THE ANALYST. 251 ORGANIC ANALYSIS. On Herzig and Meyer's Method of Cetermining Methyl. M. Bus@h. (Berichte, 1902, xxxv., 1565-1567.)-The modification of Zeisel's method devised by Herzig and Meyer affords a means of determining the methyl in combination with nitrogen, as well as of that combined with oxygen, and of distinguishing between the two in certain cases. Thus, whilst the methoxyl group is hydrolyzed by being boiled with hydriodic acid, a much higher temperature ( 2 0 0 O to 300" C.) is as a rule required to dissociate the methyl group from the nitrogen. In some cases, however, e.g., l-phenyl-4-methylanilino-urazole, the methyl is liberated from the nitrogen by the boiling acid, and the author therefore concludes that when the test gives a negative result the absence of methoxyl can be regarded as proved, but that a positive result cannot be accepted as conclusive of methylimin without further evidence.C. A. M. Separation of Cholesterin and Phytosterin from Mixtures of Fatty and Mineral Oils. J. Marousson. (Mitth. Konigl. Techn. Versuchsanst., Berlin, 1901, xix., 259 ; through Chem. Zeit. Rep., 1902, 154.)-The author has already described (ANALYST, 1901, xxvi., 302) two processes for the above purpose. The following method is specially suitable for such mixtures as contain very viscid mineral oils. It does not yield quantitative results, but if 100 grammes of the sample are taken it enables the vegetable or animal oil to be identified in presence of a mineral oil. The oil is saponified with alcoholic potash, the solution is diluted with an equal252 THE ANALYST.volume of water, and the mixture is shaken out several times with petroleum spirit till the mineral oil is extracted. The alcohol is driven off from the residual liquid, which is then agitated repeatedly with ethyl ether to dissolve the choleaterin and phytosterin. On distilling away the ether and recrystallizing from alcohol the cholesterols of the fats may be recovered pure in the usual manner. F. H. L. Tho Quantitative Estimation of Colophony in Fats, Oils, Soaps, Ceresin, Paraffin, etc. (MittheiE. aus den Konigl. techn. Versuch- saststalt., 1902,40-49.)-As the amount of resin in soaps and oils cannot be accurately estimated by either Twitchell's or Gladding's processes, the authors recommend a method which is a combination of these two processes.The results of test experi- ments, given in the original paper, show that very accurate results can be obtained in this way (a) I n Fats and Soaps containing no Unsaponifiable Matter.-About 5 grammes of the sample (correspondingly more in the case of soap containing water) are weighed out and boiled with 50 C.C. of alcoholic potash for half an hour under a reflux condenser. The alcoholic solution is then evaporated, the residue dissolved in water and decomposed with hydrochloric acid. Soaps containing no free fat may be directly dissolved in water and decomposed with acid. The separated fatty acids are removed by shaking with ether ; the aqueous acid solution is neutralized, evaporated to about 25 c.c., re-acidified and Ehaken out with ether.After distilling off the ether from the united ethereal extracts, the residue of fatty acids is dissolved in 50 C.C. of absolute alcohol, and the acids converted into esters by passing a moderately rapid current of dry hydrochloric acid gas through the solution cooled by ice-water to a temperature not above 10" C. When the operation is complete (it usually takes one to two hours) the flask and its contents are allowed to stand for half an hour at the ordinary temperature. The contents are then washed into a large Erlenmeyer flask with five times their volume of water, and boiled under a reflux condenser for half an hour. The cooled solution is now shaken out with 100 C.C. of ether in a separating funnel, and then with several successive quantities of 50 C.C.of ether until the extracts are colourless. The aqueous liquid is neutralized, evaporated to 50 c.c., acidified, and repeatedly extracted with small quantities (25 c.c.) of ether. In this way the water-soluble constituents of the colophony are obtained. After mixing the ethereal solutions they are shaken out with about 50 C.C. of potash solution (10 grammes of caustic potash, 10 grammes of alcohol, and 100 C.C. of water). A brown layer separates out between the ether and the alkaline solution, and is drawn off with the latter It contains a considerable portion of the resin-soap, which is only slightly soluble in the potash solution. The ether is then well washed with water to remove soluble resin-soaps, again with two successive quantities (10 c.c.) of the potash solution, and finally with water until the washings are colourless.In the presence of fish-oil acids or much colophony the washing with water must be very thorough, To remove any fatty esters which may have been mechanically carried into the alkaline aqueous liquors, the latter are shaken with 50 C.C. of ether. Holde and Marcusson. The method is as follows :THE ANALYST. 