Vol. 8. 1922 
PHYSICS: C. BARUS 
65 
If the case is warmer, the needle meets repulsive agencies toward the 
center from the excess of radiation inward. If the masses m at the end of 
the needle are the warmer, the needle radiates toward the case outward 
and therefore experiences an equivalent repulsive reaction. Finally 
with a non-uniform radiant field of force, increasing from needle to case, 
the effect will be virtually an increase of the torsion coefficient, of the quartz 
fibre, so that T comes out smaller. 
It would then follow that the large deflections are obtained when the 
radiant field is nearly uniform and that the larger values of the period 
T are the more nearly correct. This finally is in keeping with the 7- values 
thus obtained. 
Finally the deflections are modified and complicated by the screening 
effect of the presence of the masses M, which in the T experiments are 
absent. 
Estimate of Radiation Temperature-differences. — Let us suppose that 
the radiant repulsion from case to needle, the elastic buffer effect explained, 
can be expressed as an increment a^. of the modulus of torsion a, of the needle. 
Furthermore, if is the radiant force at each end and Ayr/4:L half the 
double deflection in radians, we get a second value of a^.. If we equate 
the two values 
/, = T'{ml/L)AT-\Ay, (1) 
To find Al/T^, the least and maximimi T found in vacuo in the last para- 
graph, i.e., = 5 minutes and T = 6.5 minutes, may be inserted, so 
that the modulus in (1) is 7.0 X 10 ~^ nearly. 
To determine the radiation forces in vacuo, it will be necessary to 
postulate the case of black body, or full radiation. If 6 and do are the 
absolute temperatures between which radiation takes place and E the en- 
ergy radiated (ergs per cm.^ per sec.) 
E = 4:Kd^A6 
The value d may be taken as 300° and K = 5.3 X lO"'*. Thus 
E = 103X5.7A(9. 
This radiation may at the outset be supposed to reach the needle from 
a single cm.^S, 5', of each plate situated on the end normal of the needle 
and plate. Thus eventually, po = E/27rc {c being the velocity of light) 
is the energy density or the light pressure at a distance of 1 cm. from the 
element 5 or 5'. If we suppose (as is nearly the case) the shot m of frontal 
area A to be situated there, the force fr upon it will be poA, where A = .16 
cm. 2 Hence 
= 10-9 X 4.8 (2) 
for the square centimeter of plate taken. 
If now we equate the values of in (1) and (2) 
= 150 A^', nearly, (3) 
