Page images

Reduce the iron to the ferrous state by means of a current of sulphurous anhydride, and boil off the excess, when cool add phosphoric acid or ammonium or sodium phosphate, in excess of that required to precipitate all the alumina, then ammonia until a faint permanent cloudiness is formed, finally excess of ammonium acetate. Should

the precipitate contain much iron (it always retains a little), it is washed, redissolved in hydrochloric acid, and again treated with sulphurous anhydride, &c. 2nd method: Add ammonia to the reduced iron solution until a slight cloudiness forms, then excess of ammonium acetate, boil, and if the precipitate contains much iron repeat the reduction and precipitation. In either case, when a satisfactory precipitate is obtained, it is dissolved in hydrochloric acid, boiled with a little nitric acid, nearly neutralised with caustic soda, then boiled with large excess of the last reagent. The alumina is then precipitated as phosphate, the precipitate being washed with a hot 1 per cent. solution of ammonium nitrate, containing about 01 gram per litre of ammonium dihydrogen phosphate, and is weighed as Al,P2O.. Test results are very satisfactory. The presence of titanium does not interfere with the process.

D. A. L.

Silica in the Estimation of Manganese in Pig-iron, and Estimation of Phosphorus in Pig-iron and Steel. By L. M. DEANE (Chem. News, 54, 174-175).- When samples of pig-iron containing more than 2 per cent. of silicon are analysed by the ordinary method, an appreciable amount of silica escapes separation, and is finally precipitated with the manganese. The ignited oxide of manganese should therefore be treated with hydrochloric acid, and the silica mixed with it determined in the usual way. For the estimation of phosphorus, the substance is dissolved in nitric acid, evaporated to dryness, redissolved in hydrochloric acid, and again evaporated to dryness. The residue is dissolved in hydrochloric acid, the silica separated, the solution evaporated nearly to dryness, a few drops of nitric acid added, and the heating continued until all nitrous fumes have escaped. The cold solution is mixed with cold water, and the phosphoric acid precipitated and weighed as ammonium phosphomolybdate. D. A. L.

Direct Separation of Manganese from Iron. By L. BLUM (Zeit. anal. Chem., 25, 519-520).- From a solution containing nickel, cobalt, zinc, manganous and ferric chlorides, with tartaric acid and excess of ammonia, potassium ferrocyanide throws down all the metals except the iron. A clear filtrate can be obtained after boiling, but the precipitate cannot be washed. A solution containing 0-00004 gram of manganese and 0.01 gram of ferric chloride per c.c. gives a distinct precipitate.

M. J. S.

Detection of Stannic Sulphide in Presence of Antimonious Sulphide. By A. GRIFFITH (Analyst, 11, 164–165).—When a solution containing antimonious sulphide is shaken with ether, and allowed to remain at rest, the sulphide rises to the surface with the ether; stannic sulphide does not behave in this manner. Therefore, when a solution containing both these sulphides is shaken with ether in

sufficient quantity, the antimonious sulphide rises to the surface and exposes the otherwise obscured stannic sulphide. The test does_not work with stannous salts. D. A. L.

Production and Measurement of Gold and other Minute Metallic Spheres to Determine their Weight. By G. A. GozDORF (Chem. News, 54, 231-232).-The minute quantity of pure gold obtained in the assay of very poor ores, &c., is dried on the aluminium plate and taken up by a red-hot boric acid bead in a platinum-wire loop; when heated before the blowpipe the gold is obtained as an almost perfect sphere. The boric acid is dissolved off, and the minute sphere of gold measured under the microscope. A sphere 0.024 mm. diameter = 0·000002178 of a grain of gold has been examined. The modes of calculating are explained. Silver spheres may be obtained and measured in a similar manner; but for copper, lead, and other metals, the boric acid is replaced by sodium carbonate, as the former would dissolve a perceptible amount of these metals. D. A. L.

Estimation of Carbon in the Organic Constituents of Water. By A. HERZFELD (Ber., 19, 2618-2621).-The method is a modification of that proposed by Degener. The substance is boiled with chromic acid in a flask provided with a reflux condenser, the upper end of which is connected with a chloride of calcium tube. In the middle of the calcium chloride is a layer of powdered antimony, which absorbs any chlorine evolved from the substance. The carbonic anhydride is estimated in an improved form of Rose's apparatus. N. H. M.


