two square inches of surface, to be 444 millionths of a grain. This he called weighing a beam of light. In his subsequent lecture at the Royal Institution, he stated it as the result of his experiments, that the radiant energy of the Sun equals that of 1,000 candles at twelve inches' distance, its mechanical power upon two square inches of surface being thus equal to 444 thousandths of a grain. This is equivalent to about 32 grains per square foot, to 2 cwt. per acre, to 57, tons per square mile, or to nearly 3,000,000,000 tons on the exposed surface of the globe-sufficient to knock the earth out of its orbit if it came upon it suddenly,' and to drive the globe into space if it were not counteracted by the force of gravitation.' But, as Mr. Crookes considerately added, ‘it must be remembered that our earth is not a lamp-blacked body enclosed in a glass case; nor is its shape such as to give the maximum of surface with the minimum of weight."

6

For comparative estimates, however, Mr. Crookes devised another form of torsion-balance,' in which one half of the pith-bar was blacked and the other left white, so that, when the whole was subjected to a broad beam of light, the bar would be made to swing by its different action on the white and on the black surfaces. This bar being made to carry a mirror and a small magnet, its slightest deflection from the N. and S. zero could be detected, by the movement of a spot of light reflected from the mirror along a graduated scale; whilst the mechanical force required to produce any degree of deflection could be very exactly estimated, the apparatus being made more

or less sensitive, as desired, by the use of a controlling magnet on the outside. The results of a long series of experiments served to show the photometric value of this apparatus, when lights of the same kind were compared; the amount of deflection produced by the light of the same standard candle at different distances being obviously conformable to the law of inverse squares, whilst the effect of two such candles placed side by side was found to be practically double that of a single candle, and that of three candles practically triple. Again, when a candle was placed on either side of the apparatus and equidistant from it, so that each would neutralise the effect of the other, the index-spot of light remained at zero; but by shading one or other of the candles, the index-spot was made to fly off to the extremity of the scale.

This arrangement afforded a ready means of comparing the Radiant Energy of different sources of light. Thus, if a standard candle' was placed on one side at a distance of 48 inches, and a gas-burner on the other was found to bring the index-spot to zero when removed to a distance of 113 inches, their relative motor powers would be as 482 to 1132 that is, the radiant energy of the gas-burner was equal to that of 5 candles. But it was as pure an assumption on Mr. Crookes's part to affirm that the mechanical action exerted by two flames of different kinds would measure their relative illuminating powers, as it would have been to say that their heating action would be proportional to their illuminating action, which we know perfectly well not to be the case,-the gas flame, as everyone knows, having a much greater heating power than the candle flame, in proportion to the light it gives.

The same form of torsion-balance' was employed by Mr. Crookes to determine the relative effects of the interposition of screens of different kinds. Thus, a deflection to 180° being produced by a candle at three feet distance, its amount was reduced to 161° by the interposition of a screen of yellow glass, to 128° by a screen of red glass, and to 102° and 101° respectively by screens of blue and green glass. A far more potent effect, however, was produced by the interposition of a glass trough containing water, which brought down the deflection to 47°, whilst a screen formed of a plate of alum reduced it to 27°. As the absorption of the luminous rays in passing through such transparent media must have been very slight, whilst they practically cut off the rays of dark heat,' this marked reduction would seem mainly attributable to the abstraction of the latter; but, strong in his conviction of the immediate mechanical effect produced by radiation, Mr. Crookes thus expressed himself in regard to it: "There is no real difference between heat and light; all we can take account of is difference of wave-length, and a ray of a definite refrangibility cannot be split up into two rays, one being heat and one light. Take, for instance, a ray of definite refrangibility in the red. Falling on a

thermometer it shows the action of heat; on a thermopile it produces an electric current; to the eye it appears as light and colour; on a photographic plate it causes chemical action; and on the suspended pith it causes motion.' Now, so far as Light, Heat, and Chemical action are concerned, this mode of expressing their relations is undoubtedly that which all physicists now accept; these agencies being regarded, not as separate and distinct, but as different manifestations of that one physical action which constitutes Radiation. This action, according to the undulatory theory, consists in the propagation, through an ethereal medium, of systems of waves of different lengths; and it is in virtue of this difference that their direction is more or less altered by refraction, the longest waves being least, and the shortest most deflected by passing through the prism. The solar beam is the composite resultant of the whole aggregate of these undulations. When falling on the eye it excites the sensation of colourless light; when falling on the hand it affects us with the sensation of warmth; when falling on the bulb of the thermometer it causes the expansion of the mercury; and when faliing on a photographic surface it produces chemical change. But when made to pass through a prism, it is decomposed not only into that succession of colour-bands, formed by rays of different degrees of refrangibility,. which constitutes the luminous spectrum; but into two other successions of rays, one of much lower and the other of much higher refrangibility, which lie beyond the two ends of the luminous spectrum.

