492 HISTORY OF SCIENCE. 



atmosphere appear in the spectrum, as they do in the spectrum of each 

 other metal under like circumstances. If hydrogen gas surrounds the 

 poles, then only the lines of the metal, together with those of hydrogen, 

 are visible, and so on. A small portion of the surrounding gaseous 

 medium is, in fact, heated to incandescence by the electric discharge, 

 and a small quantity of the metal of the poles is also volatilized, and 

 in the state of glowing gas yields its characteristic lines. Hence spark- 

 spectra and flame-spectra appear to depend upon this principle : Each 

 chemical element when in the state of a glowing gas gives off a certain 

 set of rays of definite refrangibilities. Further, it appears that some of 

 these rays are given off only when the temperature of the vaporous 

 substance is very high, and this is the condition obtained by the electric 

 discharge. 



Lines peculiar to the gases of the atmosphere surrounding the me- 

 tallic poles become visible in the spectrum of the spark, as already 

 remarked, when the discharge is intense. When the discharge is made 

 to take place in a highly rarefied gas, as in the Getssler Tubes (Chapter 

 XX.), the spectrum lines peculiar to the gas are obtained with great bril- 

 liancy. These tubes contain but extremely small quantities of gas ; but 

 when the sparks of an inductive coil are allowed to stream through them, 

 they glow with light, which in many cases presents a characteristic colour 

 even to the unaided eye. Thus a hydrogen tube seems filled with red 

 light, a carbonic acid tube with blue light, and so on. This light, like 

 that of a substance heated in the flame of a Bunsen's burner, is attri- 

 buted to the incandescence of the minute quantity of gas in the tube, 

 and if so, it represents a temperature far higher than any which can be 

 obtained in flames. The extreme smallness of the quantity of the gas 

 will explain the absence of the very marked heating effects which a 

 larger mass at a temperature of a like order would produce. 



When the light of the discharge in Geissler's tubes is examined by 

 the spectroscope, the characteristic lines of the gas are very distinctly 

 seen. Hydrogen gas, for example, gives a spectrum of three distinct 

 lines, one dark blue, one greenish-blue, and one red; the red line 

 being much more intense than the others. Hence the red light which 

 appears to fill the Geissler's hydrogen tube. The redness of the ordi- 

 nary electric spark in air is due to the hydrogen contained in the 

 aqueous vapour of the atmosphere. The spectra of nitrogen, oxygen, 

 and other gases are much more complicated than that of hydrogen. 

 Now, while the electric spark shows the spectra of the metals, the 

 method of observation by the discharge in Geissler's tubes gives the 

 spectroscopist the characteristic lines of all the gaseous elements. Thus 

 the whole of the chemical elements may be identified by the spectro- 

 scope. Many chemical compounds give, under certain circumstances, 

 lines which do not belong to the elements of which they are com- 

 posed, and these have been supposed to belong to some compounds 

 of the substances undecomposed at the temperature of the flame. 



