Subsurface Logging Methods 491 



eral background fog and to the intensities of the carbon and cyanogen 

 bands from the electrodes, but there seems to be no way to eliminate this 

 diflSculty, which becomes exaggerated when a wide slit is used. 



Excitation 



The direct-current arc is the simplest form of excitation that is used, 

 as it requires very little equipment, and, as it operates satisfactorily at 

 voltages between 100 and 250, it is comparatively safe. It has the reputa- 

 tion of being less stable than other types of arc owing to wandering of the 

 cathode "spot" and, consequently, of lack of reproducibility. Hampton 

 and Campbell,^^ however, have recently demonstrated that the direct- 

 current arc is capable of high reproducibility under carefully controlled 

 conditions. 



The high-voltage alternating-current arc was introduced by Duffendack 

 and Thompson ^^ to avoid the irregularities of the direct-current arc. It 

 has acquired a good reputation in spectrochemical work, but the writers 

 have had no experience with it. It requires more equipment and safety 

 precautions than the low-voltage arc. 



In all types of arc excitation care must be taken that the entire speci- 

 men is volatilized, otherwise serious errors will be introduced. For in- 

 stance, although there is some overlap, the order of volatilization in a 

 limestone is sodium, calcium, silicon, and last titanium. Interruption of 

 the arc before the titanium has boiled off obviously leads to error in de- 

 termination of the titanium content of the limestone. The arc draws a 

 heavy current, as is shown by an ammeter in the circuit, during the period 

 of volatilization of the major elements in the sample, and when this is 

 complete for those elements the current abruptly drops to the steady 

 value of the graphite arc. Unfortunately there is no significant corres- 

 ponding drop to indicate the completion of volatilization of what are 

 usually minor quantities of titanium or of other refractory elements. It 

 seems best under these conditions to operate the arc for some time after 

 the current drops so as to insure complete removal of the sample and to 

 design the shape of the lower electrode so that the walls are thin and burn 

 evenly down to the bottom of the cup. 



Instrumentation 



To utilize the spectrochemical method to its greatest advantage a 

 spectrograph of high dispersion, yet of relatively high speed, should be 

 available. Large quartz spectrographs of the Littrow type are advan- 

 tageous and should be equipped with glass prisms to obtain adequate 

 dispersion in the visible region. The Gaertner, Bausch and Lomb, and 

 Hilger instruments are typical of this class. In recent years the grating 



^ Hampton, R. R., atd Campbell, H. N., Measurement of Spectral Intensities in the Direct Current 

 Arc: Optic. Soo. America Jour., vol. 34, no. 1, p. 12, 1944. 



"Duffendack, O. S., and Thompson, K. B., Developments in the Quantitatiwe Analysis of Solutions 

 by Spectroscopic Means: Am. Soc. Testing Materials Proc, vol. 36, no^ 2, p. 301, 1936. 