253 After separating the ether it is extracted with 5 C.C. of the potash solution, and the aqueous alkaline extract added to the main bulk of the same. This is now acidified and shaken out with successive quantities (50 c.c.) of ether until completely extracted. The acid solution is neutralized, evaporated to a small bulk, re-acidified, and again shaken out with ether. The total ether extracts are washed with 20 C.C. of water, and the ether is distilled off.The residue of resin-acids so obtained-still con- taminated with unchanged fatty acids-is placed in a tared glass basin and evaporated with several small successive additions of absolute alcohol to remove the last traces of water, and weighed. The fatty acids still remaining in the resin-acids are removed by the employment of Gladding's process. To this end from 0.4 to 0.6 gramme of the above obtained resin-acids are placed in a 100 C.C. stoppered and graduated cylinder, and dissolved in 20 C.C. of 95 per cent. alcohol. If only a small amount of resin- acid is obtained, the following proportions of the ether-alcohol mixture must be correspondingly altered, whilst, in the case of a large yield of acids, it is advisable t o dissolve the whole in so much 95 per cent.alcohol that .20 C.C. of the solution shall contain 0.5 gramme of resin-acid. A drop or two of phenolphthalein solution are added to the alcoholic solution in the cylinder, and then concentrated caustic soda solution (1 of NaOH to 2 of water), until the reaction is just alkaline. The loosely stoppered cylinder and its contents are heated for a short time in the water- bath, then cooled, ether is added up to the 100 C.C. mark, and mixed. One gramma of dry powdered silver nitrate is now added, and the contents of the cylinder are shaken for fifteen to twenty minutes to convert the resin-acids into silver salts. When the insoluble silver salts have completely separated and settled (after standing over- night if necessary), 70 C.C.of the solution are pipetted into a second 100 C.C. cylinder and shaken with 20 C.C. of dilute hydrochloric acid (1 : 2). The ether layer is drawn ~ f f , and the aqueous portion twice shaken out with 20 C.C. of ether. The united ether extracts are washed with 20 C.C. of water, filtered into a small flask, and the ether is distilled off. The residue, amounting t o about 10 c.c., is transferred to a weighed dish, evaporated, dried for a short time at 110" to 115' C., and weighed. The weight of the resin-acids so found is calculated back into the first weight (impure acids) obtained, and then on the original substance taken. The percentage found is corrected by the subtraction of 0.4 per cent., this allowance being made for a small amount of fatty acid which is always present. As colophony contains an average of 8 per cent.of unsaponifiable matter, a second correction is therefore necessary, the true percentage of colophony in the substance under examination being found by the following equation, in which the two corrections are combined : (percentage of resin acids found - 0.4) 92 ._ _ _ -- -1 percentage of colophony. Should more than 20 per cent. of colophony be present, it is better to directly estimate the unsaponifiable matter. The latter is contained in the ethereal solution of the fatty esters after the resin acids have been removed. After saponifying the esters by the addition of 25 C.C. of N. alcoholic potash, the solution is diluted with 150 C.C. of water and shaken out twice with 150 C.C.of ether. The main bulk of the ether is distilled off, and the remainder allowed to be evaporated at the ordinary temperature. An oily residue remains, (By warming, volatile substances are lost.)254 THE ANALYST. which is freed from traces of soap by treating it with a little alcoholic potash, slowly evaporating off the alcohol, and extracting this residue with petroleum spirit. ( b ) In Fats and Soaps containing Unsaponijiable Matter.--A sufficient quantity of the sample is weighed out to yield about 5 grammes of fatty acids; benzene (free from thiophene) is added, and the sample is saponified. The unsaponified matter is separated from the soap solution in the usual way, and the resin and fatty acids in the latter estimated by the process given above.An accurate estimation of the unsaponifiable substances in this case is not possible, but they may be approximately calculated by assuming that the colophony found to be present itself contains on the average 8 per cent. of these substances. (c) Estimation of Colophony in Ceresin and Parafin.