Page images

action of hydrogen peroxide on an acidified permanganate solution in the absorption burette.

H. B.

New Mode of Testing for Nitrates. By E. BRÉAL (Ann. Agronom., 13, 322-327).—The author utilises the tendency of dissolved nitrates to accumulate in capillary spaces, or where evaporation is most active. One end of a strip of white filter-paper is immersed in the solution to be tested, the other end being freely exposed to the air. After the lapse of 12 to 15 hours, if the solution contained only traces of nitrates, these will be found entirely concentrated within a millimetre of the free end of the strip of paper. This is cut off, dried on a white surface, and moistened with a drop of phenolsulphonic acid, made by adding one part of crystallised phenol to one part of recently boiled pure sulphuric acid and diluting with two parts pure distilled water. A more or less intense red coloration indicates nitrates; when a drop or two of ammonia is added the colour changes to an intense blue or green. The filter-paper employed must be purified by repeated washing with pure distilled water. The delicacy of the test is considerable. One drop of a solution containing 1 mgrm. of potassium nitrate in 1 c.c. of water gives with phenolsulphonic acid an intense red coloration; this drop will contain about 0.025 mgrm. of nitric acid. If one drop of this solution is added to 100 c.c. of water, and the strip of filter-paper immersed in it, nitric acid may be distinctly recognised after 12 hours in the millimetre of paper nearest the free end. This method, therefore, affords the means of detecting a quarter of a mgrm. of nitric acid in a litre of water, or 1 in 4,000,000. If the strip of paper be partly immersed in moist soil, any nitrates in the soil can be very easily detected, and the method promises to become of much use in experiments on the relation of nitrates to plant growth. J. M. H. M.

Detection of Potassium by means of Sodium Bismuth Thiosulphate. By C. PAULY (Chem. Centr., 1887, 553).-The author has reinvestigated this reaction, first proposed by Carnot, and finds that it is really characteristic for potassium, and can be applied without previous separation from most other metals. It consists in adding an alcoholic solution of sodium bismuth thiosulphate, when a yellow, crystalline precipitate of potassium bismuth thiosulphate, K,Bi(S2O3)3, is obtained. A. J. G.

Determination of Alkaline Chlorides in Potashes. By H. FOCKE (Chem. Centr., 1887, 699).—In order to determine the propor tion of sodium and potassium chlorides in potashes, the dissolved sample is neutralised with dilute sulphuric acid, a few drops of normal soda added, and the whole evaporated. The residue is taken up with 85 per cent. alcohol, the water required for the dilution of the alcohol being poured on the mass, and then the absolute alcohol added gradually. After a time, the liquid is filtered, and the insoluble portion washed with alcohol; the mixed solutions are then evaporated, the residue dissolved in water, and the proportion of alkali-metal, and chlorine determined in aliquot proportions respectively. V. H. V.

Testing Copper Sulphate. By BAUDOIN (Ann. Agronom., 13, 319-321).-Copper sulphate being now largely employed for application to vines, it becomes of importance to test commercial samples, especially for the sulphates of iron, zinc, and magnesium. Iron is of course easily detected by boiling the solution with nitric acid and adding excess of ammonia. Zinc is best detected by Chancel's process; sodium hyposulphite is added to the solution of the copper sulphate until the latter is decolorised, sodium carbonate is then added, which precipitates both zinc and magnesium if present. They can be separated in the usual way. The author is engaged in ascertaining whether the density of commercial samples of copper sulphate may be utilised as indications of purity. J. M. H. M.

Analysis of Clay. By C. MEINEKE (Chem. Centr., 1887, 553).— The clay is fused with hydrogen sodium carbonate, the product treated with hydrochloric acid, evaporated to dryness, and heated for some time at 150-300°. It is then dissolved in cold water, decanted five times, and the silica washed on a filter. After ignition, the silica is again evaporated with hydrochloric acid and washed. The weighed silica is treated with hydrofluoric acid, the alumina and ferric oxide determined, and the weight deducted.

The filtrate is evaporated in a platinum dish, when a residue is obtained insoluble in hydrochloric acid, and consisting of silica and alumina.

Sand is determined in the residue obtained by decomposing the clay with sulphuric acid. The residue is dried at not too high a temperature, and exposed to air until no further increase in weight takes place during weighing. The greater part of the residue is weighed, and divided into two nearly equal portions, the one (a) is ignited, the other (b) is extracted with boiling alkali. From the weight of (a) the weight of (b) when anhydrous is calculated. The filtrate (c) with the rest of the residue is now ignited, and thus the whole weight of anhydrous residue and the proportion a: b c (a + b + c) in the dry and anhydrous state is obtained.

