because no two different parts of succession can be put together to measure one another: and nothing being a measure of duration but duration, as nothing is of extension but extension, we cannot keep by us any standing unvarying measure of duration, which consists in a constant fleeting succession, as we can of certain lengths of extension, as inches, feet, yards, &c. marked out in permanent parcels of matter. Nothing, then, could serve well for a convenient measure of time but what has divided the whole length of its duration into apparently equal portions, by constantly repeated periods. What portions of duration are not distinguished, or considered as distinguished and measured by such periods, come not so properly under the notion of time, as appears by such phrases as these, viz. before all time, and when time shall be no more.

The revolutions of the

sun and

moon, the properest

time.

$ 19. The diurnal and annual revolutions of the sun, as having been, from the beginning of nature, constant, regular, and universally observable by all mankind, and measures of supposed equal to one another, have been with reason made use of for the measure of duration. But the distinction of days and years having depended on the motion of the sun, it has brought this mistake with it, that it has been thought that motion and duration were the measure one of another for men, in the measuring of the length of time, having been accustomed to the ideas of minutes, hours, days, months, years, &c. which they found themselves upon any mention of time or duration presently to think on, all which portions of time were measured out by the motion of those heavenly bodies; they were apt to confound time and motion, or at least to think that they had a necessary connexion one with another: whereas any constant periodical appearance or alteration of ideas in seemingly equidistant spaces of duration, if constant and universally observable, would have as well distinguished the intervals of time as those that have been made use of. For supposing the sun,

which some have taken to be a fire, had been lighted up at the same distance of time that it now every day comes about to the same meridian, and then gone out again about twelve hours after, and that in the space of an annual revolution it had sensibly increased in brightness and heat, and so decreased again; would not such regular appearances serve to measure out the distances of duration, to all that could observe it, as well without as with motion? For if the appearances were constant, universably observable, and in equidistant periods, they would serve mankind for measure of time as well, were the motion away.

But not by their motion, but periodical appear

ances.

$20. For the freezing of water, or the blowing of a plant, returning at equidistant periods in all parts of the earth, would as well serve men to reckon their years by as the motions of the sun: and in effect we see that some people in America counted their years by the coming of certain birds amongst them at their certain seasons, and leaving them at others. For a fit of an ague, the sense of hunger or thirst, a smell or a taste, or any other idea returning constantly at equidistant periods, and making itself universally be taken notice of, would not fail to measure out the course of succession, and distinguish the distances of time. Thus we see that men born blind count time well enough by years, whose revolutions yet they cannot distinguish by motions that they perceive not: and I ask whether a blind man, who distinguished his years either by the heat of summer or cold of winter; by the smell of any flower of the spring, or taste of any fruit of the autumn; would not have a better measure of time than the Romans had before the reformation of their calendar by Julius Cæsar, or many other people, whose years, notwithstanding the motion of the sun, which they pretend to make use of, are very irregular? And it adds no small difficulty to chronology, that the exact lengths of the years that several nations counted by are hard to be known, they

Notwo parts of duration can be cer

differing very much one from another, and I think I may say all of them from the precise motion of the sun. And if the sun moved from the creation to the flood constantly in the equator, and so equally dispersed its light and heat to all the habitable parts of the earth, in days all of the same length, without its annual variations to the tropics, as a late ingenious author supposes*; I do not think it very easy to imagine that (notwithstanding the motion of the sun) men should, in the antediluvian world, from the beginning, count by years, or measure their time by periods, that had no sensible marks very obvious to distinguish them by. § 21. But perhaps it will be said, without a regular motion, such as of the sun or some other, how could it ever be known tainly known that such periods were equal? To which to be equal. I answer, the equality of any other returning appearances might be known by the same way that that of days was known or presumed to be so at first; which was only by judging of them by the train of ideas which had passed in men's minds in the intervals: by which train of ideas discovering inequality in the natural days, but none in the artificial days, the artificial days or vuxμeça were guessed to be equal, which was sufficient to make them serve for a measure though exacter search has since discovered inequality in the diurnal revolutions of the sun, and we know not whether the annual also be not unequal. These yet, by their presumed and apparent equality, serve as well to reckon time by (though not to measure the parts of duration exactly) as if they could be proved to be exactly equal. We must therefore carefully distinguish betwixt duration itself, and the measures we make use of to judge of its length. Duration in itself is to be considered as going on in one constant, equal, uniform course: but none of the measures of it, which we make use of, can be known to do so;

* Dr. Burnet's Theory of the Earth.

nor can we be assured that their assigned parts or periods are equal in duration one to another; for two successive lengths of duration, however measured, can never be demonstrated to be equal. The motion of the sun, which the world used so long and so confidently for an exact measure of duration, has, as I said, been found in its several parts unequal: and though men have of late made use of a pendulum, as a more steady and regular motion than that of the sun, or (to speak more truly) of the earth; yet if any one should be asked how he certainly knows that the two successive swings of a pendulum are equal, it would be very hard to satisfy him that they are infallibly so: since we cannot be sure that the cause of that motion, which is unknown to us, shall always operate equally; and we are sure that the medium in which the pendulum moves is not constantly the same: either of which varying, may alter the equality of such periods, and thereby destroy the certainty and exactness of the measure by motion, as well as any other periods of other appearances; the notion of duration still remaining clear, though our measures of it cannot any of them be demonstrated to be exact. Since then no two portions of succession can be brought together, it is impossible ever certainly to know their equality. All that we can do for a measure of time is to take such as have continual successive appearances at seemingly equidistant periods; of which seeming equality we have no other measure but such as the train of our own ideas have lodged in our memories, with the concurrence of other probable reasons, to persuade us of their equality.

measure of

motion.

§ 22. One thing seems strange to me, Time not the that whilst all men manifestly measured time by the motion of the great and visible bodies of the world, time yet should be defined to be the "measure of motion;" whereas it is obvious to every one who reflects ever so little on it, that, to

measure motion, space is as necessary to be considered as time: and those who look a little farther, will find also the bulk of the thing moved necessary to be taken into the computation by any one who will estimate or measure motion, so as to judge right of it. Nor indeed does motion any otherwise conduce to the measuring of duration, than as it constantly brings ' about the return of certain sensible ideas, in seeming equidistant periods. For if the motion of the sun were as unequal as of a ship driven by unsteady winds, sometimes very slow, and at others irregularly very swift; or if, being constantly equally swift, it yet was not circular, and produced not the same appearances, it would not at all help us to measure time, any more than the seeming unequal motion of a comet does.

Minutes, hours, days,

and years, not necessary measures of duration.

§ 23. Minutes, hours, days, and years, are then no more necessary to time or duration, than inches, feet, yards, and miles, marked out in any matter, are to extension for though we in this part of the universe, by the constant use of them, as of periods set out by the revolutions of the sun, or as known parts of such periods, have fixed the ideas of such lengths of duration in our minds, which we apply to all parts of time, whose lengths we would consider; yet there may be other parts of the universe, where they no more use these measures of ours than in Japan they do our inches, feet, or miles; but yet something analogous to them there must be. For without some regular periodical returns, we could not measure ourselves, or signify to others the length of any duration, though at the same time the world were as full of motion as it is now, but no part of it disposed into regular and apparently equidistant revolutions. But the different measures that may be made use of for the account of time do not at all alter the notion of duration, which is the thing to be