[ainslie-fuller] ultra-violet OF CERTAIN METALS 



77 



measurements down to A1872-94, but his results do not correspond 

 closely with those given below. This may be due to the fact that he 

 used a spark in air, rather than the vacuum arc as used in this work. 



From A2400 to A2000 the experimental error is probably as 

 much as one angstrom unit. This is due to the fact that the disper- 

 sion in this region is small, being about -OSmm. per angstrom unit; 

 also the lines obtained on the photographic plate are not as well 

 defined. These measurements are given for a guide for future re- 

 search work with vacuum arc sources. 



From A2000 down, the probable error is on the average, well 

 within half an angstrom unit. The dispersion varies from • 067mm. 

 at A2000 to •20mm. at A1400 per angstrom unit. 



Summary of Results. 



The vacuum arc spectra of copper, zinc, aluminium, carbon, 

 iron, tin, lead, thallium, nickel and cobalt have been studied from 

 A2400 to A 1400. The vacuum arc spectra obtained for copper, 

 zinc and aluminium were found to correspond with the results obtained 



