CONTEMPORARY ADVANCES IN PHYSICS 



301 



the classical method of the expansion-chamber (Figs. 1 and 2). 

 The particles may be tested for their charge by having a magnetic field 

 pervading the chamber. Some of the paths are then found to be 

 smoothly curved, proving beyond a doubt that the corpuscles are 



charged.^ 



The sign of the curvature of a path in a magnetic field should 

 disclose the sign of the charge of the responsible corpuscle; but here 



.^■. 







Pig 1— Two photographs (taken from different viewpoints) of a nuclear explosion, 

 probably that of a copper nucleus struck by a cosmic ray. The tracks on the right, 

 and concave to the right, are those of positive electrons; others are due to negative 

 electrons. (P. M. S. Blackett; Proceedings of the Royal Society.) 



appears a difficulty: the sign cannot be inferred unless the sense in 

 which the corpuscle described the path be known, and there is nothing 

 whatever about the aspect of an ordinary trail to indicate that sense. 

 It might be guessed that the particle is necessarily moving downward 

 rather than upward, since the cosmic rays come from above. This, 

 however, would be a bad guess, for some at least among the trail- 

 making corpuscles are secondaries set into motion by the primary 

 rays, as protons are known to be impelled by neutrons, and electrons 

 by photons; and some of these secondaries may be, and indeed certainly 

 3 Other paths seem quite straight, but there is strong reason to believe that a 

 neutral particle would not produce anywhere nearly so great a density of lon-pairs 

 as is observed along them, and it is inferred that they are due to charged particles 

 which are moving with too much momentum to be sensibly deflected. 



