ESSAY ON THE VELOCITY OF LIGHT. 151 
** According to the theory of emission, light moves in water notably faster than in air. 
According to the wave theory, it is precisely the opposite which takes place: the light 
moves faster in air than in water. Suppose that one of the rays (the upper ray, for example) 
has to traverse a tube filled with water before it strikes the mirror. If the theory of emission 
be true, the wpper ray will be accelerated in its progress; it will reach the mirror first; it 
will be reflected before the lower ray; it will make with it a certain angle, and the direction 
of the deviation will be such that the lower ray will appear in advance of the other, that it 
will appear to have been deviated more by the turning mirror. 
‘‘Circumstances remaining the same, let us admit fora moment the truth of the wave 
system. The tube of water will retard the progress of the upper ray; the ray will arrive at 
the reflecting mirror after the lower ray; it will be reflected not the first, as in the former 
case, but the second in order, and from a position of the polished reflecting face in advance 
of the position it had when it reflected the upper ray a moment before; these two rays will 
make with each other the same angie as in the other hypothesis, except (and we should well 
remark it) the deviation will take place precisely in an opposite direction; the upper ray 
will now be in advance, always indicating thus the direction in which the mirror revolves. 
“To recapitulate: two radiating points, placed near each other on the same vertical line, 
flash instantaneously before a revolving mirror. The rays from the upper point cannot reach 
the mirror until atier traversing a tube filled with water; the rays from the second point 
arrive at the mirror without meeting in their course any other medium than the air. ‘lo be 
more definite, we will suppose that the mirror, seen from the position the observer occupies, 
turns from the right to the left. Well, if the theory of emission be true; if light be material, 
the upper point will appear to the left of the lower point. It will appear to tis right, on the 
contrary, ii light results from the vibrations of an ethereal medium. ~ 
*‘Instead of two isolated radiating points, suppose that we instantaneously present to the 
mirror a vertical lunrinous line. The image of the upper part of this line-will be formed by 
rays which have traversed the water; the image of the lower part will result from the rays 
which have throughout their whole course traversed the air. In the reyolving mirror the 
image of the singie line will appear broken; it will be composed of two vertical luminous 
lines, of two lines, which will not be prelongations of each other. 
‘‘ The upper rectilinear image, is it behind the one below! Does it appear to its left? 
““ Light is « body. 
*“* Does the contrary take place? The upper image, does it show itself to the right? 
“Light ts an undulation. 
** All that precedes is theoretically, or rather speculatively exact. Now, (and here is the 
delicate point, ) it remains to prove that, notwithstanding the prodigious velocity of light, 
that notwithstanding a velocity of 190,000 miles a second, that notwithstanding the small 
length that we will be obliged to give to the tube filled with liquid, that notwithstanding the 
limited velocities of rotation that the mirrors will have, the comparative deviations of the 
two images, towards the right or towards the left, of which I have demonstrated the exist- 
ence, will be perceptible in our instruments.” 
Arago then enters into the most minute details of all the parts of the experi- 
ment: the velocity of rotation that can be given to a mirror, the visibility of 
the image formed by light after having traversed the necessary length of liquid, 
the possibility of reducing that length of liquid, or the velocity of rotation of 
the mirror by employing simultaneously several rotating mirrors from which 
the light would be successively reflected, and also in substituting for water 
bisulphide of carbon, which acts more powerfully on the velocity of the light, 
are, on his part, the objects of a thorough examination. He then terminates 
thus: 
‘* Suppose in the experiment that I propose to execute we make use of electric sparks, or 
of lights successively screened and unscreened by the use of rotating disks, as their emis- 
sions should only last during a tew thousandths of a second, it may happen that an observer, 
looking in the mirror from a given direction, and with a telescope of limited field, will only 
by chance perceive the light. To this I immediately reply that in renewing very often the 
apparitions of light—every second, for example—that if, instead of a single mirror, we rotate 
a vertical prism of eight or of ten facets, that with the concurrence of several observers, 
placed in different directions, and each with his telescope, we cannot fail to have numerous 
and clear apparitions of the reflected rays. But these are details on which I shall not dweli 
to-day. I will reserve for another communication the exposition of the system of experi- 
ments in which we will render sensible, and in which we will measure, to a certain degree, 
the absolute velocity of light without having recourse to celestial phenomena.” 
After the publication of’ this remarkable notice we could not doubt the suc- 
cess of the experiment it detailed. But, not to interrupt the chronological order, 
we should speak first of an entircly different experiment, by which the direct 
