THE ANALYST. 343 PHARMACEUTICAL CONFERENCE. We give below some abstracts of some papers read at this Conference in Dublin :- THE TITRATION OF HYDROCYANIC ACID AND CYANIDES, AND IT3 HELATION TO ALKALINETRY. BY LOUIS EIEBOLD, F.C.S. Read before the British Pharmaceutical Confwence. LXEBIQ’S method for estimating the strength of hydrocyanic acid by means of decinormal solution of silver nitrate giws perfectly accurate results if the following precautions be obaerved :-344 THE ANALYST.(1) The solution of sodium or potassium hydrate should be placed in the beaker first, and the hydrocyanic acid added to it from the pipette. If, instead of this, the acid is placed in the beaker first, and the alkaline hydrate added afterwards, there may be a slight 106s by eraporation, which becomes appreciable whenever there is any delay in the addition of the alkali.(2) The mixture of hydrocyanic acid and alkali should be largely diluted with water before the silver ritrate is added. The most suitable proportion of water, according to my experience, is from ten to twenty times the volume of the officinal or of Scheele’e acid, which is mare than twice the quantity recommended by Fresenius and other authorities.With such a degree of dilution the final point of the reaction can be observed with greater precision. (3) The amount of alkali used should be as exactly as possible that required for the conversion of the hydraoyanic acid into alkaline cyanide, as an insufficiency or an excem both affect the accuracy of the result. With an excess the results are too high ; with an insufficient quantity they are too low.The error due to the first named cause is but small and is pointed out in some of the standard analytical works, which therefore recommend the use of sufficient alkali to produce a distinct alkaline reaction and the avoidance of an undue excess. But it is just this direction which may lead tl conscien- tious yet inexperienced manipulator t o the far more serious mistake of using too little alkali, because litmus entirely fails to mark the point at which the hydrocyanic acid has been completely converted into sodium or potassium cyanide.These cyanide8 are 80 strongly alkaline to test paper, and hydrocyanic acid is so weak an acid that a mixture of the two may have a distinct or even a very strong alkaline reaction and yet contain a considerable amount of free hydrocyanic acid.Hence it follows that the use of a quantity of sodium hydrate quite sufficient to produce a strong alkaline reaction may only ensure the conversion of a portion, and perhaps of the smaller portion, of the hydrocyanic acid actually present into sodium cyanide, The inevitable result will be a serious error in the estimation, as the quantity of silver nitrate solution required to produce a permanent precipitate will only indicate that portion of the hydrocyanic acid which has entered into combination with the alkali; and this error may possibly amount to a8 much as 75 per cent.I have alluded to this source of error at one of our previous meetings, in connection with a paper on the preservation of hydrocyanic acid, and I pointed out on that occasion that the alkalinity of the mixture at the end of the reaction, e.g., after the addition of sufficient silver nitrate to produce a slight permanent precipitate, may be regarded as n sure indication that a sufficient quantity or rather an excess of alkali has been used, and that the result of the determination will be fairly correct, or in the presence of an undue excess of alkali a little too high.The alkalinity of the mixture of hydrocyanic acid and sodium hydrate completely ceases after the addition of the required amount of silver nitrate, unless some excess of alkali was used, and if it does cease, the result of the analysis will almost certainly be too low.My reason for again touching upon these points is that the neutrality of the double cyanide of sodium and silver (the product of this reaction), on which these conclusions were b a d , also forms the basis of my present communication. From what I have already stated it is clear that the titration of hydrocyanic acid with eilver nitrate cannot give results of scientific accuracy unlesg tho quantity of alkaliTHE ANALYEIT.345 used is exactly that required to combine with the acid, or unless a correction can be made for the excess of alkali employed. It is true that a slight excess of the latter does not appreciably affect the result, but then the question arises how to make certain that the exam used is but a slight one. I t will not do to start with just sufficient soda to render the mixture alkaline and then to add gradually more as the alkalinity ceases during the titration, because in that case free hydrocyanic acid would be present in the mixture during nearly the whole of the process, and under the influence of the exposure and the continual stirring a portion would inevitably be lost by evaporation, thus causing an error which, though perhaps not considerable, is certainly greater than that which would result from the ucje of even an immoderate excess of soda to start with. I find, however, that the following modus operandi will