ANALYTtCAL CHEMISTRY. 49 Analytical Chemistry. Improved Eudiometar ; its Application to Measuring I Fire- damp. Detection and Estimation of Methane and Carbon Monoxide. NESTOR GRBHANT (Compt. rend. 1906 143 813-815). -A modification of the eudiometer consists in the employment of a fine platinum wire heated by an electric current to fire the gases ; the wire is supported on two copper terminals which pass through a rubber stopper closing the eudiometer tube. The presence of lolo of methane in air can be measured with an accuracy of 92°/0 in an eudiometer of this form by passing the current 400 t o 600 times. Comparative experiments conducted on rabbits on the poisonous action of small quantities (1 5000 or 1 10,000) of carbon monoxide in air show that the amount of carbon monoxide absorbed by the blood becomes constant after five hours (compare Abstr.1900 ii 95). M. A. W. Standardisation of Normal Solutions [Acids and Alkalis] ; Note on the Zeiss Immersion Refractometer. B. WAGNER A. RIKCK and F. SCHULTZE (Ch,em. Zeit. 1906 30 1181-1183).-The acid hydrochloric acid for instance is first tested by Lunge's sp. gr. process. It is then titrated by means of standard potassium hydr- oxide free from impurities and the liquid is evaporated to dryness with the usual precautions. The residual potassium chloride is weighed and the result represents the amount of hydrogen chloride. The sp. gr. and the titration results serve a s a useful check. A further check may be made by testing the acid or alkali or the resulting potassium salt in the Zeiss immersion refractorneter.The authors have const,ructed a new table for nitric acid at 17.5" and other tables are in course of construction. When using this instru- ment it is of fundamental importance that the temperature should be 17.5" exactly and a new warming arrangement and special thermo- meters should therefore be used for a description of which the original paper and drawings must be consulted. L. DE K. Use of Alkaline Mercuric Iodide Solution as an Oxidising GYSBERT ROMIJN (Ber. 1906 39 Agent in Volumetric Analysis. 4133).-A claim for priority (compare Rupp Abstr. 1906 ii 902). A. McK. VOL. XCII ii 450 ABSTRACTS OF CHEMICAL PAPERS. Detection of Traces of Water. F. SCRIBA (Chem. Centr. 1906 ii 1458; from Zeit. phys. chem. Unterr. 19 298).-A test paper is made by soaking strips of filter paper in a 5"10 solution of ferrous ammonium sulphate; the dried paper is then sprinkled over with finely-powdered potassium ferricyanide.The most minute drop of water produces a blue spot. L. DE K. Estimation of Halogens in Carbon Compounds by Means A. STEPANOFF (Bey. 1906 39 of Sodium and Ethyl Alcohol. 4056-4057).-See Abstr. 1905 i 335 Estimation of Sulphur in Pyrites. Chern. 1906 19 1854-1855).-Polemical GEORCI LUNGE (Zeit. anyew. A reply to Drepnstedt and Hassler (Abstr. 1906 ii 896). L. DE K. Estimation of Sulphur Existing as Zinc Sulphate in Roasted Blendes. V. HASSREIDTER (BUZZ. Xoc. chim. BeZg. 1906 20 163-164).-Zinc sulphate contained in roasted zinc ores may be extracted with water and the zinc titrated as usual 1 part of zinc= 0.492 part of sulphur Insoluble basic zinc sulphate is also present in the roasted ore but in negligible quantity.L. DE K. [Kjeldahl] Ammonia Distillations with or without Cooling Arrangement. ERNST PESCHECK (J. Land. 54 [ iv] 367-384)- Equally satisfactory results may be obtained in the Kjeldahl process either with or without the use of a cooling arrangement if only care be taken to have the condenser made of glass poor in alkali (Jena glass). Loss of ammonia may be prevented as follows the liquid in the distilling flask should occupy about 300 c.c. the amount of sodium hydroxide added should not be excessive the acid in the receiver should be diluted to 100-150 c.c. and the delivery tube which a t the lower end should have a diameter of 0.5 cm.should reach almost to the bottom of the acid. If no cooling arrangement is used the acid in the receiver is considerably heated and must be allowed to cool before it can be titrated; this however depends on the nature of the indicator used 1,. DE K. Estimation of Nitric Acid in Soils. ERNST GUTZEIT (Lalndw. Yersuchs-Stat. 1906 65 2 17-2 19). -Precipitation of the humus present in soil extracts before estimating nitrates by reduction does not cause high results as stated by Buhlert and Fickendey (Abstr. 