-The resin is removed from the sample by thorough extraction with boiling 70 per cent. alcohol. The united alcoholic extracts are filtered when quite cold, and the alcohol is distilled off from the clear filtrate. Should fatty acids also be present, the residue, obtained after the distillation of the alcohol, The residue is dried at 110" to 115" C.? and weighed. must be treated according to the method described under (a). w. P. s. Larch Turpentine and Venice Turpentine.L. E. And&. (chem. Rev. Fett- u. Ham- Ind., 1902, ix., 126-128.)-Larch Turpentiw, obtained from the trunk of the larch-tree, was formerly known also as Venice turpentine from the fact that Venice was the chief emporium for it ; but this is no longer the case, and the name " Venice turpentine " is now applied to artificial mixtures of rosin, rosin oil, and turpentine used in the manufacture of sealing-wax and varnish. Larch turpentine has a pleasant odour, a bitter taste, end a yellow or brown colour. I t is readily soluble in alcohol, ether, acetic acid, and acetone, and partially soluble in carbon bisulphide. I t consists of a dextro-rotatory rosin and a laevo-rotatory volatile oil. According to Dieterich, it gives the following values : Acid value, 66.92 to 68.85 ; ether value, 46.27 to 54.94 ; saponification value, 114.5 to 12'7.71 ; acetyl acid value, 69-87 to 72.19; and acetyl saponification value, 178.95 to 190.86.The following results were obtained by Beckurts and Bruche in the examination of seven samples : Specific gravity, 1.060 to 1.190; acid value, 76 to 101 ; ether value, 0 to 9 ; and saponification value, 81 to 101. Venice Tuypentine.--In addition to the artificial product, there is also a corn- mercial French product, '( tbrdbenthirte de Venise," which is a superior variety of balsam obtained from shore pines. An artificial product examined by Dieterich had the following characteristics : Acid value, 98-79 ; ether value, 0.88 ; and saponifica- tion value, 97.66. In order to distinguish between larch turpentine and an artificial mixture, the conclusions drawn from the odour, inflammability, solubility in alcohol (90 per cent.), and analytical values may be confirmed by Hirschsohn's ammonia test.Ordinary turpentine, on treatment with five times its volume of ammonia solution (specific gravity, 0-96), gradually yields a milky emulsion, whilst in the case of larch turpentine the liquid remains clear. When the lower layer of larch turpentine is stirred it is gradually transformed into a semi-solid, opaque substance, whilst the supernatant liquid becomes slightly turbid. Ordinary turpentine, how-THE ANALYST. 255 ever, is immediately distributed, and the milky emulsion rapidly aolidifies to a gelatinous mass. In the case of a mixture of the two in equal parts, the substance is distributed throughout the ammonia, and a solid mass, which clears on heating, is obtained in about five minutes.When the mixture only contains about 20 per cent. of ordinary turpentine the milky emulsion becomes clear on heating the tube in boiling water, but no solidification occurs. Comparative tests with genuine larch turpentine may serve to detect smaller quantities. When 1 part of the sample is shaken with 30 parts of 80 per cent. alcohol a clear and permanent solution is obtained with larch turpentine, whilst in the case of ordinary turpentine about half of the substance soon separates out. C. A. M. The Use of Iodine Monoohloride in the Determination of the Iodine Absorption of Oils. (Zed. fiir Untersuch. der Nahr. und Genuss- mittel, 1902, v., 497-504.)-As regards the keeping qualities of the solution of iodine monochloride in glacial acetic acid, it is stated that the solution decreases in strength only to a, very slight extent when carefully prepared.The presence of iodine trichloride causes a loss of strength, as does also the use of acetic acid containing water and reducing substances. In the same way ordinary Hubl’s solution was found to diminish in strength according to the quality of the alcohol used in its preparation. The iodine values obtained by the employment of iodine monochloride agree with the theoretical numbers in the case of unsaturated fatty acids having only one double linking. In using Hubl’s solution various precautions must be taken in order that the results may agree with those obtained by the iodine monochloride J.J. A. Wijs. process (see ANALYST, xxiv., p. 259). w. P. s. The Alkalinity of Crude Sugar. Lauterbach. (D. Zuckerind., 1902, xxvii., 653 and 780; through Chern. Zeit. Rep., 1902, 154.)