CONTEMPORARY ADVANCES IN PHYSICS 203 



subjacent points correspond to upward-going corpuscles, and transfer 

 them across the horizontal axis. Then, the sprinkhng of points 

 extends all the way from axis to slanting line; and this is the sign of 

 fluctuations such as Anderson from the start had observed. Notice 

 however that the representative points are of four aspects: solid dots 

 and hollow circles, with or without downward-pointing barbs. The 

 dots refer to tracks which were seen in the chamber singly; the circles, 

 to particles which "entered the chamber accompanied by other par- 

 ticles." The lonely particles are prevailingly able to pass through 

 matter without suffering energy-losses nearly so great as those which 

 the others incur! Thus by itself and without any theory. Fig. 8 

 establishes a distinction between the singly-appearing corpuscles on 

 the one hand, and those which appear in company on the other. 

 Moreover the barbs are often attached to the hollow circles, bearing 

 out the inference from Figs. 5 and 7 that shower particles are likely 

 to be shower-producing particles; but rarely are they attached to 

 solid dots, never to those which lie far off from the slanting line. 



(This seems the best place for mention of the similar work now 

 being done in England by Blackett and (J. G.) Wilson, in France by 

 Ehrenfest. The Englishmen have set plates of gold, lead, copper and 

 aluminium, of various thicknesses from 3.?) mm to 2 cm, into the 

 middle of an expansion-chamber in Anderson's fashion; Ehrenfest, 

 using a pair of cloud-chambers one over the other, was able to put 

 between them a block of gold no less than 9 cm thick! Their way of 

 reducing their data for plotting is not the same as that employed at 

 Pasadena, and their diagrams therefore look very different ^ from 

 Fig. 8. Their energy-range runs much further upward, as far as 

 5000 Mev, and the great majority of the particles which they plot 

 lie beyond the limit of Fig. 8. Many of Ehrenfest's particles got 

 through the great thickness of gold without losing anywhere nearly 

 the whole of their energy, and are therefore to be classed as much 

 more penetrating than electrons should be. So did nearly all of the 

 particles of energy greater than 250 Mev observed in England, but 

 there were a few of these which lost most of their energy in 0.33 cm 

 of lead, and of these few about half seemed to belong to showers. 



^ For the benefit of those who may consult the original papers, I give the differ- 

 ence. Let £1 and £2 stand for the (quasi) energy-values of a particle before and 

 after passing through a thickness d of metal; A£ for (£1 — £2); x for \iE\ + £2). 

 What is plotted by Anderson and Neddermeyer (Figure 8) is a£/^ as ordinate and 

 £1 as abscissa. Blackett (in all his papers but the earliest), Wilson and Ehrenfest 

 begin by subtracting from a£ a quantity sd which is supposed to be the amount of 

 energy spent by the particle in detaching electrons from atoms while traversing the 

 metal (Blackett assigns the value 15 Mev/cm to 5 in lead, Ehrenfest takes 28 for gold) ; 

 they then plot (A£ — sd)jxd as ordinate and x as abscissa. Their ordinate (denoted 

 by them as R) is then more nearly ready for comparison with theory. 



