destruction of the hypochlorous acid and the formation of free chlorine and water. As the solution of sodium hypochlorite contains small amounts of chlorides, hydrochloric acid will always be present when the solution is acidified, thus insuring the destruction of the hypochlorous acid and the formation of free chlorine. We thus see that the effect of adding sodium hypochlorite to the solution is the complete destruction of the nitrous acid and the formation of free chlorine.

To remove the free chlorine in solution some compound must be added which will take up the chlorine, but will not affect subsequent operations. Such a compound is found in phenol. Under the conditions of the determination, phenol will add chlorine directly to the benzene ring, but is not affected by iodine or any of the other compounds in the solution. Chlorophenol not being ionised removes all traces of free chlorine.

This modification of the method greatly reduces the time and attention required for a determination, and, in addition, the copper solution is prepared in such a way that iodine can be liberated by copper alone.

In the determination, the copper, copper oxide, or sulphide is dissolved in nitric acid. After the addition of the sodium hypochlorite and phenol, which requires but a moment, the solution is made slightly alkaline with sodium hydroxide, and is then made acid with acetic acid, when the solution is ready for titration. Potassium iodide and starch are added, and the titration is made to the disappearance of the starch iodide colour. There is never any fear of the blue colour "flashing back," and the solutions will remain colourless indefinitely after the titration. As the ionisation constant for acetic acid is too low to allow nitrates to liberate iodine, the amount of nitric acid in solution is immaterial. Even 20 cc. of concentrated nitric acid will not affect the titration. However, too great an acidity is to be avoided, as nitrophenol will be formed. The presence of nitrophenol prevents the determination of copper, but there is no danger of its formation even in the presence of a large amount of acid if the solution is neutralised soon after the addition of the phenol. If a large amount of nitric acid is used to dissolve the copper, it should therefore be partly neutralised before addition of the phenol.

As chlorine easily oxidises phenol to compounds which prevent the determination of copper, it is essential that all of the phenol be added quickly to the solution. Under these conditions the chlorine adds directly to the benzene ring, but if the phenol is added drop by drop the chlorine will oxidise it, producing coloured compounds in solution.

In order to add the phenol quickly enough to the solution it may be poured in from a beaker, or, a more convenient way, from a pipette from which the tip has been removed so that the delivery is from an opening which is of the same bore as the rest of the tube. By forcing the phenol out of such a pipette with the breath the entire volume is added very quickly, and at the same time the phenol is well mixed with the contents of the flask.

After addition of the phenol the chlorine gas which is in the flask above the liquid is removed by blowing it out with the breath, and the sides of the flask are washed with a jet of water from a wash bottle. There should be no odour of chlorine just before the solution is made alkaline.

It should be remembered that the end-point of the titration is not pure white. Cuprous iodide has a cream colour, and when a large amount of copper is present the cuprous iodide gives a decided tint to the solution. When the end-point is nearly reached a drop of the thiosulphate is allowed to fall into the centre of the flask. If a change of colour occurs the solution is given a slight rotary motion, and after the solution is again quiet another drop of the thiosulphate is added. This "spot test" is easily recognised, and gives very accurate end-point.

The speed of reaction of the copper with potassium iodide varies with the volume. In a small volume the action is rapid, and all of the iodine is liberated at once, but in a large volume an appreciable time may be required

18,

for all of the copper to react. This is especially noticeable when a small amount of copper is present. A high concentration of potassium iodide greatly assists the liberation of the iodine. Accurate results can not be obtained unless at least 3 grms. of potassium iodide are added, irrespective of the amount of copper present, up to 500 mg. of copper.

