212 THE ROYAL SOCIETY OF CANADA 



vestigation and so making sure that the arc stream was always 

 charged with the vapour of the element. When this condition existed 

 the arc made a loud hissing sound, and also showed a colour in the 

 core, that was more or less characteristic for each of the metals used. 

 If the arc burned more quietly and showed that the undesirable 

 spectrum of carbon was being produced, the circuit was broken, the 

 carbon freshly charged, and the arc again struck. 



Under these conditions photographs were taken of the spectra 

 of tin, lead, bismuth, zinc and antimony. The exposures were all of 

 fifteen hours duration with the exception of bismuth, which was given 

 an exposure of forty hours. 



In order to make an accurate measurement of the wave-lengths 

 a photograph of the ordinary mercury spectrum in second order was 

 taken on each plate for purposes of comparison. The plates were 

 carefully measured with a vernier microscope and the lines identified 

 are gi-ven in Tables II, III, IV, V and VI. All the lines identified by 

 Walters and Randall between X = 5500A to X = 22000A are given as 

 well for comparison although in the present series of experiments 

 only the region between X = 8000A to X = 11000A was investigated. 

 The reproductions shown in Fig. 25 are typical of the results obtained. 



In conclusion acknowledgment is made to Professor J. C. 

 McLennan for his interest, encouragement and his many valuable 

 suggestions in this work, as well as to Mr. V. P. Lubovich for occasional 

 assistance in technique. 



The Physical Laboratory, 

 University of Toronto. 

 June 1st, 1922. 