1906 ii 125). It was found that results obtained with solutions from which the humus had been precipitated were considerably lower (37O/,) than those obtained with the original solutions containing humus. N. H. J. M. Eatimation of Phosphoric Acid by Woy's Method.(x. B. VAN KAMPEN (Chrn. Centr. 1906 ii 1357 ; from Chern. Weekbkctd 3,376-379),-The author confirms the accuracy of Woy's gravimetrioANALYTICAL CHEMISTRY. 51 process (weighing as 24M00,,P,05) but equally good results are obtained by the volumetric molybdate method. L. DE K. Moisture in Goal. E. E. SOMERMEIER (J. Amer. Chem. Xoc. 1906 28 1630-1 638).-This apparently simple determination only gives trustworthy and concordant results when careful attention is paid to the following points. It is absolutely necessary that care be taken t o prevent changes in moisture in the coarse sample during transit. The fine sample for chemical analysis should be in nearly an air-dry con- dition. The air of the drying-oven should be of a uniform dry con- dition and the dried sample should be cooled in a desiccator over sul- phuric acid.L. DE K. Estimation of Moisture and Volatile Matters in Goals. HENRI PELLET and ALBERT ARNAUD (Ann. Chim. anal 1906 11 428-431).-Moisture.-Having made a series of experiments the authors conclude that the moisture in coals may be determined equally well by heating 2 grams of the powder for fifteen minutes a t 100-105° and drying for an hour at go’ or by placing the sample in a desiccator over sulphuric acid for forty-eight hours a t 25-28’. Volatile Matters,-l’mo grams oE the sample are heated rapidly in a small platinum crucible provided with a suitable lid; when the flame ceases the heating is continued for another three minutes. A correction should be applied for carbon oxidised by the air inside the crucible generally 2 or 3 mg. This may be determined readily once for all by reweighing the coke several times after a six minutes’ re-ignition thus obtaining the average loss due to oxidation. L.DE K. Estimation of Carbon Dioxide alone or in Admixture with Hydrogen Sulphide or Chlorine. GEORG LUNGE and AUUUST RITTENER (Zeit. angew. Chern. 1906 19 1849-1852),- Combined carbon dioxide is liberated as usual with hydrochloric acid over brine in presence of a little aluminium the gas is collected in a gas burette and the volume of carbon dioxide estimated as usual by means of aqueous sodium hydroxide. Waste gases are passed through the burette and treated similarly. If hydrogen sulphide should be pre- sent the gas is distributed in two burettes ; in one of these the joint volume of carbon dioxide and hydrogen sulphide is found by absorp- tion and in the second one a known volume of standard iodine is introduced which is then titrated with sodium thiosulphate as usual ; the weight of the hydrogen sulphide thus found is then reduced to volume.In presence of chlorine the last traces of carbon dioxide are ex- pelled from the liquid by boiling with a little hydrogen peroxide. The joint carbon dioxide and chlorine are then found by absorption and the chlorine in the alkaline liquid is estimated by the arsenic titration process. L. 1)E K. 4-252 ABSTRACTS OF CHEMICAL PAPERS. Detection of Barium Strontium and Calcium. STANLEY R. BENEDICT (J. Amer. Chem. Xoc. 1906? 28 1596-1598).-The method is based on the fact that barium and strontium solutions are pre- cipitated completely. by excess of potassium iodate; the latter pre- cipitate however is soluble in dilute hydrochloric acid.A few C.C. of the neutral solution are mixed with half the volume of 5N-hydrochloric acid and to the mixture is added an equal volume of saturated solution of potassium iodate. An immediate precipitate indicates barium; a precipitate forming but very slowly may be due to strontium and in such a case the presence or absence of barium should be confirmed in a portion of the original solution with potassium dichromate. The filtrate is then tested for strontium by boiling the last portion of it with an equal volume of saturated ammonium sul- phate solution. Another portion of the original neutral solution is then mixed with twice its volume of potassium iodate and after a minute or so the filtrate is tested for calcium with ammonium oxalate and a drop of ammonia.L. DE K. Separation of Barium Strontium and Calcium. CARON and RAQUET (Bull. Xoc. citim. 