Estimation of Nitrates in Water by Means of Aluminium. By S. HARVEY (Analyst, 11, 181-186).-The author has adopted the following plan, after considerable experience with the method:70 c.c. of the water is mixed in a bottle with 30 c.c. of 10 per cent. aqueous soda, a strip of aluminium foil, about 0.75 gram, is put in, the bottle is closed with a stopper carrying an open tube filled with glass beads, and the whole placed under a bell-jar until the reaction (which is aided by warmth) is complete. The contents and washings are made up to a definite volume, and, if possible, Nesslerised directly, otherwise they must be distilled, &c. In cases where free ammonia has to be expelled from the original water, the 70 c.c. is only made slightly alkaline at first, the 30 c.c. of soda being added subsequently. After keeping some time, a reduction in the quantity of ammonia is observed in waters which have not been previously boiled, this is presumedly due to a re-oxidation. Accurate results are obtained. Cane-sugar, or salts of iron, calcium, or magnesium, do not affect the result. Certain sources of error are indicated.

D. A. L.

Estimation of Glycerol in Wines. By M. BARTH (Chem. Centr., 1886, 504-505).-100 c.c. of wine is concentrated to about 10 c.c., 1 gram of powdered quartz and about 3-4 c.c. of 40 per cent. milk of lime added, and the whole evaporated to dryness. The glycerol is

University of A

then extracted with alcohol and finally with ether, the alcohol and ether evaporated on the water-bath, and the glycerol weighed.

L. T. T.

Assay of Carbolic Soap. By A. H. ALLEN (Analyst., 11, 103106). In the method recommended, the hydrocarbons are removed by agitating the soap, dissolved in soda and water, with ether, and the fatty acids are precipitated by means of brine. An aliquot part of the resulting solution is acidified with sulphuric acid and titrated with bromine-water until the solution is permanently tinged of a faint yellow colour; the bromine-water is standardised immediately before or after use by a solution of phenol or cresol. The remainder of the solution may be used for preparing a larger quantity of the brominederivative for qualitative purposes. D. A. L.

Estimation of Invert Sugar. By A. HERZFELD (Chem. Centr., 1836, 603).—In the method described by the author (Chem. Centr., 85, 604), the time taken in the preliminary heating of the liquid influences the result obtained. The tables given were constructed from experiments wherein this preliminary heating occupied four minutes.

The author recommends the following precautions:-An asbestos plate with a circular opening of 65 cm. diameter is placed on the wire gauze, and the flask placed on that, the Bunsen burner being placed at such a height that the flame plays over the whole of the exposed part of the flask, and heats the liquid to boiling in from 34 to 4 minutes. With these precautions, the results are very con


L. T. T.

Titration with Fehling's Solution. By E. BECKMANN (Zeit. anal. Chem., 25, 529-530).-Many persons find a difficulty in recognising the point at which the blue colour disappears. This seems to be due to an optical illusion which causes the colourless liquid to appear complementary in colour to the yellowish-red cuprous oxide. It may to some extent be obviated by adding a drop of zinc chloride, which promotes the separation of the cuprous oxide, but the safest way is to filter and test for copper, which is easily done by placing two thicknesses of filter-paper together, touching one side with the turbid fluid and testing the liquid which soaks through to the other. With diabetic urine, however, it is still necessary to rely on the disM. J. S. appearance of the blue colour.

Detection of Salicylic Acid. By C. O. CURTMAN (J. Pharm. [5], 14, 523-524).-To 4 c.c. of liquid (wine, beer, &c.) is added 2 c.c. of methyl alcohol, or failing this, the same amount of ethyl alcohol, and then, with care, 2 c.c. of pure sulphuric acid. After mixing, the liquid is heated for about two minutes, allowed to cool for eight or ten minutes, then heated just to boiling, when, if salicylic acid is present, a distinct odour of oil of wintergreen is perceptible. If only traces of acid be present, it may be necessary to allow the liquid to stand, and to heat a third time especially if methyl alcohol has been replaced by ethyl alcohol. Although other ethereal salts are

formed in contact with beer and wine, besides methyl salicylate, the odour of the latter is the most characteristic and the most easily perceived. In examining condensed milk, fats, or other solids and semi-solids, the sample is digested in dilute alcohol at 20-30° for some hours, with frequent agitation. After filtration and concentration, methyl alcohol and sulphuric acid are added as before described; 0.001 part of salicylic acid in a food, &c., can be thus detected.

J. T.

Specific Gravity and some other Characters of Waxes and Allied Substances. By A. H. ALLEN (Analyst, 11, 223-228).— Methods are described for obtaining fragments of wax, &c., free from air bubbles, for specific gravity determinations; also a method of taking gravities of waxes, &c. (heated by a suitable arrangement to 98-99°), by means of a Westphal balance and glass-rod plummet. The observed gravities indicate that waxes are denser than fats or glycerides in the solid state, but that the reverse is the case with the melted substances. Sperm and bottle-nose oil are shown to be different in composition from spermaceti. Evidence is adduced in support of the glyceride as opposed to the wax-like character of Japan wax (see also next Abstract). The relation of high gravity to high melting point in paraffin is noted. D. A. L.

