Apeil 10, 1908] 



SCIENCE 



563 



Discharge between Metal Electrodes for 

 Potentials Less than Ionising Potentials. — 

 Hobbs has shown that discharge produced 

 by potentials less than an ionizing potential 

 depends on the character of the metal elec- 

 trode. Such discharge can be secured 

 through greater distances if Al electrodes 

 are used than if Pt were employed. 



Using electrodes of Pt, Al and Ag, in the 

 various permutations and plotting P.D. 

 and distance, I have found that discharge 

 is not determined by the character of the 

 negative electrode, but rather by the elec- 

 trode having the lower potential gradient, 

 irrespective of its polarity. 



This indicates that the discharge under 

 these conditions is not carried entirely by 

 negative corpuscles. 



The Influence of Humidity upon the Elec- 

 trical Discharge from Points in Air: 

 John Zeleny, University of Minnesota. 

 A study has been made of the discharge 

 from a point in air of various degrees of 

 humidity. A complete potential-current 

 curve was attained for each case, and from 

 the results curves were drawn showing how 

 the potential required to produce a given 

 current varies with the humidity of the 

 air. The potential required to start the 

 positive discharge was found to diminish 

 slightly (about 3 per cent.) as the humidity 

 increased from to 100 per cent. The 

 potential required to produce a positive 

 current (25 X 10"' amp. to 100 X 10"^) in- 

 creased about 6 per cent, for the same 

 change in humidity. The potential re- 

 quired to produce a negative current also 

 increased with the humidity of the air, the 

 total increase being about 16 per cent, for 

 the whole range. 



Radioactivity of a Smoke-laden Atmos- 

 phere: S. J. Allen, University of Cin- 

 cinnati. 

 This paper gives an account of a series 



of observations on the rate of decay of the 



radioactivity of the air at Cincinnati, dur- 

 ing the last eight months. The atmosphere 

 at Cincinnati is very smoky and it was 

 therefore thought interesting to observe 

 what effect the smoke particles might have 

 on the rate of decay. 



An endless copper wire running over two 

 insulated wooden pulleys was suspended 

 about 20 feet from the ground, and could 

 be charged to a high potential by means 

 of a large static machine. The length be- 

 tween the pulleys was about 180 feet, thus 

 giving 360 feet of wire for the activity to 

 be deposited on. 



The active substance deposited on this 

 wire was removed and tested for its activity 

 in the usual manner. In some of the ob- 

 servations the rate of leak method was used 

 and in others the activity was measured 

 by means of the new null reading instru- 

 ment devised by the author and described 

 in an accompanying paper. 



The active substance was collected on the 

 wire under three different conditions: 

 (1) The wire was charged negative, (2) 

 the wire was charged positive, (3) the 

 wire was uncharged. 



About fifty decay curves in all were ob- 

 tained. 



The main facts brought out in this in- 

 vestigation may be summarized as follows: 



1. "When the wire was charged negatively 

 a very large amount of activity could be 

 obtained. When charged positively only a 

 small amount. When uncharged the 

 amount obtained was quite considerable. 



2. The decay curves obtained differed 

 widely, the period of decay (time taken to 

 fall to half value) varying from 30 to 80 

 minutes. 



3. Only two simple exponential curves 

 could be obtained which extended over any 

 appreciable length of time. One gave a 

 period of about 38 minutes and the other 

 a period of 30 minutes. About one third 