The solutions required are:

A. The Sodium Hypochlorite solution is made by boiling together a mixture of 112 grms. of calcium hypochlorite and 100 grms. of anhydrous sodium carbonate in 1200 cc. of water. After the calcium is precipitated as carbonate, the solution is filtered and its strength found as follows: 5 cc. of the hypochlorite solution are added to 100 cc. of water containing 5 cc. of 30 per cent potassium iodide solution, and a few cc. of dilute hydrochloric acid are added. The liberated iodine is titrated with o'r N sodium thiosulphate. The volume of the solution is now adjusted so that 5 cc. of the hypochlorite solution are equivalent to 30 cc. of o'r N sodium thiosulphate.

B. Phenol.-A 5 per cent colourless solution of phenol.
C. Sodium Hydroxide.-A 20 per cent solution.
D. Acetic Acid, 50 per cent.

E. Potassium Iodide.-A convenient way to use this is to prepare a solution which contains 30 grms. per 100 cc. of solution. Then 10 cc. will contain 3 grms., which is the amount needed for a determination.

F. Sodium Thiosulphate.-For the accurate titration of the liberated iodine two solutions are used. One strong solution, I cc. of which equals 6 mg. of copper, and a weak solution, I cc. of which equals 1 mg. of copper. The strong solution is run in until the iodine liberated by the copper gives a light straw colour to the solution. Starch is then added, and the titration is finished with the weak solution.

(The weights given here are for calcium hypochlorite, having 35 per cent or more available chlorine).

As a thiosulphate solution loses strength it should be restandardised from time to time. A convenient way to do this is as follows: -A solution of sodium thiosulphate, approximately o'I N, is made by dissolving 24 grms. of the crystallised salt per litre of water. After the solution has stood at least twenty-four hours it is standardised against copper by the method described below. Pure electrolytic copper which has been cleaned with emery paper should be used. After dissolving 150 to 200 mg. of the copper in 6 to 8 cc. of 50 per cent nitric acid the solution is treated as described below, and the thiosulphate is then standardised with this known weight of copper. The most convenient means of re-standardising the thiosulphate is to use a solution of acid potassium iodate. Acid potassium iodate has the formula KIO,HIO3, so that a normal solution has onetwelfth the molecular weight in grms. per litre. A o'r N solution is prepared by dissolving 3 249 grms. of the salt in I litre of water, and it is standardised against a known strength of thiosulphate as follows:-Add ro cc. of the acid iodate solution to 150 cc. of water containing 0.5 to 1 cc. of hydrochloric acid. Upon the addition of potassium iodide, iodine will be liberated according to the equation HIO3+5HI=+3H2O+61. Starch is added and the titration is made to a colourless solution. From this titration the weight of copper, to which 20 cc. of this solution are equivalent, is accurately determined. A 20 cc. pipette is passed through a one-hole stopper, and is allowed to remain in the acid iodate bottle. The end of the pipette is closed with a small rubber stopper. The exact copper equivalent of a thiosulphate solution is now easily found by titrating 20 cc. of the acid iodate solution whose copper equivalent is known with the thiosulphate as described above. The acid iodate remains constant indefinitely.

G. Starch for Indicator.-The best preparation for this purpose is a o'5 per cent solution of Kahlbaum's soluble starch. This is prepared as ordinary starch, but gives a perfectly clear solution which is very sensitive with iodine. If ordinary starch must be used it should be free from all cloudiness.

CHEMICAL NEWS,

April 19, 1912

Copper.

Burn off the sulphides of gold and copper in a porcelain crucible, avoiding a heat greater than that required to burn off the filter paper. This hinders the shrinkage of the metallic gold, and leaves it more porous, so that the CuO is readily dissolved out. Too high a heat causes CuO to unite with the glaze of the crucible. Place the Au and CuO mixture in a 4-ounce beaker. Add 10 cc. of concentrated HNO3. Boil for ten minutes. Add 3 cc. of concentrated H2SO4, and boil until the HNO3 is all ex: pelled and SO, fumes are coming off freely. Cool, add 50 cc. of water and 5 grms. sodium acetate. Boil and filter off the gold (the gold residues are weighed as an approximate check, and are saved for their commercial value only; gold is best determined by fire assay). Cool the filtrate and add 5 grms. of potassium iodide; stir until dissolved. Titrate with decinormal hyposulphite of sodium: I CC.=0.0063 grm. Cu.

