of lead in tartaric and citric acid; 3. Proposed standards or purity. A few words must first be said as to the condition of the lead present. Both tartaric and citric acid crystals occasionally contain visible fragments of metallic lead. The manager at the Millwall factory, Mr. Berry, tells me that these fragments are probably fine scrapings of lead left by plumbers when making or mending the leaden vessels. M. Buchet has determined the proportion of metallic lead in the samples of tartaric and citric acid he has examined (page 103 of this paper); it is of course very irregular. To determine metallic lead, M. Buchet dissolves 200 grms. of the acid in 600 cc. of water, adds ammonia in slight excess, and after 24 hours collects undissolved matter on a small filter, and, after washing, dissolves the lead in nitric acid, and determines its quantity in the usual way with sulphuric acid and alcohol. Besides metallic lead, both tartaric and citric crystals have occasionally attached to their base, where they have been in contact with the leaden vessel, patches of an opaque whitish substance, which turns black when moistened with solution of hydrogen sulphide. This substance is evidently a portion of the corroded surface of the lead removed by the crystal. This lead compound does not entirely dissolve when the acid is placed in water, but it dissolves with great rapidity if an excess of ammonia is added to the acid solution. M. Buchet found in one bad sample of French tartaric acid 0.0082 per cent. of lead soluble in water, but when the acid was neutralised with ammonia the lead found was 0.0363 per cent. Instances of this kind are, however, exceptional; I have not found that citric and tartaric acid generally contain lead compounds undissolved by water. Metallic lead is clearly less poisonous and less objectionable than soluble lead; the determination, moreover, of its quantity presents no special analytical difficulties. I have therefore confined my attention to the detection and estimation of the soluble lead contained in the acids. Before passing to my own results it will be well to give a summary of the tests for lead in tartaric and citric acid prescribed by the various pharmacopoeias. These directions, though often of little scientific value, occupy a position of public authority, and are often referred to in legal questions as affording trustworthy definitions of purity. PHARMACOPOEIA TESTS.* a. Tartaric Acid. 1. With Hydrogen Sulphide alone.-The aqueous solution is not altered by hydrogen sulphide (British, Belgian, Italian). An aqueous solution, 1 in 10, is not coloured by solution of hydrogen sulphide (Netherlands). A concentrated aqueous solution is not blackened at the line of contact when a solution of hydrogen sulphide is floated on it (United States). The powdered acid is not altered by solution of hydrogen sulphide (Hungarian). The test under this head is the only one prescribed by the British, Belgian, and Hungarian Pharmacopoeias ; the other Pharmacopoeias have additional tests under the following heads. The strength of the tartaric acid solution is not mentioned by the British, Belgian, and Italian Pharmacopoeias. The British Pharmacopœia describes a solution of tartaric acid in the Appendix made by dissolving 1 oz. of tartaric acid in 10 oz. of liquid. 2. With Hydrogen Sulphide and Ammonia.-An aqueous solution, 1 in 10, nearly neutralised by ammonia, is not altered by solution of hydrogen sulphide (German). An aqueous solution, in 2, is not altered by solution of hydrogen sulphide even when ammonia is added (Austrian). An aqueous solution, saturated with ammonia, is not altered by ammonium sulphydrate (Italian). *The editions referred to are as follows: British Pharm. 1855; United States, 1883; German, 1899; Austrian, 1889; Hungarian, 1888; Italian. 1892; Belgian, 1885; Netherlands, 1889. The French Codex (1884) does not contain any description of tests. An aqueous solution, I in 10, saturated with ammonia, is not altered by ammonium sulphydrate (Netherlands). 3. By examination of Ash.-The ash is not to turn blue when treated with solution of ammonia, nor to blacken by a further addition of ammonium sulphydrate (United States). b. Citric Acid. 1. With Hydrogen Sulphide alone.--The aqueous solution is not altered by hydrogen sulphide (British, Belgian, Italian, United States). An aqueous solution, 1 in 3, is not altered by solution of hydrogen sulphide (Hungarian). The powdered acid is not altered by solution of hydrogen sulphide (Netherlands). As in the case of tartaric acid, the above test is the only one prescribed by the British and Belgian Pharmacopoeias, while the remaining pharmacopoeias have additional tests. The British, Belgian, Italian, and United States Pharmacopœias do not prescribe the strength of the solution which is to be tested. 2. With Hydrogen Sulphide and Ammonia.-An aqueous solution, 1 in 10, nearly neutralised with ammonia, is not altered by solution of hydrogen sulphide (German). An aqueous solution is not altered by solution of hydrogen sulphide, even when ammonia is added (Austrian). An aqueous solution, 1 in 3, is not altered by solution of hydrogen sulphide, even when saturated with ammonia (Hungarian). of The powdered acid is not altered by solution hydrogen sulphide, even when ammonia is added (Netherlands). An aqueous solution, saturated with ammonia, is not altered by ammonium sulphydrate (Italian). 