Solubility of Radium Emanation. 847 



Precaution* and Errors. 



A few preliminary experiments showed that the solubility 

 varied with temperature more markedly than former results 

 led one to suppose. They also showed that equilibrium of 

 the emanation between the water and gas-phases was fairly 

 quickly established. If the admixture of gas and liquid by 

 shaking is effective, as it must be with bulbs of the shape 

 shown, it is necessary to wait only a few minutes after the 

 absorbent has taken up the temperature of the bath. Jn one 

 series of experiments shaking was carried on ior 8 and 

 for 24 minutes ; the same results were obtained for the 

 coefficient as when the mixing bulb was left in the bath for 

 18 hours. 



In order to secure good 7-ray observations the mixing and 

 sampling bulbs should not be too large. The mixing bulbs 

 used in the experiments were of volumes 3722, 33'18, and 

 26*01 cc; the sampling bulbs of 25*77,24 98,18*87, and 

 10*97 c.c. respectively. The length of the bulbs was 8 cm. 

 Any one of the mixing bulbs could be used with any one of 

 the sampling bulbs as experiment required. 



Jn taking the 7-ray measurement four observations of 

 ionization were taken for each bulb viz. when the stopcock 

 was vertically above and also vertically below the bulb, and 

 the same when the bulb was reversed. In laying aside the 

 bulb to wait for radioactive equilibrium they were always 

 placed so that the stopcocks hung vertically below the bulb ; 

 this ensured a symmetrical distribution of the active deposit. 

 If this precaution is not attended to, but a bulb containing 

 liquid is left lying on its side, there will be less Ha C on the 

 part of the wall in contact with the liquid than on the oppo- 

 site part in contact with the gas. "When the 7-ray observa- 

 tions are taken this difference will cause, if the distance from 

 the electroscope is small, a discrepancy in ionization readings 

 on reversal of the bulb. 



There is always a slight error which is caused by the 

 small absorption of 7 rays in the liquid contained in a bulb : 

 but the error is very small, and in these experiments is 

 certainly less than half of one per cent. The 7-ray readings 

 varied a little less than one per cent, between extremes, and 

 they were always corrected, so that the emanation contents 

 of both mixing and sampling bulbs referred to the same 

 time. 



The experiments are liable to an error due to the suspension 

 of small bubbles of gas in the liquid. This error affects the 

 measurements of solubility of any gas in any liquid, and to a 

 certain extent cannot be avoided. The lubricant used on the 



