656 B1RKELAND. THE NORWEGIAN AURORA POLARIS EXPEDITION, 19021903. 



which may serve as receivers and resonators of luminous waves coming from the sun, and that here too 

 it is quite possible that the number of electrons of dispersion is proportional to the intensity of the 

 electric current emanating from Saturn in the manner admitted by us. 



An electric radiation from Saturn such as that here assumed may certainly also be imagined to be 

 accompanied secondarily by an ejection of tiny material particles resembling what CROOKES has called 

 electric evaporation or volatilisation from a cathode. 



A metallic cathode is so disintegrated during discharges that the material may be deposited in the 

 form of a reflecting layer upon the neighbouring glass wall. 



Different metals disintegrate in very different degrees when they form a cathode, circumstances 

 being equal. 



In arbitrary units, CROOKES ( J ) gives the loss of weight by disintegration in cold cathodes as follows: 



Pd Au Ag Pb Sn Pt Cu Cd Ni Ir Fe Al Mg 



108 100 83 75 57 45 40 32 ii 10 6 o o 



For incandescent cathodes it is a different matter altogether; the disintegration is then much gr ( . 

 Under the influence of magnetic forces too, there is a great difference in the amount of the disintegration. 

 I have, for instance, shown that in such a case even a cathode of aluminium can in a short time tl; 

 off a reflecting deposit upon an adjacent glass wall (C. R., Feb. 21, 1898). The disintegration from 

 a carbon cathode is very great. From one such, in a large exhausted vacuum-tube, I have seen half a 

 gramme of matter thrown off in a few minutes and deposited firmly on the glass wall of the tube. 



The cause of this cathodic disintegration has not yet been clearly determined. It is possibly to 

 some extent a kind of evaporation by which the disintegration is brought about, by the high temperature 

 that the rapid positive ions (channel rays) produce where they strike the surface of the cathode. The 

 dependence of the disintegration on the strength of the current and of the cathode-fall is in accordance 

 with this explanation; for with the same duration of current, the cathode's loss of weight by disintegration 

 is proportional to the product of the strength of the current and the potential-fall from the cathode. But 

 this product equals the electric work performed upon the positive ions between the negative column <>t 

 light and the surface of the cathode, that is to say, proportional to the kinetic energy carried by th- 

 positive ions in the time-unit to the cathode. 



This phenomenon of disintegration seems to offer a very important field for future investigation, 

 for an accurate knowledge of these things is of fundamental importance for the theories here propounded. 



HOLBORN and AUSTIN ( 2 ) have made some very interesting experiments on the amount of disinte- 

 gration of cathodes of different metals under similar electrical conditions. When the tube used was 

 filled with air, they found that y, the loss of weight in 30 minutes from circular cathodes i cm. in 

 diameter could for platinum, silver (one sample), copper and nickel, be represented by the formula 



^ 

 y = 0.0016 (V 495) 



for silver (another sample), bismuth, palladium, antimony and rhodium, the relation was 



y = 0.0018 (V 495) 



V is the cathode fall of potential in volts, A the atomic weight of the metal, and n its valency. Other 

 metals such as iron, aluminium and magnesium, do not follow either of these laws. For those metals 

 which follow the laws (i) or (2) we see that with the same current and cathode fall, the weight of 



f 1 ) See WINKELMANN, Handbuch der Physik, 2. Aufl., b. 4, p. 629. 



( 2 ) See J. J. THOMSON, Conduction of Electricity through Gases, p. 549. 



