82 ABSTRACTS OF CHEMICAL PAPERS INORGANIC ANALYSIS. Volumetric Estimation of Cadmium and Zinc. H. Enell. (Zeitsch. and. Cltem., 1915, 54, 537-546.)-A method which yields approximately correct results depends on the reaction which takes place when cadmium sulphide is shaken for some time with an exceRs of iodine ; CdS + I,= CdI, + S. The excess of iodine is then titrated and the quantity of cadmium calculated from the amount of iodine consumed.The method is unreliable. A better method is based on the reaction between cadmium sulphide and silver nitrate : CdS + 2AgN0, = Ag,S + Cd(NO,),. The cadmium sulphide, obtained by precipitation in the usual way, is washed with a warm 2 per cent. ammonium sulphate solution (to each 250 C.C. of which are added 20 drops of 10 per cent. sulphuric acid), until the wash.water no longer gives a reaction with silver nitrate. The filter and precipitate are then transferred to a stoppered cylinder, and shaken with a mixture of 50 C.C. of water end 20 c.c., or more, of & silver nitrate solution. Five C.C. of 25 per cent. nitric acid are now added, the mixture is again shaken for one minute, filtered, and the excess of silver is titrated in an aliquot portion of the filtrate.A similar method is recommended for the estimation of zinc. The zinc sulphide is obtained in a condition in which it is readily filtered and washed by rendering the zinc salt solution slightly ammoniacal, adding 2 C.C. of acetic acid, treating the solution with hydrogen sulphide, then heating the mixture on a water-bath for fifteen minutes, and adding 1 grm.of ammonium sulphate. The precipitate is washed with a warm 2 per cent. ammonium nitrate solution which has previously been rendered slightly ammoniaoal and then acidified with acetic acid. The results yielded by the method are trustworthy, from 99-5 to 100-1 per cent. of the zinc present being. obtained. w. P. s. Volumetric Estimation of Cerium by Means of Potassium Perman- ganate.V. Lenher and C. C. Meloche. (J. Amer. Chem. SOC., 1916, 38, 66-70.). -When cerium is titrated with permanganate, zinc oxide or magnesium oxide are the best neutralising agents. Fairly good results may also be obtained by using borax or sodium bicarbonate, whilst sodium carbonate is distinctly less satisfactory. The following reagents were also tried, but none were found suitable for the .purpose : Sodium hydroxide, barium hydroxide, sodium acetate, sodium silicate, trisodium phosphate, disodium phosphate, sodium tungstate, and sodium arsenate.Under proper conditions the method serves for the accurate estimation of total cerium or of trivalent cerium in presence of tetravalent cerium. G. C. J. Rapid Estimation of Iron in Presence of Organic Substances. F.Ferrari. (Anwli Chim. AppZk., 1915, 4, 341-343.) --Iron is quantitatively precipitated by ‘‘ cupferron ” (the ammonium salt of nitrosophenylhydroxylamine) in the presence of tartrates, citrates, etc. The solution of the iron salt containing an excess of hydro- chloric acid is oxidised with chlorine water, and treated, drop by drop with continual stirring, with a 6 per cent.solution of cupferron, until there is an excess of at least one-fifth of the quantity required to precipitate the iron. After standing for fifteen to twenty minutes the precipitate is separated, washed twice with cold waterINORGANIC ANALYSIS 83 I i I ~ Highest Lowest Rain - { - Highest Snow (35 samP1es) Mean .--------- { Lowest P5 Mean acidulated with hydrochloric acid, then successively with water alone, ammonia solution, and water, and ignited, and the residue of ferric oxide weighed.The method is not applicable in the presence of copper, silver, lead, mercury, tin, bismuth, titanium, or zirconium, all of which form insoluble compounds with cupferron. C. A. M. xitrogen as i Nitrogen as Sulphate. Chlorine.Free 1 Albuminoid Amnionis. Ammonia. ~ - - ----___ 38.0 0.80 18.0 0.25 0 -23 34.0 0-4 0.47 ! 