202 BELL SYSTEM TECHNICAL JOURNAL 



concavity toward the left or toward the right of an observer looking 

 into the chamber from the north-seeking pole of his magnet. Without 

 the plate, neither sense nor sign would be knowable except in the 

 rarest of cases.' Anderson in August 1932 found on one of his photo- 

 graphs the track of a particle which by this criterion was positive, 

 and which by the density of droplets along its track (we take up this 

 topic later) he identified as an electron. He thus became the discoverer 

 of the positive electron. 



Concentrating on the measuring of A£ after the excitement of the 

 positive electron had subsided, Anderson presently found that its 

 values are very fluctuating. Thus in 1934 he published the details 

 of nine traversals, made by particles assumed to be electrons, through 

 thicknesses of lead from 7 to 15 mm. (Even with a single metal plate 

 the effective thickness varies, since corpuscles traverse the plate with 

 varying degrees of obliqueness.) These were by no means identical 

 in initial energy, this ranging from 38 to 240 Mev; nevertheless one 

 might have expected the energy-loss per unit length of path in lead to 

 be about the same for all, and yet the nine values thereof were scattered 

 all the way from 18 to 120 Mev/cm! Such fluctuations suggest that 

 the energy is lost in great amounts at a few events, and not in driblets 

 at many. They did not deter Anderson and Neddermeyer from 

 making such measurements on hundreds of later particles, classifying 

 the particles into groups according to their energy-values, and 

 averaging the energy-losses within each group. What then was 

 found has a bearing upon the problem; but we pass over it for the 

 time being, and consider in Fig. 8 the record of ninety-four particles 

 which, during a later experiment, passed through a plate of platinum 

 one centimetre thick.'^ 



Plotted horizontally are the energy-values of the particles while 

 above the plate, vertically the energy-changes divided by the lengths 

 of path in the platinum. The axis of abscissae is the locus of energy - 

 losses imperceptibly small; the line slanting at 45° is the locus of 

 energy-losses which are total, the particles shown on this line having 

 been stopped by the plate. The fact that some of the representative 

 points lie below the horizontal axis means only that for every particle 

 the observers subtracted its energy below the plate from its energy 

 above, irrespective of its direction of motion. Suppose that these 



^ One might be misled by the adjective "cosmic" into believing that all cosmic- 

 ray particles come from above, their sense of motion making an angle of less than 

 90° with the downward-pointing vertical. Many, however, including Anderson's 

 first positive electron, have been found by this criterion to be moving upward (i.e. at 

 more than 90° to the downward-pointing vertical). The showers of Figs. 6 and 7 

 show that this is not a forced interpretation. 



1 1 am indebted to Dr. Anderson for a plate exhibiting data thus far unpublished. 