1906 [iii] 35 1061-1069).-The mixed carbonates separated in the usual way by the group reagent are dissolved in excess of acetic acid the solution boiled and from this the barium is precipitated by the addition of a solution containing To the filtrate ammonia solution is added until the liquid assumes a pale yellow colour and has a distinctly ammoniacal odour ; strontium chromate is then pre- cipitated unless only small quantities are present when it is necessary to add alcohol (60-70°). To the filtrate from this potassium ferro- cyanide is added when potassium calcium ferrocyanide is precipitated ; a t once if a considerable quantity of calcium is present but only on warming if the amount present is small.By this method 0.001 gram of barium chloride can be detected in pre- sence of 0.5 gram of strontium chloride or calcium chloride the two metals being precipitated in the first instance as carbonates. If the reagerhs are applied directly to a solution containing both chlorides one part of barium chloride can be detected in presence of 8000 parts of strontium chloride. A critical &sum8 of other methods in use for the separation of these three elements is given. of ammonium chromate and 5O/ of acetic acid. T. A. H. Separation and Estimation of Glucinum. CHARLES L.PARSONS and S. K. BARNES (J. Amer. Chem. Soc. 1906 28 1589-1595).-The authors have found that glucinum may be separated quantitatively from iron and aluminium by heating the neutrdised solution to boiling with excess of sodium hydrogen carbonate; lo"l0 of the solid salt should be added which causes the precipitation of the iron and aluminium and leaves the glucinum in solution. The precipitateANALYTICAL CHEMISTRY. 53 must be freed from any glucinum hydroxide by redissolving in hydro- chloric acid neutralising with ammonia and reprecipitating with sodium hydrogen carbonate. From the mixed filtrate the glucinum is recovered by adding excess of hydrochloric acid boiling and precipitating with ammonia. The precipitate is washed with a solution of ammonium acetate ignited and weighed L.DE K. Quantitative Separation of Glucinum and Aluminium. CARL FRIEDHEIM (Bey. 1906 39 3868-3869).-The method described by Glassmann (Abstr. 1906 ii 902) Bas been used previously by Joy (J. p ~ . Clwm. 1864 [i] 92 235) and also by Zimmermann (Abstr. 1888 323). J. J. 5. A Highly Sensitive Method of Precipitating Zinc. GABRIEL BERTRAND and MAURICE JAVILLIER (Compt. rend. 1906 143 900-902).-Zinc is precipitated in the form of sparingly soluble highly refractive crystalline hydrated calcium zinc oxide Ca(O*Zn*OH)2,4H,0 (Bertrand Abstr. 1893 ii l l S j when a very dilute solution of a zinc salt containing 'calcium hydroxide is boiled with excess of ammonia. and this reaction can be employed for the estimation of one part of zinc in 500,000 parts of water.For this purpose 500 C.C. of the solution is mixed with a few C.C. of dilute milk of lime or 50 C.C. of lime water and 10 to 15'10 of concentrated ammonia solution filtered and then boiled until alkaline vapours are no longer evolved; the precipitate of calcium zinc oxide and calcium carbonate is collected on a small filter dissolved in hydrochloric acid the solution evaporated to dryness the residue taken up with a little water and the calcium precipitated as the oxalate in the presence of excess of ammonia. By evaporating and igniting the filtrate in the presence of sulphuric acid the zinc is obtained and can be weighed in the form of the sulphate or in cases where the amount of zinc 'is too small to admit of estimation its presence can be detected by dissolving the residual sulphate in water and adding hydrogen sulphide or potassium ferrocyanide solution.It is essential that all the reagents employed shall be free from zinc and the result of each experiment must be corrected for errors introduced by the slight solubility of calcium oxalate and of glass in water by subtracting from the weight of the zinc sulphate the weight of the residue obtained in a control experiment. M. A. W. Influence of Ammonia and Ammonium Salts in the Schaffner Zinc Titration. ALPHONSE DECKERS (BUZZ. Xoc. chim. BeZg. 1906 20 164-166).