ANALYTICAL CHEMISTRY. Analytical Chemistry. ii. 47 Gas Interferometer Calibration. JUNIUS DAVID EDWARDS ( J . Arne?.. Chem. SOC. 1917 39 2382-2385).-A simple method of calibrating a Rayleigh-Zeiss gas interferometer is described. This differs mainly from the older method of using two gases of different refractive indices in using dry air free from carbon dioxide1 in both sides of the apparatus and changing the pressure of one side. J. F. S. The Direct Determination of the Hygroscopic Coefficient. FREDERICK J. ALWAY MILLARD A. KLINE and GUY R. MCDOLE ( J . Agric. Research 1917 11 147-165).-An investigation of Hilgarcl's method for the determination of the hygroscopic coefficient of soils (compare U.S. Dept. Agric. Div. Chem. Bull. 1893 38). If carried out exactly as described it gives trustworthy results but i t is more convenient t o replace the sheets of glazed paper by shallow aluminium 01- copper trays.The absorption boxes must not be increased in size and a larger number of samples must not be ex- posed in one box. The hygroscopic coefficient increases with rise in temperature (cornpare Lipman and Sharp A. 1912 ii 84). Preliminary drying of soils at 100-llOo does not affect' their hygro- scopicity and soils may be ground in steel mortars t c pass through a 1 Em. sieve without affecting this coefficient. It is preferable to expase the soil for twenty-four rather than for twelve hours and i t is essential to use only a very shallow layer of soil. Transference of the exposed soil from the boxes to weighing bottles must be) per- formed rapidly or the results obtained will be too low.W. G . Estimation of Sulphur in Urine. H. J. HAMBURGER (Zeitsch. physiol. Chem. 1917 100 221-240. Compare A. 1916 ii 641). -Application of the method previously described to the estimation of inorganic and ethereal sulphate and of neutral sulphur in urine. H. W. B. Gravimetric Estimation of Sulphuric Acid and Barium as Barium Sulphate. Z. KARAOGLANOW (2eit:ch. anal. Chem. 1917 56 417-439. Compare J. SOC. Chem. Ind. 1918 37 Feb.). -An investigation of the influence of various substances on the precipitation of barium sulphate Low results .are obtained when n large excess of barium chloride is added but when hydrochloric acid or nitric acid is also present the results are too high. Potassium salts cause the results to be low unless hydrochloric acid is also present when they become too high ; hydrochloric acid how ever increases the low results obtained in the presence of both potassium chloride and potassium sulphate.Sodium ammonium zinc magnesium and aluminium chlorides have no effect on theii. 4% AESTI<ACTS OE’ CHEMlCAL YAI’ICRS. l)recipitation ferric chloride makes the figures for the iiariurti sulphate t o be about 3% too low aiid chromium chloride has a similar effect . w. P. s. Estimation of Nitrogen in Explosives of the Type of Nitric Esters. BERNARDO ODDO (Gnzzcttn 1917 47 ii 145--15S). -In the method of estimation here described the reaction between nitrates sulphuric acid and mercury is employed but the process is madel gravimetric instead of volumetric.The explosive is dis- Folved in sulphuric acid in a small conical flask in which a small test-tube containing the mercury is placed. The flask is closed with a rubber stopper which carries an inlet tube and (as an outlet) a U-tube containing sulphuric acid. Purified carbon dioxide is passed through until all air is displaced (about forty minutes). The apparatus is then disconnected from the gas supply closed a t each end with a short rubber tube and glass rod and weighed. The end of the U-tube is then opened the mercury is poured into the flask by inclining the latter and the whole is shaken for a few minutes. Wlien the reaction is completz carbon dioxide is passed as before. After reweighing the percentage of nitrogen can be calculated from t!ic loss in weight.The results are satisfactory. R. v. s. Estimation of Metallic Iron in Ferrum Reductum. AUGUST EBERHARD (A rvh. I’lmrni. 1917 255 357-38t).-Ferruln reducturn used to be prepared by means of pure hydrogen at a not Loo high temperature genuine ferrurn lzydrogeiiio reducturn being thereby obtained. In recent years however impure hydrogen (COIL- taining carbon monoxide) and higher i eilzperatures have beeii einployed and these changes have so altered the quality and purity of the product t h a t the old methods of estimating the iron par- ticularly the iodometric methods no longer yield trustworthy resulls. For the technical method recommended see J . Aoc. Chew. Ind. 1918 37 3 9 A . c. s. Iso- and Meteropoly-acids. XV. Analytic and Synthetic Methods €or the Investigation of Heteropoly- acids. ARTHUR ROSENHEIM and JOEIANNES JAENICKE ( Z e i t s c h . auorg. Chenz. 1917 101 215-224) .-A critical sumiiiary is given with numerous references of the analytical methods employed for these acids in- cluding. estimations of water boric acid silica phosphoric arsenic molybdic and tungstic acids and alldis. A short outline of the general methods for preparing heteropoly-acids and salts is also given. E. 13. R.