1906.] Some Asirnnnmical Consequences of Pressure of Light 839 



WEEKLY EVENING MEETING, 

 Friday, May 11, 1906. 



His Grace the Duke of Noethu:mberland, K.G. P.C. D.C.L. 

 F.R.S., President, in the Chair. 



Professor J. H. Poynting, Sc.D. F.R.S. 



Some Astronomical Consequences of the Pressure of Light. 



[abstract.] 



The experiments of Lebedew and Nichols and Hull have proved 

 conclusively that light presses against any surface upon which it 

 falls, and the extraordinarily accurate experiments of Nichols and 

 Hull have f uUy confirmed Maxwell's calculation that the pressure per 

 square centimetre is equal to the energy in the beam per cubic 

 centimetre. 



A clearer idea of the effect of light or radiation pressure is 

 obtained by thinking of a beam of light as a carrier of momentum. 

 We then see that not only does it press against a receiving surface, 

 but also against tlie surface from which it started. 



Some experiments by Dr. Barlow and myself appear to bring to 

 the front this conception of light as a momentum carrier. If a beam 

 falls on a black surface at an angle to the normal, there should be a 

 tangential stress along the surface. Ai\ experiment was described 

 in which light fell on a blackened disc at the end of a torsion arm, 

 the disc being at right angles to the arm.* The disc was pushed 

 round by the tangential stress. The experiment was carried out in a 

 partially exhausted vessel, but the residual air was a source of dis- 

 turbance by convection and radiometer effects. A better experiment 

 was made by suspending a disc of mica blackened beneath, about 

 2 inches in diameter, by a quartz fibre, the disc being horizontal and 

 suspended from its centre. When a beam of light fell at 45° on a 

 part of the disc, the horizontal component of the beam being at right 

 angles to the radius to the part where it fell, the disc moved round 

 through the combined effects of convection, radiometer action, and 

 the tangential stress. When the beam was allowed to fall on the 

 same place at 45° on the other side of the vertical, convection and 

 radiometer action were very nearly as before, but the tangential stress 

 was reversed. The difference in torsion in the two cases was twice 

 that due to the tangential stress. An experiment with prisms t was 

 also described. 



* Phil. Mag., ix. (1905) p. 169. t Ihid., p. 404. 



z 2 



