CH. xxvii.] CORNU'S RESULTS. 379 



vapour, there are round the incandescent central portion cooler 

 vapours, which, in consequence of their reduced temperature, 

 absorb the light radiated by the poles or by the core of the arc. 

 There are two limits to this absorption. First, the absorbing 

 matter may enter into combination with the gases in the 

 surrounding air, they may be oxidised for instance ; or, failing 

 this, they may become more complex and absorb continuously 

 in the blue or red, or finally become liquid or solid by virtue of 

 falling temperature ; either of these cases is sufficient to destroy 

 line absorption. 



In a journey to Paris I had the satisfaction of discussing 

 these results with my friend Professor Cornu. He afterwards 

 published a list of cases in which he had photographically 

 recorded these phenomena in the vapours of several substances. 1 



When I began to photograph the violet and ultra-violet 

 portions of the spectrum for all the metals easily obtainable, I 

 soon saw that the most striking instances of line reversal were 

 really to be found in this part of the spectrum. Magnesium, 

 aluminium, calcium, strontium, manganese, barium, silver, 

 lead ; represents, I think, the order in which these phenomena 

 struck me, now fourteen years ago. 



If the method of throwing an image of the arc upon the 

 slit be employed, a method which I suggested and utilised 

 in 1870, 2 for laboratory experiments, there is no difficulty in 

 seeing these reversals. 



The longer arc given us by the Siemens' machine enables 

 us, however, not only to note the ordinary phenomena already 

 given at some length in chapter xvi., but to study where the 

 reversals take place. 



Further, with the Siemens' machine the arc is not only much 

 longer, but when some substances are introduced into it, it is 

 accompanied by a flame sometimes three or four inches long, of 



1 Comptcs Rendus, July 21, 1871. 



2 Phil. Trans. 1873, p. 254. 



