Fluorescent X-ray spectrometers bave a high inherent reproducibility. It is possible to achieve results, on a routine basis, which have a coefficient of variation of less than 0.1%. Given extremely good chemical standards and the absence of interelemental effects, these readings can be interpreted into percentage concentrations with the same order of accuracy. Results, with this accuracy of analyses of copper in a straight brass will be given. With samples of greater complexity and more elements the interelement effects are large and have to be tackled. A correction method has been established and will be explained in detail.

As an example of the method it was decided to tackle the determination of chromium at high concentrations (7-32 %) in high-alloy steels containing additions in the following ranges: Si, 0.2-2.4%; Al, 0,07-5%; Cu, 1,3-3.6%; W, 0.25-3%; V, 0.1-0.3%; Nb, 0.1-0.3%; Mo, 0.1-4.4%; Ti, 0.05-1%; Mn, 0.35-10%; Ni, 0.2-24%. These very wide alloy specifications were a severe test on the accuracy of the chromium determination & the correction for interelement effects.