between it and the rampart, one of them, with a calico wrapper over his head and shoulders, so close that Yorke could have kicked him with his foot. After watching the scene for a few seconds, till even in the darkness the whole grew clear, he stole back to the covered way to tell Falkland, anxiously awaiting his return, what he had seen.
THE PLANETS PUT IN LEVERRIER'S BALANCE.
Leverrier has recently completed the noblest work in pure astronomy which this age has seen. Five-and-thirty years ago he began to weigh the planets of the solar system in the balance of mathematical analysis. "To-day," said he, addressing the Academy of Sciences at Paris, on December 21 last, "I have the honour to present a paper completing the ensemble of work the first piece of which goes back to the 16th of September, 1839." At that time he had only seven leading planets to deal with; it affords some idea of the nature of his work that the discovery of the eighth planet, Neptune, was a mere incident in the progress of his labours. Perplexed by peculiarities in the motions of one particular planet of the set he had undertaken to weigh, Leverrier quietly undertook to calculate the cause of those peculiarities, and so found Neptune. It was a matter of small moment that another great mathematician almost simultaneously accomplished the same task. With Adams the discovery of the unknown planet was the ultimate object of inquiry; with Leverrier it was a mere step in a long series of investigations. To the outside world indeed it was the achievement of all others most deserving of notice in Leverrier's work, just as the discovery of Uranus by Sir W. Herschel attracted attention which labours altogether more important both in their nature and in their results had failed to secure. Leverrier himself can hardly have so regarded the discovery of Neptune. For him, its chief interest must have resided in the confirmation of his method of procedure afforded by the discovery of a planet through the careful study of perturbations due to that planet's attraction. Such confirmation was afforded at other steps of the work. In fact, the whole series of Leverrier's labours affords perhaps the noblest illustration of the value of deduction guided by and suggesting observations since Newton's "Principia" first proved the superiority of that method over mere induction.[1]
We propose to give such a sketch of Leverrier's method and results as would alone be suited to these pages. It need hardly be said, perhaps, that his work is essentially mathematical — nay, his methods, though not belonging to the very highest developments of modern mathematics, require (even to be understood) a higher degree of mathematical skill than would be implied by mere familiarity with more recent methods in mathematics. Yet it is possible to exhibit the general principles and the results of Leverrier's work in a manner which every one can understand.
In the solar system, we see first a mighty central ruler, whose mass so enormously exceeds that of all the planets taken together, that he is capable of swaying their motions without being himself disturbed. He is not indeed quite fixed. Whatever force he exerts on any planet, precisely that same force the planet exerts on him; but then he is so massive that the pull which compels the planet to circle around the sun scarcely displaces him at all. "If he pulls the planets," says Sir John Herschel, "they pull him and each other; but such family struggles affect him but little. "They amuse them," he proceeds quaintly, "but don't disturb him. As all the gods in the ancient mythology hung dangling from and tugging at the golden chain which linked them to the throne of Jove, but without power to draw him from his seat, so, if all the planets were in one straight line and exerting their joint attractions, the sun — leaning a little back as it were
- ↑ According to Bacon, science was to be advanced by making great collections of observations and classifying them — sorting and sifting until the grains of truth were winnowed out. No great discovery has ever been effected in this manner. The real use of observation and experiment has been found in their application to test the deductions from theories formed long before materials sufficient for Bacon's inductive method had been gathered. The question is one of fact. Theoretically, Bacon's method is perfect; it has hitherto failed in practice. Take any of the great discoveries of science, and it will be found that observations and experiments merely gathered together had no part in leading to the discovery; but that observations and experiments suggested by the deductions from theory were all-important. The moon might have been observed at Greenwich for all time without the observations leading to the discovery of gravitation. But Newton's deductions from the theory (when as yet the theory was but a guess) at once showed what observation might do; and it was by observation so made that the theory was established. In spectrum analysis a perfect heap of experiments had been collected without any useful results. Kirchhoff is led by a single observation to think of a theory, deduces certain consequences, tests these by three experiments, and the great discovery is to all intents and purposes effected.
