their orbits, far from being
circular, would be very eccentric. It is true that a mass of matter
driven from the sun cannot be exactly compared to a globe which touches
its surface, for the impulse which the particles of this mass receive
from one another and the reciprocal attractions which they exert among
themselves, could, in changing the direction of their movements, remove
their perihelions from the sun; but their orbits would be always most
eccentric, or at least they would not have slight eccentricities except
by the most extraordinary chance. Thus we cannot see, according to
the hypothesis of Buffon, why the orbits of more than a hundred comets
already observed are so elliptical. This hypothesis is therefore
very far from satisfying the preceding phenomena. Let us see if it is
possible to trace them back to their true cause.

"Whatever may be its ultimate nature, seeing that it has caused or
modified the movements of the planets, it is necessary that this cause
should embrace every body, and, in view of the enormous distances which
separate them, it could only have been a fluid of immense extent.
In order to have given them an almost circular movement in the same
direction around the sun, it is necessary that this fluid should
have enveloped the sun as in an atmosphere. The consideration of the
planetary movements leads us then to think that, on account of excessive
heat, the atmosphere of the sun originally extended beyond the orbits of
all the planets, and that it was successively contracted to its present
limits.

"In the primitive condition in which we suppose the sun to have been, it
resembled a nebula such as the telescope shows is composed of a nucleus
more or less brilliant, surrounded by a nebulosity which, on condensing
itself towards the centre, forms a star. If it is conceived by analogy
that all the stars were formed in this manner, it is possible to imagine
their previous condition of nebulosity, itself preceded by other states
in which the nebulous matter was still more diffused, the nucleus being
less and less luminous. By going back as far as possible, we thus
arrive at a nebulosity so diffused that its existence could hardly be
suspected.

"For a long time the peculiar disposition of certain stars, visible
to the unaided eye, has struck philosophical observers. Mitchell
has already remarked how little probable it is that the stars in the
Pleiades, for example, could have been contracted in