in the Atmosphere near the Earth's Surface. 727 



when heated, gave emanation measuring '22 division a 

 minute, of which *11 was due to the natural leak and the 

 emanation in the air drawn through. Thus, 0"95 cb.m. of 

 air contained, the emanation from *025 x 10 _10 /'ll gram of 

 radium, or 1 cubic metre from about 24 x 10 -12 gram of 

 radium. My result by the active deposit method was 3-J- times 

 as great as this. 



It will be seen that the results in the above table for a 

 "run" were always greater than for a "rest," proving the 

 existence of the emanation in the atmosphere ; but the differ- 

 ence between them on which the result depends is about "025 — 

 a small quantity, about half the magnitude of the. natural leak. 

 It was, therefore, desirable to work on a larger scale. 



Three iron pipes were prepared, 37 cms. long and 4 cms. in 

 diameter. Each contained 220 grams of finely divided 

 charcoal, Air from outside the building was drawn by a 



Fig. 1. 

 CoMb 



water-pump through H 2 S0 4 and CaCl 2 , and was passed 

 through the three iron tubes arranged in parallel. By this 

 means the speed could be increased threefold, and yet the 

 volume and cross-section of the absorbing charcoal was 

 sufficiently large for good results. A convenient speed was 

 found to be *09 cb.m. per hour, about 8 c.c. per second 

 through each tube, or 8 cb.m. altogether in 3*7 days. 



When the current of air had been flowing for a definite 

 time, usually 3'7 days, the iron cylinders were removed and 

 heated in succession to a dull red heat over a combustion 

 furnace, and the expelled gases were collected in large vessels 

 over water. 



The 660 grams of carbon gave off 15,000 c.c. of gases, and 

 these were drawn in 18,000 seconds through three small 

 glass tubes in series, each containing 7 grams of charcoal. 

 The small tubes were heated over a Bunsen burner, and the 

 gases driven off were collected over water and introduced 

 into the electroscope. Two or three hours later, the rate of 

 leak of the gold leaf was determined. A complete observation 

 occupied about nine hours. 



3 C 2 



