The establishment of the terrestrial science of spectrum analysis was due, as we have seen, equally to Kirchhoff and Bunsen, but its celestial application to Kirchhoff alone. He effected this object of the aspirations, more or less dim, of many other thinkers and workers, by the union of two separate, though closely related lines of research--the study of the different kinds of light _emitted_ by various bodies, and the study of the different kinds of light _absorbed_ by them. The latter branch appears to have been first entered upon by Dr. Thomas Young in 1803;[389] it was pursued by the younger Herschel,[390] by William Allen Miller, Brewster, and Gladstone. Brewster indeed made, in 1833,[391] a formal attempt to found what might be called an inverse system of analysis with the prism based upon absorption; and his efforts were repeated, just a quarter of a century later, by Gladstone.[392] But no general point of view was attained; nor, it may be added, was it by this path attainable. Kirchhoff's map of the solar spectrum, drawn to scale with exquisite accuracy, and printed in three shades of ink to convey the graduated obscurity of the lines, was published in the Transactions of the Berlin Academy for 1861 and 1862.[393] Representations of the principal lines belonging to various elementary bodies formed, as it were, a series of marginal notes accompanying the great solar scroll, enabling the veriest tiro in the new science to decipher its meaning at a glance. Where the dark solar and bright metallic rays agreed in position, it might safely be inferred that the metal emitting them was a solar constituent; and such coincidences were numerous. In the case of iron alone, no less than sixty occurred in one-half of the spectral area, rendering the chances[394] absolutely overwhelming against mere casual conjunction. The preparation of this elaborate picture proved so trying to the eyes that Kirchhoff was compelled by failing vision to resign the latter half of the task to his pupil Hofmann. The complete map measured nearly eight feet in length. The conclusions reached by Kirchhoff were no sooner announced than they took their place, with scarcely a dissenting voice, among the established truths of science. The broad result, that the dark lines in the spectrum of the sun afford an index to its chemical composition no less reliable than any of the tests used in the laboratory, was equally captivating to the imagination of the vulgar, and authentic in the judgment of the learned; and, like all genuine advances in the knowledge of Nature, it stimulated curiosity far more than it gratified it. Now the history of how discoveries were missed is often quite as instructive as the history of how they were made; it may then be worth while to expend a few words on the thoughts and trials by which, in the present case, the actual event was heralded.