Gold.

Scorify o'5 grm. of filings with 40 grms. of test lead, I grm. of borax glass, and I grm. of powdered silica. The borax and silica flux are placed on the mixture of alloy filings and test lead. If the lead button resulting from scorification is hard repeat the scorification, adding 10 grms. more test lead and 2 grms. of borax and silica flux. Cupel carefully and weigh Au plus Ag.

The determinations previously made make it possible to calculate quite closely the amounts of Au and Ag, &c., in the alloy.

We make up a mixture containing the same metals in approximately the same proportions; o'5 grm. of this "control" mixture is then scorified and cupelled side by side with an equal weight of the alloy. After weighing the gold and silver buttons from the assay and the "control," they are alloyed separately with 3 parts of pure Ag.

The alloys must be thoroughly melted in order to insure homogeneity. Flatten the buttons by hammering or rolling.

Part with HNO3 of graded strengths as is the usual assaying practice. Boil well with water, dry, ignite, and weigh Au. Deduct Au from Au plus Ag; the difference is Ag. Increase both Au and Ag figures by the amounts of loss shown by the Au and Ag used in the "control" assay. The fire assay for Au and Ag yields in most chemists' hands the highest and the most accurate results. The correction for fire loss (volatilisation) is far higher for Ag than for Au. Gold alloys analysed by us had the following limits of composition :

Those alloys highest in Ag were most difficult to dissolve in aqua regia.

We found the ratio 1 HNO3 to 4 HCl most efficient. Aqua regia 1 to 3 was less efficient, and 1 to 2 very unsatisfactory. Increasing the percentage of HNO3 in the mixture failed to cause more rapid solution, but increase of HCl produced unexpectedly rapid and complete decomposition. Journal of Industrial and Engineering Chemistry, iii., No. II.

PROCEEDINGS OF SOCIETIES.

CHEMICAL SOCIETY. Ordinary Meeting, March 21st, 1912.

187

Prof. PERCY F. FRANKLAND, LL.D., F.R.S., President, in the Chair.

MR. W. J. S. Naunton was formally admitted a Fellow of the Society.

Certificates were read for the first time in favour of Messrs. Arthur Anderson Bones, 640, Schoeman Street, Pretoria, S. Africa; Raymond Edwin Crowther, Edenvale, Wigton Road, Carlisle; William Dallas, Burnbank Cottage, Mount Vernon, Glasgow; Archibald Knox, 18, Newhall Terrace, Greenhead, Glasgow; Edwin Charles Lacey, B.Sc., St. Julian's Lodge, West Norwood, S.E.; Leslie Herbert Lampitt, M.Sc., Bowyer Road, Saltley, Birmingham; Robert John Milbourne, Muxton Lodge, Newport, Salop; George Walker, Stonehurst, Lancaster Road, Morecambe; Charles Reginald Wilkins, B.Sc., 40, Church Lane, Hornsey, N.

A certificate has been authorised by the Council for presentation to ballot under By-law I. (3) in favour of Mr. John Scott Thomson, Crawford Street, Dunedin, N.Z.

Of the following papers those marked * were read :— *56. "Syntheses of 3-Oxy-(1)-thionaphthen." BY ARCHIBALD MORITZ HUTCHISON and SAMUEL SMILES.

It was shown that in the presence of concentrated sulphuric acid and chlorosulphonic acid or other similar reagents, o-thiol- and o-dithio-benzoic acids with malonic acid or ethylacetoacetate yield 3-oxy-(1)-thionaphthen or derivatives thereof. that this reaction is caused by the condensation of the sulReasons were given for considering phoxylic acid (Trans., 1911, xlix., 640) with these methylene -CO2H+CH2R2=2H2O+C6H4<C>C:R2.

derivatives :

C6H4 S.OH

It was also pointed out that other examples of this type of interaction are to be found in technical processes for preparing "thioindigo."

