22 ABSTRACTS OF CHEMICAL PAPERS INORGANIC ANALYSIS. Method for Preparing a Neutral Ammonium Citrato Solution. J. M. McCandless. (J. Ind. and Eny. Chem., 1914, 6, 921-922.)-The citric acid to be used is wsll mixed and ground, and a portion weighing 0.35 grm. .is titrated with & alkali solution, using phenolphthalein as the indicator. .If the acid is pure and uneffloresced, exactly 50 C.C.of the alkali solution will be required. One thousand eight hundred and fifty grms. of the pure acid, or its equivalent in eflloresced or impure acid, are then placed in a large stoppered bottle, dissolved in about 7 litres of water, and a quantity of ammonia is added in the ratio of 1 part by weight of ammonia (NH,) to 3.765 parts of anhydrous citric acid. This is the ratio in the pure salt (NH,),C,H,07.The ammonia solution added must be titrated previously, using methyl orange as indicator, and may be used in the form of the concentrated solution or after dilution. In the latter case, less water is employed to dissolve the citric acid, The mixture is shaken until any remaining citric acid is dissolved, cooled, and diluted to a sp. gr. of 1-09, The resulting solution will have a volume of about 10 litres.w. P. s. Determination of Cadmium in Zinc. W. Cooper. (Chm. News, 1914,110, 250-251.)-The sample (5 grm8.) is covered with 150 C.C. water, acidified with 7 C.C. Rulphuric acid, and allowed to dissolve in the cold. If it does not dissolve when left overnight (which is rarely the case), the use of a piece of platinum foil is resorted to.The solution is decanted from the black residue, which contains copper and lead, but never any cadmium, and the residue is washed by decantation three times, using 20 C.C. of water each time. The acid solution and washings are made more strongly acid by addition of 20 C.C. of dilute (1 : 3) sulphuric acid and saturated with hydrogen sulphide for twenty minutes. After standing an hour or two, the precipitate of cadmium sulphide, always contaminated by zinc sulphide, is filtered off, washed three times with cold water, and redissolved in 5 C.C.hydrochloric acid and 15 C.C. bromine water, which is warmed in the precipitation flask and poured over the filter. About 5 C.C. of dilute sulphuric acid are added, the solution evaporated until fumeINORGANIC ANALYSIS 23 arise, diluted to 50 c.c., cooled, and filtered from traces of lead sulphate after standing for an hour.The filter is washed three times with cold water, the filtrate made more strongly acid by adding 20 C.C. of dilute sulphuric acid, diluted to 200 c.c., and saturated with hydrogen sulphide for twenty minutes. The precipitate is filtered off, washed, and redissolved as before, and the solution again evaporated with sulphuric acid until fumes arise, After addition of more acid and dilution, as before, it is again saturated with hydrogen sulphide.The precipitate, this time pure cadmium sulphide, is filtered off on a filter that has been washed with 200 C.C. of water con- taining 25 C.C. dilute (1 : 3) sulphuric acid, then washed free from acid, dried at 100' C., and weighed.The precipitate is washed and dried in a similar manner, and finally weighed with the tared filter. Quantities up to 0.5 per cent. can be estimated within 0.015 per cent. Duplicate determinations may differ by twice this amount. G. C. J. Rapid Method for Estimating Carbon in Iron and its Alloys. E. Szasz. (J. SOC, Chem. Ind., 1914, 33, 994-997.)-A modification of the method previously described (ANALYST, 1913, 38, 384), depending on direct dry combustion in a limited supply of oxygen, and measurement of the carbon dioxide formed.The apparatus previously described was complicated and very costly, involving the use of much platinum. It is also legs liable to injury. Its disadvantages compared with the original apparatus am that an estimation takes slightlylonger, that the turnings must be mixed with an oxidising agent, and that ferrochrome and certain other alloys rarely encountered cannot be burned in it.These are real disadvantages, but it is probable that the comparative cheapness of the apparatus will secure the adoption of the method in many cases where the expense of the original apparatus would not be justified.The new apparatus is slightly less complicated and much less costly. G. C. J. Rapid Method for Glass Analysis. E. C. Sullivan and W. C. Taylor. ( J . Ind. and Eryg. Chem., 1914, 9, 897-899.)