September 13, 1900] 



NA TURE 



471 





The memorable results arrived at by Kirchhoflf were no sooner 

 published than they were accepted without dissent. The works 

 of Stokes, Foucault and Angstrom at that period were all sug- 



Iive of the truth, but do not mark an epoch of discovery. 



Astronomical spectroscopy divided itself naturally into two 

 ain branches, the one of the sun, the other of the stars, each 

 having its many offshoots. I shall just mention a few points re- 

 lating to each. The dark lines in the solar spectrum had already 

 been mapped by Fraunhofer, and now it only needed better 

 instruments and the application of laboratory spectra with Kirch - 

 hoff's principle to advance this work still further. 



Fraunhofer had already pointed out the way in using gratings, 

 and these were further improved by Nobert and Rutherfurd. 



Kirchhoff's Map of the Solar Spectrum, published in 1861-62, 

 was the most complete up to that time ; but the scale of refer- 

 ence adopted by him was an arbitrary one, so that it was not 

 long before this was improved upon. Angstrom published in 

 I S68 his map of the "Normal Solar Spectrum," adopting the 

 niUural scale of wave-lengths for reference, and this remained in 

 use until quite recent times. 



The increased accuracy in the ruling of gratings by Rutherfurd 

 materially improved the efficiency of the solar spectroscope, but 

 it was not until Prof. Rowland's invention of the concave 

 grating that this work gained any decisive impetus. The maps 

 (first published in 1885) and tables (published in the years 

 1896-98) of the lines of the solar spectrum are now almost uni- 

 versally accepted and adopted as a standard of reference. These 

 tables alone record about 10,000 lines in the spectrum of the 

 sun, which is in marked contrast to the number 7 recorded by 

 Wollaston at the beginning of the century (1802). Good work 

 in the production of maps has also been done in this country by 

 Higgs. 



Michelson has also recently invented a new form of spectro- 

 scope called the "Echelon" {Ast. Phys. Journ., vol. viii. 

 1898, p. 37), in which a grating with a relatively small number 

 of lines is employed, the dispersion necessary for modern work 

 being obtained by using a high order (say the hundredth) into 

 which most of the light has been concentrated. 



Besides lines recorded in the visual and ultra-violet portions 

 of the solar spectrum, maps have been made of the lines in the 

 infra-red, the most important being that of Langley's, published 

 in 1894, prepared by the use of his "bolometer." Good work 

 had, however, been done in this direction previously by 

 Becquerel, Lamansky and Abney ; the last, indeed, succeeded 

 even in photographing a part of it. 



The recording of the Fraunhofer lines in the solar spectrum is 

 not all, however. The application of the spectroscope to the 

 sun has several epoch-marking events attached to it, notably 

 those of proving the solar character of the prominences and 

 corona, the rendering visible of the prominences without the aid 

 of an eclipse by the discovery of Lockyer and Janssen in 1868, 

 the photography of the prominences both round the limb and 

 those projected on the solar disc by the invention of the spectra- 

 heliograph by Hale and Deslandres in 1890. 



Success has not yet favoured the many attempts to photograph 

 the corona without an eclipse by spectroscopic means ; but even 

 now this problem is being attacked by Deslandres with the 

 employment of the calorific rays. 



Spectroscopic work on the sun has led to the discovery of 

 many hundreds of dark lines, the counterparts of which it has 

 not yet been possible to produce on the earth. 



But besides those unknown substances which reveal their 

 presence by dark lines, there were two others discovered, which 

 >howed themselves only by bright lines, the one in the chromo- 

 phere, to which the name of Helium was given, and the 

 )ther in the corona, to which the name of Coronium was 

 applied. 



The former was, however, identified terrestrially by Ramsay 



in 1895, though the latter is still undetermined. The revision 



t its wave-length, brought about by the observations of the 



clipse of 1898, may, however, result in this element being 



uansferred from the unknown to the known in the near future. 



The study of stellar spectra was taken up by Huggins, 

 Rutherfurd and Secchi. Rutherfurd {Am. Journ., vol. xxxv. 

 1862, p. 77) published in 1862 his results upon a number of 

 stars, and suggested a rough classification of the white and 

 yellow stars ; but Secchi deserves the high credit of introducing 

 the first systematic differentiation of the stars according to 

 their spectra, he having begun a spectroscopic survey of the 

 heavens for the purposes of classification {Comptes retidus. 



NO. 



161 I, VOL. 62] 



t. Ivii. 1853), whilst Huggins devoted himself to the thorough 

 analysis of the spectra of a few stars. 



