The magnetic deflection experiments of Curtiss and of Mott-Smith

Abstract

We have carried out a geometrical analysis which applies to that type of cosmic ray deflection experiment in which an extended region of magnetic field lies between the lower two of three or more Geiger-Mueller counters arranged in line for counting coincident discharges. It is apparent that charged particles of sufficiently low energy will be deviated away from the last counter and that the counting rate will therefore be reduced by application of the field. This paper presents a method for computing what proportion of electrons of a given energy will be eliminated by application of the field. The method is extended to include a constant loss of energy per unit path length as the particle traverses an iron core. The results of the experiments of Curtiss and of Mott-Smith are interpreted in the light of these calculations. Curtiss' result could be explained if all rays were electrons of 4 × 10 8 volts energy. One may conclude from his observation that if all the radiation consists of positive or negative electrons, the fraction between 2 × 10 7 volts and 1.5 × 10 8 volts is less than 30 per cent. of the total above 2 × 10 7, and that the fraction between 2 × 10 7 and 10 9 is greater than 30 per cent. A comparison of Curtiss' and Mott-Smith's experiments yields no reason for doubting the correctness of using the magnetic induction B when computing the deflection within iron. Nor does it necessarily sustain this procedure. An experiment by W. F. G. Swann and the writer is at present under way for the purpose of securing, if possible, more conclusive evidence on this point.