98 THE ANALYBT. INORGANIC ANALYSIS. Reactions distinguishing between Cobalt and Nickel. G. Gubrin. (Jown. Yharm. Chim., 1904, xix., 139.)-If a solution of a cobalt salt be treated with an excess of potassium hydroxide, and a 2 per cent. solution of iodine in potassium iodide be added until the supernatant liquid becomes yellow, the whole of the cobalt will be deposited in a few seconds in the form of the black hydrated sesquihydroxide. Nickel salts under the same conditions yield the light-green hydroxide. The same hydroxides are formed on adding potassium hydroxide in excess and then the iodine solution to the precipitates produced in solutions of cobalt or nickel salts by ferro- cyanides, carbonates, and phosphates of the alkali metals. On the other hand, when the ferricyanides are treated with potassium hydroxide both become black, yielding the respective hydrated sesquihydroxides of nickel or cobalt.The reaction is almost instantaneous in the case of nickel ferricyanide, but slower with cobalt ferricyanide. C. A. M. A New Reaction of Molybdenum. E. Lecocq. (Bzdl. de I'Ass. bdge Chim., 1903, xvii., 412-414.)-Diphenylcatrbazide, recommended by Caxeneuve (ANALYST, xxv., 331) as a very sensitive reagent for mercury, copper, iron (in the ferric state), and chromic acid, can also be employed for the identification of molybdenum. The solution of an alkali molybdate in 30 to 40 C.C. of water is acidified with 2 or 3 drops of strong hydrochloric acid, and treated with an alcoholic solution of diphenylcarba- zide prepared several days previously. On the addition of the first drop a brilliant violet-indigo coloration is produced, which increases in intensity on the addition of more of bhe reagent, and eventually a precipitate of the same colour is formed.This. reaction, which does not take place in the presence of strong acids or alkalies, is capable of detecting 0.007 gramme of ammonium molybdate in a litre of water. In the author's opinion a compound is first formed between the alcohol and the diphenyl- csrbazide, and this gives the coloration with the molybdenum. This view receives support from the facts that no reaction is obtained with a benzene solution of diphenylcarbazide, and that the colorations can be varied by replacing ordinary alcohol by other alcohols of the fatty series. Thus, with methyl alcohol the colour is bluish indigo ; with propyl alcohol, violet ; with butyl alcohol (liquids immiscible), violet red ; with isobutyl alcohol, violet red ; and with amyl alcohol the upper layer is rose and the lower colourless.None of these colorations is sufficiently marked to serve a8 a means of distinguishing between the different alcohols, except in the case of amyl alcohol, which can bo readily identified by this test. C. A. M.THE ANALYST. 99 Quantitatdve Separation of Metal$ by Heans of Pemulphatee. G. v. Knorre. (Zeit. and. Chern., 1904, xliii., 1-l4.)-The method of determining man- ganese by precipitation with ammonium persulphate (ANALYST, xxvii., 68) has now been tested by the author, to determine to what extent it can be used in the separa- tion of different metals.Experiments described in detail show that it is not possible to effect in this way a sharp separation of manganese from copper, zinc, cadmium, niokel, or magnesium. In the case of chromium, however, good results can be obtained. Thus, if a solution containing ohrome alum and manganese sulphate be treated with an excess of ammonium persulphate and heated to boiling, the chromium salt is completely oxidized to chromia acid, whilst the manganese is quantitatively precipitated as peroxide. In a test experiment 50 C.C. of a solution of chrome alum containing 0.2772 gramme of chromic oxide was mixed with 100 C.C. of a solution containing 3.102 grammes of pure manganese ammonium sulphate, and the liquid boiled for six minutes with 100 C.C.of a solution containing about 6 grammes of aamonium persulphate. The amount of the oxide (Mn,O,) left on ignition of the precipitate was 0.6058 gramme, as against the theoretical 0.6055 gramme ; whilst the chromium oxide (Cr,O,) obtained from the filtrate weighed 0-2774 gramme, instead of the theoretical amount, 0.2772 gramme. In the author's opinion this method is more rapid and accurate than any other known method of separating manganese and chromium. C. A. M. The Determination of Bromides in the Presence of Chlorides. Imbert and Dumolmd. (BaEZ.. Phum. du. Sud-Est ; through Ann. de Chim. anal., 1904, ix., 21,22.)-Part of tha solution containing the mixed salts is treated with silver nitrate, and the weight of the precipitated silver chloride and bromide determined.Another (equal) portion of the original solution is then boiled for thirty minutes with ammonium persulphate, which decomposes the bromide, the larger part of the bromine being set free, and the remainder converted into bromate. The solution is again treated with silver nitrate, and the amount of precipitated silver chloride deducted from the weight of the former to obtain the amount of silver bromide. The authors have proved that the ammonium persulphate does not liberate any chlorine or convert any trace of chloride into chlorate. On the other hand, a solu- tion of ammonium persulphate, after being boiled by itself for thirty minutes, gives a slight precipitate with silver nitrate. To obviate any error from this source 8 little nitric aoid should be added after the boiling to destroy undecomposed persulphate.C. A. M. Fractional Combustion of Gas Mixtures - contqining Hydrogen by Means of Heated Palladium Wire. F. Richardt. (Zeds. A w g . Chm., xxxviii., 65.)