3. By examination of Ash.--The ash should not turn blue by treatment with solution of ammonia, or blacken by the further addition of ammonium sulphydrate (United States). On glancing through these directions we are at once struck by their diversity. A test adopted by one pharmacopia is excluded by another; and where the same test is employed considerable differences are introduced by the variations in the strength of the solution to be tested. In short, a sample of acid which will pass the rules of one pharmacopoeia will be rejected by the rules of another; this will become still more evident further on. The pharmacopoeias concern themselves solely with combined lead, though probably the preparation of the ash directed by the United States Pharmacopoeia would result in the oxidation of any metallic lead present, and this oxide would be subsequently blackened by the ammonium sulphydrate. Most of the pharmacopoeias regard solely the lead soluble in an aqueous solution of the acid, but those of Italy and the Netherlands apparently include any lead compounds which will dissolve in ammonium tartrate. NEW INVESTIGATIONS. 1. The Detection of Lead. Several of the most characteristic tests for lead, as sulphuric and chromic acid, are not available for the detection of lead in tartaric or citric acid, owing to the solubility of salts of lead in these acids. A soluble chromate may be added to a solution of citric acid without any reduction of the chromic acid taking place; a considerable amount of lead may, however, be present without giving rise to any precipitate. The tests practically available for the detection of lead are the sulphides, employed either in an acid or alkaline solution. The whole of the trials as to the respective delicacy of different methods have been performed with solutions of pure tartaric and citric acid, to which known amounts of lead had been added. The same standard lead solution was used for the whole investigation. The Standard Lead Solution employed was made of recrystallised lead nitrate, dried over sulphuric acid; it contained 1 grm. of metallic lead in 1,000 cc. ;* 1 cc. thus contained 0.001 grm. of lead. From this solution a second was prepared of one-tenth the strength. If such solutions are to remain constant in composition it is essential that they should contain a little free nitric acid; if this is absent a basic salt is gradually deposited. Effect of Hydrogen Sulphide as Gas. Tartaric Acid. The delicacy of the test is very largely influenced by the strength of the tartaric solution. With a strength of 50 grms. in 100 cc. the gas produced no change when passed for half-an-hour, though lead equal to 0.0072 per cent. of the acid was present. When the proportion of lead was raised to 0.0080 per cent. a very distinct grey turbidity was produced. With a strength of 25 grms. in 100 cc. a considerable dark grey turbidity was produced in two minutes when the acid contained 0.0036 per cent. of lead, and a slight grey turbidity was produced in half an hour when the lead amounted to 0.0030 per cent. No reaction was obtained with 0.0024 per cent. of lead. With a strength of 10 grms. in 100 cc. a distinct grey turbidity was produced in ten minutes when the acid contained 0-0020 per cent. of lead, and a very slight reaction (quite distinct, however, when compared in a favourable light with a similar solution free from lead*) was produced in half an hour when the acid contained 0.0015 per cent. There was perhaps a trace of reaction when the lead was still further reduced to 0·0010 per cent. Citric Acid.-The reaction with hydrogen sulphide in the presence of this acid is far more delicate than in a tartaric solution; the strength of the acid has, as before, a great influence upon the result. With a strength of 50 grms. in 100 cc. a dark purple turbidity having a red reflection was produced in three minutes when the acid contained 0.0032 per cent. of lead, and a distinct purple turbidity in 20 minutes when the lead was 0.0020 per cent. A very slight turbidity appeared in 25 minutes when the lead was reduced to 0·0016 per cent. With a strength of 10 grms. in 100 cc. a distinct turbidity was produced in ten minutes when the acid contained 0.0010 per cent. of lead, and a very slight reaction appeared in half an hour when the lead was reduced to 0.0005 per cent. Effect of Hydrogen Sulphide in Solution. An aqueous solution of hydrogen sulphide is a much more delicate test for lead in tartaric and citric acid than hydrogen sulphide used as gas, and the appearance it produces is often much more striking, the solutions being generally deeply coloured, while the gas gives rise in most cases to a finely-divided precipitate. The coloured lead solutions produced by a solution of hydrogen sulphide probably contain the lead as a sulphydrate. The solution of hydrogen sulphide employed was always saturated with the gas, aud quite free from suspended sulphur. This reagent is best prepared in bottles nearly filled with water. When the water is saturated, the air is displaced with hydrogen sulphide, the stopper introduced, and the bottle laid on its side. Tartaric Acid.