432 



Profs. G. D. Liveing and J. Dewar. [Mar. 16,. 



arc in a magnesia crucible. The portions selected are of special 

 interest, because in these regions a remarkable outburst of broad 

 Fraunhofer lines, not usually visible, is recorded by the Astronomer 

 Royal as having occurred in a sun-spot ("Monthly Not. Ast. Soc," 

 1881). 



Fig. 1 shows the principal (not all the) lines which were in the 

 same field of view when the spectrum of the 4th order produced by 

 a Rutherford grating (17,296 lines to the inch) was observed, the light 

 being that of the arc of a Siemens machine in a magnesia crucible. A 

 small piece of copper was first put in and then some nickel, and by 

 the lines of these metals the portion of the spectrum under examina- 

 tion was identified. The iron lines were as usual also present. The 

 symbols affixed to the several lines show those which came out when the 

 metals indicated were introduced. Thus, when chromium was dropped 

 in, a very brilliant line came out near the middle of the field, a little 

 below the iron line wave-length 5090*4 ; titanium cyanide brought 

 out a line at about wave-length 5086, cobalt one at 5094, uranic oxide 

 one at 5087, and cerium (which may have contained lanthanum and 

 didymium) a number of lines. These lines were very bright for a 

 second or two, and soon became much less brilliant, but were revived 

 when more of the metal was put in. Lead brought out a very 

 evanescent diffuse band represented in the figure by dotted lines. 

 The distances of the several lines from the extreme nickel lines 

 were measured hastily by a micrometer, and are here reproduced to 

 scale. One line at wave-length 5096, though constantly present, did 

 not seem to be affected by any of the metals introduced. An iron 

 line is indicated on Angstrom's map at this place, but the introduc- 

 tion of iron, which expanded the neighbouring line at wave-length 

 5097'3, had no effect on it. It is remarkable that this region in 

 Angstrom's normal solar spectrum is particularly bare of lines, though 

 Vogel gives several faint lines between those marked by Angstrom. It 

 is, however, a region in which many lines have been observed in sun- 

 spots (Greenwich Spectroscopic and Photographic Results, 1880), and 

 the most prominent of these lines seem to correspond to lines developed 

 by cerium, chromium, and cobalt, though more exact measures than 

 we were able to take at the times that those observations were made 

 are needed in order to establish an exact coincidence. 



Fig. 2 represents the lines brought out in a similar way in another 

 short portion of the spectrum, which is also remarkably bare of lines 

 in the solar spectrum. 



Pig. 3 shows lines brought out in another place by the several 

 metals indicated. Other lines were visible in this region but were 

 not specially developed by the metals introduced. 



The line at wave-length 4923, which occurs so often in the chromo- 



