74 THE CHEMISTRY OF THE SUN. [CHAP. 



obtained by an automatic make-and-break from the surface of 

 the metal under examination. 



A small oval disc was made to revolve, and in its revolution 

 to act on one arm of a lever, beneath which a spring was 

 placed like that to a flute key ; this lever was fixed on a plate 

 of sheet iron, having at its extremity a small cup to hold the 

 mercury or other molten metal, and the other end of the 

 lever was bent at right angles, so that as the oval disc revolved 

 the bent end of the lever was alternately dipped into the metal 

 and removed, and the spark always occurred precisely at the 

 same point. 1 



In this way Wheatstone examined and mapped the spectra of 

 mercury, zinc, cadmium, bismuth, tin, and lead. 



This method was afterwards superseded by the introduction of 

 the induction coil, an arrangement which produces high tension 

 sparks of exceedingly high temperature, and has now come into 

 general use for obtaining the spectra of metals. If necessary 

 the temperature of the spark can be still further increased by 

 introducing a Leyden jar into the circuit as shown in Fig. 32. 

 This was the method of research employed by Kirchhoff. 



As a specimen of Kirchhoff's work, we may consider his 

 researches on the spectrum of iron. The point to determine 

 was which of the bright lines corresponded with the dark 

 Fraunhofer lines. Over the whole visible reach of the spectrum 

 Kirchhoff carried this comparison. I will give one or two 

 extracts from his paper. He says, 2 " It is especially remarkable 

 that, coincident with the positions of all the bright iron lines 

 which I have observed" (that is the bright lines from the 

 vapour of iron, using two iron poles with an induction coil), 

 " well-defined dark lines occur in the solar spectrum. By the 

 help of the very delicate method of observation which I have 



1 Paper read at the fifth meeting of the British Association, 1835, and printed 

 in the Chemical News, vol. iii. p. 198, 1861. 



2 Researches on the Solar Spectrum,. Roscoe's translation, part i. p. 18. 