-When 50 cc. of tartaric acid solution are mixed with 50 cc. of solution of hydrogen sulphide, the final strength being 25 grms. of acid in 100 cc., a slight grey turbidity is generally produced on standing when the acid contains 0.0012 per cent. of lead, and a more distinct turbidity is immediately produced when the lead amounts to 0.0016 per cent. As the proportion of lead increases the appearance produced by the reagent changes. The smallest visible result is a grey turbidity; with more lead the turbidity appears as a very finely-divided purple-grey preeipitate; with a further increase in lead a bronze tint appears; with still more lead the solution acquires a dark smoky-brown colour, and there is no longer any suggestion of a precipitate. To detect a very slight turbidity comparison must be made with 3 similar mixture of tartaric acid and solution of hydrogen sulphide containing no lead: the solutions must stand between the observer and the light, and be viewed against a dark background. When the solution of tartaric acid is of such a strength that after adding its own volume of solution of hydrogen sulphide the mixture contains 10 grms. of acid in 100 cc., a distinct grey turbidity is almost immediately produced when the acid contains 0.0010 per cent. of lead. With 0.0007 per cent. of lead a slight distinct turbidity is produced after some time. If 10 grms. of the tartaric acid are dissolved in 20 cc. of water and the liquid brought to 100 cc. by the addition of solution of hydrogen sulphide, a slight distinct turbidity may be seen when the lead in the acid is only 0.0005 per cent. Citric Acid.-As in the experiments with the gas, the delicacy of the test is found to be much greater with citric than with tartaric acid. The series of tints produced is different in the case of citric acid. When the citric solution is mixed with its own volume of solution of hydrogen sulphide, and the mixture finally contains 25 grms. of the acid in 100 cc., a clear, bright brown colour is immediately produced when the acid contains 0.0016 per cent. of lead. With 0.0008 per cent. of lead a cinnamon-tinted turbidity is immediately produced. With 0.0004 per cent. the solution soon acquires a slight pinkish turbidity. A solution of acid containing 0.0002 per cent. of lead gave no distinct reaction. When the solution of citric acid receives its own volume of solution of hydrogen sulphide, and the final mixture contains 10 grms. of acid in 100 cc., an immediate clear brown tint is produced when the acid contains 0.0010 per cent. of lead. With 0.0005 per cent. of lead a distinct brownish turbidity soon appears. With 0.0003 per cent. a slight turbidity appears after some time. With 0.0002 per cent. no appreciable reaction took place. When, however, the 10 grms. of citric acid were dissolved in water to 20 cc. and the whole diluted to 100 cc. with solution of hydrogen sulphide, a slight turbidity was produced when the acid contained 0.0002 per cent. of lead, and even with 0.0001 per cent. of lead a reaction was thought to be appreciable. Hindrance to the Action of Hydrogen Sulphide.-While engaged in the present work I came across a singular fact. When tartaric liquors have been purified from lead with hydrogen sulphide, the liquor concentrated, and a crop of crystals obtained, the mother-liquor contains a soluble sulphur compound which is capable of preventing the precipitation of lead by hydrogen sulphide. To such a mother-liquor, which was of a dark straw colour, slightly turbid, and had a marked garlic odour, I was able to add lead up to 0.0048 per cent. of the tartaric acid present without being able to detect it by subsequent treatment with a solution of hydrogen sulphide. If this amount of lead was exceeded the hydrogen sulphide produced a brown colour. When this mother-liquor was treated with a small quantity of an oxidising agent its peculiar characters disappeared, and it behaved in a normal manner with lead and hydrogen sulphide. This fact is of little importance to the public analyst, as the crystal from such liquors contains but a trace of this compound, and we shall presently see that by testing in an alkaline solution its prejudicial influence can be almost or entirely removed. Effect of Hydrogen Sulphide in a nearly Neutralised Solution. The German Pharmacopoeia directs that both tartaric and citric acid shall be tested for lead with a solution of hydrogen sulphide after the acid has been nearly neutralised with ammonia, and German analysts apparently lay great stress on this point as essential to a searching application of the test. We have just seen that 0.0005 per cent. of lead in tartaric acid, and 0.0002 per cent. of lead in citric acid may be detected without any neutralisation of the acids, and we shall presently see that still sinaller amounts may be plainly detected when the acids are treated with ammonia in excess; what, then, is to be gained by the intermediate treatment of incomplete neutralisation? The object of leaving the solution feebly acid is, apparently, to