-The novel feature of these methods is the use of oxalic acid to decompose fluorides. As the oxalates can in turn be decom- posed by heat, no bases or acids are introduced to interfere with the determination of the glass constituents.The method is especially useful with lead glasses, con- taining little eke save silica, soda, and potash, and also with glasses containing antimony or arsenic, but it can be adapted to other gIasses with suitable modification of the subsequent operations. The powdered glass (1 grm.) is placed in a platinum crucible and moistened with water.Oxalic acid (2 grms.) is next added, and about 20 C.C. of 48 per cent. hydrofluoric acid. The mixture is evaporated at a temperature just high enough to expel the excess of oxalic acid, an apparatus described by Hillebrand (27.8. Geol. Survey, Bull. 422, 31) being used for the purpose. .When all the acid has been expelled, the crucible is cooled and the evaporation repeated twice more with water and oxalic acid.About 5 grms. of oxalic acid are used in all, and the quantity must be roughly weighed unless the preparation is free from non-volatile impurities. The only impurity likely to be present in a good sample is soda, but 6 grma. of oxalict acid may contain more than 1 mgrm. of this. With their lead24 ABSTRACTS OF CHEMICAL PAPERS glasses, the authors treat the residual oxalates with hot water, filter off the lead oxalate (and a trace of calcium oxalate, which they neglect), wash this, and titrate it with permanganate.The filtrate from the lead oxalate is evaporated to dryness in a platinum basin, and the residual oxalates are decomposed by heat. The carbo- nates are taken up with water and hydrochloric acid, and the solution evaporated to remove a, trace of sillca.The salts are taken up in water and hydrochloric acid, and iron, aluminium, and manganese are precipitated by means of ammonia and bromine water, and removed by filtration. In half the filtrate a trace of magnesia is preci- pitated as phosphate, whilst the other half of the filtrate is evaporated to dryness, and the residue freed from ammonium salts by heat, and weighed.From this weight, that of the magnesia found in the other portion, the alkali in the reagents used, and the potash subsequently estimated, the percentage of soda is calculated. In applying the method to other glasses, the above procedure must necessarily be modified. I n addition to lead and calcium, zinc always remains as insoluble oxalate, whilst the omlates of aluminium, chromium, antimony, and arsenic are soluble.Iron is wholly soluble except in presence of zinc, when a small proportion may be found with the latter. Manganese up to 2 per cent. passes into solution, but larger amounts tend to be distributed between the soluble and insoluble oxalates. Magnesium in quantity tends to remain with the insoluble portion, as do copper, cobalt, nickel, and barium; but the separation in these cases is not sharp, and depends partly on the presence of other elements.Arsenic and antimony are removed from the soluble oxalates with hydrogen sulphide before the oxalates are decomposed. This necessitates acidifying with hydrochloric acid, which must be expelled before decomposing the remaining oxalates, as otherwise alkali chlorides would be lost by volatilisation.With more than 3 per cent. of alumina present, the method is not trustworthy. Borates do not interfere. On the contrary, they make it possible to estimate accurately somewhat larger amounts of alumina than 3 per cent. A method described by Wherry and Chapin for the estimation of boric acid in complex mineral silicates (ANALYST, 1909, 34, 34) has been lound very useful in the analysis of borosilicate glasses.For the latter purpose one or two minutes' fusion (instead of fifteen minutes) suffices. I t is pointed out that the washing of the considerable precipitate of calcium carbonate must be effected with the aid of suction, or it is impossible to wash out all the boric acid without using more than 100 C.C.of water, as directed by Wherry and Chapin. Borosilicate glasses containing zinc or lead give unsatisfactory results by this method, but a modification which successfully meets such cases is described in the paper. G. C. J. Precise Standardisation of Hydrochloric Acid Solutions. L. W. Andrews. (J. Amer. Chem. Soc., 1914,36, 2089-2091.)