*57. "Behaviour of Metallic Alloys when heated in a Vacuum." By CLARENCE RICHARD GROVES and THOMAS TURNER.

The authors have heated a number of binary alloys in a porcelain boat under a pressure not exceeding 1 mm. of mercury. The boat was contained in a porcelain tube, and heated in an electric resistance furnace to temperatures ranging from 500° to 1200°.

It is concluded that the alloys may be divided into five groups as follows:

in a vacuum for a moderate time at or below 1200°. 1. The metals are not appreciably volatile when heated

of copper with iron, aluminium, tin, or nickel. It may As examples of this class may be mentioned the alloys also be assumed that any mixtures of these four metals in any proportions would also be non-volatile.

2. The constituent metals are quantitatively separable. Thus the alloys of the copper-bismuth, copper-lead, and copper-zinc series all separate at the melting-point of copper. The bismuth, lead, or zinc volatilise, whilst pure copper remains. Iron-zinc alloys can be quantitatively separated at 500°.

3. Any excess of the more volatile metal is removed, and a chemical compound remains.

The gold-zinc series yields AuZn; the copper-antimony, Cu3Sb; the gold-cadmium, AuCd; and the magnesiumzinc, MgZn2.

4. The excess of the more volatile metal is removed, but the residue is not a chemical compound.

Thus in the copper-arsenic series the arsenic retained diminishes as the temperature rises. At 1200° the com

position remains constant with about 20 per cent of arsenic, which does not correspond with any simple atomic proportions.

5. Two (or more) metals may volatilise together. Thus ead and zinc tend to pass over together. In the iron-zinc series also there is an increasing proportion of the iron carried over as the temperature rises from 500°. In the silver-zinc series, although separation is nearly quantitative at 700°, there is an increased loss of silver with higher temperatures.

DISCUSSION.

Dr. HODGKINSON inquired if the authors had observed any gas evolution during the process, although he doubted whether that would have been possible owing to the character of the vacuum which had to be maintained against a sealing-wax closure.

His reason for asking was that recently he had melted some very pure electrolytic copper in a silica tube in a Sprengel vacuum, and the evolution of hydrogen had not ceased when the tube cracked after more than twelve hours' heating. The globule of copper was found to be full of large bubbles when cut open after cooling.

The total separation of zinc from copper in brass at 1100° in a vacuum appeared very extraordinary, as some eight or nine years ago he, with Major Howarth, R.A., had made brass of the composition Cu2Zn by passing zinc vapour, with the aid of a stream of hydrogen, over copper

heated to a little over 1000°.

Other metals, such as platinum, nickel, and palladium, also retained definite amounts of zinc in similar circumstances.

Dr. J. F. SPENCER pointed out that Prof. Turner's experiments indicated that alloys could be divided into pairs, depending on their behaviour on distillation, which were analogous to the various pairs of completely miscible liquids; thus, for example, with some of the alloys a complete separation was possible, whilst with others one metal distilled away, leaving apparently a definite compound. This supposed compound corresponded exactly with the constant-boiling liquid mixture of minimum vapour presDr. Spencer asked the author if he had made any experiments to see if the composition of the supposed compound was changed in any way, for example, by change in the pressure under which the distillation was carried out.

sure.

Mr. EGERTON drew attention to the construction of the

electric resistance furnaces used by Prof. Guntz in his work on the preparation and distillation of metallic barium, and asked Prof. Turner whether he had had trouble with the platinum at high temperatures. Another point which had interested him was the mention of the vapour pressure of metals at the ordinary temperature; it was questionable whether a crystalline metal had any vapour pressure below a certain temperature.