- The author shows that methane is not oxidized in its mixture with air when led over heated palladium wire if the temperature is kept below 450° C., and that if the time of oontact is short practically no combastion bakes place even at 600" to 650" C. Ethane behaves in a similar way, but burns a little more readily. Ethylene is burnt to some extent at as low a temperature as 300' C. It follows from them observations that in mixture8 of methane or ethane, but not of ethylene, with hydrogen the latter100 THE ANALYST. may be burnt completely without attacking the hydrocarbon, by leading a slow current of the gas, mixed with air, ov0r palladium wire kept at a temperature approaching 450" C.A. G. L. Fractional Combustion of a Mixture of Methane and Hydrogen by Winkler's Method. I(. V. Charitohkof. (Joum. SOC. Phys. Chim. R., xxxiv., 710 ; through Chem. News, lxxxviii., 282.)-The author finds that the best way of removing hydrogen from mixtures containing methane is by means of palladium sponge at a temperature of 70" C., not 120" C. Even at this low temperature, and when the quantity of hydrogen present is small, some of the methane is burnt, and hence the carbon dioxide formed should be determined. To mixtures containing much hydrogen the method is inapplicable. A. G. L. A New Volumetric Method of determining Nitric Nitrogen.L. Debour- deaux. (BUZZ. SOC. Chim., 1904, xxxi., 3-6.)-This method is based on the oxidation of oxalic acid by nitric acid in the presence of sulphuric acid and manganese sulphate, with the formation of nitrous oxide and carbon dioxide. When the solution contains more than 20 per cent. by volume of sulphuric acid the reaction takes place in accordance with the equation whilst with less than that proportion of sulphuric acid, nitrous oxide is produced- The most favourable conditions for the reaction are when the solution contains from 4 to 6 grammes of manganese sulphate and 11 to 14 C.C. of sulphuric acid per 100 C.C. The temperature should be gradually raised to about 94' C. In making a determina- tion, about 0.5 gramme of thenitrateis mixedin a flask with 50 C.C.of a solution con- taining 35 to 40 grammes of crystalline oxalic acid, 50 grammes of manganese sulphate, and 120 C.C. of sulphuric acid, specific gravity 1.84 ( 6 6 O Be.) per litre. The flask is placed in a cold-water bath, and slowly heated under a reflux condenser. When the water in the bath boils (which should take at least thirty minutes) and all evolution of gas has ceased, the reaction is complete. The residual oxalic acid in the solution is then titrated with standard potassium permanganate solution (about 20 grammes per litre), and the difference between this result and that given by a preliminary titration of 50 C.G. o€ the oxalic reagent shows the amount of oxalic acid destroyed by ihe nitric acid present. The method is shown to yield closely concordant results under the conditions described above.An addition of from 1 to 5 grammes of potassium sulphate per 100 C.C. is without noticeable effect, whilst potassium bisulphate does not act in the same way as free sulphuric acid. Ammonium salts in the proportions in which they occur in artificial manures have also no perceptible influence on the reaction, which is also the case with small amounts of chlorides, such as occur in commercial nitrates. Large quantities of chlorides, however, lead to some destruction of oxalic acid, and ought to be removed by precipitation with silver. 3H2C20,.2H20 + 2KN0, + H,SO, = 10H,O + &SO, + 6C0, + 2 N 0 , 4H2C204.2H20 +3KNO, + H,SO, = 13H20 + K,SO, + 8C0, + N,O.THE ANALYST. PO1 Vanadium sulphate can replace manganese sulphate in this reaction, the con- ditions then being as follows: Amount of sulphuric acid, 2 to 6 C.C.; and amount of vanadium sulphate, 0.15 to 0525 gramme per 100 c.c, I n this case also the tern- perature must be raised very gradually. C. A. M. The Determination of Nitrites in Absence of Air. I. IC. Phelps. (zeits dnorg. Chem., xxxviii., 113.)-The method described is a modification of that of Dunstan and Dymond (Phrm. Journ., xix., 741), and consists in allowing potassium iodide to act cm the acid solution of the nitrite, afterwards determining the iodine liberated. It is carried but in a 250 C.C. flask provided with a rubber stopper, which carries a tap-funnel rand a wide exit-tube; this last is expanded into a bulb just above the stopper, and then bent twice at right angles ; the bulb serves to prevent mechanical loss during boiling, For the determination, an excess of standard arsenious acid solution and 25 c,c.of a concentrated sodium carbonate solution are introduced into the flask; the stem of the tap-funnel is completely filled with water, and the liquid in the flask is boiled for five to eight minutes in order to expel all air. The end of the exit-tube is then plunged into a deep column of mercury, the boiling in- terrupted, and the flask cooled by immersing it in water or ice. Through the tap- funnel 7 C.C. of sulphuric acid (1 : 4) are allowed to flow into the flask; the flask is raised so that the exit-tube no longer dips into the mercury, but into a layer of water placed above it ; the funnel is carefully washed free from acid, and the solution of the nitrite to be examined addtea. 'Enough sulphuric acid (about 5 C.C. of 1 : 4) is then added to make the contents of the flask acid, followed by an excess of potassium bicarbonate. The liquid is then again boiled for five minutes to completely expel all nitric oxide, aIlowed to cool, and titrated with standard iodine solution. The method was checked against that of Kinnicut and Nef (Awzer. Chem. Jouru., v., 388). It gives satisfactory results, slightly above the theoretical. A. G. L.