-When crude sugar is stored for any length of time, bicarbonates are formed in it, while the water it contains becomes saturated with carbon dioxide. It is therefore necessary before titrating a, sample for alkalinity with phenolphthalein to boil the liquid thoroughly. If this is done the results are perfectly accurate, and Herberger’s objections (this volume, p. 196) to the use of that indicator are without point. F. H. L. Identification of’ Sugars. E. Votocek. (Chem. Listy, 1902, xxvi., 122 ; through Chem. Zeit. Rep., 1902, 141.)-The author finds that it is as easy to test for pentoses in any suitable hydrazone by distillation with 12 per cent.hydrochloric acid as in any other compound, the test being rendered additionally delicate by col- lecting a distillate of 180 c.c., and then redistilling it with sodium chloride, a method which enables, for example, 0.06 gramme of the methylphenylhgdraaone of methyl- pentose to be recognised. Tested with phloroglucinol, 0.33 gramme of the benzyl- phenylhydrazone of arabinose gives a large greenish-black precipitate, 0.2 gramme of256 THE ANALYST. the corresponding compound of xylose behaving identically. 0.3 gramme of the same hydrazone of galactose yields no precipitate with phloroglucinol, while those of rhamnose and rhodeose from jalapin or convolvulin give cinnabar-coloured precipitates. In presence of furol, methylfurol may be discovered by means of resorcinol, as the gray furol resorcide does not mask the carmine colour of methyl- furol resorcide.F. H. L. - -__ _ _ A New Method for the Determination of Cellulose. S. Zeisel and M. J, Strittar. (Berichte, 1902, xxxv., 125%1255.)-The authors' method is based upon the fact that cellulose in the narrover sense (dextro-cellulose) is not converted to any large extent into products soluble in dilute ammonium hydroxide, when oxidized by potassium permanganate in the presence of nitric acid, whereas the non-cellulosic constituents of wood readily yield oxidation products, part of which are soluble in water, whilst part can be extracted with 2-5 per cent.ammonium solution. From 1 to 1-5 gramme of the finely divided material is stirred with dilute nitric acid, and a 3 per cent. solution of potassium permanganate introduced, 1 C.C. at a, time, until the red colour of the liquid is plainly perceptibls after thirty minutes, the oxidation taking about two hours in all. The excess of permanganate and the precipitated manganese dioxide are then destroyed by the addition of sulphur dioxide or sodium bisulphite, and the liquid filtered. The residue is washed with water, then treated for forty-five minutes with a 2.5 per cent. solution of ammonia at 60" C., and again washed successively with hot water, alcohol, and ether, The results were concordant, but were much lower than those obtained by Schulze's chlorate process, although showing a satisfactory agreement with those given by the tedious process of Schulze and Henneberg, in which the wood is extracted with water and alcohol, and oxidized for days with a solution of potassium chlorate in nitric acid.About 30 per cent. of the cellulose is converted into oxycellulose by oxidation with permanganate, but the slight error thus introduced can be corrected by extraot- ing the residual cellulose with a boiling 10 per cent. solution of sodium hydroxide, in which the oxycellulose dissolves. The amount of soluble compounds formed from the cellulose in the oxidation does not exceed 4 per cent. Hemi-celluloses are com- pletely converted into soluble oxidation products. I n this way oak raspings were found to contain 37.2 per cent.of cellulose. C. A. M. Detection of Pentoses in Urine in the Presence of Glycuronic Acids, K. Von Alfthan. (Arch. Experiment. Pathol., 1902, xlvii., 417 ; through Chem. z&. Rep., 1902, 154.)-On treatment with benzoyl chloride and sodium hydroxide both pentoses and glycuronic acids are converted into benzoyl esters. When these are saponified with sodium ethoxide the glycuronic acids separate out, the pentogeg remaining in solution. If, then, the filtrate yields the phloroglucinol or orcinol test, pentoses must be present in the urine. F. H. L.THE ANALYST. 257 A Reagent for Albumin in Urine. Pollacci. (Giorn. di fawn. d i Trieste, 1901, 235 ; Ann. de Chim. anal., 1902, vii., 195.)-The reagent consists of 1 gramme of tartaric acid, 5 grammes of powdered mercuric chloride, and 10 grammes of pure sodium chloride dissolved in 100 C.C.of water, and mixed with 5 C.C. of formalin. Two C.C. of this reagent are placed in a test-tube, and 3 to 4 C.C. of the urine under examination carefully poured down the side of the tube so as to avoid mixing the liquids. I n the case of pathological urine, a white zone is immediately produced at the junction of the liquids, whilst normal urine does not give any reaction until after a lapse of ten to fifteen minutes. The limit of sensitiveness of the reaction is 1 in 370,000. C. A. M. Valuation of Rubber Goods. 