The introduction of photography marks another epoch in the 

 study of stellar spectra. Sir William Huggins applied photo- 

 graphy as early as 1863 (Phil. Trans., 1864, p. 428), and 

 secured an impression of the spectrum of Sirius, but nearly 

 another decade ejapsed before Prof. H. Draper (Am. Journ. of 

 Soc. and Arts, vol. xviii. 1879, p. 421) took a photograph of 

 the spectrum of Vega in 1872, which was the first to record any 

 lines. With the introduction of dry plates this branch of the 

 new astronomy received another impetus, and the catalogues of 

 stellar spectra have now become numerous. Among them may 

 be mentioned those of Harvard College, Potsdam, Lockyer, 

 McClean and Huggins. The Draper Catalogue (Annals 

 Hatvard Coll., vol. xxvii. 1890) of the Harvard College, which 

 is a spectroscopic Durchmusterung, alone contains the spectra 

 of 10,351 stars down to the 7-8 magnitudes, and this has 

 further been extended by work at Arequipa, whilst Vogel and 

 Miiller of Potsdam [Astro-Pkys. Obs. zu Potsdam, vol. iii. 

 1882-83) made a spectroscopic survey of the stars down to the 

 7*5 magnitude between - 1° and -f 20° declination. This has 

 again been supplemented by Scheiner {ibid., vol. vii. 1895 = 

 " Untersuchungen liber die Spectra der helleren Sterne"), and 

 by Vogel and Wilsing {ibid., vol. xii. 1899 : " Untersuchungen 

 iiber die Spectra von 528 Sternen "). Lockyer {Phil. Trans., 

 vol. clxxxiv. A, 1893) in 1892 published a series of large- 

 scale photographs of the brighter stars, and more recently 

 McCiean {Phil. Trans., vol. cxci. A, 1898) has completed a 

 spectroscopic survey of the stars of both hemispheres down to 

 the 1% magnitude. For the study and investigation of special 

 types of stars, the researches of Duner on the red stars, made at 

 Upsala, and those of Keeler and Campbell on the bright-line 

 stars, made at the Lick Observatory, deserve mention. For the 

 study of stellar spectra the use of prisms in slit or objective- 

 prism spectroscopes has predominated, though more recently 

 the use of specially ruled gratings has been attended by some 

 degree of success at the Yerkes Observatory. 



Several new stars have also been discovered by their spectra 

 by Pickering in his routine work of charting the spectra of the 

 stars in different portions of the sky. The photographic plate 

 containing their peculiar spectra was, however, not examined 

 in many cases until the star had died down again. 



Spectrum analysis also opened up another field of inquiry, 

 viz. that of the motion of the stars in the line of sight, based on 

 the process of reasoning due to Doppler, and accordingly named 

 Doppler's Principle ("Ueber das farbige Licht der Doppel- 

 sterne," . . . Abhandl. der K. Bohmischen Ges. d. IViss. V. 

 Folge, 2 Bd. 1843.) 



The observatories of Greenwich and Potsdam were among 

 the first to apply this to the stars, and more recently Campbell 

 at Lick, Newall at Cambridge, and Belopolsky at Pulkowa have 

 made use of the same principle with enormous success. 



It was also discovered that there are certain classes of stars 

 having a large component velocity in the line of sight, which 

 changes its direction from time to time, and in many such cases 

 orbital motion has been proven, as in the case of Algol. 



Another class of binary stars has also been discovered spectro- 

 scopically and explained by Doppler's principle. I refer to the 

 stars known as spectroscopic binaries, in which the spectrum 

 lines of one luminous source reciprocate over those from the 

 other source of light, according as one is moving towards or 

 away from the earth. This displacement of the spectrum lines 

 led to the discovery of the duplicity of /3 Aurigse, and C Urs^e 

 Majoris by Pickering {Am. Jour. [3], 39, p. 46, 1890). 



Several other such stars have now been detected, notably 

 /3 Lyrre, and lastly Capella, discovered independently by 

 Campbell {Astro- Phys. Jour., vol. x. p. 177) at Lick, and 

 Newall {Monthly Notices, vol. Ix. p. 2, 1899) at Cambridge. 



The progress of the new astronomy is so closely bound up 

 with that of photography that I shall briefly call to mind some 

 of the many achievements in which photography has aided the 

 astronomer. 



Daguerre's invention in 1839 was almost immediately tried 

 with the sun and moon, J. W. Draper and the two Bonds in 

 America, Warren de la Rue in this country, and Foucault and 

 Fizeau in France, being among the pioneers of celestial photo- 

 graphy ; but no real progress seems to have been made until 

 after the introduction of the collodion process. Sir John 

 Herschel in 1847 suggested the daily self-registration of the 

 sun-spots to supersede drawings ; and in 1857 the De la Rue 