-The following is preferred to any method which involves the transference and washing of a precipitate, or requires the use of any standard substance containing water of crystallisation.It depends on the loss of weight caused by conversion of silver nitrate into the chloride. Since hydrochlorio acid kept in glasEi vessels usually contains traces of non-volatile impurities, the method includes compensation for any error due to this cause, the partioulm form of control adopted also compensating for errors whichINORGANIC ANALYSIS 25 might arise from consecutive weighings of comparatively large objects under diverse atmospheric conditions.Two similar silica dishes are provided with watch-glass covers, and one of them with a stirring-rod, short enough to lie under the cover.Into this dish about 2 grms. silver nitrate are put, and both dishes are placed in an oven at 160° C., the temperature being subsequently raised to 244' C. After cooling in a dessicator, both dishes are weighed. hydrochloric acid to be standardised are run into each dish, the contents of which are then evaporated on the water-bath, and finally dried at 240" C. Fifty C.C.of The normality of the solution is given by the expression- N = w- .- W1+w,-w 0.02655 v--' in which V=volume of acid delivered by the 50 C.C. pipette, W=weight in air of silver nitrate and dish, Wl=weight of dish with mixed chloride and nitrate, w =weight of control-dish before, and w1 its weight after the experiment. G. C. J. Sulphate Method for Standardising a Magnesium Salt Solution.C. W. Foulk, and 0. R. Stewart. (J. Amer. Chem. SOC., 1914, 36, 2360-2372.)-As a standard preparation of magnesium-for use, for example, in an investigation of methods for the estimation of that metal-the most convenient is a solution of the pure chloride, standardised by treating measured portions with excess of sulphuric acid, evaporating, and igniting the residue.The magnesium chloride must of course be free from non-volatile bases other than magnesia, and the method assumes that magnesium chloride can be converted to anhydrous sulphate without loss. In the course of an investigation on the estimation of magnesium, the authors even found it necessary to test this assumption. Their work, conlparable in technique to the determination of atomic weights, shows it to be well founded.The above work involved the preparation of anhydrous magnesium chloride of the highest possible degree of purity. The methods used are described, but, for general analytical purposes, the preparation of the anhydrous salt is unnecessary. The preparation of a salt free from heavy metals, calcium, and sodium-to name the commonest impurities-is, however, a condition precedent to the preparation of a standard solution of magnesium.Occasionally a supply of the salt can be purchased which gives, even in 50 per cent. solutions, no visible precipitate with hydrogen sul- phide nor with ammonium oxalate. As very small quantities of the heavy metals or of calcium can be detected by these means, such a sample may be considered satis- factory in this respect.Sodium, at least in spectroscopic traces, will almost certainly be present. To purify the salt, the authors used the following methods, but it is to be noted that these methods were adopted to give the greatest possible weight to their results on the conversion of chloride into sulphate : The concentrated solution was treated with purified carbon dioxide and ammonia gas, and the double carbonate of magnesium and ammonium was washed forby to fifty times by decantation with specially redistilled water.I t was then washed on a platinum cone contained in a platinum funnel for four hours, transferred to a platinum dish, and heated in a hot-26 ABSTRACTS OF CHEMICAL PAPERS air oven until the odour of ammonia could no longer be detected.This preparation was free from even spectroscopic traces of sodium, as well as from platinum, and would presumably answer any ordinary analytical purpose as a source of magnesia, free from other non-volatile bases. For their purpose the authors dissolved it in purified hydrochloric acid, added approximately the theoretical amount of specially prepared, pure, ammonium chloride, purified the double chloride by recrystallisation twice, and from this prepared anhydrous magnesium chloride by heating in a current of dry hydrochloric acid gas and subsequent fusion.On account of the hygroscopic nature of the anhydrous chloride, the last operations were conducted in a modifica- tion of Richards’ apparatus (J. Chem. Soc., 1911, 99, l20l), which is described, and, unlike Richards’ apparatus, could be assembled in any laboratory. G.C. J. Perchloric Method of Determining Potassium, as Applied to Water Analysis. (J. Amer. Chm. SOC., 1914, 36, 2085-2089.)