-Experiments showing that the accuracy of the Schaff ner process (titration with sodium sulphide) is seriously affected by the amount of free ammonia and ammoniacal salts present in the liquid. I n order to obtain trustworthy results the comparison liquid should therefore contain the same amount of ammonium54 ABSTRACTS OF CHEMICAL PAPERS.compoundp. Owing to the action of these salts on the zinc sulphide they should not be introduced too liberally. L. DE K. Quantitative Estimations of Metals by Electrolysis. FRITZ FOERSTER (Zeit. arzgew. Chern. 1906 19 1842-1849 1889-1895).- A lengthy review of the subject dealing principally with the electrolytic estimation of nickel zinc cadmium and copper. A large number of tables and formula are given. The use of wire gauze electrodes is strongly recommended. L. DE K. Volumetric Estimation of Copper by means of Potassium Iodide. H. CANTONI and M. ROSENSTEIN (BUZZ. Soc. chim. 1906 [iii] 35,1069-1073).-1t is shown that the results obtained by this method are not influenced by the quantity of potassium iodide added and are but little affected by the presence of potassium sodium ammonium zinc or magnesium sulphate whilst the results are considerably modified by the presence of varying amounts of water or acetic acid.The method gives excellent results when carried out under certain well-defined conditions and it is indispensable that the solution of sodium thiosulphate employed should be standardised under conditions similar to those which will hold when the actual determination is being made and that the solution of iodine should be standardised with reference to the particular type of copper salt which i t is proposed to estimate with it. T. A. H. Solubility of Iron in Vinegar.W. HOFFMANN (Chem. Cent?'. 1906 11 1368-1369; from Deut. Essigind. 10 306).-Iron is decidedly soluble in vinegar communicating a green colour to maIt vinegar and a yellow or red colour to distilled vinegar. This iron interferes with the ordinary titration of samples which should there- fore be submitted to distillation L. DE K. [Estimation of] Chrome and Acid in Chrome-tanning Liquors. F. W. ALDEN (Chern. Centy. 1906 11 1460; from J. Amel.. Leather Chem. ASSOC. 1 174).-A quantity of the liquor sup- posed to contain 2-3 grams of chromic oxide is diluted with water to 500 C.C. Ten C.C. are then diluted with 15 C.C. of water and 2 grams of sodium peroxide are added. After boiling for a very-short time excess of hydrochloric acid is added and the chromate formed is then titrated as nsua.1 with potassium iodide and sodium thiosulphate.The combined acid is estimated by adding 50 C.C. of the liquor to a boiling solution of 50 C.C. of N/2 sodium carbonate diluted with 150 C.C. of water and diluting when cold to 500 C.C. Two hundred C.C. of the fil- trate are then titrated back with N/2 hydrochloric acid. L. DE K. Conditions of Quantitative Precipitation of Uranium Per- oxide. ARRIGO MAZZUCCHELLI ( A t t i R. Accad. Limei 1906 [v] 15 ii 429-438 and 494-500 Compare Fairley Trans. 1877 127; Melikoff and Pissrtrjewsky Abstr. 1898 ii 165).-The author hasANALYTICAL CHEMISTRY. 