projecting clouds of missiles with sufficient vehemence to set them finally free from the globe from which they spring. It is not easy to state the question simply, but we must make the attempt. I shall suppose that the speed which the missile receives from the volcano is compounded with that derived from the orbital movement of the planet. We have already seen that if this total speed is less than eight miles a second, then no matter what the direction of the movement of the projectile may be, it must fall short of the earth's track, and can therefore never possibly reach our globe. If, on the other hand, the volcano on Ceres were so powerful that the speed it imparted, when combined with that which the missile derives from the orbital movement, exceeded sixteen miles a second, then the path in which the body starts on its voyage through space would take the form of a hyperbola. In this case, although the missile might cross the earth's track once, it would never do so again, for the attraction of the sun would not be sufficient to recall it. Should the total speed of projection lie between eight miles and sixteen miles a second, then the orbit would be elliptical, and the body would move round and round with the same regularity as a planet. But among all the different possible orbits of this kind comparatively few will actually intersect the earth's track.
To take an illustration, let us suppose the case of those missiles which start with a total speed intermediate between the two extremes we have just named. Let us imagine that they have a velocity of twelve miles per second; it can be demonstrated that projectiles launched forth at the speed just named, but in all directions, will assume all sorts of orbits, and of these
