where they now are, for in such no relative motion, I believe,
has as yet been detected by the spectroscope. All this, too, is
in keeping with the nebular hypothesis of Kant and Laplace so
long as this does not assume a primitive infinite dispersion of
matter, but the gathering of matter from finite distances first
into nebulous patches which aggregating with each other have
given rise to our system of stars. But if we extend this
hypothesis throughout an infinite past by the supposition of
aggregation of infinitely remote particles we replace the
simultaneous approach required in order to accotnt for the
simultaneous phenomena visible in the heavens, by a succession of
aggregative events, by hypothesis at intervals of nearly infinite
duration, when the events of the universe had consisted of fitful
gleams lighted after eternities of time and extinguished for yet
other eternities.

Finally, if we seek to replace the eternal instability involved
in Kant's hypothesis when extended over an infinite past, by any
hypothesis of material stability, we at once find ourselves in
the difficulty that from the known properties of matter such
stability must have been

296

permanent if ever existent, which is contrary to fact. Thus the
kinetic inertia expressed in Newton's first law of motion might
well be supposed to secure equilibrium with material attraction,
but if primevally diffused matter had ever thus been held in
equilibrium it must have remained so, or it was maintained so
imperfectly, which brings us back to endless evolution.

On these grounds I contend that the present gravitational
properties of matter cannot be supposed to have acted for all
past duration. Universal equilibrium of gravitating particles
would have been indestructible by internal causes. Perpetual
instability or evolution is alike unthinkable and contrary to the
phenomena of the universe of which we are cognisant. We therefore
turn from gravitating matter as affording no rational account of
the past. We do so of necessity, however much we feel our
ignorance of the nature of the unknown actions to which we have
recourse.

A prematerial condition of the universe was, we assume, a
condition in which uniformity as regards the average distribution
of energy in space prevailed, but neterogeneity and instability
were possible. The realization of that possibility was the
beginning we seek, and we today are witnesses of the train of
events involved in