of times those which have thus been recognized. In the former case the capturing-places in the solar system are very limited in extent compared with the dimensions of the system; but in the latter case it is the whole domain of the sun which we have to compare with that mere thread of space traversed by our earth within the solar system.
I might have arrived at the same result, however, in entirely different ways—a consideration which is at once the most marked characteristic and the surest test of truth in a general theory of this sort.
Suppose, for instance, I had begun with the discovery by Professor Young that the sun has tremendous ejective might. I have shown, from the circumstances attending the formation of the eruptive prominences, that they do not indicate the ejection of glowing hydrogen and helium, but of small masses of denser matter through those gases. (I note, in passing, that Tacchini's observations during the last eclipse have practically demonstrated the justice of this view.) I have further proved that such masses of ejected matter have in some cases had velocities exceeding the three hundred and eighty-two miles per second which the sun can master, and therefore must have passed forever away from him. From this demonstrated fact, as surely as from M. Daubrée's demonstrated facts about meteorites, we can work out the whole theory of cometic and meteoric ejection. For our sun, being one of the stars, we may infer that what he does each star also does. Again, what he does now he must have done (perhaps once with even greater energy) during all the millions of years that he has been a sun and doing sun-work. So also must all the suns which people space, during the past millions of years of their sun-work, have expelled from time to time flights of small bodies (whose nature we have yet, so far as this discussion of our theory is concerned, to determine). We may conclude that from the total matter ejected at any outburst many millions of small bodies would be formed as the originally vaporous matter vomited forth condensed into the liquid form and then into the solid—perhaps quite close to the parent orb. But the total mass ejected would bear to the ejecting body some such relation as the total mass of the dust ejected at Krakatoa bore to the six hundred millions of millions of millions of tons of the earth's mass; a hundred millions of years of such ejective work from an orb like our sun might well be unable to eject a total mass from out of which such a globe as even the least of the asteroids could be formed. Moreover, what we have thus inferred about each sun during the whole of its career up to the present time, we must infer also of each one of the bodies attending on each of those suns, during the sun-like portion of the career of each such attendant orb.
Hence, taking an average meteor-flight to represent the number of bodies at each ejection, ten effective ejections per annum for each sun-like orb, an average of a million years only for the sun-like duration of each orb in space, a thousand millions of suns in our galaxy
