1884.] Spectroscopic Studies on Gaseous Explosions. 4:71 



iron wire was burnt, show, as might be expected, the same iron lines 

 as are developed in the flash of the detonating gases, and of the same 

 relative intensities. These intensities are not quite the same rela- 

 tively as they are in the arc spectrum. Thus the lines X 3859, 3745, 

 3737, 3735, and 3719 come out in great strength, much stronger than 

 the lines X 3647, 3631, 3618, which are remarkably strong in the arc. 



German-silver wire, burnt in the flame of coal-gas and oxygen, 

 gave the same nickel lines as were given by nickel in the detonat- 

 ing gases, as well as those of copper and lead. 



Copper wire gave, besides the lines X 3272, 3245*5, a set of bands in 

 the blue, which correspond with those given by copper salts in 

 flames, and are probably due to the oxide. 



The greater part of the lines which we have observed in the flames 

 of the exploding gases have been observed by us to be reversed when 

 the several metals were introduced into the arc in a crucible of lime 

 or magnesia ; which is quite in accord with the supposition that the 

 metals experimented on are volatile, and emit as well as absorb these 

 particular rays, at temperatures lower than that of the arc. 



That iron is volatile at a temperature below the fusing point of 

 platinum, which is about 1700° C, has been pointed out by Watts 

 (" Phil. Mag.," vol. xlv, p. 86), who observed in the flame of a 

 Bessemer converter almost all the green and blue lines of iron which 

 we have seen in the exploding gases, besides one or two lines which 

 we have not observed or identified. Having regard to this volatility 

 of iron, it does not seem so surprising that iron lines should be observed 

 accompanying those of hydrogen to great heights in the sun's atmo- 

 sphere as that they should not be always seen there. 



It is interesting to note that Copeland and Lohse (" Copernicus," 

 Dec. 1882) observed in the spectrum of the great Comet of 1882, 

 amongst other lines, four lines, X 5395, 5369, 5326, 5269, which are 

 nearly identical with iron lines observed in the flash of the detonating 

 gas. 



It is remarkable that such volatile metals as mercury, zinc, and 

 cadmium should give no lines in the flame of the exploding gases. 



The gases exploded in the tube were generally mixed in nearly the 

 proportions in which they combine chemically ; but experiments 

 were made with oxygen in excess and also with hydrogen and car- 

 bonic oxide in excess. A small excess of any one of the gases did not 

 seem sensibly to affect the result, but on the whole the metallic lines 

 were more certainly developed when there was not much excess of 

 oxygen, and more constantly developed when hydrogen was used than 

 when carbonic oxide was used. 



The absence of any metallic lines in the flame of the exploding 

 gases which are more refrangible than T may be in part due to a 

 falling off in the sensibility of the photographic plates for light of 