Saponification of Fixed Oils. By A. H. ALLEN (Analyst, 11, 145-147). The author has collected, grouped, and tabulated results obtained by various investigators in the examination of a large variety of fatty substances by Kottstorfer's saponification method. Oils consisting of olein mixed with comparatively small quantities of stearin or palmitin, whether of animal or vegetable origin, neutralise about the same quantity of potash-from 18.93 to 19-66 per cent. Oils from cruciferous plants require 17.02 to 17.9 per cent. of potash to neutralise them. Vegetable drying oils require 187 to 19.6 per cent. These numbers are not characteristic, but show that linoleic acid must have a higher atomic weight than is generally supposed. With marine animal oils, also, the numbers obtained are not characteristic, varying from 18:51 in cod-liver oil to 21.88 in porpoise oil, which contains much valeric acid; marine waxes, however, require only from 12:30 to 14.74, much less than the marine oils. The butter class contains-butter-fat requiring 22:15 to 23-24, cocoa-nut oil 24-62 to 26.84, palm-nut oil 22:00 to 24.76 per cent. of potash. The various mixtures of palmitin, stearin, and olein require from 19 to 20 per cent. of potash for their neutralisation. Beeswax requires 92 to 97; Chinese wax 65; castor oil 17.6 to 18·15; Japan wax 21:01 to 22.25. D. A. L.

Adams' Method for Milk Analysis. By A. H. ALLEN and W. CHATTAWAY (Analyst, 11, 71–73); and W. THOMSON (ibid., 73— 75; compare also Abstr., 1886, 583).—Allen rolls up a piece of string with the paper to keep the folds of paper from touching one another; hence ensuring exposure of a more extensive surface. He also ties a cap of filter-paper over the bottom of the coil. With this modified coil, suspended by a loop in the string, 5 c.c. of milk may be run direct from a pipette on to the coil without fear of loss, and then the objec

tionable two weighings may be dispensed with. Thomson recommends distributing the 5 c.c. of milk on an extended strip of filter-paper, drying quickly, coiling and then extracting the fat in usual manner. D. A. L.

Separation of Morphine and Strychnine from Fatty Matters. By FOCKE (J. Pharm. [5], 13, 360-361).-According to the author, the following method gives good results. Exhaust the suspected matter with hot alcohol charged with tartaric acid. Filter after cooling, and evaporate on the water-bath. Take up the residue with 10 times its weight of water and add an excess of baryta-water. After some hours, add a slight excess of sulphuric acid, allow to remain for some time, filter and remove excess of acid by means of barium chloride. Filter and evaporate again so as to completely expel the hydrochloric acid of the barium salt. The residue is taken up with absolute alcohol, and the solution is evaporated to dryness on the water-bath. The new residue, which is slightly acid, is dissolved in water and extracted with ether, which removes the fatty matter taken up by the water. The aqueous solution, after being made alkaline, is again treated with ether, the ethereal solution evaporated, and the residue treated with water acidified with hydrochloric acid, which only dissolves the alkaloïds.

J. T.

Detection of Artificial Colouring in Red Wine (Claret). By SAMUELSON (Chem. Zeit., 10, 998).-When mixed with an aqueous solution of sodium nitrate, white wine remains clear, but the colour becomes darker. In genuine red wines, a precipitate forms and the supernatant liquid becomes yellow, sometimes only after some time; this is not the case with artificially coloured wines. In a mixture

of red and white wines, the amount of precipitate formed is inversely proportional to the quantity of white wine present. White wines coloured red with bilberry, mallow, red poppy, or orseille colouring matter do not give any precipitate. Red wines mixed with coloured white wines yield, in addition to the precipitate, the following reactions: with bilberry or mallow colours, a violet liquid; with orseille, a cherryred liquid; with red poppy, a bright red liquid. The addition of cider to white wine can be detected by sodium nitrate, as cider is coloured dark-brown by this reagent and after some time gives a slight precipiD. A. L.


Estimation of Tannin. By H. DIEUDONNÉ (Chem. Zeit., 10, 1067). -This simple and, when carefully conducted, accurate method for estimating tannin is based on reading very small differences of density with a very delicate hydrometer. The hydrometer indicates 1o Baumé, and is divided into hundredths; with this the solution, or extract of the substance to be examined, made with distilled water, is tested at 22°, both before and after treatment with powdered skin, the difference in the hydrometric readings is due to tannin. A table is given showing the quantity of tannin corresponding with solutions of densities varying from 15 to 105 of the centesimal Baumé degrees at 22°. The standard solutions used in constructing the table con

« PreviousContinue »