THE ANALYST. 275 CORRESPO NDE AT CE. [The Editors are not responsible for the opinions of their Correspondents.1 T H E S E L B Y F L O U R C A S E . To THE EDITOR OF ‘‘ THE ANALYST.” Sxa,--In March last I condemned a sample of flour for containing “ about half the usual quantity of alum.” Before the case was heard at Selby, the defendant induced the magistrates to order a portion of the remaining part of the sample to be sent to Somerset House.The result was that the gentlemen occupying the position of chemists at Somerset .House sent a certificate stating that they found the sample to contain an amount of alumina equivalent to 21 grains of ammonia-alum per 41b. of the flour, that their experiments on pure flour had shewn an amount of alumina natural to flour varying from an equivalent of 2 to 10 grains of alum, and that, with regard to the excess of alumina found, ‘‘ the results of their experi- ments did not enable them to affirm the existence of alum.” This certificate is interesting to chemists, as throwing light on the views of the Inland Revenue Chemists on a subject on which they had hitherto been silent.I n the sequel some further information was elicited.Mr. Bell received a subpcena to attend and explain himself, but sent his colleague, Mr. Bannister, instead. Mr6 Bannister explained that the alumina had been determined by DuprC’s method. The opinion that the excessive portion of alumina was not due to alum was based on an experiment made in some way with logwooG, Mr. Bannister shewing a residue dried up at the bottom of a porcelain dish276 THE ANALYST.in confirmation of his statement. I t also appeared that some of the Somerset House Chemists had made a series of experiments with pure flours with and without the addition of alum. On the other hand, I had found an amount of alumina corresponding to 19g grains of alum per 4lb., a result with which the determination at Somerset House agreed well.But there was this difference. As stated in evidence. I had dissolved the ash in HCl, and filtered, before evaporation to dryness to separate silica. By proceeding thus, and omitting the fusion with alkaline carbonate, I believe I leave any silicate of alumina insoluble. I have repeatedly analysed flours and wheat by this method, and have never obtained more alumina in solution than corresponds to 3 grains of alum per 41bs., although on fusing the insoluble matter with alkaline carbonate more alumina may be extracted.This view I expressed at a meeting of the Society of Public Analysts, an account of which appears on page 8, of vol. i., of the Prcceedings but my observa- tions are reported as if I were at the moment referring to Dr.Duprk’s process, whereas, in quoting these figures, I was referring to a process which I have had in frequent use before and since, and in which the fusion is omitted, I fully believe that Dr. Duprb’s method is the best for the estimation of the totalalumina, but by using acid alone, followed by filtration, I think it is possible to discriminate between added and natural alumina, or between soluble and insoluble, Such a plan has the disadvantage that incomplete solution of the alumina of the alum may cause an error of deficiency.In addition to the estimation of alumina 1 had observed the absence of excess of iron, and had determined the ash. I had also had a portion of the flour made into bread, and tested by the logwood and ammonia test. I t was a weak point that I had not been present when the bread was made, but the servant who made it was present ready in Court to swear that she used yeast and water only.(The water is a pure moorland water, containing 5 or 6 grains per gallon of solid matter.) The result of my logwood test on the bread was the production of a distinct, but not very well developed, blue colour. That the Somerset House Chemists failed on the flour, I couttnd, is due to one of two things-either faulty manipulation and inex- perience, or the occasional tendency of the test to fail when applied to flour.I t has certainly failed in my hands, and Hassall condemns it on the same account. Hence, as the Somerset House Chemists faii to obtain the logwood reaction, and ignore the teachings of the alumina, any sample of bread or flour, how- ever much alum it contains, is liable to get reported pure if referred to Somerset House.AB to the amount of alumina natural to wheat, Dr. Dupr6, L)r. Muter, Dr. Stevenson, Mr. Wanklpn, Mr. Wigner, and others agree in finding a maximum even less than that of Somerset House, thougb, pro- bably owing to differences of process, they found more than I have done.d Mr. Reynolds, a druggist, of Leeds, asserts that he found a considerable proportion of alumina in Egyptian wheat some fifteen yeara since, Without wasting time by criticking this statement, I may say that another recent case, in which a high amount of alumina was found, is undoubtedly to be explained by the process employed, Thus, if the calcium and magnesium phosphates are at once precipitated by soda, and the phosphate of aluminium merely recovered from the filtrate, the weight of the latter will be seriously in excess of the truth.If the analysis be conducted in that way the phosphate of aluminium ought always to be redissolved, and precipitated from a cold solution containing free acetic acid. The result of the Selby flour case was that the magistrates dismissed the summons but allowed no cosb.I may add that the same defendants were summoned for another sale of flour, in which I certified to 10 grains of alum per 4lbs., described this amount ia my certificate as insignificant, and wrote a letter to the authorities stating that they must not prosecute, for though my own opinion was decided as to the presence of alum, the proportion was too small for a conviction.I presumed this recommendation had been acted on, until I was informed that the sample had been sent to Somerset House. I write this letter chiefly with the view of raising a discussion on the detection of alum in flour and bread, especially with reference to the logwood test. I am, &c., ALFRED H. ALLEN. SHEFFIELD, May, 1878.T H E B R O M S G R O V E M I L K C A S E . To THE EDITOR OF THE ANALYST.” SIR,--T did not use porcelain dishes for (‘ economy ’’ as my report shows. I used platinum for total solids and ash, but porcelain for the determination of fat. The reason why I considered the use of platinum as an injury to the dealer was that the residue in platinum soon became very hard and dry, and I found I could obtain a larger percentage of fat by the more slow evaporation in porcelain dishes. The whole matter has been a great grief to me, but I have the satisfaction of knowing that the pub- licity given to the case (which cannot but be of injury to myself) has completely exonerated the milkseller from blame.THE ANALYST. 277 My laboratory is detached from my house and under lock and key, and no one is permitted to be there except in my presence.My friend, who unfortunately made the mistake in the tare of the dish had assisted me some time before in an extensive series of milk analyses, and I thought him quite competent in the use of the balance. I made two analyses which agreed with each other of the milk before I made my report, but unfortunately used the same tared dish. Yours faith fu I1 y, HORACE SWETE, M.D., May 23rd, 1878. Analyst, Worcestershire.