of view of the Kinetic Theory of Matter. 29 



volumes on the two sides of the diaphragm (D) are considered. 

 The whole exhausted system of the mercury pump is in connec- 

 tion with the region (A) of the tube. The volume of the pump 

 system is two or three times greater than the whole volume of 

 the tube. So that the indication of the second radiometer 

 merely shows that no measurable reduction in the gaseous 

 pressure on the A-side of the tube occurred during discharge. 

 This experiment was repeated many times, and at different 

 pressures. In every case the indication of the radiometer on 

 the B-side of the tube showed a decrease of pressure when its 

 stopcock was open, so that it was put in communication with 

 the rest of the exhausted system. In no case was there any 

 considerable change of deflection noticed in opening the 

 cock of the radiometer on the A-side of the tube. 



When the gaseous pressure in the tube is very much higher 

 than that at which the radiometers can be used to advantage to 

 measure differences of pressure, if a discharge be sent through 

 the tube by means of the above described uni-directional po- 

 tential-difference, a discharge stream is seen to proceed from the 

 cathode and continue on its course to within a short distance 

 of the openiug in the funnel-shaped diaphragm. IJere the dis- 

 charge stream seems to end in a bright band, and from this point 

 on, in the region (B), striae are seen to succeed one another, as in 

 the stratification phenomenon. The tube in this state trans- 

 mits discharge through the region (A) by the discharge stream, 

 and through the region (B) by the stratification process. As 

 discharge stream and stratification phenomena are charac- 

 teristic of different gaseous pressures, this last experiment is in 

 agreement with the experiments described in the preceding- 

 paragraph. 



Not being provided with a McLeod gauge, or other reliable 

 means of measuring low gaseous pressures, it is impossible to 

 state with accuracy the absolute value of the change of pres- 

 sure observed by means of the radiometers in the previous ex- 

 periments. However, the pressure at which the radiometers 

 gave their maximum deflection could not have been far from 

 50.10" 8 Ats., and the pressure at which the first set of the 

 above experiments was made could not have been far from 

 10 . 10" 6 Ats. The spot of light from the radiometer con- 

 nected with the B-side of the tube at the above pressure, fell 

 at SO or 90 cm from zero. It seems probable, therefore, that 

 the change of pressure observed amounted to more than 10 

 per cent of the uniform pressure of the tube. 



The above experiments would seem to be in good agree- 

 ment with the considerations in § 3 on the pressure resulting 

 from the motion of a gaseous medium. So that in the above 

 experiments, region (B), into which a discharge stream passes, 



