SPECTROSCOPY IN INDUSTRY — HARRISON 207 



length, which makes identification of the lines easier and uses the 

 photographic plate more effectively. The new aluminum gratings on 

 glass, ruled by R. W. Wood at Johns Hopkins, have done much to 

 remove prejudice against the grating because of the comparatively 

 low intensity of the spectra it produces. Several of Wood's gratings 

 produce spectra more intense than those produced by prisms, and 

 with such a grating exposures in which thousands of lines from an 

 iron arc are recorded in 5 seconds are not uncommon. 



Some industrial users of spectographs who express great satisfac- 

 tion with their prism instruments consider the grating spectrograph 

 too fragile for use in their laboratories. This speaks well for the 

 prism type of instrument, but usually bespeaks lack of familiarity 

 with the grating. Wliere good prism and grating instruments are 

 both available, the grating instrument is commonly the favorite, 

 and that it cannot be the delicate apparatus sometimes supposed is 

 borne out by its very satisfactory use in such metallurgical labora- 

 tories as those of the Watertown Arsenal and the Cleveland Lamp 

 Works. 



As a typical example of all-around use of the spectrograph in 

 attacking an industrial problem, I shall discuss the spectroscopic 

 procedure connected with what we may call "The Case of the Haf- 

 nium Crystals." (The real crystals were not hafnium, nor were they, 

 in fact, crystals.) The spectroscopic procedure used was so varied 

 that the case serves as an excellent illustration. The problem was to 

 locate minute deposits of hafnium in a certain transparent crystalline 

 mineral and to measure the distribution and amount of the hafnium 

 in typical crystals. The spectrographic method was selected as one 

 of the best ways of doing this. 



First, emission analysis was used to determine the amount of 

 hafnium present in the crystal material. Several milligrams of the 

 material were burned in an electric arc, and the light emitted was 

 photographed with a spectrograph. The wave lengths of all lines 

 appearing on the plate were then determined with an accurate meas- 

 uring engine, and from published catalogs of wave lengths all lines 

 appearing on the plate were assigned to their parent atoms. A 

 qualitative analysis of all metallic elements present in amounts 

 greater than about one part in a million was thus obtained. 



A spectrographic quantitative analysis was next made. From the 

 qualitative analysis the elements could be grouped into three classes 

 by observing the intensities of their spectrum lines. Those which 

 were present in large concentrations (above about 1 percent) were 

 listed as major constituents, those present in amounts lyuig between 

 1 and 0.001 percent were listed as minor constituents, and those 

 present in smaller amounts as traces. 



