1879.] On the Reversal of the Lines of Metallic Vapours. 405 



the appearance of these two lines, or bring out a reversal of either 

 of them. We failed to see any line of silver either bright or re- 

 versed with wave-length about 4240, as mentioned by Cornu, whether 

 the silver were nsed pure or alloyed with zinc or with lead. With 

 the carbons -arranged vertically as shown in our last communication 

 on this subject, p. 472 (vol. xxviii), and the light viewed through the 

 perforated carbon, silver gave a channelled spectrum as described by 

 Lockyer and Roberts. As this channelled spectrum was not seen with 

 silver in any other arrangement of our crucibles, we are led to attri- 

 bute it to a comparatively cool condition of the silver vapour ascending 

 the carbon tube, a condition of near approach to a state of liquefaction. 



Having observed that lines frequently came out with mixtures 

 which were not visible when the separate ingredients were used, we 

 tried a few amalgams. None of these showed any reversals of the 

 mercury lines. But an amalgam of bismuth gave readily the 

 reversal of the bismuth line, wave-length 4722, and with more difficulty 

 that of the line, wave-length 4119. 



Antimony did not appear to give any lines, or none easily dis- 

 tinguishable, in the arc. 



With copper we observed the reversal of two lines only, wave- 

 length 5105, 5153. 



Iron introduced as metal, or as chloride, in the usual way, gave us 

 no reversal ; with an iron rod used as positive pole instead of one 

 of the carbons, we succeeded in getting the reversal of one line, wave- 

 length 4045, which expanded and showed a fine dark line in its middle ; 

 bat by passing an iron wire into the arc through the positive carbon, 

 which was perforated in the manner described in our last communica- 

 tion on this subject (p. 472), for the introduction of gases into the 

 arc, and pushing in the wire slowly as the end burned away, we 

 succeeded in reversing several of the brightest of the iron lines. The 

 three violet lines, wave-lengths 4045, 4063, 4071, were the first to be 

 reversed They all expanded before showing reversal, and the order 

 of reversal was that of refrangibility. Besides these the lines, wave- 

 lengths 4957, 4920 (the stronger and less refrangible of a pair), 4404, 

 4383, 4325, 5040, 4307 (G) were reversed. These last two lines could 

 not be certainly distinguished from calcium lines, as the last is almost if 

 not quite coincident with a calcium line, and the other is too close to 

 a calcium line to be distinguished without more exact measurement 

 than the time during which it could be observed would allow, but 

 we judged from their relative intensity that they were the iron lines. 

 The line, wave-length 4415, was not seen reversed when the other 

 bright lines near it were reversed. 



riickel, whether put into the crucible in the old way, or fed into the 

 arc in small fragments filling a platinum tube which was passed 

 through a perforated carbon pole, gave no definite reversal of any of 



