Compounds by homogeneous X-radiation. 185 



Methyl Formate 2C + ±H + 20 = 1*64 



Ethyl Oxide (Ether) ... 4 C + 10 II + = 1-42. 



Ethyl Formate 3 C + 6 H + 2 = 1-84 (1-84) 



Methyl Acetate 3C+ 6 B + 2 = 1'79 (1-84) 



Benzene 6C+ 6 // = M5 (1*20) 



Ethyl Acetate 4 C + 8H + 20 = 2*03 (2*04) 



Ethyl Alcohol 2 C + 677+ = 0*83 (1-00) 



In each case the effect of hydrogen can he neglected, and 

 two equations suffice to obtain values of O and C, the 

 '* atomic " corpuscular radiations. 



The equations chosen are those for ether and methyl 

 formate on account of the much greater experimental 

 accuracy obtainable in these cases, due to the larger pres- 

 sures used, and consequent larger ionizations obtained. 



These equations give values : = 0*62, (7=0*20. 



On substituting these values in the other equations, the 

 values given in brackets are obtained, the agreement being 

 remarkably close in some cases, and well within the 

 limits of accuracy expected in all cases except that of 

 alcohol. 



The discrepancies in the cases of benzene and alcohol were 

 expected to be much greater than in any others, as the total 

 ionizations obtained were extremely small, being less than 

 3^ that of air. Of this, about 50 per cent, was due to 

 corpuscular radiation from the sides and from the hydrogen, 

 so that the corrections amounted to half the total ionization 

 observed. 



In order to obtain accurate results, the mean of ten 

 observations was taken in each case, but even with this 

 precaution, an error of 5 per cent, of the whole ionization 

 was reckoned possible, which is equivalent to a 10 per cent, 

 error on the result. 



The smallness of the ionizations in these two cases is 

 partly due to the low molecular weight of these compounds 

 (and to the absence of oxygen from the benzene molecule), 

 but is chiefly due to the low^ pressures used. In the case of 

 ethyl acetate, the low vapour-pressure at 0° 0. is made up 

 for to a great extent by the high molecular weight and the 

 large proportion of oxygen in the molecules, hence the 

 observations on ethyl acetate were much more accurate than 

 those on benzene and alcohol. 



The agreement obtained between the two values for 



