In his special line as a celestial explorer of the most comprehensive type, Sir William Herschel had but one legitimate successor, and that successor was his son. John Frederick William Herschel was born at Slough, March 17, 1792, graduated with the highest honours from St. John's College, Cambridge, in 1813, and entered upon legal studies with a view to being called to the Bar. But his share in an early compact with Peacock and Babbage, "to do their best to leave the world wiser than they found it," was not thus to be fulfilled. The acquaintance of Dr. Wollaston decided his scientific vocation. Already, in 1816, we find him reviewing some of his father's double stars; and he completed in 1820 the 18-inch speculum which was to be the chief instrument of his investigations. Soon afterwards, he undertook, in conjunction with Mr. (later Sir James) South, a series of observations, issuing in the presentation to the Royal Society of a paper[111] containing micrometrical measurements of 380 binary stars, by which the elder Herschel's inferences of orbital motion were, in many cases, strikingly confirmed. A star in the Northern Crown, for instance (Eta Coronæ), had completed more than one entire circuit since its first discovery; another, Tau Ophiuchi, had _closed up_ into apparent singleness; while the motion of a third, Xi Ursæ Majoris, in an obviously eccentric orbit, was so rapid as to admit of being traced and measured from month to month. It was from the first confidently believed that the force retaining double stars in curvilinear paths was identical with that governing the planetary revolutions. But that identity was not ascertained until Savary of Paris showed, in 1827,[112] that the movements of the above-named binary in the Great Bear could be represented with all attainable accuracy by an ellipse calculated on orthodox gravitational principles with a period of 58-1/4 years. Encke followed at Berlin with a still more elegant method; and Sir John Herschel, pointing out the uselessness of analytical refinements where the data were necessarily so imperfect, described in 1832 a graphical process by which "the aid of the eye and hand" was brought in "to guide the judgment in a case where judgment only, and not calculation, could be of any avail."[113] Improved methods of the same kind were published by Dr. See in 1893,[114] and by Mr. Burnham in 1894;[115] and our acquaintance with stellar orbits is steadily gaining precision, certainty, and extent.