Compounds by homogeneous X-radiation. 179 



electroscope exposed to the same radiator. By replacing the 

 mixture with air, the ionization produced in air at 76*0 cm. 

 by the same beam o£ rays could be measured immediately 

 after the experiment with the mixture, the leak with air 

 being compared with the same standard electroscope as before. 



From these observations the ratio of the ionization in the 

 mixture to the ionization produced by the same beam in air 

 was calculated for each mixture used, these ratios being 

 given in column II. Table I. 



To determine what part of the ionization was due to the 

 hydrogen in the mixture, separate experiments were per- 

 formed as indicated above, using hydrogen alone at atmo- 

 spheric pressure. The ionization in the hydrogen alone 

 includes the effect of the corpuscular radiation from (he ends 

 of the chamber, and, deducting this from the total ionization 

 produced in the mixture, the remainder gives the ionization 

 due to the vapour alone at its saturation-pressure for 0° 0. — ■ 

 at any rate as a first approximation. This method of calcu- 

 lation also eliminates any other constant error, such as 

 insulation leak, &c. 



The values for the relative ionizations in the hydrogen are 

 given in column III. Table I., column IV. giving the ioniza- 

 tion due to the vapour alone at its saturation-pressure at 

 0° C. in terms of the ionization produced by the same beam 

 in air at atmospheric pressure. From the values tabulated 

 iu column IV. the ionizations which w T ould be produced in 

 the vapours alone at 76*0 cm. were calculated, assuming the 

 ionizations to be proportional to the pressures : these are the 

 numbers given in column V. of the same Table. 



It was found impossible to obtain any "atomic" ionization 

 values which would satisfy these observations at all accurately. 

 Values for the ionizations due to an atom of carbon and to 

 an atom of oxygen were obtained which would satisfy any 

 two of the substances used, but when these values were 

 substituted for in the molecular composition of any of the 

 other vapours, the values obtained differed by as much as 

 25 per cent, from the observed values. 



That the absorption and ionization produced in the same 

 vapour at different pressures is proportional to the pressure 

 — provided the variation in the secondary radiation escaping 

 is negligible — has been amply shown by Crowther *, Owen f, 

 and others and therefore this method of calculation would 

 have demonstrated the existence of " atomic " ionization if 

 ionization had been truly atomic. The amount of secondary 



* Camb. Phil. Soc. Proc, Feb. 1009. 

 t Kov. Soc. Proc, Mar 1912. 



N 2 ' 



