g exerted on D would be a maximum. Let D now
begin to move in obedience to the infinitesimal attraction exerted
upon it. Motion being once set up, the idea of _vis viva_ arises. In
moving towards F the particle D consumes, as it were, the tensions.
Let us fix our attention on D, at any point of the path over which it
is moving. Between that point and F there is a quantity of unused
tensions; beyond that point the tensions have been all consumed, but
we have in their place an equivalent quantity of _vis viva_. After D
has passed any point, the tension previously in store at that point
disappears, but not without having added, during the infinitely small
duration of its action, a due amount of motion to that previously
possessed by D. The nearer D approaches to F, the smaller is the sum
of the tensions remaining, but the greater is the _vis viva_; the
farther D is from F, the greater is the sum of the unconsumed
tensions, and the less is the living force. Now the principle of
conservation affirms _not_ the constancy of the value of the tensions of
gravity, nor yet the constancy of the _vis viva_, taken separately, but
the absolute constancy of the value of both taken together. At the
beginning the _vis viva_ was zero, and the tension area was a maximum;
close to F the _vis viva_ is a maximum, while the tension area is zero.
At every other point, the work-producing power of the particle D
consists in part of _vis viva_, and in part of tensions.

If gravity, instead of being attraction, were repulsion, then, with
the particles in contact, the sum of the tensions between D and F
would be a maximum, and the _vis viva_ zero. If, in obedience to the
repulsion, D moved away from F, _vis viva_ would be generated; and the
farther D retreated from F the greater would be its _vis viva_, and the
less the amount of tension still available for producing motion.
Taking repulsion as well as attraction into account, the principle of
the conservation of force affirms that the mechanical value of the
_tensions_ and _vires vivae_ of the material universe, so far as we know
it, is a constant quantity. The universe, in short, possesses two
kinds of property which are mutually convertible. The diminution of
either carries with it the enhancement of the other, the total value
of the property remaining unchanged.

The considerations here applied to gravity apply equally to chemical
affinity. Ina mixture of oxygen and hydrogen the ato