AbstractIn an attempt to use the equipment available for the determination of the exchangeable cations in soils, u statistical analysis has been made of the errors in the spectrochemical results obtained by means of an air‐acetylene flame and a large quartz, spectrograph. Even with an elementary technique, an average standard error of less than 4% was obtained. Factors contributing to the total error, such as the error of a microphotometer reading, fluctuations of the flame, and variations of the photographic emulsion, have been studied; two types of plate, sensitised and non‐sensitised, have been examined, It has been shown that most of the total error was due to variations in the photographic plates, and the significance of this in relation to the placement of the standard spectrograms on a plate is discussed. It has been found that, in general, a large quartz spectrograph is not suitable for use with an air‐acetylene Hume because the very low light‐collecting power of the instrument necessitates very long exposures or the use of solutions of impracticably high concentrations, or sometimes both.