12 



KNOWLEDGE, 



[Jantjakt 1, 1900. 



among the vapours which surround the bright shell from 

 which most of the sun's light proceeds. This method 

 was adopted by Sir Norman Lockyer about twenty-five 



f K j^- k 1 ^ 



" p f 



a 





Fig. 2. — A portion of the Solar Spectrum sliowiug intensification of 

 lines due to aqueous vapour in our atmosphere wlien the air is 

 moist. (Crewe.)* 



years ago, and more recently Prof. Rowland has, in this 

 way, compared the spectrum of the sun with that of 

 every known element except gallium. The majority of 

 the atronger lines have now been identified with respect 





t a 

 o o 



,' 3 



- -O 



Sel 



J; ^1 



to the elements which produce tlieni. To the substan- 

 p.es jrecognised by Rowland, Messrs. Runge and Paschenf 



* Astrophysioal Journal, Vol. IV., 1896, page 324. 



■ + Astrophysioal Journal, Vol. IV., 1896, p. 318; Vol. VIII., 1898, 

 p. 73. I 



ines, however, still 

 lines. Here, then. 



have added oxygen, the presence of which may now be 

 considered as demonstrated (fig. 4), and Hartley and 

 Ramage have added gallium. | 



Some thousands of the Fraunhofer 

 belong to the category of " unknown 

 is a great field for further enquiry, for it is. perhaps, 

 too early to conclude that these " unknown " lines have 

 no terrestrial equivalents, or even that they represent 

 the dissociated products of our terrestrial elements 

 The latter view in fact appears to some extent nega- 

 tived by the recent researches of Sir Norman Lockyer, 

 which show that the first stage in the dissociation of a 

 metal is indicated by the appearance of enhanced lines 

 (lines which are brighter in the spark than in the arc 

 spectrum), and in the case of iron and other well-known 

 metals these enhanced lines probably do not appear as 



Fig. 4. — The Hues of oxygen in the Solar Spectrum. (Runge and 

 Paschen.) (1) Solar Spectrum. (2) Oxygen vacuum tube. 



such among the dark lines of the solar spectrum. The sub- 

 stances which we can best compare with the Fraunhofer 

 lines agree in indicating that the absorbing vapours 

 which produce them exist under conditions very similar 

 to those which exist in the electric arc. It may be, 

 therefore, that some of the unidentified lines, which are 

 mostly feeble, represent lines in the arc spectra of 

 known substances which are so faint as to escape de- 

 tection unless photographs are taken with very long ex- 

 posures. In fact, the tables of lines recently published 

 by Hasselberg, and by Prof. Rowland himself, for vana- 

 dium, chromium, and other elements, leave little doubt 

 that many of the unidentified lines in Rowland's solar 

 tables are to be accounted for in this way. 



Another important point also appears to have received 

 insufficient attentiou. It is by no means impossible that 

 among the constituents of the earth's crust are many 

 still unrecognised elements which exist in such small 

 quantities as to evade the ordinary processes of chemical 

 analysis, but which may yet be revealed to the delicate 

 eye of the spectroscope. As the usual practice in the 

 matching of solar lines is to deal with elements in as 

 pure a state as possible, it would appear important to 

 make a spectroscopic comparison with the sun of sub- 

 stances as they occur naturally in the form of minerals 

 and rocks. Prof. Hartley has, in fact, already found 

 that some of the rarer metals, especially lithium and 

 gallium, are very widely diffused in mineral substances, 

 and this furnishes an excellent illustration of the deli- 

 cacy of the spectroscopic method. Until such mineral 

 comparisons have been made, it would be unwise to 

 suppose that all unidentified lines of the solar spectrum 

 owe their origin to non-terrestrial matter. 



In the investigation of the constituents of the sun, as 

 already remarked, we are not limited to the dark line 

 spectrum. The bright line spectra of the chromosphere 

 and prominences may be examined any time the sun is 

 visible, and by taking advantage of total eclipses, the 

 outlying parts which constitute the corona are opened 



X Asfnyphi/sical Journal, Vol. IX , 1899, p. Li4. 



