﻿Absorption of Rontgen Rays in Air. 687 



the 7 rays from the radium C in a known quantity of radium 

 bromide * at a measured distance from the electroscope f. 



Now at a distance of 40 m. the hard bulb A produced 

 2340 ions per cubic centimetre per second ; hence 



h 



OQ4.0— .,-'00029x4000 



w ^ U -(4000^ 

 Therefore 



&=1-17 xlO 11 , 

 and 



N=2<7r/f/\ = 2-54xl0 1 \ 



For the bulb B at 40 metres 

 &=2-14xl0 u , 

 N= 2-36 x 10 15 . 



These values for N were found by integration over a hemi- 

 sphere. As a comparison, it may be pointed out that, 

 integrated over a sphere, the total number of ions per 

 second, due to a gramme of radium, with RaO in equilibrium, 

 is l'l X JO 15 . 



The power supplied to the Rontgen-ray bulbs was of the 

 order 100 watts or 10 9 ergs per second. As the energy 

 required to make an ion is about 2 X 10 -11 erg, it follows 

 that the power spent in ionizing is 2 x 10 ~ n x 2*4 x 10 15 , or 

 about 5 X 10 4 ergs per second. This is only 1/20000 part of 

 the power passing to the bulb, the greater part of which is 

 transformed into heat. The whole apparatus of induction- 

 coil and bulb is exceedingly inefficient as a producer of 

 Rontgen rays, which are, as it were, a minute side product. 



If the ionization is integrated to the surface of the bulb 

 instead of to the anticathode, the correction is less than 1 

 per cent. 



Summary, 

 (1) The coefficients of absorption by air have been found, at 

 short ranges (4-10 m.), as follows :— 



Soft bulb (C), '0018 to -0010. 



Moderate bulb (B), -0004. 



Hard bulb (A), '00033 to '00025. 



These values agree well with previous determinations 

 by Rutherford, and by Rutherford and McOlung. 



* Phil. Mag-. Sept. 1906 and Oct. 1911, 



f It will be necessary, for more accurate determinations, to investigate 

 the influence of the thickness and material of the walls of the electro- 

 scope on the internal ionization, and to compare the values obtained 

 with those found for a very thin- walled electroscope. This question has 

 received attention for y rays (Phil. Mag. 1911, p. SO and p. 554) but not 

 apparently for llontgen rays. 