avoid any darkening of the solution due to the formation of iron sulphide. If a small quantity of a ferrous salt be added to a solution of pure citric acid, and the whole nearly neutralised with ammonia, a solution of hydrogen sulphide may be added without any darkening taking place. The case is quite otherwise with tartaric acid. If a solution of tartaric acid is taken containing a salt of iron in distinct quantity, and ammonia is gradually added, a precipitate is first formed of the acid ammonium tartrate, which dissolves as more ammonia is introduced. If the solution is to be left acid, the addition of ammonia must stop before the resolution of the acid tartrate is completed. If now a solution of hydrogen sulphide be added to this feebly acid liquid, the characteristic green and black tints of iron sulphide at once appear. It is of no use to attempt to increase the acidity of the liquid by adding hydrochloric acid, for the only effect of this acid is to throw down a great precipitate of the acid ammonium tartrate, the solution remaining no more acid than before, and the iron sulphide being left unaffected. The formation of iron sulphide can only be prevented by having the solution of tartaric acid so weak that no precipitation of the acid ammonium tartrate can occur. With a solution so weak no gain in delicacy when testing for lead can be expected. The direction of the German Pharmacopoeia is thus valueless, so far as tartaric acid is concerned. If soda be substituted for ammonia, there is no precipita tion of iron when solution of hydrogen sulphide is added to a nearly neutralised tartaric solution. A considerable practical difficulty is met with in attempting to test for lead in neutralised solutions; this consists in the great scarcity of pure reagents. I have been unable to obtain sodium hydrate which did not give a brown colour with solution of hydrogen sulphide; the soda from sodium, prepared by Messrs. Hopkin and Williams, was useless from this cause. Ordinary ammonia is equally useless. I have obtained ammonia ("puriss") from Messrs. Hopkin and Williams which gives no colour with hydrogen sulphide; and from Messrs. Burgoyne and Burbridges and Co. I have obtained sodium carbonate which is almost equally pure. When looking for traces of any substance an analyst should, of course, test all his reagents. The presence of iron in tartaric or citric acid is best ascertained by adding potassium ferrocyanide to the aqueous solution of the acid. The quantity of iron present is usually very minute; we shall see presently that its presence need not interfere with the testing for lead even when the solution is made alkaline with ammonia. Effect of Sulphides in an Alkaline Solution. The pure tartaric and citric acids I have employed, when dissolved and treated with an excess of pure ammonia, remain entirely unaffected by ammonium sulphydrate. This is the most extreme test for lead in these acids with which I am acquainted. If to either of these acids lead equal to 0.0001 per cent. (1 per million) be added, the addition of 1 drop of ammonium sulphydrate to a strong ammoniacal solution produces at once a very slight, distinct, clear brown tint. It is possible to detect even half this amount of lead. The greater delicacy of this mode of testing is, in part, due to two causes: (1) the greater concentration of the solution, as but one drop of ammonium sulphydrate is needed instead of dilution with an equal volume of solution of hydrogen sulphide; (2) the clear brown colour produced, which admits of easier recognition than a faint grey turbidity. The amount of iron present in commercial tartaric and citric acid is, as far as my experience goes, too small to interfere with the test for lead just described; ammonium sulphydrate is not in fact by any means so delicate a test for iron as potassium ferrocyanide. If iron should be present in considerable quantity this will be at indicated by the abnormal blackish colour given by the ammonium sulphydrate. In this case the solution of tartaric or citric acid, made alkaline with ammonia, should be treated with a few drops of a solution of potassium cyanide, and heated to near boiling; when cool the testing once is made with ammonium sulphydrate, the iron now giving no reaction having been converted into ferrocyanide.* In testing by this method the presence of copper is revealed by the blue colour produced when the acid is treated with excess of ammonia; this blue colour is discharged, and the subsequent formation of copper sulphide prevented, by adding a little potassium cyanide* to the ammoniacal liquid. The testing for lead in an alkaline solution has one further advantage, namely, that the sulphur compound which may possibly be present in small quantity in acid which has been purified with hydrogen sulphide does not in this case prevent its detection. 