-The method was found accurate, the maximum error recorded being less than 1 mgrm. with 500 mgrms. potassium present. With smaller weighings the absolute errors are less, but the percentage error tends to increase.The experiments included separations from 200 mgrms. of calcium chloride, 100 mgrms. magnesium chloride, 200 mgrms. sodium chloride, 100 mgrms. sodium nitrate, and 100 mgrms. sodium phosphate, all present together, as well as separations from 1 grm. of sodium chloride. Ammonium salts must be eliminated and large quantities of sulphates must not be present, though small quantities do not interfere.The author eliminates sulphates by adding barium chloride to the strongly acid solution (10 C.C. hydrochloric acid in 150 c.c.). The filtrate is evaporated to dryness, and the residue heated to expel ammonium salts, and then taken up in 20 C.C. water. More than enough perohloric+acid is added to combine with all the bases present, and the solution evaporated to dryness.The residue is taken up in 10 C.C. water and evaporated with more perchloric acid until fumes arise. The cooled residue is treated with 25 C.C. of 97 per cent. alcohol containing 0.2 per cent. of perchloric acid, and the precipitate filtered off on a Gooch filter, washed with the acid alcohol, and dried at 120-130’ C. for an hour. NOTE BY ABSTRACTOR.-DrtViS (ANALYST, 1913,38, 47) has shown that the best liquid with which to wash the precipitate is alcohol saturated with potassium perchlorate.C. Scholl. G. C. J. Ashes of Hedge Clippings and Trimmings as a Source of Potash. E. J. Russell. (J. Board of Agric., 1914, 21, 694-697.)-The ash of bonfires com- posed of threshing-waste, grass, weeds, dead and green wood, etc., was found to contain 11 per cent.of potash (K,O), a percentage nearly equal to that of kainite, which contains about 12.5 per cent. of K,O. Thus some of the normal processes of farm routine give rise to ash rich in potash, which would in normal times be worth about 40s. a ton, and is now worth much more. Since the potassium is present in the ash as carbonate, and therefore soluble in water, care must be taken to prevent loss due to rain, as experiments showed that a, single night’s rain of less than 0.1 inch diminished the potash in a heap of ash by 50 per cent.The cleaning out of hedgeINORGANIC ANALYSIS 27 bottoms yielded about 5 pounds, and hedge-trimmings on the average 15 pounds of ash per 100 yards of hedge (one side only). A 20-acre field with 1,300 yards of hedge would, therefore, on this basis yield ash equivalent to more than 3 cwt.of kainite from the hedge bottoms, and to nearly 2 csvt. from the total trimmings. It was found that the cost of labour for obtaining and burning such clippings worked out at from 3d. to 8d. per pound of ash when the material was mainly grass, and about 2d. a pound when more wood than grass is present.Even when the ash could be obtained for Id. a pound this is equivalent to kainite at 3 9 10s. a ton, SO that potash obtained in this form would be very expensive if charged with the whole cost of the process, but where the trimming, etc., has to be done in any case, would well repay the trouble of collection. H. F. E. H. Radium : Uranium Ratio in Carnotites.(Estimation of Uranium.) s. C. Lind and C. F. Whitternore. (J. Amer. Chenz. Xoc., 1914, 36, 2066-2082.)- More than twenty samples of various origin were examined, and in every case in which the sample was carefully drawn from a considerable bulk of ore (several hundred pounds up to 25 tons) the ratio was found to approximate closely to the value 3.33 x 10-7 found by Heimann and Marckwald (Jahrb.d. Radioakt. 