PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. VI. 



717 



iat the particles expelled from the cathode must now be assumed to move more or less at right angles 

 > the magnetic lines of force, and that the field-strength was about 1800 lines of force per square centi- 

 ictre, it was not possible to prove any turning aside of the particles, first of all because the field was 

 ot strong enough, but also because it was not possible in these preliminary experiments to obtain sharp 

 mdows of the screen in the metallic deposit upon the glass wall. 



On the other hand it appeared that the deposit came not only on the front of the screen, but also 

 nmdantly on the back of it, especially if it stood near the cathode, a fact which indicated that the par- 

 ries could acquire a retrograde motion after they had retired from the cathode. 



From the appearance of the deposit upon the back of the mica screen farthest removed from the 

 Lthode, I received the distinct impression that the particles had struck almost parallel with the surface 

 the screen. I therefore made, on each side of the long side of the mica screen, a raised edge one or 

 millimetres in height. Both edges turned away from the cathode. It then appeared that there was 

 ) longer any deposit upon the back of the screen, although on the front and on the protecting side 

 Iges there was an abundant deposit of palladium. 



a Fig. 267. b 



The next arrangement was as follows. A long, rectangular cathode of palladium was attached to a 

 tick brass wire that passed through a quartz tube with walls 2 millimetres in thickness. Only the palla- 

 cam plate reached beyond the quartz tube, which was placed axially in the vacuum-tube. The anode 

 MS annular in shape, and was placed 10 cm. behind the cathode, which again was only 2 mm. above 

 t: sheet of plate-glass that was cemented to the end of the vacuum-tube. In order to prevent the 

 ucking of the sheet of plate glass with the heat from the cathode-rays, a small square of mica was 

 cmented to the sheet of glass just under the cathode. 



There were further cemented to the sheet of glass some half-cylinders at various distances, with 

 t.'ir convex side towards the cathode. 



By these means it was shown that the palladium particles to a very great extent made their way 

 i o the concave side of the half-cylinder, if it was placed near the cathode, whereas if it was far from 

 tt cathode, the particles hardly entered it at all, although they abundantly covered the convex side with 

 flladium. Figs. 267 a and b show how these little half-cylinders were arranged, and also that the most 

 c.tant half-cylinder has cast an almost straight shadow behind it on the sheet of glass (farthest to the 

 rht in fig. 267 b), where therefore the palladium has not been deposited. The nearest half-cylinder has 

 t -own no distinct shadow, or at any rate it is only by careful examination that there is seen to be less 

 r.lladium deposited just behind it than beside it. It seemed as though some of the expelled palladium 

 Articles had a tendency to return to the cathode again, just as if they were positively charged. 



In a subsequent experiment, the cathode was a fairly thick platinum plate with a surface of a few 

 saare millimetres, while the anode was a brass plate with a surface measuring a couple of hundred 



Birkeland. The Norwegian Aurora Polaris Expedition, 1908 1903. 91 



