The Principles of Spectrum Analysis. 365 



metal in a state of vapour has the power to arrest particular 

 rays with a constancy that can be relied on. The arrangement 

 best suited for these experiments is either Dubosq's electric 

 lamp., or the Drummond light, but many of the spectra may be 

 conveniently studied by using Crookes's spectroscope, as made 

 by Spencer Browning and Co., and now too well known to 

 need detailed description. This instrument is well adapted for 

 ordinary purposes, but to appreciate the full beauty and delicacy 

 of the various spectra, we should need an apparatus as perfect 

 as that constructed for Kirchhoff by Steinheil of Munich. 



When artificial light is employed — as that of gas or lamp — 

 the dark lines may be brought out by interposing a glass trough 

 or bottle containing nitrous acid gas between the light and the 

 instrument. This gas may be obtained by the action of a small 

 quantity of nitric acid on a piece of copper ; and, as we have 

 before mentioned, it acts as an absorptive medium. If a piece 

 of sulphur be introduced into the flame of a spirit-lamp, a good 

 view of its dark bands may be obtained. If the subject for 

 examination be an alkaline metal, the spirit-lamp may be used, 

 or better still, a flame of hydrogen mixed with air and burnt on 

 the top of a tube covered with wire gauze. We thus obtain 

 a flame of high temperature with little light, except what is 

 derived from the substance employed. The metal in a state 

 of chloride is the most convenient form — it being more easily 

 volatilized. It may be introduced into the cotton wick ; or if 

 the gas-burner be used, then a loop of platina wire sliding on 

 an upright support is the easiest to manage. The copper 

 spectrum may be readily obtained by dipping a coil of fine 

 wire into pure hydrocloric acid, and immediately inserting it 

 into the gas-flame. When iron, silver, etc., are operated upon, 

 wires of these metals should form the electrodes of a powerful 

 voltaic battery, and be brought by its agency to an incandescent 

 state, when a portion of their substance is volatilized, and ex- 

 hibits its characteristic action through the prism. 



Kirchhoff and Bnnsen, while pursuing their researches on the 

 composition of some mineral water, obtained from the combus- 

 tion of the solid matter a series of bands in two spectra, which 

 did not correspond with those produced by any of the known 

 metals. This led them to infer the presence of some new 

 elements which the eye of man had never yet seen. After 

 evaporating several tons of the fluid, their labours were re- 

 warded by obtaining two new metals, which they named 

 Ccesium — (greyish blue) that being the colour of the bands — 

 and Eubidium (dark red) . No sooner had they done this than 

 they were off into the depths of speculation, conceiving that 

 they had in their power a means of analysis capable of much 

 higher application. 



