20 



NA TURE 



[May 6, 1897 



Prof. Asaph Hall was elected vice-president ; Prof. Ira 

 Remsen, home secretary ; and Prof. A. Graham Bell, treasurer. 

 New members elected were Messrs. Wm. H. Dall (of Washing- 

 ton) ; Frank A. Gooch (of Vale) ; Chas. S. Minot (of Boston) ; 

 and E. W. Morley (of Cleveland). 



The autumn meeting of the Academy will be held at Boston 

 on November 16 next. 



CONTINUA TION OF EXPERIMENTS ON 

 ELECTRIC PROPERTIES OF URANIUM} 



IN a paper read before the Society on March i, we had the 

 honour to communicate some preliminary results on the 

 electric properties of uranium. We propose now to give other 

 results on the same subject, bearing on the conductance induced 

 in air by uranium. 



To measure the leakage in air at ordinary pressure at different 

 voltages, we used in our first experiments the two-Leydens method 

 described in a former paper. We found that the leakage was not 

 proportional to the electro-motive force. It was not perceptibly 

 increased when the uranium was heated, or when the sunlight 

 fell on it. 



We also observed the leakage in hydrogen, oxygen, and 

 carbonic acid. The experimental arrangements necessary for 

 this are described in a paper published by the Royal Society of 

 Edinburgh. We found that the rate of leakage is greater in 

 oxygen than in air. The ratio of the rates depends on the 

 voltage chosen. The leakage in hydrogen is less than in air. 

 In carbonic acid it is less for four volts per two cms., but 

 greater for ninety volts per two cms. than it is in air ; for the 

 latter voltage the leakage in carbonic acid is greater even than 

 the corresponding leakage for oxygen at ordinary pressure. 

 We also made experiments with air, hydrogen, oxygen, and 

 carbonic acid at different atmospheric pressures. We found 

 that the leakage in air at pressures ranging from 760 mms. to 

 23 mms. was very nearly proportional to the atmospheric 

 pressure. The rate of leakage for lower pressures was so slow as 

 to make the results not very trustworthy. At pressures under 

 2 cms. no appreciable leakage with 4 or with 90 volts per two cms. 

 was observed. With hydrogen, oxygen, and carbonic acid the 

 rate of leakage at higher pressures was somewhat approximately 

 proportional to the pressure, at lower ones to the square root of 

 the pressure. 



We found that at ordinary atmospheric pressure, sparking 

 took place in air at 4800 volts, between a rough fragment of 

 uranium and a metal tube around it, connected to the two elec- 

 trodes of a vacuum-tube within which they were fixed. At 232 

 mms. pressure, the potential necessary to produce a spark fell 

 to between 1500 and 2000 volts. At 127 mms. it had lallen to 

 between iioo and 1300 volts. At 54 mms. it was 700 volts ; at 

 7 mms. 420 volts ; at 2 mms.- about 400 volts. At i/iooo mm. 

 the voltage necessary to produce sparking rose again to 

 2000 volts. 



To measure the potential difference between two mutually 

 insulated metals when the air between them is rendered con- 

 ductive by the presence of uranium, we used two methods, which 

 are described more particularly in the paper above referred to. 

 The steady reading obtained when the quadrants of an electro- 

 meter were in metallic connection we shall call the metallic- 

 zero. The deviation fromthe metallic-zero, when the quadrants 

 were insulated, to a steady 'point — the uranium-conductance-zero, 

 as we shall call it — depended on the volta difference between 

 the two opposed surfaces of metals, more or less tarnished as 

 they generally were. This deviation took place gradually in 

 about half a minute with one arrangement of apparatus, and in 

 about four minutes with a second arrangement. On the other 

 hand, if the insulated metal had a charge given to it of such an 

 amount as to cause the electrometer reading to deviate from the 

 metallic zero beyond the uranium-conductance-zero, the reading 

 quickly fell to this conductance-zero, and there remained 

 steady. 



The following table gives the potential differences between 

 the electrometer wires, when one of them is connected with 

 uranium, and the other with a plate of one or other of the 

 named metals opposed to it : — 



1 By Lord Kelvin, Dr. J. Carruthers Beattie, and Dr. M. S. de S.-nolan, 

 Read before the Royal Society of Edinburgh, April 4. 



