Determination of Minute Amounts of Lead in Water, with Notes on Certain Causes of Error

IN the determination of very small amounts of lead, the usual analytical methods are far from being satisfactory. Small amounts of lead may be present in the reagents empioyed, and it is customary to run a blank experiment to correct the error arising in this way. When the correction is of small magnitude compared with the amount of lead in the sample, this procedure may be adopted; but when very small amounts of lead are being determined it may happen that the lead found in the blank is as much, or even more, than the lead actually present in the sample, and the analysis has to be rejected. Moreover, the blank experiment cannot check errors which are not common to the blank and the analysis itself eg, the apparatus used in the parallel experiments, etc. It seemed to be desirable, tl;l.erefore, to make an investigation of the problem to ascertain, as far as possible, the nature of the errors arising from the introduction or loss of lead during the analysis, and how far these could be eliminated or corrected for; also, to determine the limits of accuracy of the improved method and the minimum amount of lead that could be reliably estimated.In the investigation, reliability and sensitiveness were regarded as essential features. Speed of operation, on the other hand, was considered of less importance. The amounts of lead dealt with being usually less than 1 mg, there is no alternative to the use of a colorimetric method. The latter, depending as it does upon the development of a brown coloration when ammonium sulphide is added to a very dilute -solution of a lead salt, is far from being a characteristic test for lead. Indeed, it is well known that the salts of most of the commonly occurring metals interfere with this test. Hence it is of vital importance that the lead should be isolated in the form of a pure compound possessing quite definite and characteristic properties before the colorimetric estimation is proceeded with...