224 Mr. W. Wilson on the Effect of Pressure on 



At the higher pressures, such as those above 20 atmospheres, 

 the gold-leaf did not come to rest for some time after filling. 

 At 30 atmospheres it moved irregularly to and fro, even 

 after 8 hours. In such cases, if the leaf were not moving 

 too irregularly, its position on the scale was read every 

 quarter minute for two or three minutes and the mean was 

 taken as the reading at the middle of this period. 



If we reduce the 7-ray curve to the same dimensions as 

 the natural ionization curve, as shown in the curve dotted in 

 G.g. 4, we see that the two curves are quite dissimilar, and 

 therefore the natural ionization in the vessel cannot be wholly 

 due to external 7 rays. The natural leak, however, was 

 found to be about 40 per cent, lower when in the pressure- 

 chamber than when out, showing that the ionization in an 

 unscreened vessel is partly due to this cause. 



The number of ions produced per c.c. per second was 

 deduced as follows : — Eve has shown that the number of 

 ions produced per c.c. per second in an aluminium vessel by 

 the y rays from 14*3 m.g. radium bromide placed at a 

 distance of 303 cms. is 297. In the present case 3*7 m.g. at 

 a distance of 67 cms. produced an ionization of 5*50 scale- 

 divisions per minute. The natural ionization at one atmo- 

 sphere on this scale is *046 division per minute. 



This corresponds then to a number of ions 



•046 297 x 3-7 x (303) 2 1Q1 



-^-z~^r . — ttt> — in-w =lo'l per c.c. per second. 



5-50 14-3 x (6 () 2 r F 



This agrees very well with the corrected value of the 

 number found by Cooke, who obtained 9*9 to 13*8, and 

 confirms the test of the insulation. 



The curve obtained if the ionization were due to the 

 admixture of an emanation with the air, such as that of 

 radium which emits rays of low penetrating power, would 

 be a straight line passing through the origin within the 

 limits of pressure examined ; while if the walls of the vessel 

 emit rays of low penetrating power, either intrinsic or due 

 to radioactive impurities, the ionization should not vary with 

 the pressure. 



The experimental curve can be made up of two parts : — 



(1) A constant part due to radiations from the sides of 

 the vessel. 



(2) A part which increases proportionally with the pressure, 

 due for the most part to an emanation mixed with the gas, 

 but in some part very likely also to penetrating radiation. 

 The latter part must be very small, however, or the curve 

 would be concave towards the pressure axis. The emanation, 



