Line Spectrum emitted by Iron Vapour. 283 



lias not enabled me to resolve these doublets, but according 

 to Dr. King's observations of the high- temperature furnace 

 spectrum of iron, made with a high dispersion, both com- 

 ponents of each pair show in the furnace spectrum with 

 approximately similar relative intensities as in the arc*. 

 Hence the behaviour of these lines in the furnace is such as 

 to suggest their emission being governed by electric rather 

 than thermo-chernical or chemical actions. 



§ 3. Probable Mode of Excitation. 



The forcing of heavy electric currents through the carbon 

 tube resistance-furnace entails the establishment between its 

 extremities of a certain potential difference, the value of 

 which depends upon the resistance of the tubes and the tem- 

 perature to which it is to be raised. Now, if with the rise in 

 temperature the gases or vapours enclosed in the furnace 

 became progressively ionized to a high degree, a stage should 

 be reached at which part of the heating current will be 

 carried by the ionized vapours according to the fundamental 

 laws of electric conduction. The well-known experiments 

 by Drs. Harker and Kaye t on thetionization in tube resistance- 

 furnaces have furnished most important data on this point, 

 and their results leave no doubt as to the relatively high 

 conductivity of the ionized gases within the furnace-tube. 

 In a first experiment these physicists showed that it was 

 possible to send an electric current across a gap between two 

 carbon rods held concentrically in the middle of the tube. 

 With small potentials (up to 6 or 8 volts) measurable values 

 of the current were obtained when the temperature rose 

 above 1400° 0., and at 2000° furnace temperature it reached 

 the value of several amperes. In their own words: "The 

 magnitude,of the ionization currents indicated that, although 

 the pressure was atmospheric, the atmosphere of the furnace 

 was ionized to an unusual degree at high temperatures."" In 

 a further experiment, and one of still greater importance 

 from our point of view, Drs. Harker and Kaye measured the 

 current which leaks across the highly ionized space sur- 

 rounding a heated carbon rod passing concentrically through 

 a carbon tube. In this case the ionization current is part of 

 the heating current supplied to the rod, and it flows across 

 the ionized space when one end of the rod is joined to the 



* A. S. King, Astrophysical Journal, vol. xxxvii. p. 250 (1913). 

 t J. A. Harker and G. W. C. Kaye, Proceedings Royal Society, 

 Series A, vol. Ixxxvi. p. 379 (1912) ; ibid. vol. lxxxviii. p. 522 (1913). 



U2 



