THEORY OF RELATIVITY 
277 
extension of the principle of relativity that will establish the 
equality of the real nature of the two concepts. 
A simple hypothetical experiment pointing to this extension 
will make the point made in the last paragraph clear. Let us 
consider a cubical box of sufficient size so that an observer pro- 
vided with appliances for experiment finds room in the same. We 
further conceive of this box so far away from all bodies or masses 
so that it is not influenced by them in a gravitational way. The 
observer will experience great difficulty to stand upright. When 
lying on the floor the least Impulse will send him to the ceiling. 
When he lets go of an object in his hand it will not fall to the 
floor. A body attached to a string fastened to the ceiling will not 
stretch the string unless pulled down by the observer which he 
would find hard to do on account of the instability of the bodies 
involved. 
Let us now imagine a rope attached to the outside of the box 
and some being begin to pull at the same upward with a con- 
stant force, giving the box a uniform acceleration. The man in- 
side the box will at once be able to stand upright in the box. Let- 
ting go of an object it will fall to the floor; looking at the body 
on the string he will see it hanging vertical with the string taut. 
In view of the observed facts the man has every reason to be- 
lieve that he is in a “gravitational field.” 
Now let us further imagine an observer located at rest in space 
some distance to one side of the box. This observer would be of 
different opinion from the one in the box ; he would say, the rea- 
son the string hangs taut and hangs in a vertical direction is due 
to the “inertia” of the upward accelerated box ; and give a simi- 
lar explanation with regard to the other phenomena occuring 
there. 
The observer in the box then attributes to the body on the 
string “gravity mass” ; the observer outside the box, “inertia 
mass.” The two concepts are identical and, moreover, inde- 
pendent of the nature of the body. 
Within the box the actual acceleration of bodies is strictly 
“equivalent” to a gravitational field. 
We may now substitute for the box a frame of reference K' 
limited, however, to the confines of the box; and for the region 
outside, where the second observer is located, a frame of refer- 
ence K. K' is uniformly accelerated with respect to K. Relative 
to K' or as seen from K', K is uniformly accelerated. K', as seen 