0.59 17.0 1.7 ---- 5.0 0.02 0-13 i 0.12 I Estimation of Nickel with Dimethylglyoxime. K. Wagemann. (Ferrzcnt, 1915, 12, 126-129 ; through Zeitsch. angezo. Chem., 1915, 28, 590.) -The following procedure is recommended i n order to render this method mme repid. The metal is precipitated with an aqueous dimsthylqlyoxime solution, the precipitate is collected, dissolved in dilute sulphuric acid, and the solution boiled after the addition of hydrochloric acid.Tbe latter treatment serves to destroy the dimethylglyoxime. The solution is now rendered ammoniacal, and the nickel is deposited electrolytically, and weighed. w. P. s. Nitrogen, Chlorine, and Sulphates in Rain and Snow.B. Artis. (Chem. News, 1916,113, &)-The work is a, continuation of that by Knox (Chem. News, 1914, 111,61), and covers the period from October, 1914, to June, 1915. Fifty-nine samples were investigated, of which 35 were of rain and 15 of snow, or 15.35 inches of rain and 37.75 inches of snow, equivalent in all to 18.49 inches of rain. The author con- siders 12 inches of snow as equivalent to 1 of rain, and 101 tons 54 cwt.as the weight of 1 inch of rain upon an acre. The chlorine was the most constant con- stituent in amount, showing 7.1 parts per million in 30 out of the 59 samples examined. The following values were obtained as parts per million : Nitrate. 21.3 3.5 8.7 14.0 7.0 83.0 - --- H. F. E. H. (Cf. AKALYST, 1914, 39, 190.) Perchlorate and Gravimetric Cobalti -nitrite Methods for Determination sf Potash.T. D. Jarrell. (J. Assoc. Ofic. Agric. Chemists, 1915, 1, 29-32.)- Experiments are described which support the opinion that the gravimetric cobalti- nitrite method must be greatly improved before it can be successfully used. In the experiments with the perchlorate method, the precipitate WBS washed first with a 0-5 per cent. solution of perchloric acid in 95 per cent.alcohol, and finally with 10 C.C. of84 ABSTXACTS OF CHEMICAL PAPERS absolute alcohol. All the results were low, and this is attributed to the fact that the precipitate was partially redissolved whilst being washed, t is stated that if it were possible to find a 6‘ wash ” in which potassium perchlorate is insoluble, results might perhaps be obtained which would compwe favourably with those obtained by the platinum method.Note by Abstractor.-Such a ‘‘ wash ” is a saturated solution of potassium perchlorate in 95 per cent. alcohol (Davis, ANALYST, 1913,38, 47 ; Gumming and Thin, ANALYST, 1915, 40, 296). G. C. J. Qualitative Detection and Separation of Platinum, Arsenic, Gold, Selenium, Tellurium and Molybdenum.P. E. Browning. (Clzem. News, 1915, 112, 825.)-In the method of Noyes and Bray (J. Amw. Chem. Soc., 1907, 29, 137) compounds of gold, tellurium and selenium may, under certain conditions, be precipitated together with the arsenic, on adding magnesium chloride in ammo- niacal solution immediately after the separation of the platinum. It is there- fore advisable to modify the method as follows: The platinum is precipitated as platinichloride by evaporating the solution with a potassium salt.The filtrate is heated with oxalic acid or with hydrogen peroxide to predpitate the gold, and the filtrate from this evaporated nearly to dryness and treated with strong hydrochloric acid and sodium sulphite to remove the selenium. The filtrate is diluted and the tellurium precipitated by the addition of potassium iodide and sodium sulphite.On now evaporating the filtrate from the tellurium with bromine or nitric acid the arsenic and molybdenum are oxidised, and the arsenic may be precipitated by magnesian mixture. The filtrate from the arsenic is treated with zinc and potassium thioeyanate, which forms the red molybdenum thiocyanate.C. A. M. Titration of Thiosulphates in the Presence of Sulphides and the Estima- tion of Thiosulphates in the Presence of Sulphites, Bisulphites, and Sul- phides. A. Sander. (Chem. Zeit., 1915, 39, 945-947.)