Radiation Pressure. 403 



in fig. 6, and at an angle of incidence about 55°. After two 

 internal reflexions it emerged centrally as EF from the other 

 end. Thus a stream of momentum AB was shifted parallel 

 to itself into the line EF, or a counter-clockwise couple acted 

 on the beam. The reaction was a clockwise couple on the 

 block. Using mirror, telescope, and a millimetre- scale about 

 184 cm. distant, a very small deflexion could just be detected 

 with the strongest light and in the right direction. But the 

 quartz fibre was rather coarse, indeed needlessly strong ; and 

 as the time of vibration was only 39 seconds, the deflexion 

 was very minute. To render the effect more evident we used 

 intermittent passage of the beam, sending it in during the 

 half-period of vibration while B was moving from A, and 

 shutting it off while B was moving towards A. The swings 

 then always increased. When the beam was sent in during 

 the approaching half and shut off during the receding half, 

 the swings always decreased, and always rather more rapidly 

 than they increased during the first half. For in the first 

 case the natural damping acted against the light couple, and 

 in the second with it. In one experiment the average increase 

 was '55 scale-division and the average decrease '61 per period, 

 and was fairly regular in each case. The mean was *58. The 

 steady deflexion is half this, or 0*29 division, giving a couple 

 11 X 10 -6 cm. -dyne. We made a measurement of the energy 

 in the beam by means of the rate of rise of a blackened silver 

 disk ; but it was necessarily very inexact, as we had no means 

 of securing constancy in the arc used in this experiment. This 

 energy measurement gave as the value of the couple 6 X 10 -6 , 

 and the agreement is sufficient to show that the order of the 

 result is right. 



An analysis of this experiment shows that the couple was 

 really due to the pressures at the two internal reflexions ; for, 

 as we have seen, the forces at incidence at A and emergence 

 at E are normal and produce no twist. 



Another experiment which we have made is, I think, more 

 interesting, in thatjit brings into prominence the pull outwards 

 or push from within occurring on refraction. Two glass 

 prisms, each with refracting angle 34°, another angle being 

 a right angle, and with refracting edge 1*6 cm. long, were 

 arranged as in fig. 7 (which shows the plan) at the ends of a 

 thin brass torsion-arm suspended at its middle point from a 

 quartz fibre in the same case as that used in the last experi- 

 ment. The two inner faces were 3 cm. apart, and their width 

 was 1*85 cm. A mirror gave the reflexion of a millimetre- 

 scale 171*4 cm. distant. The moment of inertia of the system 

 was 48 gm.-cm. 2 , and the time of vibration was 317 seconds. 



