418 Prof. E. Wiedemann on the Behaviour of Gases 



cases, the discharge of the battery, according to the conditions 

 existing in the battery itself, is either altogether discontinuous, 

 or takes place in such a way that stronger discharges cause 

 the chief discharge, upon which follows a continuous discharge 

 gradually becoming weaker, and so on. 



But even if the discharges are completely continuous, I do 

 not believe that they can be employed in spectroscopic inves- 

 tigations, since the final temperature of the gas depends on so 

 many circumstances which cannot be exactly determined — 

 for example, upon the radiation-coefficient, which varies with 

 the temperature. 



Theoretical Considerations. 



If we endeavour, from the above facts and others yet to be 

 mentioned, to represent to ourselves the mode in which the 

 electric discharge takes place through a gas, we see, in the 

 first place, that the theory proposed by G. Wiedemann and 

 B. Kuhlmann is not supported by these facts. According to 

 this theory, the discharge is carried on by molecules charged 

 with electricity which are driven off from the electrode, and 

 which, upon collision with other molecules, yield up their 

 electricity to them. The same theory has been adopted, in a 

 somewhat different form, by Crookes to explain the pheno- 

 mena described by him, but most of which, though no doubt 

 without his knowledge, had been long ago made known by 

 Hittorf, and then by Goldstein and others. According to this 

 theory, the molecules must possess velocities corresponding 

 to the velocity of propagation of electricit}^ in gases. But we 

 know, from Wheatstone's* experiments on the discharge in 

 gases, that this is certainly greater than 200,000 metres, or 30 

 (German) geographical ( = 124 British statute) miles. But 

 it is quite certain that the molecules do not possess so great a 

 velocity of translation measured in the direction of the cur- 

 rent. Dr. von Zahn has observed the lines of the spectrum 

 given by a Geissler's tube, first when the axis of the tube and 

 the axis of the collimator of his spectroscope were parallel, 

 and then when they were at right angles, and observed no 

 displacement of the lines, although the dispersion was so 

 great that a displacement of -^ of the distance between the 

 D-lines could have been observed. A displacement of this 

 magnitude would, in accordance with Doppler's principle, 

 have corresponded to a velocity of 4 geographical miles ; and 

 a velocity of 30 geographical miles would have produced a 

 displacement nearly equal to the entire interval between the 



* Wheatstone, Pogg. Ann. xxxiv. p. 464, 1835. 



