284 
PAUL BIEFELD 
ture depending on the field strength in the region of the path. 
Again a ray of light or electromagnetic radiation possesses 
energy and hence the equivalent of 'mass/ consequently it is acted 
upon by a gravitational field, the amount of bending depending 
on the strength of the field from point to point of the path, and 
varying besides inversely with the distance from the centre of 
the Sun. At the surface of the Sun the deflection is predicted 
to be 1".75 
Why there should be a deflection due to the gravitational field 
of the Sun may be seen by referring back to the principle of 
equivalence in connection with the hypothetical box-=experiment. 
Let us imagine the man to one side of the box firing a bullet 
percenticular to the vertical wall of the box. The bullet will pass 
through the box and out on the other side. In the space out side 
it will fly in a straight line being in a field free of force; inside 
the box, however, it will be deflected downward. Interpreted by 
the man in the box as due tO' a gravitational field and by the man 
outside due to the upward acceleration of the box. In the case 
of a ray coming from a star the ray takes the place of the mass 
of the flying bullet, by virtue of energy of radiation having the 
equivalent of mass ; and the accelerated box being equivalent to a 
gravitational field, taking the place of the Sun's gravitational 
field. The analogy is oi course only approximate, the 'equiva- 
lence' holding only within the confines of the box. 
We will now give the results of the observational test of the 
Light Deflections Observed at the 1919 and 1922 Eclipses 
No. of 
No. of 
Ang. dist. 
Range of 
Eclipse 
Expedition 
Telesc. 
plates 
stars 
fr. Sun 
deflect 
Res.' 
P. E. 
Observers 
1919 
Sobral 
4", 19' 
7 
7 
0.°5-l.°5 
0."88-0."32 
l."98 
±."12 
Dyson 
May 29 
Sobral 
8", 11' 
16 
6-12 
0. 5-1. 5 
0. 88-0. 30 
0. 86 
±. 10 
Eddington 
Principe 
8", 11' 
2* 
5 
0. 5-1. 5 
0. 86-0. 30 
1. 61 
±. SO 
Davidson 
1922 
Wallal 
6". 10' 
2 
18 
0. 5-2. 6 
0. 85-0. 18 
1. 74 
±. 30 
/Chant \ 
\ Young / 
Sept. 21 
/ Cordillo\ 
\ Eowns / 
3". 5' 
2 
14 
0. 6-2. 9 
0. 83-0. 16 
1. 77 
±. 30 1 
fDodwell ^ 
1 Davidson/ 
Wallal 
5". 15' 
4 
62-85 
0. 5-3. 4 
0. 85-0. 15 
1. 72 
-f-. 11 
Campbell \ 
Trumpler / 
Wallal 
4", 5' 
(3)* 135-139 
0. 5-10. 4 
0. 85-0. 04_ 
— 
-- ( 
* 4th nlate not yet reduced. 
Prediction by Einstein's Theory l/'75 
The agreement with prediction is especially good in connection 
with the 1922 eclipse. The last 4 plates, only 3 being reduced so 
far, were especially taken to test the 'distance law/ For this 
