1879.] On the Reversal of the Lines of Metallic Vapours. 403 



horizontally, so that the axis of the collimator of the spectroscope was 

 directed immediately on to the arc without the intervention of a mirror, 

 and a small vertical opening was made for the introduction of the 

 substances under observation, and was closed when not in use by a 

 piece of lime. 



With sodium carbonate the green pair wave-lengths 4983, 4982 were 

 reversed, showing dark lines in the middle of the bright ones, the less 

 refrangible of the two giving the stronger dark line. The sodium line 

 given by Lecoq de Boisbaudran at wave-length 4670 showed as a 

 diffuse blue baud with a pair of fine dark lines in it, of which the 

 stronger and more lasting was the less refrangible. The diffuse blue 

 band resolved itself into two diffuse lines as the sodium carbonate 

 evaporated, and the measurement of their positions in comparison with 

 a conspicuous titanium line, which lies between them, and was made to 

 show at the same time by introducing a fragment of titanic oxide into 

 the crucible, gave for this sodium pair the wave-lengths 4667, 4664. 

 The orange pair, wave-lengths 6160, 6154, were also seen reversed in 

 like manner, but we failed to detect any difference in the strength or 

 continuance of the dark lines in this case. These reversals of sodium 

 lines were many times observed, and we noticed that the reversals of 

 the other pairs continued after that of the red pair had ceased to be 

 visible, that next after the disappearance of the dark lines in the red pair 

 those in the diffuse blue pair ceased to be visible, then the dark lines 

 in the green pair, and then those in the yellowish-green (5687, 5681). 

 In some cases when a large quantity of sodium carbonate was put into- 

 the crucible a curious double reversal occurred. In the middle of an 

 enormous dark expansion of D a bright yellow band appeared, which 

 in turn had a narrower dark band, or a pair of dark lines, in its middle. 

 A similar double reversal of the lithium blue line occurred so far as to 

 show a bright line in the middle of the dark one. Of the two violet 

 lines of potassium we observed that the more refrangible remained 

 reversed longer than the other. 



In addition to the reversals of calcium lines before observed by us r 

 Ave have noticed the reversal of all the more conspicuous calcium lines 

 of the G group, viz., those wave-lengths— 4282, 4289,4298, 4397, 4318,. 

 besides 4302 before noted, also that of 4379, and all three lines of a triplet 

 4585, 4581, 4578. The finer lines, wave-lengths 4434/3, 4454'5, slightly 

 less refrangible than the strong lines 4434, 4454, were reversed, but 

 only when one of the poles was a bar of iron, instead of carbon. The 

 strong lines just mentioned were expanded so as to cover their neigh- 

 bours, and all four lines were seen black against the bright back- 

 ground in the positions and of the same relative strengths as when 

 bright. 



When strontium chloride was put into the crucible the lines with 

 the following wave-lengths, besides those before noted, were observed 