55 studied the conditions under which uranium peroxide UO is pre- cipitated quantitatively from solutions of uranpl salts by means of hydrogen peroxide.The solubility of uranium peroxide expressed in grams of UO per litre of solution is (1) in water 0*0061 at 20' and 0.0084 at 90° and (2) -in N/10 ammonium chloride 0.0053 at 20° and 0*0088 a t 90'. I n presence of free acetic hydrochloric or nitric acid the precipita- tion of uranium peroxide from solutions of uranyl salts by hydrogen peroxide is incomplete the amount remaining in solution depending on the extent of dissociation of the acid. When free sulphuric acid is present the precipitation is slight in amount and slow probably owing to abnormal ionisation of the uranyl sulphate. Hence the acidity of the solution should only be slight and sulphuric acid should be absent. I f the solution contains ammonium chloride (best in about N-concentra- tion) the precipitated uranium peroxide is easily filtered and washed ; the washing liquid should be N/10 ammonium chloride.A tempera- ture of not above 70" and an excess of the hydrogen peroxide accelerate the precipitation. Moderate amounts of the chlorides sulphates and nitrates of the alkali metals or ammonium do not interfere appreciably with the pre- cipitation of uranium peroxide but acetates fluorides oxalates or tartrates hinder it to a greater or less extent. Large proportions of the chlorides of the alkaline earth metals prevent the precipitation in some degree which is however diminished by the addition of ammonium chloride. The anion present in uranyl salts may be estimated by adding a slight excess of hydrogen peroxide and titrating in presence of the precipitate with sodium hydroxide or ammonia solution using phenol- phthalein as indicator; the liquid assumes an orange colour when alkaline T.H. P. Nitrogenous Compounds and Silica in Sea-water. WILHELM E. RINGER (Chem. Centr. 1906 ii 1459; from Chem. Teekbkad 3 585-608).-Raben's process is recommended. The sample must be examined without delay; otherwise it should be preserved by addition of 0*1"/ of mercuric chloride. One hundred C.C. are distilled with a few drops of acetic acid and the distillate is examined colorimetrically for nitrous acid (phenylenediamine test). One gram of magnesium oxide is then added and the free ammonia collected and determined by the Nessler test. The residue may then be tested for nitrate by reducing it with aluminium foil and sodium amalgam and Nesslerising the ammonia formed.Silica is estimated in the usual manner by evaporating 3 litres of the sample with hydrochloric acid in a platinum dish and taking up the residue with dilute acid. L. DE K. Estimation of Manganese in Drinking Waters. JOHANNES PBESCHER (Chem. Centr. 1906 ii 1458 ; from Pitarm. Centr.-Ha& 799-802).-The residue obtained from 1 litre of water is treated56 ABSTRACTS OF CHEMICAL PAPERS. with nitric acid and potassium chlorate (Hampe’s process for estimat- ing manganese in iron) and the resulting manganese dioxide is titrated as usual with standard oxalic acid and potassium permanganate. L. DE I(. Detection of Small Quantities of Dextrose in Urine. CE. PORCHER (Chem. Centr. 1906 ii 1362; from Bull. Assoc. Chim. Sucr. Dist.24 155-159).-The sample is precipitated with mercuric nitrate then neutralised and diluted to ten times its volume. The excess of mercury is removed by zinc dust or hydrogen sulphide the filtrate is evaporated in a vacuum to a small bulk the residue is mixed with excess of alcohol and the filtrate evaporated in a vacuum. The residue is then tested for dextrose by the usual phenylhydrazine method (osazone reaction). An excess of this reagent should be avoided. L. DE K. Laevulosuria. Detection of Lmvulose in Urine. ADOLF JOLLES (Arch Pharm. 1906 244 542-549).-L;levulose may occur in urine alone as well as in conjunction with dextrose. Lmdose is detected by adding to 10 C.C. of the urine a little resorcinol and about 3 C.C. of lOo,l0 hydrochloric acid and heating to boiling ; an immediate red coloration indicates kevulose (Seliwanoff’s reaction).For the estimation of lzevulose Ost’s method (Abstr. 1891 125) is the best. The urine is boiled with a solution of copper sulphate in aqueous potassium carbonate and potassium hydrogen carbonate and the precipitate is reduced to metallic copper and weighed. If dextrose is present a polarimetric determination is made also the urine being first clarified with normal lead acetate; from the results of the two experiments the amounts of dextrose and lzevulose can be calculated. The method of calculation is exemplified in the paper and some test analyses are given. C. F. B. Separation of Starch and Glycogen. EMIL BAUR and EDUARD POLENSKE (Chem. Centr. 