These dark rays are not recognisable by the eye, because the retina is no more sensible to them than the ordinary cutaneous surface is to luminous impressions; but their heating power can be measured by a thermometer or a thermopile, and their chemical power by the action they excite on a photographically prepared surface. The Heating power is thus found to attain its maximum a little outside the red end of the colour-spectrum; and from that point it progressively diminishes towards the violet end of the luminous spectrum, beyond which it is scarcely traceable; whilst it diminishes in the contrary direction also, until it dies out at about the same distance from the maximum on one side, as that at which the violet lies on the other. The Illuminating power has its maximum in the yellow band of the spectrum, and shows a gradual reduction towards the violet end, a more rapid towards the red. The Chemical power, on the other hand, has its maximum in the violet band; and whilst it gradually diminishes towards the red end of the luminous spectrum, beyond which it is scarcely traceable, it diminishes at about the same rate in the opposite direction; dying out at nearly the same distance from the maximum on one side, as that at which the red lies on the other. Thus, while the rays of low refrangibility, whose wave-length exceeds 812 millionths of a millimetre, are characterised almost exclusively by their heating power, and those of high refrangibility, whose wave-length is less than

400 millionths of a millimetre, are characterised almost exclusively by their chemical power, the rays of medium refrangibility, whose wave-lengths are between 400 and 800 millionths of a millimetre, combine these with illuminating power, in proportions varying with their respective wave-lengths. But there is no more reason, as Mr. Crookes has justly remarked, for attributing these several effects to different rays, than there is for hypothetically splitting up the element iron (for example) into a number of components, of which one gives its specific gravity, a second its chemical reactions, a third its magnetic properties, and so on.

But to the three attributes of Radiation universally recognised by Physicists, Mr. Crookes proposes (in the passage already cited) to add a fourth, the power of producing an electric current in a thermopile; and a fifth, the power of producing mechanical motion when acting on light bodies freely suspended in a vacuum. Now the notion that radiation directly excites the electric current of a thermopile, is one (I apprehend) which no well-informed Physicist would endorse; for (as the name of the instrument implies) it is by the disturbance of the thermal equilibrium between the two metals of which it is composed, that the electric current is produced. And since this disturbance may be produced in a variety of ways (as by friction or conduction), and the potency of the electric current is strictly proportional to the amount of that disturbance, there is no reason whatever for attributing to radiation any other power of exciting an electric current, than that which it exerts mediately through its power of heating the thermopile. And the question which, after the first shock of novelty passed off, has greatly exercised the minds of Physicists, is whether the mechanical motion, also, is not an intermediate effect of some one of the previously known forms of radiant energy-that which first suggests itself being the action of Heat upon that residual vapour of which it is impossible to get rid entirely by any means at present known.

This idea very early occurred to some of the distinguished Physicists who took most interest in the experiments first communicated by Mr. Crookes to the Royal Society. I more than once conversed with Wheatstone on the subject; and he expressed a very strong belief that the swinging round of the pith-bar was due to the disturbance of the thermal equilibrium in the residual vapour, dwelling very strongly upon the impossibility of obtaining a perfect vacuum, since even glass,' he said very emphatically, 'would give off a vapour, if all other vapour were withdrawn.' On the other hand, two of the most distinguished among British mathematical Physicists, whose opinions I had the opportunity of learning from themselves, were disposed to concur with Mr. Crookes in regarding the repulsion of the heated end of the pith-bar as an immediate effect of Radiant Energy; dwelling especially on the fact, that the

6

repulsion was stronger in proportion to the completeness of the vacuum, from which it seemed fair to infer that it would be most strongly manifested in a perfect vacuum (if such could by possibility be obtained), in consequence of the entire removal of mechanical resistance (save the friction of the pivot) to the rotation of the mill.

Now, since our belief in Newton's First Law of Motion has no other experiential basis than the fact, that, the more completely we can eliminate friction and the resistance of the air, the longer is the persistence of motion in a body once put in movement, provided that no opposing force be brought to bear upon it, this argument for the directness of Radiant Repulsion seemed alike valid and cogent. We shall presently see, however, that it proves fallacious when brought to an experimental test of greater delicacy.

The doctrine propounded by Mr. Crookes was first explicitly called in question in a communication made to the Royal Society on the 18th of June, 1874, by Professor Osborne Reynolds; who maintained that, on the kinetic theory of gases (which represents any gaseous substance as consisting of molecules constantly in motion at great velocities), the effect on the torsion-balance is really due to alternate evaporation of vapour from, and its condensation at, the surface of the pith; evaporation producing a reactionary force equivalent to an increase of pressure on the heated surface, whilst condensation must be attended with a force equivalent to a diminution of pressure over the cooling surface. Thus, when the heat radiated from the lamp falls on the pith, its temperature will rise, and any moisture on it will begin to evaporate, thus generating a mechanical force which will drive the pith from the lamp. Conversely, when a piece of ice is brought near, the temperature of the pith will be reduced, causing a condensation of vapour which will cause the pith to move towards the ice. When the two arms of the pith-bar are unequally exposed to heat, the evaporation will be greatest in that which is nearest the lamp; and this is driven away, therefore, until the force on the other arm becomes equal, after which the bar will come to rest, unless the momentum it has acquired in swinging carries it further.

In a subsequent communication (March 23, 1876) Professor O. Reynolds applied a similar doctrine to the continuous rotation of the 'light-mill;' maintaining, from theoretical considerations, the existence of reactionary forces, or heat reactions,' whenever heat is communicated from a surface to a gas, and vice versâ; and showing that there is enough residual air in the best exhausted globe to enable an amount of force to be thus developed, which is sufficient to keep up the rotation of its contained mill. This explanation obviously implies the existence of a reactionary force, communicated by the intervening gas, between the discs of the mill and the enclosing glass; and the existence of such a reaction was experimentally proved by an ingenious arrangement first devised by Dr. Schuster and subse