N. H. M.

Determination of Manganese by means of Mercuric Oxide and Bromine. By C. MEINEKE (Chem. Centr., 1887, 554).-Volhard's method is modified as follows:-The solution in nitric or sulphuric acid containing iron as well as manganese to be determined, is precipitated with zinc oxide and made up to a definite bulk. An aliquot part of the solution is treated in the following manner:-Mercuric oxide rubbed with water and bromine-water are alternately added until the former is in considerable excess, when a sudden separation of manganese occurs. The manganese peroxide adhering to the sides of the vessel is dissolved in hydrochloric acid and treated separately, The precipitate is mixed with the oxides of iron and zinc, but is free from manganous oxide. The pure manganic oxide is determined volumetrically by reducing with oxalic acid mixed with hydrochloric acid (eq. mols.), and determining the excess of oxalic acid with permanganate. N. H. M.

Estimation of Metallic Iron in Slags. By G. NEUMANN (Zeit. anal. Chem., 26, 530-534).—The usual process, which consists in digesting the slag with copper sulphate and ascertaining the quantity of copper deposited, is accurate in the absence of other reducing agents but is tedious. The measurement of the hydrogen evolved by the substance when treated with a dilute acid gives equally accurate results, and is much more rapid. The ferric oxide present seems to remain inert. A special apparatus for the purpose (for which the unfortunate name "hydrometer" is proposed) consists of a graduated tube with a stopcock-funnel at the top, and a wide-bore (5 mm.) stopcock at the bottom. A narrow tube inserted at the side in an upward direction just above the lower stopcock communicates with a pressure tube like that of the nitrometer. The flask containing the substance is connected directly with the lower end of the graduated tube which is widened to the size of the neck, and is filled to the lower stopcock with water. The graduated tube is then filled through the pressure tube with acid which, on opening the stopcock, finds its way down to the slag. The gas rises through the stopcock into the burette. Any carbonic anhydride or hydrogen sulphide can be absorbed by running in potash through the stopcock-funnel until the contents of the burette are alkaline. M. J. S.

Rapid Methods for the Estimation of Silicon, Sulphur, and Manganese in Iron and Steel. By J. J. MORGAN (Chem. News, 56, 82-83).-Silicon is estimated by the so-called "roasting" method, heating in a muffle at a bright-red heat for 20 minutes, treating with hydrochloric acid, and igniting the insoluble silica, &c.

For the estimation of sulphur, the sample is treated with sulphuric acid, and the gases evolved are passed into a measured quantity of a dilute solution of lead acetate. The colour of this solution is then compared with that obtained by treating a steel having a known amount of sulphur in a similar manner, and so on.

To estimate the manganese, the sample is dissolved in nitric acid, cooled, treated first with a small quantity of water, then with lead peroxide and a few drops of strong nitric acid, boiled for four minutes, and finally cooled. The manganese is calculated from the permanganate formed, which is estimated by comparing the colour with a standard as in the case of the sulphur. For phosphoric acid, the author prefers the molybdate method. D. A. L.

Estimation of Silicon in Iron and Steel. By T. TURNER (Chem. News, 56, 49-50).-It is shown by experiments that silicon can be correctly estimated in cast irons of very good quality by evaporation with dilute sulphuric acid; but with phosphoric irons the residues obtained are white but impure, and when in addition an iron contains titanium, the residue, although very nearly white, may have 20 per cent. of impurity, and contains iron, phosphoric acid, and titanium. Evaporation with sulphuric acid gives fairly good results with steels containing only a small quantity of silicon. In the aqua-regia method, the colour of the residue is usually an indication of its purity.

D. A. L.

Determination of Sulphur in Iron. By P. PLATZ (Chem. Centr., 1887, 579). The barium chloride method for the determination of sulphur in iron is modified as follows:-The iron and sulphur are oxidised by nitric acid; on prolonged boiling, the nitric acid is displaced by hydrochloric acid by gradually adding the latter. The whole is then evaporated until vapours of hydrogen chloride are no longer evolved, the residue diluted with water, and the insoluble silica filtered off. To the filtrate, barium chloride is added, and the precipitate collected and heated to oxidise any barium sulphide which may have been formed.

The precipitate is finally digested with hydrochloric acid to dissolve out the iron oxide, and weighed. Control experiments gave satisfactory results. V. H. V.


Volumetric Estimation of Cobalt in Presence of Nickel. N. MCCULLOCH (Chem. News, 56, 27-29).-The author finds that cobalt "peroxide" is reduced by free iodine, iodides, hypochlorites, hypobromites, and hydrogen peroxide and ammonia to a lower oxide or oxides, possibly CoO19, hence the author confirms Bayley's results (Abstr., 1879, 507), but he gives no hope of founding a method of estimating cobalt in presence of nickel, depending on the higher oxides of cobalt. D. A. L.

