﻿118 Mr. E. W. B. Gill on the Variation with Distance 



one and the large detector, were always the same ; and the 

 error in the second experiment of radiation from the left 

 passing through the right or vice versa is negligible, the 

 aluminiums being '2 mm. thick. 



Some experiments at the outset, however, proved that it 

 was unnecessary to evacuate the tube each time. 



In these experiment* the large detector was kept at the 

 longer distance (about 105 cm. from anode), and the charge 

 received by it per minute compared with that received by the 

 small detector for various pressures in the tube. 



The small detector in these experiments after the first 

 minute gave the same reading per minute throughout, proving 

 that the bulb was emitting constant radiation ; and this was 

 generally the case in further sets of experiments, till the bulb 

 in any set had been running for some time. The electro- 

 meter deflexion due to the large detector, as a mean of several 

 experiments, was found to be the same when the air in the 

 tube was at atmospheric pressure, half the atmospheric 

 pressure, and at a pressure of 4 cm. of mercury; thus proving 

 that the air absorption was negligible. The total length of 

 the tube was about 76 cm., and in general this would produce 

 an appreciable absorption ; for instance in some experiments 

 of Rutherford * the rays lost 1 per cent, on traversing 10 cm. 

 of air. 



The above results, however, are explained by the fact that 

 the bulb used was rather hard, producing a very penetrating- 

 radiation, and any easily absorbed radiation would probably be 

 stopped at the first aluminium. Further experiments showed 

 that the rays lost about 27 per cent, in going through *4 mm. 

 of aluminium. 



The method adopted in this first set of experiments was to 

 make two observations with the large detector, then two with 

 the small, then four with the large, followed by two more 

 with the small, ending up with two with the large. 



The electrometer gave about 200 mm. deflexion for 1 volt, 

 and the readings due to the large in 1 minute were about 

 40 mm., and to the small 35 mm. The voltage applied to 

 the parallel plates, 80 volts, was always sufficient to produce 

 saturation. 



The above method of experiment was adhered to with 

 slight variations in all cases, and any change in the con- 

 stancy of the radiation could be at once detected. 



It was thought worth while at this stage, to try whether 

 water vapour had an exceptional absorbing effect for Rontgen 



* Phil. Mag. xliii. p. 241 (1897). 