2. The Quantitative Estimation of Lead. Gravimetric methods for the determination of lead in tartaric and citric acid have been described by M. Buchet (loc. cit.), and by M. Guillot, Jour. de Pharmacie et de Chimie, 1892, 541; these methods are apparently identical. Dr. W. R. Smith (loc. cit.) has also described his method. M. Buchet dissolves 200 grms. of the acid in 600 cc. of water, and adds ammonia in slight excess. After standing for 24 hours, insoluble matter, containing the metallic lead, is separated by decantation and filtration, the solution is slightly acidified with hydrochloric acid, and hydrogen sulphide gas passed through the liquid to saturation. After standing 12 hours the precipitated sulphide is collected on a filter, washed, redissolved in nitric acid, concentrated to a small volume, and the lead precipitated by sulphuric acid, with the addition to the liquid of twice its volume of alcohol. No experiments are mentioned proving that hydrogen sulphide gas will precipitate all the lead under the conditions named. Dr. Smith is content to work on 50 grms. of the acid; this is burnt to a carbon-free ash in a muffle, the ash is dissolved in nitric acid, the solution evaporated to dryness, the residue dissolved in dilute acetic acid with sodium acetate, the lead precipitated as chromate, collected on a "Gooch" asbestos filter, and weighed. There is said to be "a possible error of 2 per cent. of the total amount of lead present in the sample." How the latter fact is arrived at is not mentioned; there is no reference to any experiment in which a known amount of lead was present. In approaching the subject with the view of arriving at a good commercial method, I desired that the method selected should, if possible, be volumetric instead of gravimetric, hoping thus to insure the economy of time, which, if accuracy can be preserved, is always so desirable. A good many chemists have tried to find a volumetric method for the determination of lead in tartaric and citric acid by matching the tint produced in these acids by a solution of hydrogen sulphide with the tint produced by the same reagent in a similar volume of water containing a known quantity of lead. When such methods are tested by employing tartaric and citric acid containing known amounts of lead they are found to be entirely fallacious; the effect produced by a solution of hydrogen sulphide in a dilute solution of a lead salt is indeed profoundly altered by the presence of tartaric or citric acid. Of this fact any chemist may assure himself by performing the following simple experiments: : To Prepare solutions of pure tartaric and citric acid containing 25 grms. in 50 cc., place these solutions in glass cylinders, and in a third cylinder place 50 cc. of water. each cylinder now add 1 cc. of the standard solution of lead nitrate already mentioned, and mix thoroughly; the tartaric and citric acid now contains 0.0040 per cent. of lead. To each cylinder 50 cc. of solution of hydrogen sulphide is next added, and the whole rapidly mixed. The contrasts presented by these solutions, each containing the same quantity of lead, are certainly astonishing; their may be thus described: appearances Water-Quite pale olive-brown, clear. Dr. F. L. Teed has already called attention to this use of potassium cyanide in The Analyst, August 1892, 141. If now the experiment is repeated, but the lead reduced to one-fourth of the previous quantity, the appearances will then be Water-Very slight brown tint. Citric acid-A deeper brown, more opaque. Tartaric acid-No colour, clear, or a slight grey turbidity. It is quite obvious that any matches of colour made among solutions in which the reaction proceeds so differently are worthless as evidence that an equal quantity of lead is present in each case. Dr. W. R. Smith has published (loc. cit.) determinations of the soluble lead present in 11 samples of tartaric and citric acid; he thus describes the method which he employed : "A convenient quantity of the sample for. examination was dissolved in 50 cc. of water, a little dilute hydrochloric acid added, and then a measured quantity of saturated sulphuretted hydrogen water. The tint thus produced is matched by a standard solution of acetate of lead, just as in Nesslerising. The conditions are kept rigorously parallel, whether they apply to the volume, to the quantity of dilute hydrochloric acid, or the amount of sulphuretted hydrogen water. By these means the amount of soluble lead present is ascertained." In this method we have an additional element introduced -the addition to each solution of an equal volume of dilute hydrochloric acid. Nothing is said as to the object of this addition, or as to what is to guide the analyst in selecting a proper quantity of hydrochloric acid. I presume that the object in view is to equalise the tints produced by the hydrogen sulphide. If a certain amount of hydrochloric acid is mixed with each of the comparative solutions in the first of the two series of experiments just mentioned, the quality of tint produced on the addition of the solution of hydrogen sulphide will be the same in every case; each of the three cylinders will in fact yield a lavender-grey precipitate. Unfortunately the improvement stops here. It is possible by adding hydrochloric acid in equal quantity to each solution to make the precipitates of lead sulphide in water and in solutions of tartaric and citric acid comparable in the quality of their colour, but it does not help us to make them equal in quantity, although the same amount of lead may be present in each case. It is clear from the foregoing remarks that the only way in which the appearance of the lead sulphide in a solution of tartaric or citric acid can be accurately used as a means of determining the quantity of lead present is by comparison with the appearance in solutions of pure tartaric or eitrie acid of the same strength, to which known amounts of lead have been added. I have employed two methods for the estimation of lead; I will describe them in the order of adoption. The Glycerol Method.