1906 ii 1360-1361 ; from Arbb. Kais.Ges. A. 24 576-580).-Estimution of the Slarch.-A few decigrams of the mixture of starch and glycogen obtained from sausages are dissolved in 30 C.C. of water and 11 grams of powdered ammonium sulphate are added. This precipitates the starch which is washed first with a solution of ammonium sulphate of the same strength and then with proof spirit. The aqueous filtrate (60 c.c.) contains the glycogen which may be recovered by adding 300 C.C. of water and 500 C.C. of alcohol. Estimation of Glycogen in Meat.-A modification of Mayrhofer’s process. Sixty grams of meat are treated with alcoholic potassium hydroxide the hot liquid is mixed with 100 C.C. of prJof spirit and the crude glycogen is collected. It is washed first with alcoholic potassium hydroxide (80 grams in 1 litre 90% alcohol) and then with 90% alcohol until the filtrate is no longer rendered turbid by hydro-ANALYTlCAL CHEMISTRY I 57 chloric acid.The precipitate is then dissolved in 50 C.C. of N-aqueous potassium hydroxide and when cold acidified with acetic acid and diluted to 110 C.C. From the filtrate the glycogen is precipitated by addition of alcohol collected in a Gooch crucible and washed first with 70°/0 alcohol then with absolute alcohol and finally with ether. After weighing any mineral impurity is found by incineration. L. DE K. Methods of Determining Esters Aldehydes and Furfur- aldehyde in Whisky. LUCIUS M. TOLMAN and T. C. TRESCOT (J. Amer. Chem. Xoc. 1906,28,1619-1630)-A lengthy investigation as to the merits of the various procasses recommended for the estima- tion of esters aldehydes and furfuraldehyde in whisky.All these processes give unsatisfactory results unless applied t o the distillates when the whisky is coloured with caramel. For further particulars the original paper and tables should be consulted. L. DE K. Estimation of Soluble and Insoluble Volatile Fatty Aoids [in Butter]. J. DELAITE and J. LEGRAND (Bull. Soc. chim. Belg. 1906 20 230-233).-The glycerol-soda saponification process is not safe in practice as the soluble volatile fatty acids decrease on slightly prolonged heating whilst the insoluble volatile acids increase. When saponifying with alcoholic potassium hydroxide the boiling should not be prolonged to more than half an hour as shown by the following experiment. After boiling for fifteen minutes the Reichert-Meissl figure of a butter was found to be 30.03; after thirty minutes 30.14; after forty-five minutes 30.58 ; after an hour 31.98 ; after two hours 34.89 and after five hours 46-53.L. DE K. Rapid Method for Estimating Citric Acid in Lemon Juice. CELSO ULPIANI and A. PARROZZANI (Atti R. Accad. Zinc& 1906 [v] 15 ii 517-518).-This method is based on the fact that in a cold solution of citric acid containing calcium chloride sodium hydroxide produces a precipitate when the total acidity is neutralised whilst if the liquid is hot precipitation takes place when only one-third of the acidity is neutralised. After the total acidity of the juice has been estimated approxi- mately by titration with N-sodium hydroxide 50 C.C. of the juice are placed in a 200 C.C.flask together with sufficient N-sodium hydroxide solution to neutralise about one-tenth of the acidity ; the latter value represents the maximum proportion of tartaric and oxalic acids which are hence precipitated after the liquid is boiled (wide irLfra). About 17 grams of calcium chloride and 5 grams of animal charcoal are added to the solution which is then boiled cooled made up to volume and filtered. The following estimations are made in the filtrate (1) to 60 c.c. N-sodium hydroxide solution is added in the cold until a faint permanent turbidity appears; (2) another 50 C.C. is boiled and58 ABSTRACTS OF CHEMICAL PAPERS. N-sodium hydroxide solution is run in until a permanent turbidity appears. The difference between the volumes of sodium hydroxide solution used in (1) and (2) represents two-thirds of the citric acid free and combined present in the juice. T.H. P. Estimation of Salicylic Acid in Canned Tom&oes Cat- sups &c. WILBUR L. DUBOIS (J. Arner. Chem. SOC. 1906 28 1616-1619).