0. Mayer. (Chem. Zed., 1902, xxvi., 481.)-In this article the author points out that there is no method which can be considered universally applicable to the examination of indiarubber goods.Heintz has not submitted sufficient evidence to prove that the substance left after the’various foreign ingredients have been removed by appropriate solvents agrees with the formula (CloH16)~z; indeed, in view of the way in which caoutchouc may be decomposed during its incorporation with ‘‘ substitutes,” etc., and of its liability to change during the extraction of those substitutes, such a method of ultimate analysis as is recom- mended by Heintz (this volume, p. 200) is not likely to yield trustworthy results. Even the Henriques process cannot be regarded as a routine method, for it needs modification to suit different grades of material; and there are, in fact, on the market now certain descriptions of rubber wares which are not amenable to the Henriques treatment without considerable alteration in the usual procedure.Mayer considers it unfortunate that manufacturers have been led to believe that reliable methods of valuation have already been elaborated. F. H. L. A Reaction distinguishing bet ween a- and P-Naphthols. A. Jorissen. (Ann. de Chim. anal:, 1902, vii., 217-219.)-A pinch of the naphthol is treated with about 2 C.C. of an aqueous solution of iodine and potassium iodide, and the liquid shaken with an excess of an aqueous solution of sodium hydroxide. A clear, colourless liquid is obtained with P-naphthol, whilst in the case of a-naphthol there is turbidity and an intense violet coloration. When mixtures of the two naphthols are thus tested the liquid assumes a more or less pronounced violet colour.C. A. M. The Volumetric Determination of Thymol. E. Zdarek. (Zeit. anal. Chenz., 1902, xli., 227-331.)-This is an application of Koppeschaar’s method of determining phenol, the thymol being precipitated by bromine, and the excess of the latter titrated with sodium thiosulphate after the addition of potassium iodide and starch solution. The solutions employed by the author are the same as those of Koppeschaar, with the exception of the bromine solution, which is of twice the strength. They are as follows : (1) An aqueous solution of sodium thiosulphate (9.76 grammes per litre),258 THE ANALYST. corresponding to a solution of iodine containing 5 grammes per litre ; (2) an aqueous solution of potassium iodide containing 125 grrtmmes per litre; (3) an aqueous solution containing in a litre 3.571 grammes of dry sodiuni bromate and 12.178 grammes of dry sodium bromide.On mixing 25 C.C. of this solution with 10 C.C. of the iodide solution (2) and 5 C.C. of concentrated hydrochloric acid, the liberated iodine is equivalent to 90 C.C. of the sodium thiosulphate solution (1). In making a determination, the dry thymol is weighed into t ~ , flask provided with a stopper, and for each 0.1 gramme 20 C.C. of the bromide solution (3), and 4 C.C. of concentrated hydrochloric acid are added. The flask is well shaken for about five minutes, after which 10 C.C. of the potassium iodide solution and a little starch solution or paste are introduced, and the contents immediately titrated with the thiosulphate. According to the results of the author's analyses, 1 molecule of bromo- thymol contains 4 atoms of bromine.I n ten test determinations thus made, the amounts of thymol found varied from 98.6 to 100-2 per cent. of the theoretical quantity. The bromine compound gradually loses bromine when kept in a desiccator over sulphuric acid, until finally a stable compound corresponding to dibromo- thymol is left. On treating thymol with it large excess of bromine an oily compound was formed, which, when kept for some days in a sealed glass tube, yielded a deposit of nearly colourless crystalline needles, melting at 71" C., and having the composition of a thymol containing 5 atoms of bromine. The author has used this method to determine the solubility of thymol, and has found that 1 gramme dissolves in 11764 C.C.of water at 19.4' C. At 15" to 20° C. 1 part of thymol is soluble in 0.24 to 0.28 part by weight of strong alcohol (90 to 91-2 per cent. by volume), in 0.22 to 0-26 part of ether, and in 0.67 to 0.7 part of chloroform. C. A. M. Colour Reaction for Thiophen. H. Kreis. (Chew. Zeit., 1902, xxvi., 523.)