-The smoky, opaque tints produced by a solution of hydrogen sulphide in solutions of tartaric acid containing a considerable amount of lead are very unfavourable for comparisons of colour. If, however, a certain amount of glycerol be mixed with the tartaric acid before the addition of the hydrogen sulphide the tint becomes lighter in tone, and the liquid much more transparent. For the purpose of this method there is required a stock of solution of tartaric acid free from lead, containing 50 grms. of acid in 100 cc. Also the lead nitrate solutions already described, contained in small 10 cc.. burettes graduated to tenths of a cubic centimetre. Of the tartaric acid to be examined 75 grms. are dissolved in water, made up to 150 cc., and filtered. In the ease of obstinate turbidity in the acid it is well to introduce a little filter-paper pulp before filtering. A preliminary experiment is first made to ascertain what volume of the tartaric solution will yield a suitable depth of tint, the error unavoidable in colourmetric work being a much smaller proportion of the matter to be determined when a full but not excessive depth of colour is observed. The desired mixture is prepared in a graduated cylinder. 20 ee. of glycerol are first introduced, and then 30 cc. of the tartaric acid solution; the mouth of the cylinder is then closed with an india-rubber stopper, and the contents well mixed. An equal volume of a clear saturated solution of hydrogen sulphide is then added, the whole again mixed, and transferred to a cylinder of clear glass standing on a sheet of white paper. The appearance of the liquid enables the chemist to decide whether more or less of the tartaric acid should be taken to obtain a convenient tint; 10 cc.50 cc. is the range which I have found necessary. When the proportion of lead is small, 10 cc. of glycerol will suffice. The selected mixture is made exactly as just described, the volume of solution of hydrogen sulphide employed being equal to the volume of liquid already in the cylinder. When the selected mixture has been prepared, the next step is to match its colour with a similar mixture containing a known amount of lead. The same amount of glycerol and of a pure tartaric acid solution are introduced into the cylinder as in the former experiment, with a few tenths of a cubic centimetre of the stronger of the two solutions of lead nitrate, and the whole is well mixed; the solution is then brought to the same volume as the former mixture by the addition of solution of hydrogen sulphide. A comparison of the tints of the two mixtures will at once show whether too much or too little lead has been used, and new experiments are made with varying quantities of lead until an equality of colour is reached. When an approximate equality is attained it is best to start three mixtures as simultaneously as possible, one a repetition of the selected mixture of the acid under examination, the other two containing slightly varying quantities of lead, the object being that the colour yielded by the acid under examination shall lie somewhere within the range of colour in these two experiments. The observer, having now a scale before his eye, is enabled to make a very close estimate of the quantity of lead corresponding to that contained in the acid to be analysed. One reason for making a fresh mixture of the acid to be examined is that the tint obtained becomes more turbid on standing. In using this method care must be taken as to the quality of the glycerol employed. The first glycerol I made use of proved worthless, as it contained some metal (tin or arsenic) giving a yellow precipitate with hydrogen sulphide. The glycerol was unaffected by a solution of hydrogen sulphide unless an acid was also added, but the presence of tartaric, acetic, or especially of hydrochloric acid at once determined a yellow turbidity, which became in time an actual precipitate. I obtained at last from Messrs. Hopkin and Williams a glycerol nearly free from this impurity. I have found the glycerol method a satisfactory one for determining lead in tartaric acid; it cannot, however, be used with advantage if the proportion of lead falls below 0.0020 per cent. We have already noticed that in the case of tartaric acid a remarkable break occurs in the series of tints produced by a solution of hydrogen sulphide with rising quantities of lead, the appearances produced by the smallest quantities of lead being mere grey turbidities, which change to a bronze and brown colour as the amount of lead increases. A break occurs in the same part of the scale when glycerol is present, and below this point varying amounts of lead have no