-The following method has been adopted finally by the author fifty grams of the pulped tomatoes are transferred t o a 200 C.C. flask containing 50 C.C. of water the mixture is neutralised with ammonia and 15 C.C. of milk of lime (1 lOj are added. The whole is diluted to the mark well shaken and filtered. 150-160 C.C. of the filtrate are acidified with hydrochloric acid and extracted four times in succession with 75-100 C.C. of ether. After washing the mixed ethereal extracts twice with 25 C.C.of water the ether is recovered by distilling slowly until about 20-25 C.C. are left which are then left to evaporate spontaneously. The residue is dissolved in hot water and when cold diluted to a definite bulk. The salicylic acid is then estimated colorimetrically in the usual manner with ferric alum solution. L. DE I(. [Assay of] Sodium Salicylate. FRANK H. ALCOCK (Pharrn. J. 190G 597).-0.5 gram of the sample and 0.5 gram of ammonium chloride are dissolved in 10 C.C. of water and evaporated to dryness in a platinum dish. During this operation ammonia is evolved and also salicylic acid together with any phenol if present. The residue is now gently ignited and the mass which consists of sodium chloride is then dissolved in water and titrated with N/10 silver nitrate of which 31.2 C.C.should be required. An allowance should of courge be made for any chloride pre-existing in the sample L. DE K. Estimation of Humus Acids in Soils by Tacke’s Method. C. I(. VAN DAALEN (Chem. Centr. 1906 ii 1468. Compare Tacke Abstr. 1898 ii lO3).-The difficulty of the method is attributed t o the slow evolution of carbon dioxide and to the complexity of the reactions which cannot be represented as the result of a single reaction. N. H. J. M. Estimation of Fatty M a t t e r in Butter. A. FROEHNER (Chem,. Zeit. 1906 30 1250-1251).-0*63 gram of the sample is melted preferably in a graduated tube with 10 C.C. of water t o remove the salt 10 C.C. of alcohol are added then 25 C.C. of ether and the whole is well shaken. The shaking is now repeated after addition of 26 C.C.of benzene. When the aqueous layer has separated 25 C.C. of the ethereal layer are withdrawn and evaporated in a weighed dish. I n calculating the amount of fat the total volume of the ethereal solution may be taken as 52.8 C.C. (52.3 C.C. + a correction of 0.5). The salt may be titrated in the aqueous solution in the usual manner. L. DE K.ANALYTICAL CHEMISTRY. 59 The Examination of Wool-fat. W. HERBIG (Chem. Rev. Feett. Ham. Ind. 1906,13,303-304).-Concordsnt results may be obtained in the estimation of the saponification number of wool-fat by dis- solving the latter in light petroleum before boiling with the alkali. The figures obtained tend to be higher than those yielded by the ordinary process but are constant between themselves for each parti- cular sample.The increase in the saponification number observed on boiling the fat and alkali mixture for a long time is not due to the action of the alkali on the higher alcohols present. w. P. 8. Constants of' Carnauba Wax. LIONEL G. RADCLIFFE (Phurm J. 1906 596-597).-A sample of the pure unbleached wax gave the following result m. p. 84' ; acid number 2.9 ; saponification number 88.3 ; ether number 85.4; iodine number 13.17. The free acidity was found as usual by boiling with absolute alcohol and titrating with Nj20 aqueous potassium hydroxide. The saponifica- tion number was found by boiling with a mixture of redistilled amyl alcohol and alcoholic potassium hydroxide for several hours and titrating the excess of the latter. The iodine number was obtained by dissolving the sample in chloroform and treating with Wijs's reagent (iodine monochloride in glacial acetic acid) ; the original Hub1 solution was almost inactive.L. DE K. Estimation of Camphor [in Celluloid]. ALOIS ARNOST (Zeit. Nahr. Genussm. 1906 12 538-539).