194 THE ANALYST. NOTES ON THE E8TIMATION OF LEAD IN BBATED WATERS. BY ALFRED H. ALLEN. I RECEXTLY received from an inspector under the Sale of Food and Drugs Act, a sample of lemonade, which I certified to contain 1Q grain of lead per gallon. The estiniation was made colorimetrically with sulphuretted hydrogen, and the presonce of load was confirmed by chromate of potassium, which gave an immediate turbidity in the uncon- centrated sample. In consequence of my certificate, the vendor was snmmoned before the magistrates at Otley Petty Sessions, when his solicitor produced a certificate from Mr. F. Riinmington, of Bradford, stating that the sample contained *05 grain of lead per gallon. I n consequence of the discrepancy between our certificates, the remaining portiou of the sample was refemeci to Somerset House, whence, in due course, a certificate was received stating that the sample contained 1/55 grain of lead in 10 oz., and that this proportion mas within the limits of accidental impurity.Calculated on the gallon, the amount of lead found by Somermt Rouse is 0.30 grain per gallon, but, of course, the Bench did not understancl this,-and the defence took care not to tell them--while I, the imfortunate analyst, had not even been infoimed that my certificate was in dispute. The result was that the case was dismissed, together with another in which the facts wei’e similar, but the magistrates decided to reserve the question of costs till they leaimeci whether any explanation was forthcoming. Thus, at len@h, 9: have heaid of the case, and have had an opportunity of calling attention in writing to the following facts :- The samples were never diiided at all.Three closed bottles of eachmere 1lurchasecl, sealed by the inspector, and duly distributed between the vendor, the analyst, and himself. It is evident that the contents of the three bottles should have been mixed (in a jug), and then divided, if so required by the defendant. Seeing that are-examination of the remaining portions of my samples has proved the substantial accuracy of my certificates, it is clear that there W I L ~ no accidental inistake or transposition, and, as the estimation of lead in water is too simple a matter for an error of chemistry to occur, I prosume that the mnountR of lead found by Mr. XCirnmington, the Somerset House chemists, and myself, really reprefiented the proportions of metal present in the various2HE ANALY8T.195 -- bottles examined by us. If this be the case, it is certainly rather stai4ling to find that bottles of arated water, from the same manufactory, and of presumably nearly contem- poraneous manufaature, should be apt to contain amounts of lead varying so much as the figurks of Mr. Rimmington and myself show, but the probable cause of the vayiation in the amount of lead will be evideiit to the readers of the ANALPST. If we assume the 0.3 grain of lead per gasLon found by the Somerset Homse chemists in their poi%ion of one of the samples to represent the general extent of the contamination by lead, it is clear that the case was not one to be pooh-poohed or dismissed, for, although 0-3 grain per galloii may be within the limits of accidental impurity, people w i l l gene~ally object to be poisoned, even accidentally.Another point worthy of notice in the examination of mated waters for lead is the tendency of the contents of a bottle to become contaminated from contact with the leaden alloy which foims part of the stopper aiwmgement in a. certain description of patent bottle. 111 a recent instance I found 0. I7 grn. of lead in a sample of lemonade analysed a few days after it was received, but after standing some three weeks, with the leaden portion of the stopper immersed, the proportion of lead had increased to 3-36 grains of lead per gallon. In all cases in which I test for lead in mratccl waters I a531 in the habit of eonfilm- ing the result by the chromate test.'When carefully managed, chromate of potassium. aill indicate any proportion of lead greater than one-third of a grain per gallon, without it being necessai'y to concentrate the water. The sample should be placed in a Nesaler cylinder and a drop of potassium cboniate solution added, in snch a manner that the yellow solution gradually sinks through the clear and colourlcss liquid. The faintest cloud of lead chomate can thus be recognised. Addition of acetic acid seriously mars the delicacy of the test. I n testing =rated waters for load with sulphurstted hycbogen, the possible presence of tin and coppel* must not be lost sight of. Copper, if present, may at once be recognised by the ferrocyanide reaction, but traces of tin are not readily identified.The plan I have found best is to precipitate 200 C.C. of the water with sulphuretted hydrogen and clissolve the precipitate in strong hydrochloric acid. When the sulphwetted hydrogen is expelled, the solution is diluted somewhat, and boiled with metallic iron to insure that the tin exists in a stannous condition. The liquid is then decanted from the undissolved iron and tested with mercuric cliloi*ide, when any formation of a dky-looking cloud of niercurous chloride will bc readily recognised. Although not closely connected with the detection of metals in arated. drinks, I may take this opportu&y of calling attention to the fact that the ordinary teat for zinc, with an alkaliiie sulphide, is far fro111 delicate. A much inore satisfactory test is one which I described many years ago iu the Chen&,nZ Xws, Vol. ~XXLII,, page 290, but it has never €ound its way into the text books. The solution to be tested for zinc is rendered ammo- uiacal, heated to boiling, aiid potassiuui fcrrocyanicle added, when a white precipitate be produced if the mei'cst trace of zinc b~ present,