SPECTROCHEMICAL ANALYSIS 



391 



co-workers,* who use a "stepped diaphragm" to calibrate each plate 

 with an intensity pattern in order to restrict their density measure- 

 ments to the straight-line portion of the characteristic curve of the 

 plate or to correct for deviations from the straight line portion. In 

 this laboratory a step sector (Fig. 6) has recently been used to put the 

 intensity pattern on each plate and the characteristic curve (Fig, 8) 

 determined from this, using a photoelectric densitometer built in this 

 laboratory. It is obvious from this curve that the measurement 



0.8 



.9- 



0.6 



0.4 



0.2 



I 2 3 4 5 6 



LOG(.5 RELATIVE EXPOSURE 



Fig. 8 — A typical characteristic curve (exposure = intensity X time). 



of the ratio of intensities of a given pair of lines varies with the densities 

 of the lines as recorded on the plate. In this way errors of considerable 

 magnitude may be introduced unless a calibration is made of the 

 response of the plate to various levels of intensity of light of the wave- 

 lengths concerned. Results to date have shown an increased pre- 

 cision in all cases where plate calibration has been used, and a pre- 

 cision of better than ± 5 per cent of the amount determined is not 

 uncommon. E.xamples from our experience include tin in lead- 

 antimony cable sheath and other lead alloys (Fig. 7), magnesium in 



^ Duffendack, Wolfe, and Smith, Indus, cf Engg. Client. {Anal. Ed.), 5, 226 (1933) 

 and others. 



