212 Hackxert— The Secondary Radiation excited by y Rays. 
interest; but the particular physical expression we give to it depends on the 
view taken of the mechanism of the absorption of 8 rays, and the emission of 
secondary radiation. If we assume the y rays are ether-pulses, it is highly 
improbable that the B particles emitted acquire their velocities under the 
influence of the electric forces in the y rays. The energy of the y rays absorbed 
by the atom must be regarded as acting like a detonator. It upsets the stability 
of the atomic grouping, causing the expulsion of one or more B particles, carrying 
a part of the atomic energy with them. Since yp, is the fraction of the incident 
energy absorbed per unit volume, and p,«, the amount of energy emitted as 
B rays, x, 1s an index of the explosiveness of the atom in producing an expulsion 
of electrons by the absorption of a given amount of the energy from the incident 
y rays. From the reasoning given above, it would seem as if «, were constant 
for elements of moderate atomic weight, but had a marked increase for such 
elements as platinum, lead, and uranium. 
THe SECONDARY RADIATION PER ATOM. 
The expressions deduced for s, can be discussed from another standpoint which 
is of interest. The term f,e"7u,«, is the amount of secondary radiation excited 
per unit volume in the surface layer. If we write Rye“? Pkt _ ¢ where 
M is the atomic weight and p is the density, then o is proportional to the 
secondary radiation per atom when a substance is traversed by y rays. We 
have then 
gee Roe "v7 w,K, LB GE ley HD o/M (A) 
SW ewiam ip vle/lam) Orit aue 
Ry € "v7 pK o/M 
Go = OL tie) — Lie C) 
i Xp X3/p \ 
It has already been stated that we know i, and x, only for the B rays 
of radium. But if we use the values given by McClelland, we find that 
d,/p (1 + /1 = «,) 18 almost constant (Table II.), and the change in A,/p is not 
great compared with the change in atomic weight. 
If these results hold true, as they most probably do, for the B rays in the 
secondary radiation we are discussing, it follows that the denominator 
[(1 + /1 = kg) As/p or A,/p] in formula (A) or (C) must be nearly constant; and 
whichever formula we adopt, we conclude that o/M is approximately proportional 
to S,, the secondary radiation. 