- If a very weak solution of thalline base (tetrahydro-p-oxyquinoline methyl ester) in petroleum spirit is shaken with a liquid containing thiophen and with some 1.4 nitric acid, a distinct, but transient, violet colour is produced, which gradually becomes reddish, and finally yellow. Addition of water destroys the colour, so that weaker nitric acid must not be employed; stronger acid is also to be avoided. The test can be used to show either thiophen or thalline in very minute quantities.It is very convenient for the detection of thiophen in benzene, and it serves equally for the recognition of methylthiophen in toluene. Other alkaloids shaken with benzene containing thiophen and nitric acid do not yield noteworthy colour reactions. Bellier's test for sesame oil (ANALYST, 1900, xxv., 50), which is somewhat similar, does not, however, depend on the presence of thiophen. F. H. L. The Volumetric Determination of Disodium Methyl-amenate. E. Falibres. (Joz~m. Phzrm. Chim., 1902, xv., 466-469.)-Silver methyl-arsenate is sufficiently soluble in water to prevent an exact gravimetric determination of the sodium salt by precipitation with silver nibrate. It is, however, completely insoluble in a, & or & solution of silver nitrate, and on this fact the author bases his method of estimation.THE ANALYST.259 Disodium methyl-arsenate crystallizes with 6 molecules of water, which it loses at 120" to 130" C., but regains to a large extent on exposure to the air. The author prefers to take the formula, As(CH,)O4Na,.6H,O, as typical of the salt, since this does not alter in composition after an exposure of several days to the air. In the volumetric method 0.2 gramme of the crystalline salt (from which chlorides, sulphates, arsenites, arsenates, phosphates, carbonates, and iodides have been proved to be absent) is dissolved in 10 C.C. of water, and the solution mixed with 40 C.C. of i> silver nitrate solution, and rapidly filtered. The filtrate is used to titrate 10 C.C.of & sodium chloride solution diluted with 30 C.C. of water, and the amount of silver nitrate still present is thus found. The amount of disodium methyl-arsenate corresponding with the amount of silver nitrate used in the precipitation is then found by means of the formula 292 x N x 5 x 100 340 where N represents the weight of silver nitrate consumed, 292 the molecular weight of disodium methyl-arsenate, and 340 t,wice the molecular weight of silver nitrate. I C. A. M. The Prezipitation of Alkaloids, Metallic Salts, and Proteids by Extracts of Coffee and Tea. (Journ. Med. Research, Ohio, 1902, 43-53; through Amer. Jourrz. Pharm., 1902, lxxiv., 299-303.)-According to the results of the author's experiments, the reactions given by tea tannin closely resemble those of gallotannic and quercotannic acid, and differ greatly from those given by coffee tannin (a diglycosyl ester of cinnamic acid).The extracts used in the experiments were prepared by boiling the tea or roasted coffee for forty-five minutes with 10 parts of water, and filtering the liquid. Neither extract gave any precipitate with Mayer's reagent or with hydrochloric acid in the dilution in which they were used in the tests. The proportions usually employed in the case of alkaloids were 1 b.c. of the decoction to 2 C.C. of a 1 per cent. aqueous solution of the alkaloid or its salt, or 5 C.C. each of the alkaloidal solution (1 : l,OOO), and of the decoction. Alkaloids.-Atropine, coniine, morphine, and pyridine were not precipitated, even from a moderately concentrated solution by coffee, but were precipitated by tea from strong solutions. Aconitine, brucine, cocaine, lobeline, nicotine, and pilo- carpine were sparingly precipitated from weak solutions by tea,, but not by coffee, even when they were present in strong solutions. Apomorphine, the cinchona alkaloids, hydrastine, strychnine, and veratrine were precipitated by but h tea and coffee from weak solutions. None of the alkaloids except apomorphine was com- pletely precipitated. Metallic Salts.-Tea was found to effect a more complete precipitation than coffee. Aluminium, lead, and silver were precipitated practically completely by both, whilst mercury was partially precipitated by tea, but not by coffee. Proteids (Egg-white, Albumose, Gelatin).-A sharp differentiation between the tannins of tea and coffee could be effected by means of these proteids. The tea decoction gave a large precipitate with them, whereas the coffee decoction was unaffected, or, at most, rendered slightly turbid. T. Sollmann. C. A. 14.260 THE ANALYST. The Kjeldahl Process. K. Neuberg. (Beitr. chem. Physiol. zc. Pathol., 1902, ii., 214; through Chem. Zeit. Rep., 1908, 141.)-In order to destroy the amido- mercuric sulphate which is produced in the Kjeldahl flask when mercury is employed, the author recommends sodium thiosulphate. This has the advantage of being fit for use in the solid form, so that dilution of the liquid is avoided. F. H. L.