the star the place should be found vacant, this evidence would be almost conclusive that the supposed star was actually the planet. By an examination of this kind it was found that the planet Uranus had been observed, and its place carefully recorded by no less than three astronomers, each of whom had seen it several times, without any suspicion of its planetary character. The astronomer Flamsteed was the first who had mistaken this planet for a star nearly ninety years before its discovery by Sir William Herschel. It was subsequently observed by Bradley, by Mayer, and by Le Monnier, who fixed its place no less than twelve times during the period from 1750 to 1771. These ancient observations furnished an opportunity to test the accuracy of the computed elements of the orbit of the new planet, and to correct these elements, in case they were found to be sensibly in error. This work was executed in a most faithful and exact manner by M. Bouvard, who also computed tables predicting the places of Uranus for many years in advance. It was supposed with reason that these tables would point out the places of Uranus with the same certainty as those of Saturn and Jupiter-computed by the same astronomer-gave the places of these planets. In this the hopes of the astronomical world were disappointed, and this extraordinary discrepancy between computation and observation gave rise to the discovery of an exterior planet, as we shall now relate.

CHAPTER XIII.

NEPTUNE, THE NINTH AND LAST KNOWN PLANET IN THE ORDER OF DISTANCE FROM THE SUN.

Uranus Discovered by Accident.-Ceres by Research with the Telescope. -Rediscovered by Mathematical Computation.-The Perturbations of Uranus.-Not due to any known Cause.-Assumed to arise from an Exterior Planet.-Nature of the Examination to find the Unknown Planet.-Undertaken at the same time by two Computers.-Computation assigns a Place to the Unknown Planet. - Discovered by the Telescope. - Discoveries resulting.-A Satellite detected.-The Mass of Neptune thus determined. Neptune's Orbit the Circumscribing Boundary of the Planetary System.

THE discovery of Neptune is undoubtedly the most remarkable event in the history of astronomical science—an event without a parallel, and rising in grandeur pre-eminently above all other efforts of human genius ever put forth in the examination of the physical universe.

The planet Uranus was discovered by the aid of the telescope, not exactly by accident, but still without any expectation on the part of the discoverer that his examination of the fixed stars would result in the addition of a primary planet to the system. Indeed, as we have seen, so little did the astronomical world then anticipate the discovery of a new planet, that the announcement by Sir William Herschel that he had detected a most remarkable comet was accepted on all hands, and it was only continued observation that finally compelled astronomers to accept the new object as a planet. In the case of the discovery of the first asteroid we find a systematic organization of astronomical effort to detect a body whose existence was conjectured, on the single ground of the harmony of the universe, or that the law of interplanetary spaces, interrupted between Mars and Jupiter, would be restored by finding a planet. revolving within that vast interval. Hence a search was

commenced which consisted in examining every star in the region of the ecliptic, to ascertain whether its place was already laid down on any known map or chart of the heavens. Now, it is evident that if it were possible to make a perfect daguerreotype of any region of the celestial sphere, say to-night, and the same could be effected on the following night, the comparison of these two pictures would exhibit to the eye any change which may have occurred in the interval from the one picture to the other; and hence, if a star were found on the second and not on the first picture, this star might fairly be suspected to be a planet, or the same suspicion would attach to a star found on the first, but missing on the second picture. Now, a map of the heavens, so far as it includes the correct places of the stars, answers our purpose quite as well as the daguerreotype, and any star found in a region well charted, but not laid down on the map, may be fairly suspected to be a planet. A few hours of examination will show it to be at rest or in motion. If in motion, then its planetary character is decided.

This method of research has been employed in the discovery of all the asteroids, and there is but one example in which a more powerful and searching examination became necessary. This was in the case of the asteroid Ceres, which, as we have seen, was discovered by Piazzi, at a time when but few observations could be made previous to its being lost in the rays of the sun. For a long time it seemed almost a hopeless task to undertake the rediscovery of the planet, as the telescope would be compelled to grope its way slowly round the heavens, in the region of the ecliptic, comparing every star with its place in the chart. In this dilemma mathematical analysis essayed to erect a structure on the narrow basis of the few observations obtained by Piazzi, whereon the instrumental astronomer might stand and point his telescope to the precise point occupied by the lost planet. The genius of Gauss succeeded in this herculean task, and when the telescope was pointed to the heavens in

the exact place indicated by the daring computor, there, in the field of view, shone the delicate and beautiful light of the long-lost planet.

This was certainly a most wonderful triumph of analytic reasoning; yet in this case the planet had been discovered, was known to exist, and had been observed over 4° out of the 360° of its revolution round the sun. On this basis of 4° it was possible to rise to a knowledge of the planet's position at the end of a few months of time.

The case of the discovery of Neptune is entirely different. Here no planet was known to exist; no telescopic power, however great, had ever seen it. For ages it had revolved round the sun in its vast orbit, far beyond the utmost known verge of the planetary system, unfathomably buried from human gaze and from human knowledge. No sage of antiquity had ever dreamed of its existence. The fertile

brain of even Kepler had failed to imagine its being, and the powerful penetration of Newton's gigantic intellect had failed to pierce to the far-off region inhabited by this unknown and solitary planet.

Indeed, with the knowledge which existed prior to the discovery of Uranus, no human genius, however mighty, could have passed the tremendous interval which separates the orbits of Saturn and Neptune from each other. The discovery of an intermediate planet was requisite to furnish a firm foothold to him who would adventure to pass a gulf of not less than 2,000 millions of miles at its narrowest place.

We shall now proceed to relate the circumstances which led to the discovery of Neptune. As already stated, a careful and elaborate study of the orbit of Uranus had been accomplished by M. Bouvard, and tables giving the computed places of this planet had been prepared by the same astronomer. It was not anticipated that these tables would be absolutely perfect, even if based on perfect observations. We must remember that each body of the solar system affects every other, and hence no single set of observations

are sufficient to give a perfect orbit. In case all the other worlds were blotted out of existence, and there remained only the sun and Uranus, then three perfect observations of the planet's place would suffice to determine positively all the elements of its orbit, and fix for ever all the circumstances of its motion. We shall call the figure of the orbit of Uranus, obtained under the above hypothesis, the normal figure, and the ellipse which it would describe about the sun, under the above circumstances, the normal ellipse. If now we introduce another planet into our system, as, for example, Saturn, it is possible, as we have already seen, to compute the exact amount of power exerted by Saturn to disturb the movements of Uranus, and to change the figure of its orbit. In like manner, by adding successively all the interior planets, it is possible to compute the perturbations that each produces upon the orbit of any particular one, until, finally, by using all the power of analytic reasoning, the human mind may reach to a complete knowledge of all possible derangements produced by the combined action of all existing known causes of perturbation.

Supposing our knowledge in this way to become perfect as to the movements and orbit of Uranus, we can then predict its places in all coming time, and these predictions, being arranged in tabular form, may be verified by comparison in after years with the observed places of the planet. If now a new planet were added to the system, revolving in an orbit exterior to that of Uranus, perturbation would arise from the introduction of this disturbing body into our system, which would at once cause the planet to deviate from its predicted track, and the observed and computed places would no longer agree.

We can perceive at once, from this statement of the problem, that these very discrepancies between the old track and the new one, pursued by the planet, would give us a clue whereby it might become possible to determine, in space, the position of the disturbing body.