Atmospheric-Electric Observations, 1915-16 



387 



"■7 







fact that if we employ an electroscope which is to measure 200 volts on its scale, it obviously 

 can not show very much movement for a small alteration of potential of say 1 volt. 



In the apparatus which was devised to overcome the above difficulties^ the central 

 cylinder is connected to the fiber of a single-fiber electroscope which can be adjusted to any 

 convenient sensitivity. For land observations a sensitivity of about 20 divisions per volt is 

 convenient, but at sea a somewhat smaller sensitivity (about 5 to 10 divisions per volt) 

 is more desirable. The potential of the fiber is never allowed to depart far from zero poten- 

 tial, and the necessary field is obtained by insulating and charging the outer cylinder to 

 about 150 volts. On releasing the fiber from earth it, of course, starts to move, and the 

 rate of movement can be noted. The leakage correction becomes reduced to a very small 

 amount owing to the small departure of the system from zero potential during the experi- 

 ment. In order to prevent the charge on the outer cylinder from affecting the number of 

 ions coming to the apparatus, the cylinder is shielded by another cylinder separated from it 

 by a thin hard-rubber ring. This latter cylinder is earthed, and as it takes a charge equal 

 and opposite to that on the cylinder next to it, it to a great extent annuls the effect of that 

 cylinder. It by no means does so com- 

 pletely, however, even when the two cylin- ^ ^_ 



ders in question are separated by no more 

 than 1 mm. We can easily see why this 

 is so, for though it is difficult to estimate 

 the exact distribution of forces around the 

 mouths of the cylinders, we can easily see 

 that since a point such as F, Figure 23, 

 which is inside the charged cylinder, is at 

 — 150 volts, and a point such as D, which is 

 outside, is at about zero, the ions which get 

 inside will have to do so in opposition to a 

 field corresponding to a fall of potential of 

 150 volts in a comparatively short dis- 

 tance. If the velocity of the air-current 

 falls below a certain minimum value, ob- 

 viously no ions will get inside. For the 

 usual air-currents employed, however, the 

 effect is to diminish the number of ions 

 entering. This difficulty is one which 

 shows itself very materially in practice, as 

 appeared when experiments were made to test it; it is, however, completely overcome by 

 the device shown in Figure 2S, A. 



The figure shows the top end of the apparatus, B and C are the two outer cyUnders, and 

 E is the top of the central rod. E carries a collar attached to three wires which in turn 

 support a thin hollow cylinder G. A cap H fitting over the entire end of the apparatus is 

 fitted with a piece / which just goes inside the top of G, from which it is electrically sepa- 

 rated by an amber ring R. This ring is fixed at the top of G on the inside, and is prevented 

 from falling down by a little shoulder turned on the brass. By this arrangement G is kept 

 from shaking, and it will be seen that air at zero potential outside can now get well inside 

 without having to pass through a point different from zero. Any repulsive influence of 

 the cylinder B on the ions as they come near the lower end of G can only result in their 

 being turned sideways and caught by G, and this produces no error, since G is connected 

 electrically to E. 



A u 



Fig. 23.- 



C B 



Diagram lUuatrating Attachment to Upper End of 

 Central Rod of Ion-Counter. 



'See W. F. G. Swann. Terr. Mag., vol. 19, pp. 171-176, 1914. 



