SPECTROSCOPY IN INDUSTRY — HARRISON 



209 



concentrations below 0.1 percent is ordinarily greater than that of 

 chemical methods. 



Having determined the actual amount of hafnium present, we are 

 now ready to determine its location in the crystals. The first method 

 tried was to pick out small particles of the crystal and burn them in 

 the arc, observing visually the intensities of the visible hafnium lines 

 with a spectroscope. A sample the size of a pinhead was found to 



.0021 

 .0001 



.001 .01 



PERCENT 



0.1 1.0 I O 



CONCENTRATION 



100 



FiGUEB 1. — A typical workiug-curve, for deterniiuing the conceutratioii of an element present 

 in a sample burned in the electric arc, In terms of the intensity of one of its spectrum 

 lines photographed under standard conditions. The logarithm of the percent concen- 

 tration of the element is plotted against the logarithm of the intensity of the line, as 

 determined from a number of samples of known concentration. At low concentrations 

 the curve is theoretically a straight line having 45° slope, which greatly simplifies 

 quantitative spectrographic analysis at low concentrations. 



give fairly consistent results, but particles less than 1 mm. in diameter 

 were too likely to be lost from the arc. A series of runs was made 

 to determine the minimum amount of hafnium which could be de- 

 tected spectroscopically. A weighed sample of the material, known 

 to contain 10"° gm. of hafnium, was introduced into a 2-amp. arc 

 between pure graphite electrodes. Strong hafnium lines were found 

 to be still visible in the spectroscope after 1,000 seconds of burning. 

 The spectrum was then photographed with a high-speed spectrograph, 

 and an exposure of 0.0001 seconds was found sufficient to record the 



