Page:Williamherschel00simegoog.djvu/237

Rh this when he stood in wondering awe before the hole in the heavens.

That Herschel fell into mistakes regarding double stars cannot and need not be denied. It was unavoidable that the first traveller in an unexplored region, billions of miles distant from our earth, should err in tracing paths, measuring time, and estimating distances. He failed in his calculations with γ Virginis, which he represented as two companions that revolved round a common centre in 708 years. His son by a careful discussion of the observations made since 1718 showed that the time of revolution was not 708 years but 513. It was also predicted that the smaller of the two companions would reach the point where it is nearest the larger in the beginning of 1834. Even these revised calculations proved to be incorrect, for it did not reach that point till two years later. Observations of the star were then renewed for several years; new calculations were made, and the time of revolution of the lesser companion round the greater was found to be 182 years. But it came out that the orbit of 1834, with the time 513 years, was nearly the same, in part of its course, as the true orbit, and was "a curious example, and by no means the first in the history of the progress of discovery, where of two possible courses, each at the moment equally plausible, the wrong has been chosen."

But Herschel's study of the fixed stars and of the unity of plan in nature went farther than we have yet traced. A paper read by him in 1814 contains the following facts, that might almost have been prophecies of wonders in store for men—"Stars although 15