August 4, 1910J 



NATURE 



141 



The reflecting surface was made by depositing a 

 silver film on the outside of the compound disc by 

 means of the discharge from a silver kathode in an 

 exhausted receiver. 



Four holes the size of the discs were cut in a plate 

 of mica ABCD, the centres of the holes being at the 

 corners of a 2-cm. square (Fig. 2). The discs were 

 fixed in these holes by a minute amount of celluloid 

 varnish. Below the mica plate were the two observ- 

 ing mirrors M,, M„. This system was suspended by 

 a quartz fibre G, 9 cm. long, in the centre of a glass 

 flask of 16 cm. diameter. The upper end of the quartz 

 fibre was fixed to a brass collar H, held by friction in 

 the neck of the flask. 



The exhaustion of the flasl-: was carried out to a 

 very high degree, but an account of the method adopted 

 and the precautions taken is here unnecessary. In the 

 final stage of the process the residual gas was ab- 

 sorbed bv a charcoal bulb kept surrounded by liquid 

 air, which was boiled off continuously at a reduced 

 pressure of about 2 cm. of mercury, for several hours 



before, and dur- 

 ing the whole of 

 the measure- 

 ments. 



The source of 

 light was an 

 Ediswan 50-volt 

 "focus lamp," 

 which was fed 

 from accumula- 

 tors. By a suit- 

 able arrangement 

 of achromatic 

 lenses a uniformly 

 illuminated image 

 of a circular dia- 

 phragm was 

 focussed on the 

 disc to be worked 

 with. The flask 

 was mounted on 

 a turn-table, so 

 that by rotation 

 through 180° ex- 

 periments could 

 be made on the 

 reverse sides of the 

 discs. 



For reading the 



deflections the 



image of an elec- 



Fr . 2- trie lamp on a 



millimetre scale 



at_ a distance of 113 cm. was used, the deflection 



being read to o'2 mm. 



.\s the mean of a number of determinations we have 

 the final values (in scale divisions) for the pressures 

 on the four discs : — 



BE BS ss SB 



l6-i 22-3 2S7 28-0 (Observed) 



.\ determination of the energy of the beam was 

 made by allowing it to fall on 'a blackened disc of 

 silver and observing the initial rate of rise of tempera- 

 ture by means of a constantan-silver thermoelectric 

 junction soldered to the disc. The energv was found 

 to be 33 X 10-' ergs per cm. length of the' beam used. 

 This would be the force in dvnes on a fullv absorbing 

 surface. Had the BB disc "been fully absorbing the 

 beam should have dcflqcted it 136 "scale divisions. 

 Assuming that the asphaltum disc reflects 5 per cent. 

 and the silver 95 per cent, of the incident beam— an 

 estimate which cannot be seriouslv in error — it can 

 easily be shown that the deflection's of the four discs 

 should have been : — 



o 



NO. 2127, VOL. 84] 



14-3 220 26-5 26-1 (Calculated) 



The general agreement of these values with the 

 observed values given above appears to afford satis- 

 factory evidence tor the existence of the recoil efi'ect. 

 In the case of the BB disc, however, there is a marked 

 difference between the observed and calculated values, 

 and this discrepancy is probably to be ascribed to 

 residual radiometer action. There was reason to 

 suspect that this action was not sufficiently reduced 

 to make it quite negligible, and it was obvious that 

 this disc should be more affected than the others by 

 radiometer action, as the difference in temperature 

 between the two sides of the disc is greatest in 

 that case. 



The forces due to light are so small, and the dis- 

 turbances due to convection are relatively so great, 

 that we cannot expect to find any effects due to light 

 pressure here on the surface of "the earth surrounded 

 by a dense atmosphere. But out in interplanetary 

 space, where the vacuum must be far higher than 

 anything to which we can attain, the light forces may 

 have uninterrupted play, and in the course of ages 

 they may produce great effects; but even then only 

 small bodies will be seriously affected. Take, for 

 instance, a sphere; the pressure of sunlight upon it 

 varies as the square of the radius, and the mass as 

 the cube of the radius. Thus the acceleration pro- 

 duced is inversely as the radius for spheres of the 

 same density. The whole pressure of sunlight upon 

 the earth is only a forty-billionth part of the sun's 

 gravitative pull. If we reduce the radius the pressure 

 becomes more important in proportion, and on a sphere 

 of one forty-billionth of the radius of the earth — or 

 16x10-' cms. radius — and of the earth's density, if 

 diffraction did not come into play, the pressure of 

 sunlight would balance gravitation. Still smaller 

 spheres would be pushed awav. 



But turning to the case of bodies somewhat larger 

 than those in which gravitation is neutralised by 

 light-pressure, bodies for which gravitation is still 

 much greater than the light pressure, Vi'e will now 

 consider an effect on them due to the pressure of radia- 

 tion against the source. Let us suppose that a small 

 spherical absorbing body is circling round the sun. 

 It is receiving radiation from the sun on its bright 

 half, transmuting it to heat, and then giving out this 

 energy as radiation again all round. If the sphere 

 is sufficiently small — say of i cm. diameter or less — 

 it will be practically of the same temperature through- 

 out, a small difference of temperature from front to 

 back sufficing to carry through the energy which it 

 radiates from the dark half. It virtually receives from 

 the sun on its diametral plane, and it radiates out 

 from its whole surface, which is four times as great. 

 So that its rate of radiation per sq. cm. is one-fourth 

 the solar radiation per sq. cm. passing the sphere. 

 But we suppose that the sphere is moving round the 

 sun. As it moves forward it crowds up the waves in 

 front and opens out the waves behind it. It follows, 

 then, that in consequence of the motion, the pressure 

 is slightly greater against the radiation emitted in 

 front, and slightly less against the radiation emitted 

 behind. The negative acceleration, or retardation, 

 works out to be 



_ S J' 

 4ap U-' 

 where S is the solar stream of radiation, a the radius 

 of the sphere, p its density, v its velocity, and U the 

 velocity of light. As the sphere moves against this 

 resisting force its energy is gradually abstracted, and 

 it tends to fall into the sun. 



If a sphere i cm. in diameter, and of the density of 

 the earth, is moving round the sun at a distance equal 