Separation of Nickel and Cobalt from Iron. By J. B. MACINTOSH (Chem. News, 56, 64-65).-Proceeding at first by the old method of precipitating all three metals by ammonium sulphide, and treating with dilute hydrochloric acid; a solution with much iron and little nickel and cobalt, and a precipitate rich in nickel and cobalt but poor in iron are obtained. These are then treated separately in the ordinary way, by precipitating the iron as basic acetate, and perfect separation of the nickel and cobalt from iron is effected much more readily than by the "basic acetate" treatment alone.

D. A. L.

Direct Precipitation of Nickel Oxide in the Presence of Iron. By T. MOORE (Chem. News, 56, 3).-The solution containing the two metals, freed from acids, is treated first with glacial phosphoric acid or sodium pyrophosphate until the precipitate begins to dissolve, then with potassium cyanide which dissolves the remainder of the precipitate. This solution is now boiled for a couple of minutes, and the addition of potassium cyanide continued until potassium hydroxide does not give a precipitate. When cool, the solution is made alkaline with potassium hydroxide, excess of bromine solution in potassium hydroxide added, and the whole warmed; nickel and manganese (if present) are precipitated, iron and cobalt remain dissolved. The precipitate is dissolved in sulphuric acid and the nickel deposited electrolytically. Good results have been obtained with this inethod. D. A. L.

Water Analysis. (Chem. News, 56, 113.)-The Chemical Section of the American Association for the Advancement of Science recommends the following plan for a uniform method of stating results of

water analysis. They are of opinion that two distinct schemes should be in use, one for mineral and one for potable waters. Mineral water results should state in parts per 1000, by weight, each basic element, each acid element in combination or supposed combination with the bases, the remaining acid elements being given in connection with all the oxygen of their salts (CO3, SO1, &c.). Volumes of gases expelled on boiling to be in cubic centimetres per litre. Constituents should be arranged in electropositive order, positive ones first. Potable water results to be stated in parts per million, to include— Total solids, chlorine, nitrogen expelled on boiling with sodium carbonate, and nitrogen as free ammonia, nitrogen expelled by boiling with alkaline permanganate and albuminoïd nitrogen, nitrogen as nitrite and as nitrate; organic matter; hardness. D. A. L.

Estimation of Methyl Alcohol in Presence of Ethyl Alcohol. By O. HEHNER (Analyst, 12, 25-29).—Since 1 part of ethyl alcohol (oxidised to acetic acid) reduces 4-278 parts of potassium dichromate, whilst I part of methyl alcohol requires 9.224 parts (oxidising to carbonic anhydride and water), the proportions of the two in a mixture can be calculated from a determination of the amount of dichromate reduced by 1 part of the mixture. This was confirmed by several experiments on mixtures of known composition. The oxidising solution contained 80 grams of dichromate, and 150 c.c. of concentrated sulphuric acid in the litre. Of the diluted alcohol, a quantity containing about 0.2 gram is mixed with 25 or 30 c.c. of the chromate solution in a bottle, which is corked and heated in boiling water for at least two hours. The unreduced chromate is then estimated by iron solution and standard dichromate. M. J. S.

Dimethyl Ethyl Carbinol. By B. FISCHER (Arch. Pharm. [3], 25, 777-779).—This compound (amylene hydrate) has recently been strongly recommended as an opiate. As it may be contaminated with fermentation amyl alcohol, which is poisonous, it should be subjected to the following tests. 1 gram is dissolved in 15 c.c. of water and tinted a faint red with permanganate solution; the colour ought not to change within 15 minutes (ethyl and amyl alcohol). 1 gram is dissolved in 15 c.c. of water, and slightly warmed with potassium dichromate and dilute sulphuric acid, no green colour should appear within half an hour (as above). 1 gram dissolved in 15 c.c. water is warmed with some drops of silver nitrate and a trace of ammonia. The silver should not be precipitated (aldehyde, with which most of the primary alcohols are contaminated). As an opiate this compound is more powerful than paraldehyde, but less so than chloral hydrate. Its therapeutic value depends on the fact that it affects neither the breathing nor the action of the heart. J. T.

Estimation of Glycerol in Fermented Liquids. By L. LEGLER (Analyst, 12, 14-16).-The ordinary process, in which the crude glycerol is purified by extraction with ether-alcohol, is unsatisfactory, owing to the retention of some of the glycerol by the impurities. The author dilutes the crude glycerol with water, adds an excess of

« PreviousContinue »