358 Prof. E. Wiedemann on the Behaviour of Gases 



It has been shown by these investigations that the usual 

 theory, which refers the emission of light by a gas under the 

 influence of the electric discharge to an increase of tem- 

 perature up to the point of incandescence, is no longer tenable, 

 and that this point requires fresh experimental examination. 



Spectrum analysis, so far as it employs the electric dis- 

 charge for the production of spectra, has to examine the 

 question of the dependence of the spectra produced on the 

 quantity of electricity transmitted. For this purpose we 

 must pass discharges of given quantities of electricity through a 

 given gas of definite pressure, observe the spectral phenomena, 

 and at the same time determine the magnitude of the quan- 

 tities of energy or of heat given off by the discharges ; for 

 these alone can furnish a measure of the resultant phenomena 

 of motion in the gas. As we cannot deal with a single dis- 

 charge, we must determine the number of discharges which 

 correspond to a given quantity of electricity. It is abso- 

 lutely necessary that the gas should return to its original 

 condition before the entrance of each new discharge, so that 

 any previous heating may have ceased to be sensible, and 

 especially that the gas should have become dark again. 



The present research is intended to be a first contribution 

 to the solution of this unusually complicated question. 



In the first place I have thoroughly examined the thermic 

 relationships of the discharge of the induction-machine 

 under different conditions, and in doing so have observed a 

 peculiar behaviour of positive and negative electricity. Then 

 follow experiments to determine numerically the conditions 

 under which the transformation of the band spectrum into 

 the line spectrum occurs in the case of hydrogen, as well as 

 some contributions to our knowledge of the discharge taking 

 place from the negative electrode in highly rarefied gases ; 

 then follows a discussion of the employment in spectrum ana- 

 lysis of the other sources of electricity — induction coils, large 

 galvanic batteries, and Leyden jars, and of the continuous and 

 discontinuous discharge in gases. 



Theoretical considerations on the discharge in gases and on 

 the nature of spectra form the conclusion. There are parti- 

 cular subjects more briefly treated in this first paper, which 

 shall be more fully investigated later. 



2. Apparatus. 



In correspondence with the problems for investigation, the 

 apparatus employed consisted of three parts : — (1) the source 

 of the electricity, and the arrangements to measure the inten- 

 sity of the current; (2) the discharge-tube and its electrodes, 



