this it will have for a considerable distance around the vertices of all the polyhedrons, the negative value always covering as much as two fifths of the spherical surface about the centre of the molecule, and compensating even in this case for its less extent by a greater intensity, as the mean of this coefficient over the whole surface is always exactly zero. Within this distance of no action, for some space about the centre of the prevailing polyhedric face, attraction would prevail till the higher powers became sensible, and even (as it would seem) quite up to the centre in the case of a single envelope, the repulsive action of which, when combined with the slight force of the centre, would apparently be limited to quasi-ellipsoidal spaces extending out from each vertex, and having a longer axis equal to this outer distance of no action. But this limitation of the repulsive action will be still greater if the excess of the absolute attractive power in the molecule is more considerable, as long as the distribution of the force in the different envelopes remains unaltered; and though the molecules can approach within tolerably short distances of each other in certain directions, this is not objectionable, since such an approach may even be required for chemical union and cohesion. Introsusception would hardly be probable, unless they were very different in size. The compound molecule once formed, whether its components were of the same or of different substance, might exercise a repulsive force at a considerable distance in all or nearly all directions; nevertheless, it might still admit of further increase or of disruption by an agitation among the molecules, due to heat, light, or electricity. Of course, even on this theory, for the maintenance of physical equilibrium the mean distance of the molecules would have to be considerably less than that of no action, in order that a repulsion should be produced to balance the attraction of those beyond this distance. Still, if the excess of attractive force in each molecule, and consequently the size of each, be made small enough, their dimensions may still be small compared even with this mean distance; so that in no case, except that of chemical union, would it be necessary to take account of the higher powers. Any motion communicated from one molecule to another would then probably be by means of an actual relative movement of the centres of gravity, instead of by internal vibrations.