A revision of the atomic weight of thallium

Abstract

Since T. W. Richards in 1888 first devised the method of nephelometric titration and applied it to the precise determination of the ratios of soluble halides to silver, it has been the mainstay of atomic weight determinations in all countries. This is proof enough of the great advance he thus made in the accurate measurement of the fundamental ratios of chemistry. Yet, valuable as the method of nephelometric titration has been, it is, as many who have used it will admit, both extremely tedious and liable to uncertainties much greater than would appear from the results which are published. Despite the most careful observance of the conditions laid down by Richards and subsequent workers, it is a common experience to find that even after standing for the normal time, 7-14 days, a liquor showing on the first test approximate equivalence of silver and chloride, will, on adding small quantities of precipitant behave quite erratically, so that in extreme cases addition of chloride may, in subsequent tests, produce apparently an increase in dissolved silver, and vice versa. Thus, even where nephelometric tests proceed normally, there is always a fear that apparent equivalence of silver and halide may be illusory; and a considerable experience of such tests has given us the impression that in a normal titration, involving 5-10 grams of silver in, say, 4-6 litres of solution, the resulting uncertainty in the weight of silver is of the order of 0•0005-0•0002 gm. It is, further, no slight disadvantage that a nephelometric titration can hardly be completed in less than 2 1/2-3 weeks and usually extends over a period of 4-6 weeks. One of the principal objects of the present work has been, therefore, to devise and test a method of titrating halides with silver which shall be more rapid and precise. A good trial of any such method is afforded by the determination of the atomic weight of a heavy element of small valency, for which, evidently, the ratio of the halide to silver is unfavourable to the attainment of precise results. It was in part for this reason that the case of thallium was chosen. Apart from any question of method, moreover, there was good reason for a new determination of this atomic weight.