90 
experiment due to Fizeau, in which two rays of light were sent 
along the same path , one with and one against a stream of water, 
that is, one with and one against the resulting aether drift. 
If 1' is the speed at which a swimmer travels through the 
water of a river flowing at velocity v, then the swimmer will he 
able to travel up, down, and straight across the river at speeds 
which are respectively ( V — v), (F 4- v) t and \/( V 2 — v 2 ). 
Hence if the river is of width s , the time T ± taken to swim across 
and back is 2s/ v' (F 2 — #1, and the time T 2 to swim distance s up 
or down the river and back is2Fs/ (F 2 — v 2 ). That is, we have 
T t : T x : : F : v ( F 2 - f). 
But F is greater than y (F 2 v 2 ) for all possible values of F 
and r, and thus the time for a certain journey up arid down stream 
is always greater than for the same length of journey across 
stream. For example, a person who swims 2 miles per hour will 
take less than 3 mins. 28 set's, to cross and re-cross a river four 
chains wide (lowing at one mile per hour. He will, however, take 
4 mins, to do the same length of double journey up and down the 
river. 
Now, as the Earth is moving relatively to the Sun, and the 
Sun relatively to other members of the sidereal universe, our 
Earth is evidently in general travelling through the aether, or, the 
aether has a drift relative to the Earth. And for light travelling 
at velocity F through an aether drift of magnitude v, the time for 
the double journey along the line of aether drift must be greater 
than the time for a path of equal length at right angles to, that 
is athwart, the aether drift. Michelson and Morley attempted to 
test this by experiment. They sent a ray of light along a certain 
path and reflected it hack to the point from which it set out. 
Another ray was sent an equal distance along a path at right 
angles, and any minute difference in the times taken by the two 
rays to return could be easily ascertained by a delicate interfer- 
ence test. The two rays were found to take precisely the same 
time. As this was contrary to theory, it was clear that the path 
which was across the aether drift must really have been longer 
than the path which lay along the aether drift. Now these paths 
were along rigid iron arms attached to a vertical stand floating in 
mercury. The apparatus could thus be rotated through a right 
angle so that the path which formerly was along the aether drift 
was now across it, and vice versa. But on repeating the experi- 
ment in this new position there was found again to be no differ- 
ence in the times taken by the rays to cover the two paths. Only 
one conclusion appears possible — the rigid arms altered in length, 
shortening when turned into the direction of the aether drift and 
lengthening when turned at right angles to it ! Remarkable as 
is this conclusion, there is no escape from it, and scientists now 
accept the fact that our standards of length — the standard yard 
