A834 REPORT-—1904. 
find a difference in the velocity of light according to whether we make the ray 
go in the same direction as the earth or in the opposite. 
Such a measurement of the velocity of light would be only possible when we 
use two mirrors at a great distance from each other which rotate with the same 
angular velocity. 
Let A, be a source of light going through a diaphragm D and falling on the 
mirror M,, whence it is reflected to the second mirror M,. The ray goes after 
reflection from this mirror to D,. Another ray coming from the source A, takes 
the same path, and, since both mirrors remain at rest, comes to A,. But if the 
mirrors rotate in the same phase the ray requires time to traverse the long distance 
a and finds the mirror M, in a changed position, so that the ray does not come to 
A,, but makes an angle with the direction M, A,. The observation of this angle 
gives us the velocity of light as it is well known in Fourault’s method. Reversely 
the ray coming from A, will not go to A,, but makes an angle with the direction 
M,A,, depending also upon the velocity of light. 
But both the values of this velocity could not be the same if the earth moves 
in the direction M, M.,. 
For in the time that the light needs to go from M, to M,, the mirror M, has 
changed its place and increased the distance from a to a+ a. Let cbe the velocity 
of light, » that of earth, then we have 
The time of the ray between M, and M, is therefore “_, and for the opposite 
ray between M, and M, eet Thus we should find a difference in the time which 
the two rays need to travel over the distance a in the amount of 2” ofthe value, 
ce 
that ig Bv0D 
It seems that the best observations of the velocity of light have obtained a 
greater accuracy than we need here But we ought to remember that the diffi- 
culties of these proposed experiments are far greater, because the mirrors M, and 
M, have to rotate with the same velocity. On the other hand, this agreement has 
to last only a short time, and every difference of the angular velocity could be 
detected by a change in the direction of the ray. That is the reason why I 
hope the difficulties may be overcome, like many others, by the experience of a 
physicist like Mr. Michelson, and I am glad to draw attention to this problem. 
3. Preliminary Note on the Tangential Stress due to Light incident 
obliquely on an Absorbing Surface. By Professor J. H. Poynrine, 
D.Se., LBS. 
The existence of pressure on a surface, due to the incidence of a normal beam 
of light, first deduced as a consequence of the electromagnetic theory by Maxwell, 
has been fully confirmed by the experiments of Lebedew, and quite independently 
by the exact work of Nichols and Hull. These experiments show that the 
pressure exists, and that it is equal to the energy per c.c., or to the energy density 
in the incident beam. 
In so far as it produces this pressure, we may regard the beam as a stream of 
momentum, the direction of the momentum being along the line of propagation, 
and the amount of monientum passing per second through unit area cross section 
of the beam being equal to the density of the energy in it. Let E denote this 
energy density. If the beam is inclined at @ to the normal to a surface on which 
it falls, the momentum stream on to unit area of the surface is Ecos 6 per second, 
and this is the force which the beam will exert in its own direction. If the beam 
is entirely absorbed, the result is a pressure Ecos? @ along the normal, and a 