meet the difficulty and en~ure results agreeing perfectly with those grsvimetrio determinations :-The acid is allowed to run from the pipette into an excess of solution of sodium hydrate; decinormal solution of silver nitrate is then added drop by drop until a slight opalescence is produced, and this point being attained, standard normal hydrochloric or sulphuric acid is added until the opalescence begins to increase, which does not take place until the whole of the free alkali is neutrdized. From experience I find that for each C.C.of standard mineral acid thus required 0.01 C.C.should be deducted from the volume of the silver solution used, and the remainder calculated for HCy. It will be seen that in this process the cyanide of sodium and silver acts as an acidimetric indicator, and indeed, it answers well for the purpose, for a single drop of free acid produces with it a very distinct precipitation of silver cyanide.The fact, that in the abscence of a suflicient quantity of soda the volume of silver solution required to produce a permanent precipitate only indicates that portion of the hydrocyanic acid which has been used up in the formation of sodium cyanide, and that thia determination of NaCy is in no wise affected by the presence of free hydrocyanic acid, renders this method applicable for the analysis of mixtures of the free acid and alkaline cyanides.Supposing the solution to be analysed contained free hydrocyanio acid and potassium cyanide, the volume of silver solution required to produce a per- manent opalescence would show at once the quantity of ECy present. On now adding NaHo in slight excess and continuing the titration until the opalescence ia again produced we find the quantity of free HCy.Before quitting thia subject I wish to refer to a very handy process for the estima- tion of cyanides recently communicated to the Chemical Society by Mr, J. B. Hannay. It consists in the addition of decinormal solution of mercuric chloride to the hydrocyanic acid or cyanide rendered previously alkaline with ammonium hydrate, until a permanent precipitate is formed, which does not occur until the whole of the cyanogen has been used up in the formation of mercuric cyanide, as alkalies have no action on the latter.I have tried the prooess repeatedly with most satisfactory results, and believe that it will find much favour with pharmacists in the testing of hydrocyanic acid, especially But it cannot be used like the other for the analysis of mixtures of free HCy and cyanides.I now come to the second part of my report, viz., the relation of the titration of cyanides to alkalimetry. It stands to reason that if an alkaline cyanide can be correctly estimated in the presence of free hydrocyanic acid by silver nitrate, this titration must answer as well for the estimation of a caustic alkali as for that of hydrocyanic acid.For that purpose the quantity of KCy or NaCy found, or the volume of ailver eolution The results thus obtained are quite exact, an excess of alkali does not affect its accuracy.346 THE ANALYST. used, is simply calculated for KEO or NaHO instead of HCy. Now if the applicability of this test for alkalimetric purposes were confined to the determination of caustic alkalies, I feel certain that nobody would think of using prussic acid and silver nitrate in preference to the customary sulphuric acid and litmus ; but I find that it answers equally well with the alkaline carbonates, and here I consider it decidedlypreferable to the process in general use, for the following reasons :- (1.) The solution of alkaline carbonate does not require boiling, as the carbonic acid does not interfere.(2.) The change from perfect clearness to an unmistakable turbidity, as produced by a single drop of the silver solution, is more striking than that of the colour of litmus brmght about by one drop of standard sulphuric acid. (3). As a decinormal solution is used the results are more accurate than those obtained by normal H2S0, or HC1.(4). The results may be readily checked, without the necessity of operating on a fresh portion of the sample ( 5 ) . The chloride present in commercial alkaline carbonates can be estimated by the same process with but little additional trouble. It is well known that hydrocyanic acid does not decompose alkaline carbonates at an ordinary temperature. But in the presence of silver nitrate the decomposition takes place in accordance with the following equation- IC2C03 + 2HCy + AgNO,= KAgCy, f KNO, + C02 + H2O The first drop of silver solution added in excess precipit,ates silver cyanide.The weak solution of the carbonate to be tested (about 0.5 to 1 gram in 100 C.C. of water) is mixed with 10 to 20 C.C. of hydrocyanic acid of Scheele’s strength (a decided excess), and the decinormal solution of silver nitrate added drop by drop, stirring well all the time until a permanent turbidity is produced.Each C.C. of the silver solution required corresponds to 0.138 gram E$03 and to 0.0106 Na2C03. I quote the results of a few determinations to show the accuracy of the process. XAgCy, + AgNO, = 2AgCy + KNO, Pure &CO3 used Found 0.5850 .. . . . . . . . . . . . . . 