21. Elekronik, 1913,10,299; Physik. Zeitsch., 1913,14,303) for pitchblende, the mean value for all these large samples, ten in number, being within 1 per cent. of Heirrialan and Mrtrckwald’s figure. Small parcels of a few poucds showed values in some cases very close to the normal pitchblende ratio, but in other cases widely different, the extreme values recorded being 2.4 x 10-7 and 4-6 x 10-7.It is suggested that these low and high ratios are due to local transposition of radium within the ore bed. For the purpose of this work it was necessary to improve on previous methods for the estimation of uranium in minerals containing only 2 to 3 per cent. of uranium and twice as much vanadium. The ore (2 to 5 grms.) is treated with 10 C.C.of hydrochloric acid, and, after fifteen minutes, 5 C.C. of nitric acid are added, and the mixture heated on the water-bath until effervescence ceases, when it is evaporated to dryness. Hydrochloric acid (3 c.c.) and water (5 c.c.) are added to the warm residue., still on the water-bath, and, after a few minutes, 25 C.C. of hot water are added, the mixture filtered, and the residue washed.The residue, which may retain vanadium, is ignited in platinum, treated with 5 C.C. hydrofluoric acid, and the solution evaporated to dryness on the water-bath. Hydrochloric acid (3 c.c.) is added and the solution evaporated to dryness, and this treatment repeated. The residue is treated with 2 C.C. hydrochloric acid and 2 C.C. water until any red crusts are dissolved and tho solution filtered into the main liquid.This liquid is freed from copper, etc., by means of hydrogen sulphide and filtration, freed from hydrogen sulphide by boiling, and concentrated to 100 C.C. Iron is oxidised with hydrogen peroxide, and then precipitated by neutralising with dry sodium carbonate and adding 2 to 3 grms. in excess. After boiling for fifteen minutes, the ferric hydroxide is filtered off, washed, redissolved in the least possible quantity of nitric acid (1 : l), the solution treated with 10 C.C.hydrogen peroxide and iron precipitated with sodium carbonate as before. The precipitate is now free from uranium, but may contain a little vanadium, and is reserved if vanadium is to be28 ABSTRACTS OF CHEMICAL PAPERS estimated.The filtrate from it is added to the first filtrate from iron, the mixture concentrated to 200 c.c., treated with 10 C.C. nitric acid, and boiled to expel carbon dioxide. Ammonia is added until a slight permanent turbidity appears, and then about 4 per cent. of nitric acid, followed by 10 C.C. of 20 per cent. lead acetate solu- tion and enough ammonium acetate to reduce the hydrion concentration approxi- mately to that of acetic acid.About 20 C.C. of a mixture of 80 parts strong ammonia, 100 parts water, and 70 parts acetic acid, effect this purpose with sufficient exactness. After heating on the water-bath for an hour, the lead vanadate is filtered off, washed, dissolved in the least possible quantity of dilute (1 : 3 or weaker) nitric acid, the solution neutralised with ammonia as before, treated with 3 C.C.nitric acid, then with 2 C.C. lead acetate solution and excess of ammonium acetate. The lead vanadate, after digestion on the watier-bath, is filtered off, washed, and reserved for the estimation of vanadium. The united filtrates from lead vanadate are concentrated to 400 c.c., treated with 10 C.C. sulphuric acid to separate most of the lead as sulphate, which is filtered off and washed with cold water.The filtrate is neutralised with ammonia and treated with freshly prepared ammonium hydrogen sulphide until the solution is yellow, and the uranium and what lead is present,are separated as sulphides. When these have been settled by warming on the water-bath, they are filtered off, washed slightly with warm water, redissolved in hot, dilute (1 : 2) nitric acid, and the solution mixed with 5 C.C.sulphuric acid, and evaporated until dense fumes arise. After cooling and dilution, the solution is boiled, allowed to cool again, and filtered from lead sulphate, which is washed with very dilute sulphuric acid. The filtrate is nearly neutralised with ammonia., cooled, and about 2 grms.ammonium carbonate are added. The precipitate of alumina is filtered off, washed with hot water, and, if bulky or at all yellow, is redissolved in dilute sulphuric