--Ti be different behaviour of alkali thiosulphates and sulphides towards mercuric chloride affords a means for their estimation when they are present in the same solution.If mercuric chloride is added to sodium sulphide solution, a black precipitate at first forms, which, on the addition of more mercuric chloride, changes to white mercuric sulphochloride : 3HgC1, + 2Na,S = 4NaCl+ (2HgS.HgC1,). On the other hand, sodium thiosulphate reacts with mercuric chloride according to the equation : 2Na,S,O, + 3HgC1, + 2H,02 = Na,SO, + 4HCl+ (2HgS.HgCIJ.For the estimation, an aliquot portion of the solution containing the two salts is titrated with & iodine solution; another portion of the solutian is then treated with an excess of mercuric chloride, ammonium chloride is added, and the mixture is titrated with & sodium hydroxide solution, using methyl orange as indicator. The number of C.C. of the alkali solution used divided by 2 gives the quantity of & iodine solution required for the thiosulphste, and the difference between the two titrationsINORGANIC ANALYSIS 85 corresponds with the iodine required for the sulphide.For the estimation of thio- sulphate in the presence of sulphites a method described by Bosshard and Grob (ANALYST, 1913, 38, 297) may be employed. An alternative method described by the author for this purpone consists in titrating a portion of the solution containing the two salts with iodine solution ; another portion of the solution is then treated with mercuric chloride and the acidity formed from the thiosulphate is titrated with alkali solution.The reaction between sulphites and mercuric chloride proceeds according to the equation : Na,SO, + HgCI, = NaCl + HgCl.Na80,.Thiosulphate, sulphide, and sulphite, when present in the same solution, may be estimated as follows : A definite volume of the solution is added to an excess of FT iodine solution previously acidified with 10 C.C. of & hydrochloric acid, and the excess of iodine is titrated with thiosulphate solution. Let the number of C.C. of iodine solution consumed be a. The same solution is then titrated with & sodium hydroxide solution, using methyl orange as the indicator; after deducting the 10 C.C.of Fv hydrochloric acid added previously, the remaining quantity of alkali solution used is equivalent to the amount of iodine solution ( b ) required for the sulphite. Another equal portion of the original soiution is now treated with an excess of mercuric chloride, ammonium chloride is added, and the acidity of the mixture is titrated with & alkali solution ( c ) ; the ammonium chloride is added to prevent precipitation of mercuric oxide by the alkali during the titration. Then, the amount of iodine solution used by the thiosulphate is f, by the sulphite, b, and by the sulphide, a - ( b +;).A mixture consisting of Na,S,O,, 0,2471 grrn., Na,SO,, 0.1249 grm., and Na,S, 0,1207 grm., when analysed by the above method yielded Na,S,O,, 0.2483 grm., Na,SO,, 0.1261 grm., and N+S, 0.1201 grm.(See also ANALYST, 1914, 39, 279 ;1915, 40, 254.) w. P. s. APPARATUS, ETC. Sources of E r r o r in Viscosity Measurement. E. C. Bingham, H. I. Schlesinger, and A. B. Coleman. (J. dmer. Chern.SOC., 1916, 38, 27-41.) -The average pressure to be used in the calculation of a viscosity measurement is not strictly one-half the sum of the initial and final pressures as ordinarily assumed, and the use of this value may lead to considerable error. The method of calculating the correct value has been developed and tested experimentally. I n the flow of a liquid through a capillary tube it is immaterial whether the capillary is in a horizontal or in a vertical position. When a liquid flows out from a capillary with a trumpet-shaped opening, there is a definite loss of kinetic energy, but whether this kinetic energy correction is the same as in the well-studied case of a capillary of uniform bore, is riot proved. The capillaries should in consequence have ends as nearly right angles to the axis as possible, so that the kinetic energy correction may be calculated. I t is also shown that the bulbs should be as short as possible. This is best accomplished by making each bulb approximate the shape of two cones placed base to base. G. C. J.