NO. 1436, VOL. 56] 



Metal. Volt. 

 Polished aluminium (i) immediately after being 



polished .. -i'i3 



Polished aluminium (i) next day ... ... ... - 0*90 



Polished aluminium (2) -I'oo 



Amalgamated zinc ... .. ... ... ... - o'8o 



Polished zinc -071 



Unpolished zinc -o'55 



Polished lead -0*54 



Tinfoil -049 



Unpolished alumininum (i) ... ... ... -o'4i 



Polished copper - 0"I7 



Silver coin ... ... ... ... ... ... -f0'05 



Unpolished copper -f 0*07 



Carbon ... ... ... ... ... .. -f0'20 



Oxidised copper (a) -fO"42 



Oxidised copper (3) ... ... ... ... ... -f0'90 



It will be noticed that the difference of potential observed 

 depends very much on the state of polish of the metal con- 

 cerned. With a third specimen of oxidised copper a potential 

 difference of -l-0"35 of a volt was obtained. This specimen 

 was afterwards connected to sheaths ; a piece of polished 

 aluminium was placed opposite it, and connected to the insulated 

 terminal of the electrometer. The uranium disc, insulated on 

 paraffin, was then placed between them, and the deviation 

 observed was equivalent to a potential difference of - i "53 volts ; 

 that is, we obtained an effect equivalent to the sum of the effects 

 we had when the metals were separately insulated in air opposite 

 to uranium. 



We observed also the effect of various .screens on the rate of 

 reaching the conductance-zero. For e.xample, when a sheet of 

 lead about 2 mms. in thickness was used as screen, no deviation 

 from the metallic-zero was obtained. In other words, lead 



2 mms, thick is not transparent to the uranium influence. Glass 



3 mms. thick did not entirely stop the deviation ; it reduced the 

 deviation in the first minute, however, to ^ of the amount 

 obtained with no screen. A copper screen, 0*24 mm. in thick- 

 ness, reduced the rate to 3 ; two copper screens, total thickness 

 o'48 mm., reduced it to ^^t; ; three copper screens, 072 mm., 

 reduced it to ^^. A mica screen did not reduce the rate at all. 

 A zinc screen, 0*235 ^^- thick, reduced it to i. Two zinc 

 screens, total thickness 0*47 mm., reduced it to -f. Paraffin, 

 3 mms. thick, when placed between the two mutually insulated 

 metals, stopped the deviation from the metallic to the con- 

 ductance-zero. 



The final difference of potential observed between the electro- 

 meter wires connected to two mutually insulated metals, when 

 the air between them was made conductive by uranium, was 

 found to be independent of the distance between the metals 

 through distances ranging from less than ij cm. to 8 cms. 



The difference of potential observed when two mutually 

 insulated metals were brought into electric connection with 

 one another by a drop of water, was in the same direction as 

 the uranium conductance-zero between the two surfaces when 

 dry, and was smaller in magnitude. On the other hand, when 

 the uranium surface was covered with water to the depth of 

 about a millimetre, and an air space left above the water, 

 between the submerged uranium surface and the opposed 

 insulated metal, so that we had uranium-water-air-metal, the 

 rate of deviation from the metallic-zero was reduced so much 

 as to be scarcely observable. 



We found that the uranium-conductance-zero between zinc 

 and uranium was the same in air, hydrogen, and oxygen. And 

 that the final steady reading did not depend on the atmospheric 

 pressure, though the rate at which this steady reading was 

 reached did largely depend on the atmospheric pressure. 



UNIVERSITY AND EDUCATIONAL 

 INTELLIGENCE. 



Cambridge. — Dr. Nansen has made a contribution of ^^50 

 towards the teaching of Geography in the University. 



The voting of the Senate on the resolutions respecting degrees 

 for women will take place from i to 3 p.m. on Friday, May 21, 

 in the Senate House. 



The University of Madras is to be added to the list of 

 Indian Universities which are affiliated to the University ot 

 Cambridge. 



