ATOMIC WEIGHT O F POTASSIUN
639
Even when every conceivable precaution is taken, a single salt is not an adequate basis for the certain decision of an atomic weight. For this reason a parallel investigation on potassium bromide was simultaneously in progress a t the Chemical Laboratory of Harvard College. T h e next communication, describing this other research, must be considered in connection with the work which has just been described. As will be seen, excellent confirmation of the present work is afforded by the work with the bromide.
Summary. This investigation concerning the quantitative composition of potassium chloride resembled in many respects the recent investigation of Richards and Wells on sodium. I n several details, however, improvements were introduced which effected a considerable saving of time and a perceptible gain in accuracy. T h e precautions necessary for the accurate use of the Gooch-Munroe perforated crucible were ascertained ; its employment was found to be advantageous. Platinum weighing bottles with conical ground-platinum stoppers were used instead of boats and glass tubes for weighing the potassium salt. Occlusion of silver nitrate by the precipitated chloride was diminished by allowing the latter to stand for a long time in a solution containing neither excess of silver nor excess of soluble chloride, and by adding more silver nitrate only after the precipitate had assumed a fairly permanent condition of aggregation. T h e nephelometric estimation or" small amounts of suspended silver chloride was increased in accuracy by redissolving in amnionia both of the opalescent precipitates to be compared, and reprecipitating, in order to equalize the conditions. As final results, the outcome of twelve experiments, IOO.OOO parts of silver were found to correspond to 52.01 18 parts of potassium chloride, and IOO.OOO parts of silver chloride were found to correspond to 69.1073 of this salt. T h e corresponding values for the atomic weight of potassium (if silver is assumed to be 107.930 and chlorine 35.473) are 39.1134 and 39.1145, in unusually close agreement. [CONTRTBUTIOH FROM T H E C H E M I C A L
LABORATORY O F HARVARD COLLEGE].
A REVISION OF T H E ATOMIC WEIGHT OF POTASSIUM. YSIS O F POTASSIUM BROMIDE
THE ANAL-
Bs THEODORE WILLIAMRICHARDS A N D EDWARD MUELLBR. Received February 15, 1907.
Introduction T h e foregoing quantitative study of potassium chloride by Arthur Staehler and one of the present authors affords strong evidence that
640
THEODOKE W I L L I A ~ I R I C H A R D S ASD
EDWAXI) IIYELLER
the atomic weight of potassium is ahout 39.I 14, slightly l o n e r thaii the \.alue based upon the work of Stas. T h e authors fully appreciated, Iiowever, that the investigation of a single compound is not enough to establish a clieniical coiistaiit so important as this, and, accordingly, tlie present investigation was prosecuted simultaneously. It was expected that tlie two investigations niiglit either support one another or else, by affording incoiiipatible resiilts lead to the discorery of a constant error i n one or the other, ailti thus pave the ivay for further advaiice i n kiion.ledge. X s will be seeii, tlie \ w r k oii tlie chloride w a s satisfactorily coiifiriiied by the work 011 the bromide. In the present case tile careful study of potassiuni bromide Ivas particularly iitcessary, liecause there already exist tn-o concordaiit series of experiments upoii this siihstance, performed b y tlie old masters hIai.igi:ac and Stas, pointing to a v a l u e i r i the rieigliborliood of 39.14) instead of t h e 1,efore-nieiitioiied new value 39. I I . 111 this case the higher value for potassiuiii is not diniinislicdj as in the case of tlie chloride, b y an additi1.e correctioti i n the atoiiiic \wight of the lialogen, because Baxter h a s sliown Stas's estiiiiate for bromine to have beeii nearly correct.' Helice the discrepancy reiiiaiiis one too serious to lie tolerated "lie careful study of this work of Stas arid Marignac affords convincing evideiice that the potassium lironiide used by t h e m for aiialysis w\.as not sufficiently pure for the purpose. S t a s admitted t h a t some of his i n water, niid his iiiethod of preparatioiis were not eveii ~ ~ l i o l solulAe ly procedure was sncli that some of tlieiii probably contained platinum and liytlroside. Because speculatioiis of this kiiid conceriiiiig work so lorig past are of but little value, it was clearly necessary to repeat this work lvitli moclerii care ; and t h e follon-ing pages recount the details OC t h e repetition. Iii ~0111111011iritli others of the same type, the iiivestigation easily resolves itself into se\.eral sectioiis and will be discussed iiudrr the follo\viiig heads : T h e prepJratioii of materials ; the drying anti \vcigliiiig of potassiuni I>roiiiide ; the deteriiiination of tlie ratio of silver broiiiide to potassium bromide ; the determiiiatioii of the ratio of silver to potassium bromide ; and the discussion of the results. Of the laboratory tasks tlie preparcition of the materials was by far tlie most laliorious and puzzling, n.iiile the aiialytical worl; was comparatively siiiip:e. The Preparation of Materials The Sorswe ( I / f h c Poinssi7s7ir. -The task of preparing pure potassiuni bromide, simple as it may appear, is 11). 110 means an easy one. In viclv 1 ; I brief rt.viei\- of St:ii's ai111SI;,rigii:\v'.. \vork II~:L!. lie i o i i i l i l I I I C1;1r!