acid and reprecipitated as described. The filtrate from alumina is acidified with sulphuric acid, boiled to expel carbon dioxide, made slightly alkaline with ammonia while it is hot, and heated on the water-bath until the ammonium uranate settles.The latter is filtered off, mashed with 2 per cent. ammonium nitrate, dried, and ignited to U,O,. After weighing, the precipitate should be dissolved in nitric acid and tested with hydrogen peroxide for vanadium, and with ammonium carbonate for aluminium. The most prevalent source of error in the estimation of uranium is co-precipitation of the oxides of these two metals and of silica.For the estimation of vanadium, the lead vanadate is dissolved in nitric acid, and the solution mixed with 10 C.C. sulphuric acid, and evaporated until dense fumes arise. To the cooled and diluted solution, which need not be filtered from the lead sulphste, is added the aqueous extract resulting from fusing, with sodium carbonate, the precipitate of ferric hydroxide obtained at an early stage of the analysis, and extract- ing with water, Vanadium is reduced to the tetravalent condition by treatment with 10 C.C.of a saturated solution of sulphur dioxide, the excess of which is eliminated by boiling for ten minutes after its odour can no longer be detected. The hot solution is then titrated with permanganate. G. C. J.INORGANIC ANALYSIS 29 Improvement in the Electrical Method of Determining Salt in Soil. W.Beam and G. A. Freak. (Cairo Sci. J., 1914, 8, 130-133.)-The resistance of the soil solution is measured by means of a compact circular slide-wire bridge, of which the part supporting the bridge wire should be made of vulcanite, and not wood, when used in dry climates, The resistance of a given salt concentration is affected by temperature and by the texture of the soil, and also by the nature of the salts present, especially sodium carbonate.Humus and organic matter are prejudicial in that these substances increase the resistance. Error, due to the nature of the salts, may be avoided by constructing special tables for the particular combination of salts present in the soil.This is readily done by using a mixture extracted from the soil in question. A more serious error is due to the inclusion of calcium salts, especially the sulphate, with the soluble injurious Ralts. Calcium sulphate is not only harm- less, but actually neutralises the effects of other salts. Thus, in the case of the cotton plant, it has been shown that a, sufficiency of calcium sulphate increases the resistance of the plant to sodium chloride as much as thirty-two times.The best way to eliminate the influence of calcium sulphate is to employ, in place of water, diluted alcohol (40 to 50 per cent. by volume) for the extraction of the salt, com- parison being then effected with a table of resistances of salt in the same solvent. Further, it is possible to extend the method to the determination of the calcium sulphate present, since it is only necessary to make another extraction with water on a, fresh sample, and from its conductivity and that of the salt known to be present to calculate the calcium sulphate. The authors give curves of the resistances of sodium chloride in water, calcium sulphate in water, and sodium chloride in 40 per cent. alcohol. Sodium sulphate, which occurs in soil in the Nile, has so nearly the same resistance as the chloride that the one curve suffices. The usual method, when only a moderate amount of salt is present (about 0.2 per cent.), is to take an amount of the soil, roughly equivalent to 20 grms. of the dry soil, place it in a stoppered bottle, and shake up with 100 C.C. of 40 per cent. alcohol for ten to fifteen minutes. Subsequent filtration is unnecessary. Slight variations in the time of shaking and the strength of the alcohol are immaterial. The following is an example of the method : Twenty grms. of soil treated with 40 per cent. alcohol gave a resistance of 1,400 ohms, equivalent to 0.10 per cent. of soluble salts (0.10 per cent. of salt in water has a resistance of 630 ohms). The resistance found for the water extract of the soil was 325 ohms. Then, as resistance is the reciprocal of conductivity, &- = &* ; therefore the resistance due to gypsum is 670 ohms, equivalent to 0.131 per cent. H. F. E. H.