106 M. E. Pringsheim on the Radiometer, 



on both sides. The light-source was placed so that its centre 

 was in the normal to the centre of the mirror ; so that the 

 cone of rays issuing from it was tangent to the glass sphere 

 in the same vertical plane in which the mica vane hung. 

 Therefore, if our assumptions are correct (which shall subse- 

 quently be still more closely investigated) , and if it be further 

 assumed that any possible absorption within the rarefied air 

 is without influence (a question which will also be afterwards 

 discussed), any effect that may be produced must be ascribed to 

 the heating of the glass side alone. • 



First some experiments were made with sunlight reflected 

 from the metallic mirror of a Duboscq heliostat upon the 

 apparatus. With these there was not the slightest motion of 

 the spot of light. . 



The thing took another shape, however, when the flame of 

 a Bunsen burner was employed as the source of heat. There 

 were now energetic deflections which were all positive — that 

 is, in which the irradiated side recedes — amounting, on the 

 average, to 65 millim. 



We see therefore that repulsion of the vane by the more 

 strongly heated side of the glass case takes place, if our above 

 suppositions are correct. 



But now it might still be possible that we should have to 

 distinguish in the thin mica plates a front and a back side, 

 and that the front face absorbs a greater portion of the rays 

 than the back, hence becomes warmer, and, in consequence of 

 this, recedes. 



In order to test this possibility the glass case was irradiated 

 laterally and at the same time half of it shaded so that the 

 mica plate was in the shade, and its plane formed exactly the 

 plane of division of light and shade. If now motion again 

 occurred, it certainly could not arise from direct absorption 

 by the mica plate, since this itself was outside of the cone of 

 light. 



On employing sunlight, again no trace of motion appeared; 

 and consequently, as directly follows from the action of the 

 gas-flames, it was proved that the thin glass absorbed only 

 vanishingly little of the sun's rays. 



Lateral irradiation by a gas-flame, on the contrary, effected 

 a deflection in the same direction as before — namely, so that 

 the mica plate was always repelled from the illuminated side 

 of the glass. 



The results were : — 



I. With lateral irradiation from the vane side (so that the 

 part of the glass case nearest to the vane was first struck bv 

 the light) :— 