considerable influence upon the colour produced. It is evident that, with the method just described, the matter derived from the corroded surface of the leaden vessels will, if present, partially escape detection. It would probably be easy, after pouring off the solution of tartaric acid, to act on any residue with a little sodium carbonate, then to dissolve the lead carbonate in some of the tartaric solution, and finally add this to the main bulk. I have not, however, experimented in this direction, as the next method I have to mention gets over all difficulties of this description. The glycerol method has not been employed for citric acid, though I know of no reason against its application. The need for the use of glycerol is, however, less in the case of citric acid, as the tints of lead sulphide in citric acid are far clearer than in tartaric; the scale of tint is also more continuous. The Ammonium Sulphydrate Method.-We have already mentioned that ammonium sulphydrate applied to an ammoniacal solution is a far more delicate test for lead in tartaric and citric acid (especially in the former) than solution of hydrogen sulphide. The tint produced is also of the same quality throughout a long scale, and is very free from turbidity. The reaction is, in short, admirably adapted for use as the basis of a colourmetric method. The depth of tint produced, for the same quantity of lead present, is far greater in an ammoniacal tartrate or citrate solution than in the same volume of water; it is quite essential, therefore, if equality of tint is to interpreted as equality of lead, that all comparisons should be between two citrate or tartrate solutions, and not between one of these and water. To carry out the method it is therefore necessary to have in stock solutions of lead-free tartaric and citric acid supersaturated with pure ammonia; these solutions should develop no colour when treated with ammonium sulphydrate. A convenient strength is 100 grms. of acid in 300 cc. of final solution. Of the tartaric or citric acid to be examined 40 grms. are taken and dissolved in a little water; warm water is most convenient for crystal and cold for powder; the solution is preferably conducted in a flask. To the cold solution pure strong ammonia is gradually added till it is in slight excess; the final point is indicated in the case of tartaric acid by the solution of the acid ammonium tartrate first formed; in the case of citric acid it is conveniently shown by a fragment of turmeric paper floating in the liquid. When an excess of ammonia is reached the liquid is cooled, if necessary, diluted to 120 cc., and filtered. As a preliminary experiment 10 cc. are taken, diluted to 50 cc. in the measuring cylinder, and placed in a Nesslerising glass, one drop of ammonium sulphydrate solution added, and the whole well stirred; the colour developed indicates what volume of solution should be taken for the determinaIf less tion, this volume may range from 5 cc. to 50 cc. than 50 cc. are taken the volume is brought to 50 cc. with water, and one drop of ammonium sulphydrate is then added. The tint thus adopted has now to be matched with the pure solutions. A volume of the pure ammoniacal tartrate or citrate, identical with that taken of the acid under examination, receives a measured quantity of lead solution from the burette, the volume is brought to 50 cc., it is placed in a Nesslerising glass, and receives one drop of ammonium sulphydrate; the experiment is repeated till a match is obtained. As in the previous method, the best comparison of tints is obtained by making finally three simultaneous experiments, one with the acid under examination, the other two with slightly varying amounts of lead, the aim being that the tint given by the acid to be analysed shall lie within this narrow scale. In following this method con siderable use has to be made of the weaker of the two lead solutions already mentioned. The whole time required for a determination of lead by either of the methods now given is about 1 hour; this time will be somewhat shortened as the operator becomes familiar with the tints produced by varying proportions of lead. We have already noticed the steps which should be taken if copper is present, or if iron is in sufficient quantity to disturb the result. In a few cases the same sample of tartaric acid has been analysed both by the glycerol and ammonium sulphydrate method; the agreement of the results was good. The Amount of Lead present in Commercial Acids. In Table I. will be found determinations of lead in commercial tartaric acid, and in Table II. similar determinations in citric acid, made by the methods just described; the figures obtained by the glycerol method are marked g. The aim has been to analyse a sample of the acid made by each manufacturer; the collection made of foreign acids is, however, very imperfect. Where more than one sample has been obtained from any maker the results are bracketed. The object has been to show the composition of unpurified acid; all samples known to have been purified from lead have been excluded; it is possible, however, that some of the best of the foreign acids have undergone some purification. |