ASTROPHYSICS, or the new astronomy, the science dealing with the physical nature of the heavenly bodies, has revealed in a re markable manner, through its discoveries during the last half-century, the wonderful ability and resourcefulness of the human mind. By means of the spectroscope we have been able to investigate the physical constitution of the sun, planets and far-off stars; it has be come possible to measure motions, not athwart the sky as the older astronomy was able to do, but to or from the observer in the line of sight; and it has been possible to arrange the stars in orderly series, tracing their evolu tion from the primaeval nebula, till now we are well on the way toward the solution of the grandest problem of human investigation, whence came we and whither are we going? Since its birth in 1859, when Kirchhoff dis covered the principles of spectrum analysis, astrophysics has advanced with great rapidity, due to the improvement of instruments and to the application of the photographic plate. This important acquisition has a two-fold ad vantage over the human eye: the eye can re ceive and retain an impression for a small fraction of a second only; the photographic plate accumulates impressions, no matter how faint, with the result that by long exposures there is brought to view objects the eye could never hope to see; and secondly, the photo graph gives a permanent record that can be examined and studied at leisure.
Instruments.— A spectrum is produced either by a prism or a grating. With prisms, increased dispersion is obtained by additional prisms. In work on the sun, before the use of gratings, it was quite common to employ 20 or more prisms. Gratings were used as early as 1815 by Fraunhofer and were made by winding fine wire over two exactly similar screws. Rutherfurd of New York ruled some very satisfactory gratings on speculum metal. A diamond is used as the ruling point and the plate to be ruled is moved by means of a screw. The character of the grating depends on the excellence of the screw. Rowland of Johns Hopkins University made a dividing of such accuracy that he was able to rule stx-inch gratings with 20,000. lines per inch. The splendid definition of Rowland gratings has never been surpassed. Rowland
made the brilliant discovery that if the grat ing were ruled on a concave surface the spec trum could be focussed without the use of collimator or camera lenses. By the elimina tion of these lenses not only was the sharpness of definition increased but the region of short wave-lengths could be investigated. This vast increase in accuracy, coupled with other prop erties of the grating, namely, normal spectrum, overlapping spectra and astigmatism, whereby comparisons are rendered easy by coinci dences, have wonderfully augmented the power of the astrophysicist to determine accurate wave-lengths. The concave grating thus marked a new epoch in the history of spec troscopy. The formula for resolving power which is defined as Y
, where aA is the difference of wave-length of two lines of mean waive-length A which can just be di vided, can easily be expressed for a grating and is the product of the total number of lines of the grating and the order of the spec trum. Ordinary six-inch Rowland gratings have a resolving power of about 400,000. Rowland's dividing engine was not capable of ruling gratings greater than six inches in diameter. Quite recently Michelson of the University of Chicago and Anderson of the Carnegie Solar Observatory have succeeded in ruling 10-inch gratings with an accuracy quite comparable with that attained by Row land. A grating of different form is the echelon invented by Michelson, made by put ting together glass plates of uniform thick ness. By this means Michelson has obtained a resolving power as great as 900,000. The precision of wave-lengths has been still fur ther increased by the use of interferometres. Four different kinds have been invented by Michelson, Perot and Fabry, Hamy, and Lum mer and Gehrcke. Michelson has been able to compare directly the wave-lengths of the prominent cadmium lines with the Internation al Metre (Michelson,