No doubt if some of the alloys examined had been subjected to a higher temperature, different and very interesting results might have been obtained. He hoped someone would carry out such experiments. The separation of metals in a vacuum was usually effected at the melting-point of the less fusible metal, but in some cases, as in the zinc-iron series, the alloy was solid throughout the experiment. They had tried various substances for making the joints at the cold ends of the porcelain tube, and had found ordinary red sealing-wax gave the most satisfactory results.

*58. “Oxidation of Atmospheric Nitrogen in Presence of Ozone." By THOMAS MARTIN LOWRY.

When air under the ordinary pressure is driven over a series of spark-gaps a product is obtained which gives an increased yield of nitrogen peroxide if mixed with ozone. It is suggested that this product may contain a chemically active form of nitrogen. DISCUSSION.

Dr. FEILMANN remarked that it would be comparatively easy to test Dr. Lowry's hypothesis on the formation of active nitrogen by passing pure nitrogen, free from oxygen, over a spark-gap, and then mixing it with ozonised air. Dr. ROBERTSON stated that in extending the spectroscopic method for the quantitative estimation of nitrogen peroxide to mixtures containing that gas in very low concentrations, he had found it advisable to make the parts of the apparatus of non-reactive substances; thus for strict quantitative work, observation tubes should be made of silica rather than of glass.

*59. "Method of Producing a Steady Thallium Flame.” By THOMAS MARTIN LOWRY.

Thallium chloride is vaporised by heating it gently in a silica bulb, and is carried into the flame by a current of zontal tube, and passes into the flame through a second The oxygen is admitted to the bulb by a horitube, which terminates in a vertical jet.

oxygen.

*60.

"Electro reduction of Alkylnitrosoamides." By HILMAR JOHANNES BACKER.

In continuation of previous researches on the electroLeyden, 1911), the author has studied the reduction of reduction of nitroamino-compounds to hydrazines (Diss., alkylnitrosoamides of common organic acids, which has been unsuccessfully attempted by various investigators.

It was shown that oxalylbismethylnitrosoamide (m. p. 66°, decomp.) and succinylbismethylnitrosoamide (m. p. 110, decomp.), when reduced in slightly acid solution with a tinned cathode, yield the corresponding hydrazides, which may be isolated as benzylidene derivatives.

The above-mentioned nitrosoamides, especially that of oxalic acid, are unstable. Their decomposition by alkalis and organic bases was described; in each case diazoBy

methane is formed.

*61. "Model of an Asymmetric Carbon Atom." WILLIAM EDWARD GARNER.

Prof. TURNER said, in reply to Dr. Hodgkinson, that their experiments showed that the quantity of gas evolved with the relatively small quantities of alloys employed was usually small. When any considerable quantity of gas was evolved, they found that some was occluded in the tube, but practically the whole of this was evolved if the tube were cooled and re-heated. With a second cooling and re-heating, the residue of gas was very small indeed. Porcelain tubes were less permeable to gases at high temperatures than fused silica tubes. It was possible to make brass by heating bright copper in an exhausted glass tube containing zinc at a temperature 50° below the meltingpoint of zinc, or, say, 375°. The platinum in the resistance furnace was in the form of a spiral of thin sheet. It lasted well so long as the temperature did not exceed 1200°, but deteriorated rather rapidly with greater heat.

The asymmetric carbon atom is represented by a cube, which is composed of four wooden blocks (" × 1′′ × 1"). The blocks are placed vertically, and bolted in a horizontal direction, leaving a space of " between the blocks (Fig. 1). The bolts are placed as near the centre of the model as possible. Fig. 3 gives a cross-section, showing the manner in which the bolts are employed. Four iron rods are attached to the bolts in such a way that they can move freely in a vertical direction. Each rod is connected to the two adjacent rods by thin cord, which passes through rings at the top and the bottom of the cube. The method by which this is accomplished may be seen by reference to the diagram (Fig 2). This represents the top of the cube, but the arrangement of the cords which pass through the lower ring is similar to this if the model be viewed from underneath.