-In the method described the camphor is obtained in alcoholic solution which is shaken with light petroleum and acidified water; the increase of the volume of the light petroleum is equivalent with certain corrections to the quantity of camphor present. Ten grams of the finely-rasped celluloid are agitated for one hour in a flask with 100 grams of 25O/ sodium hydroxide solution. The flask is then connected with a condenser and its contents heated to a temperature of about 80' until homogeneous By increasing the heat the camphor is volatilised into the condenser and is carried over into the receiver (a 100 C.C.flask) by adding Eeveral successive small quantities of alcohol to the distillation flask. The distillate is diluted to 100 C.C. with alcohol and the camphor estimated as follows 90 C.C. of water acidified with sulphuric acid are placed in a bulb-shaped vessel the neck of which is graduated showing each 0.05 C.C. from 98 C.C. to 100 C.C. ; the bulb holds exactly 90 c.c. and the part of the neck between the 90 C.C. and 98 C.C. marks is widened out into a small bulb. Above the graduated neck is an upper pear- shaped bulb provided with a short neck which may be closed with a stopper. Ten C.C. of the alcoholic camphor solution are next intro- duced into the apparatus and then 50 C.C.of light petroleum D 0.64-0.67. After closing the apparatus with an indiarubber stopper the contents are shaken for two minutes allowed to separate for half an hour and the increase of the volume of the petroleum read off all the readings being taken a t 15'. This increase is due to the extracted camphor and to the contraction caused by the mixing of the alcohol and aqueous layers. The quantity of alcohol actually present60 ABSTRACTS OF CHEMICAL PAPERS. is ascertained by transferring 50 C.C. of the aqueous portion to a flask adding an excess of sodium hydroxide and distilling the alcohol. The sp. gr. of the distillate when the latter has been diluted to 50 C.C. indicates the percentage of alcohol and the contraction due t o its presence is then found by reference t o tables The corrected increase of the volume of the light petroleam gives the voZume of the camphor in the 10 C.C.of solution taken for the estimation. Camphor has D:; 0.993 Contraction Tuble. Alc o h 01. c. c. 3 s o 0 4-00 5 -00 6 .OO 7.00 8.00 9 *oo 10.00 Water C.C. K 90. 91. 92. 93. 95. 96. 97. 98. 2 176 176 175 175 175 175 174 174 174 174 173 242 242 242 241 241 241 241 240 240 240 240 311 311 311 310 310 310 310 309 309 308 308 384 384 383 382 381 381 380 379 879 378 378 469 459 458 468 457 456 455 454 453 451 450 540 539 538 537 536 534 533 532 531 530 529 628 626. 624 621 618 616 614 613 612 611 610 717 715 713 711 709 706 704 702 700 698 696 The contractions are expressed in cubic millimetres. W.P. 5. Direct Estimation of [Quebracho] Tannin. H. FRANKE (Chem. Centr. 1906 ii 916-917 ; from Phurm. Centy.-Halle 47 599-604). -One hundred C.C. of an aqueous decoction of Quebracho wood or a solution of 0.2 gram of the tannin obtained from it by Korner's process (Ledermurkt 1897 No. 37) in 100 C.C. of water are heated to boiling with 50 C.C. of formalin; 25 C.C. of 25 per cent. hydro- chloric acid are added and the whole is heated for ten minutes longer. After half an hour the precipitate is collected washed with water alcohol and ether and dried a t 110'. Its weight x 0.9834 = que- brachotnnnic acid. L. DE K. Separation of Proteoses and Peptones from the Simpler Amino-compounds. WILLARD D. BIUELOW and F. C. COOK (J. Amer. Chem. Xoc. 1906 28 1485-1499).-Acomparison has been made of the action of the proteid precipitants tannin-salt solution and phospho- tungstic acid on amino-compounds such as glycine alanine crentine phenylenediamine and sarcosine both singly and in mixtures. It has been found that a satisfactory separation of proteosee and peptones from the simpler amino-compounds can be effected by means of a solu- tion containing 4 to 5 grams of sodium chloride and 15 grams of tannin in 100 C.C. The re- agent should be kept in a cool place and should not be left more than a few days before being used. An error is occasioned by the precipi- tation of a portion of the creatine by the reagent but this may be corrected by estimating the creatine in a portion of the solution before applying the reagent and in the filtrate from the precipitate given by that reagent. E. G. The precipitation is best carried out at 12".