434 On the Disappearance of some Spectral Lines, fyc. [Mar. 16, 



sphere, according to Young and Tacchini, and is assumed to be due 

 to iron, is so near to lines which come out in our crucibles on the in- 

 troduction of other metals, that we cannot help feeling some doubt as 

 to its absolute identification with the iron line ; the more so as in 

 Young's catalogue bright lines are sometimes assigned to two metals, 

 of which the real lines differ by nearly a unit of Angstrom's scale. 

 This is the case, for example, with the line at wave-length 501 7*6, 

 which is ascribed to iron and nickel. And where lines are broadened, 

 a,s in sun-spots, the identification with either of two very close lines 

 becomes very difficult. 



Fig. 4 shows the lines which come out in the neighbourhood of 

 wave-length 4923. A pair of lines are developed by iron close to this 

 line, and a very bright but evanescent line comes out at about 4923" 5, 

 on the introduction of cerium. This is an exceedingly brilliant line 

 for the time, and might easily be mistaken for the iron line unless 

 examined under high dispersion, and it seems to show that metallic 

 cerium is readily volatile under these conditions. The iron line at 

 4923 seems to disappear on the addition of titanium, which, on the 

 other hand, brings out the lines marked titanium in the figure. 

 Nickel brings out the cerium line strongly. The line which comes 

 out at 492T 3 on the addition of chromium and titanium is most likely 

 the line seen by Young in the chromosphere thirty times, which up to 

 the present time has not been recognised as due to any element but 

 sulphur. 



Both the nickel line at 5016*5 and the adjacent iron line at 501 7' 5 

 are seen in the arc in our crucibles, but the nickel line is much the 

 stronger and more persistent. Cerium when put into the crucible 

 brightens the titanium lines, as well as the line at 501 7"5. An . alloy 

 of manganese, iron, and titanium had the effect of making the nickel 

 line broad and diffuse, without strengthening the 501 7'5 line. 



These are but samples of the large amount of work which remains 

 to be done before we can pronounce that any of the solar lines are not 

 due to terrestrial elements, or can draw any safe inferences from 

 observed variations in their relative strengths or apparent coincidences ; 

 and no real scientific advance can be made by attempting generaliza- 

 tions with the knowledge which we at present possess. 



