Radium Emanation +at Low Temperatures. 725 



marked its presence there by causing an increase of ionization. 

 In this way the ionization measurements indicated roughly 

 the amount of emanation in the upper part of the tube cor- 

 responding to the various temperatures, and thus gave a 

 means of following the progress of the volatilization. 



The condensing arrangement finally used is shown in fig. 1. 

 The free end of the glass tube which contained the emanation 

 was dipped into a very small glass bulb containing just 

 enough pentane to cover the end of the tube and the junctions 

 of the thermo-couples which were used for determining the 

 temperatures. Surrounding this small bulb was an outer 

 bath of pentane, which was itself surrounded by a bath of 

 liquid air contained in a 6-inch, silvered, Dewar cylinder. 

 The Dewar cylinder was kept filled to the top with liquid air 

 until it was desired to allow the temperature to rise. The 

 liquid air was then allowed to evaporate slowly, and the 

 gradual lowering of the level caused a slow variation of 

 temperature at the point where the emanation was condensed. 

 The temperature usually rose at a rate of C, 5 C. per minute. 

 Without the double bath arrangement it was found that the 

 rate of temperature rise was not sufficiently uniform. 



Some trial experiments with moderate quantities of emana- 

 tion showed, after the baths were removed, the presence of a 

 bright point of light at the extreme end of the glass tube, 

 and a uniform fluorescence over the rest. This bright spot 

 was due to the active matter which had been deposited by 

 the condensed emanation. Its concentration at this one 

 point showed that the emanation had condensed not over any 

 considerable area but at the very tip of the tube. Conse- 

 quently the junctions of the thermo-j unction were placed in 

 the inner pentane bath exactly at the tip of the glass tube 

 which contained the emanation. The thermo-couple was a 

 double, copper-constantan element of number 30 double-cotton 

 covered wires. The warmer junctions were maintained at 

 the temperature of! melting ice. 



It has already been mentioned that there was a difficulty 

 in obtaining saturation in the ionization chamber. Since the 

 ionization-temperature curves afterwards shown could not be 

 used to determine the actual amount of emanation volatilized 

 at a given temperature, complete saturation was not essential; 

 nevertheless, in all experiments saturation was approximately 

 attained. For this purpose the ionization vessels employed 

 were made very small. The one mostly used was a brass 

 cylinder, 1 cm. in diameter and 5 cms. long, fitted with the 

 usual central electrode and ebonite insulation. 



Since the quantities of emanation employed in different 



