SPECTROSCOPY IN INDUSTRY — HARRISON 211 



in plate 4, figure 1 (right). Magnifications up to 2,000 diameters 

 could be used, and thus extremely small amounts of hafnium could be 

 located, since a layer of "hafnium" atoms only 5 atoms thick will give 

 an observable degree of opacity. 



Most industrial spectroscopic problems are simpler and more 

 readily solved than the one sketched above, but it is described as 

 illustrating a number of different methods of using a spectrograph. 



For many industrial purposes it is important that speed be com- 

 bined with precision. Two hundred tons of molten copper in a 

 furnace represents a large investment which can be lost through im- 

 proper treatment. The copper must be removed from the furnace 

 at the proper degree of purification, and several hours may be con- 

 sumed in determining by chemical methods the proper time for re- 

 moval. Often recourse is had to rule-of-thumb methods, for lack 

 of anything better. An ideal solution would be a spectroscopic ar- 

 rangement such that the furnace operator could periodically dip 

 some molten copper from the furnace, burn it before a spectroscope 

 and decide from the appearance of certain impurity lines exactly 

 when the proper time for removal was reached. There is no funda- 

 mental reason why this operation should take more than 2 minutes if 

 performed visually or with a photoelectric cell. 



Eye methods can be used in certain cases only, since many of the 

 important lines for analysis lie in the ultraviolet region. It is, to 

 be sure, possible to use fluorescence to observe ultraviolet lines, but 

 this method has not yet been well developed. It awaits an increased 

 demand from industry, which in turn awaits knowledge of the feas- 

 ibility of such methods. 



The photoelectric method of spectroscopic analysis has been tried 

 with encouraging results in Germany and at the Watertown Arsenal 

 in this country and shows considerable promise, though it must be 

 considered as in only a preliminary state of development. The limi- 

 tation on this technique, which is simple and should be capable of 

 precision, is in getting enough light to operate photocells from a 

 single line of an element which is present in the sample in only 

 small amounts, when this line is in the midst of a bewildering array 

 of strong lines from other elements. Spectroscopes of high speed, 

 combined with improved phototubes and amplifiers, and improved 

 sources will probably bring ultimate perfection of this rapid method. 



The photographic method can be speeded up greatly, and in at least 

 one foundry, that of the Campbell, Wyant & Cannon Co., of Muske- 

 gon, Mich., most of the routine chemical tests on the output of the 

 foundry have been superseded by rapid spectrographic methods. 



197855 — 40 15 



