180 Proceedings of the Royal Society 
mg from the object with the velocity found, being seen to be in- 
consistent with the phenomena of interference, we can scarcely be 
said to make use of a hypothesis when we conclude that it is an 
action transmitted through a medium bodily at rest, it may be, but 
whose component molecules act upon one another in such a way as 
to propagate the effect in question. By the term light we mean 
this action considered as a physical fact, separate from our percep- 
tion of it by the eye, and independent of its arrival or non-arrival 
at our organs of vision. 
The propagation of light from a luminous point with the same 
velocity in all directions (in a homogeneous medium), implies that 
the action originating at any instant in the source is diffused over 
a spherical surface whose radius, measured from the luminous point 
as centre, constantly increases at the rate of the velocity of light ; 
and the constancy with which this propagation is kept up, implies 
that there are an infinite number of such spherical surfaces, over 
each of which is diffused an action which originated in the source at 
a preceding instant. Next the question presents itself whether all 
these actions originating in the source at successive instants, and 
occupying successive spherical surfaces, are similar and equivalent. 
The phenomena of interference answer, that if we imagine a series 
of these spherical surfaces separated from each other by a very 
small constant distance A, the action propagated upon each of these 
surfaces is the same, and that midway between each pair of the 
series is a surface propagating an action capable of destroying that 
of its neighbour of the first series, from which it is separated by the 
constant distance ^ . Now, that is equivalent to saying that each 
thin spherical shell of the medium through which the action is 
transmitted, vibrates between opposite phases, and as it is im- 
possible for us to conceive or believe that any finite change can 
take place in the material world that does not involve an infinite 
number of intermediate infinitesimal changes, we are authorised to 
say that light consists in periodic vibrations, propagated with very 
great velocity, and decomposable in an infinite number of ways 
into half vibrations exactly contrary to one another. 
Thus far we have arrived without having recourse to any hypo- 
thesis, having assumed nothing regarding the nature of these 
