THE PEESSUKE DUE TO KADIATION. 137 



The observed angle of deflection of the spores from the vertical was 

 roughly that given from the computation, and the observers believed 

 that the efi'ects shown must be due almost entirely to light pressure, 

 with possibly a slight gas action. The action of gases upon heated 

 bodies of this size had, so far as we know, never been studied, but 

 one of the writers" had studied the gas action on larger bodies down 

 to a pressure of permanent gases of 0.0005 mm. of mercury, as shown 

 by a McCleod gauge, and had observed that for this pressure the gas 

 action had begun to fall off sharply. The pressure of the permanent 

 gases in the hourglass must have been well below this value, and it 

 was thought that nearly all pressure due to vapor had been frozen out. 



Later, a review of the preliminary computation was made and an 

 error discovered which had the effect of bringing out the computed 

 light pressure on bodies of this size and densit}' far too large. It was 

 plain, therefore, that the force of deflection due to gas action, proba- 

 bly of the character of rocket action, was at least ten times as large as 

 the effect attributable to radiation pressure. Radiation pressure 

 alone would produce a measurable effect under the conditions of 

 observation, but would have been far less pronounced than the effect 

 obtained. 



The experiment had unfortunately to be tried under circumstances 

 nuK'h more unfavorable for a pronounced effect of radiation pressure 

 than exists in comets, for the deflection produced by repulsion must 

 be measured in terms of terrestrial gravitation, which is over 1,600 

 times as great as solar gravitation at the distance of the earth. To 

 approach cometary conditions, therefore, it would have been necessary 

 to use a light beam 1,600 times as intense as sunlight at the earth. 

 In the experiment, beams from twenty to forty times as intense as 

 sunlight were used. 



Because of the meagerness of present knowledge concerning the 

 actual conditions in comets' tails it is im23ossible to say how closely 

 the foregoing experiment fulfilled the purpose for which it was tried. 

 It would be difficult to prove from present astronomical data that the 

 hydrocarbon vapors know^n to exist in comets' tails exert no radio- 

 metric repulsion upon the small reflecting particles present. Still 

 more difficult w^ould it be to show that nothing which corresponds to 

 what has been called rocket action occurs. This latter repulsion does 

 not require the presence of any generally diffused atmosphere what- 

 ever, but simply that the particles send off gases toward the sun under 

 the action of the sun's heat. Thus, in passing from the era where no 

 adequate physical causes which would meet the required conditions 



«A result gained in a series of unpublished experiments on gas forces by W. v. 

 Uljanin and E. F. Nichols. See also W. Crookes, Phil. Trans., p. 300, 1878. 