0.5861 0.1670 . . . . . . . . . . . . . . 0.1672 0.8775 . . . . . . . . . . . . . . . 0.8779 0.2088 . . . . . . . . . . . . . . . 0.2085 If after the addition of the required quantity of si:ver nitrate the mixture is boiled down t o less than half its volumeor until the excess of free HCy has been completely expelled, then mixed with a few drops of solution of potassium chromate and the additionof silver nitrate now proceeded with until the colour of the mixture changes to red, the volume of the test thus used will be found equal t o that uscd in the first titrrtion.This may serve to check the previous result. I n the presence of chloride, however, the number of C.C. used in the second titration will be greater than that used in the first.The difference between the two exactly indicates the chloride I f 40 C.C. were used in the first and 45 C.C. in the second experiment the difference of 5 c.c must be calculated for chloride. A few of my results will show the value of the method. Used. Found. Pure NaCl Of680 NaCl 0.0683 &GO3 0‘9750 Pure NaCl 0-1825 NaCl 0,1830 (I). Pure KzC03 0.2000 I(zc03 0.2005 (2).Pure &c03 0,9750THE ANALYST. 347 Hence I believe, that this method merits the attention of those who are much engaged in alkalimetric estimations. I have also employed silver nitrate with success i n the analysis of mixturesof hydrocyanic acid and mineral acids, and indeed I find that these processes may be advantageously extended to other applications, but my experiments in this direction are not yet completed. THE VOLUXETRIC ESTIJIATION OF SONE IRON CONPOUNDS OF THE PHARMACOPCEXA. BY H.N. DRAPER, F.C.S. TEE author said he brought forward his notes rather as queries than as results. Profcssor Tichborue had questioned to him the correctness of the figures given by the British Pharmacopceia in the volumetric estimation of arseniate of iron.The quantity of bichromate solution said to be necessary for the conversion of 2 grammes of ferrous arseniate seemed t o Professor Tichborne too small. According to the British Pharmacopceia 100 C.C. of the solution are capable of converting from proto to per salt 1.68 gramme of iron. This statement is theoretically correct, and Mr. Draper found by actual experiment, taking the mean of three made with piano wire, that the actual quantity required was 97 C.C.Now regarding arseniate of iron, the Pharmacopceia states that 2 grammes require 17 C.C. I f it were possible to prepare the arseniate so that all the iron should be in a ferrous state the conversion would require 44.84 C.C. This shows that the British Pharmacopceia standard is that of a salt containing only 37.9 per cent.of absolute ferrous arseniate, a constitution which is but inadequately described by the phrase I‘ partially oxidised.” But on making the arseniate according t o the Pharmaccepia instructions even this proportion of ferrous arseniate was not realized. The mean of three experiments gave only 21.7 per cent. Other specimens purchased in Dublin gave respectively 3.34, 5.2, 6.64 and 13-6 per cent, The British Pharma- copceia requires a standard of 44.8 per cent.of absolute ferrous phosphate. By preparing the salt according to its instructions a proportion of 60.9 per cent. was obtained, and from four purchased specimens the following proportions were respectively found : -24.1 29.6, 31.3, 49.2. I n magnetic oxide the British Pharmacopceia requires, by its volumetric standard, only a proportion of 28.8 per cent. of true ferroso ferric oxide.A specimen prepared according t o directions, howzver, was found to contain only 19.4. This is because the iron is in a ‘< partially oxidisetl” condition. A different result was obtained with the ferrous phosphate. M, TEISSERENC DE Bow has just charged the Agronomic Institute to mako a chemical analyses of all the wines sent to the Paris Exhibition.The number of samples to be examined is more than 10,000, coming from France, Spain, Italy, Austria, America, Africa, &o. Mr. J. H. Collins, Public Analyst for the County of Cornwall, and for the Borough of Penzance, has been appointed Analyst for the City of Truro and for the Borough oi Launceston,548 THE ANALYST. ANALYSES OF BERLIN BEER. THE following analyses are from the dllgerneine Zopfeelz Zeitzcfig, Beer brewed in the celebrated Brauerie Konigstadt, of Berlin, was found to contain a8 follow8 :- LIGHT COLOURED BEER. Alcohol . . . . . . . . . . . . . . . . . 4.501 per cent, by weight Saccharine . . . . . . . . . . . . . . . . . . 1'893 .. 9 ) Dextrine . . . . . . . . . . . . . . . . . . 0,861 .. ,, Albuminoids . . . . . . . . . . . . . . . 0,630 .. 1, Hop-bitter, extractive andsaline matter . . . . . . 2,296 .. 9 ) Acid... . . . . . . . . . . . . . . . . . . 0.006 .. t, Unfermented extract 6.680 per cent. DARK COLOURED BEER. Alcohol Saocharine . . . . . . . . . . . . . . . . . . 1'950 .. $ 9 Dextrine . . . . . . . . . . . . . . . . . . 1.053 .. $1 Albuminoids . . . . . . . . . . . . . . . 0.621 .. 9 ) Hop-bitter, extractive, and saline matter . . . . . . 3.366 .. 1, Acids . . . . . . . . . . . . . . . . . 0.005 I 1 . . . . . . . . . . . . . . . . . . 4.260 per cent. per weight. Unfermented extract 7.010 per cent.