172 THE ROYAL SOCIETY OF CANADA 



of the quartz spectrograph was directed at the incandescent plate G. 

 A short tube of asbestos cloth was attached to the quartz tube directly 

 in front of this plate so that the radiation from the arc passed through 

 it to the slit of the spectroscope. This arrangement was found 

 necessary to cut ofï the radiation from the Bunsen flame itself. 



With the arrangement just described, it was found that when 

 the direct current 110-volt circuit with suitable resistance in series 

 was applied to the terminals A and B, and the plate G brought to 

 incandescence strong arcs could be maintained for hours with both 

 cadmium and zinc. With the 220-volt circuit applied the arcs of 

 these two metals could be made most intense and could be maintained 

 for long periods. With the 220-volt circuit it was found that when the 

 arc was once struck it could be easily maintained for a considerable 

 time without the continued use of the oxy-cathode G. With low 

 voltages, however, it was always necessary to maintain the plate G at 

 incandescence in order to keep the arc established. 



A photograph of the cadmium arc spectrum taken with a lamp of 

 this form is shown in the upper row of Fig. 6. As the illustration 

 shows, there is a marked diflference between the arc and spark spectra 

 of cadmium, numerous lines coming out in the one which do not appear 

 in the other. Another point of interest in connection with these 

 spectra is that in the arc spectrum there was a reversal at A = 2288*79 

 A°.U. and also another though less clearly marked at A = 3100 A°.U. 

 In the arc spectrum no reversal at X = 3260-17 A°.U. was observed. 



IV. The Absorption Spectrum of Zinc Vapour. 



Although a number of observers had looked for absorption bands 

 in the absorption spectrum of zinc vapour none was observed until a 

 short time ago, when a well defined band was noted by one of us at 

 A = 2139-3 A°.U.^ The reason that this band had not been found 

 before was that it was far down in the ultra-violet beyond the range 

 examined by the other investigators. It is shown in Fig. 8. The 

 upper spectrum is that of the zinc spark in air and the lower two are 

 the spectra obtained when the light from the zinc spark in air was 

 passed through zinc vapour in a heated evacuated quartz tube. It 

 will be seen that with increasing vapour density the band developed 

 symmetrically. Photographs of the absorption spectrum of zinc 

 vapour were also taken with the larger spectrograph. One of these is 

 shown in Fig. 9. The upper spectrum is that of the zinc spark in air 

 taken directly after it had passed through zinc vapour. The absorp- 

 tion corresponding to A = 2139-3 A°.U. as will be seen is extensive and 

 clearly defined. It will also be seen that there is a narrow band at 



îMdLennan, Phil. Mag., Vol. XXVIII, Sept. 1914. 



