38 ABSTRACTS OF CHEMICAL PAPERS. Analytical Chemistry. Estimation of Ozone. By OTTO BRUNCK (Ber., 1900, 33, 2999--3000).-A reply to Ladenburg (Abstr,, 1900, ii, 721). T. M. L. [Estimation of Persulphates and Chromium.] By RODOLFO NAMIAS (L’Orosi, 1900, 23, 218--223).-See this vol., ii, 15. Estimation of Nitric Acid in Water. By N. N. KOSTJAMIN (Chem. Centr., 1900, ii, 878-8‘79; from Pharm. Zeit., 45, 646)- Five C.C. of the sample are slowly mixed with a freshly prepared solu- tion of 1 part of brucine in 3000 parts of sulphuric acid of sp. gr. 1.84 until a permanent rose coloration is formed. The greater the amount of nitrate the less of the reagent will be required. If nitrites should be present, these must be expelled by boiling with dilute sul- phuric acid, A table is given showing the number of C.C.of the reagent required for samples containing from one to twenty milligrams of nitric Volumetric Estimation of Phosphorous Acid. By OTTO KURLINU (Bey., 1900, 33, 2914--2918).-Phosphorous acid may be estimated volumetrically by the aid of potassium permanganate solution but the reaction takes some time for completion. Zinc sulphate is added to the solution of the acid and the liquid is then heated on the water-bath and treated with permanganate solution. The zinc sulphate reacts with the alkali produced from the permanganate, yielding zinc hydroxide which carries down the manganese dioxide rapidly and com- pletely. The titration is eithm carried out until the liquid remains permanently coloured for 10-15 minutes, or an excess of perrnangan, pentoxide per litre.L. DE I(.ANALYTICAL CHEMISTRY. 39 ate is added, the liquid heated for 1--1*5 hours, and filtered, the precipitated manganese dioxide being then treated with potassium iodide and sulphuric acid and the liberated iodine estimated with thio sulphate. The excess of permanganate may also be estimated in the filtrate. A. H. Destruction of Organic Substances by means of Chromyl Chloride in Toxicological Analysis. By C. PAGEL (Chem. Centr., 1900, ii, 784; from Pharm. Post, 33, 489--490).-The apparatus con- sists of a tubulated retort fitted with a separating funnel and connected with a balloon which is in turn connected with two wash-bottles, of which the first is half filled with water and the second with a 1 per cent. solution of potassium hydroxide The balloon is cooled by a thin stream of water.The finely divided, dried substance is put into the retort, 30-40 grams of a mixture of two parts of sodiumchloride and one part of potassium dichromate are added for every hundred grams of material, and sulphuric acid is gradually introduced through the funnel. On heating, yellow fumes of chromyl chloride are evolved which carry over all the arsenic and most of the antimony and mer- cury. Altogether 40-50 C.C. of sulphuric acid are used and the heating is continued until the charred mass has become quite clear again. The residue in the retort may contain non-volatile metallic chlorides. The process seems a particularly delicate one for arsenic, and the author has found this to be a normal constituent of some parts of the animal body (spleen, brain, testicles, ovaries, &c.).By HEINRICH GOCKEL (Zeit. angew. Chem., 1900, 1034).-Koch (Abstr., 1895, ii, 86) has described an apparatus which, with slight modification by Ledebur, has become the recognised apparatus of the Society of German Iron Smelters. The author has made another slight modi- fication to render it still more convenient. In its new form, the apparatus differs from the previous one by having a wider ground neck into which fits the condenser through which passes a thistle funnel tube reaching to the bottom of the flask ; the top of the neck of the flask is expanded, so that it can be filled with water to seal the apparatus after introducing the condenser. The side tube of the flask is placed somewhat higher than before. The acid mixture is poured through the funnel tube which is theu sealed by means of a glass rod.At the lower end of the condenser is fitted a small slightly bent hook supporting a little glass bucket con- taining the sample; by a slight jerk this is made to drop into the acid. Recognition of Barium Compounds as the Cause of Poisoning. By DIOSCORIDE VITALI (L'Orosi, 1900, 23, 260--263).-The detection of barium in the stomach contents or in the matter ejected by vomit- ing and the determination of the form in which it was administered may be carried out as follows. The material is first examined to see whether it gives a marked alkaline reaction, after which it is filtered and the residue extracted with boiling water so long as the latter becomes alkaline.If now the filtrate has an alkaline reaction and barium L. DE K. Flask for the Estimation of Carbon in Iron and Steel. L. DE K.40 ABSTRACTS OF CHEMICAL PAPERS. carbonate is precipitated on passing carbon dioxide through it, the poisoning was brought about by barium oxide or hydroxide. If the addition of dilute sulphuric acid to the filtrate from the barium carbonate causes evolution of hydrogen sulphide and precipitation of barium sulphate, barium sulphide must have been taken. To detect barium chloride, nitrate, or acetate, the filtrate from the carbonate is evaporated to small volume and finely divided, pure lithium carbonate added until the liquid gives a persistent alkaline reaction, when it is boiled and filtered and the precipitate washed; the filtrate is eva- porated to dryness on the water-bath, the residue extracted with a mixture of equal volumes of strong alcohol and ether, and the alcohol- ethereal solution evaporated to dryness.The presence in the solid residue thus obtained, of lithium chloride, nitrate, or acetate shows that the corresponding barium salt was the cause of poisoning. In the event of barium carbonate having been employed, tliis will have been converted to a small extent into the chloride in the stomach, but as vomiting in general occurs soon after the administration of the barium salt, it will mostly remain as carbonate in the insoluble portion of the stomach contents, and may be detected by treating the latter with hydrochloric acid, which will cause evolution of carbon dioxide and the formation of soluble barium chloride.By THOMAS ROMAN and G. DELLUC (J. €'harm., 1900, [vi], 12, 265-267).-When alcohol is stored in galvanised iron vessels, a small quantity of the zinc is dissolved. A chloroform solution of urobilin serves as a very delicate test for zinc, with which it gives a green fluorescence, even when the merest trace of the metal is present. I n making the test, it is necessary to mix the chloroform solution with 24 times its volume of absolute alcohol, in order to prevent the formation of an emulsion. T. H. P. Presence of Zinc in Alcohol. H. R. LE S. Volumetric Estimation of Copper as Oxalate, with Separation from Cadmium, Arsenic, Tin, and Zinc. BY CHARLES A. PETERS (Arner. J. Sci., 1900, [iv], 10, 359--367).-The process is based on the fact that under certain definite conditions copper may be completely precipitated as normal oxalate.After being thoroughly washed on an asbestos filter, the precipitate may be treated with 5-10 C.C. of dilute sulphuric acid (1 : I), diluted with a suitable quantity of water, heated to boiling and titrated with permanganate, or it may be dissolved in 10 C.C. of hydrochloric acid mixed with 0.5 gram of manganous sulphate and titrated with permanganate at I n order to ensure the complete precipitation of the copper, there should be present at least 0.0128 gram of its oxide in 50 C.C. of liquid, containing as nearly as possible 5 C.C. of strong nitric acid; two grams of oxalic acid will completely precipitate the copper after 12-16 hours.Ammonium nitrate interferes with the complete precipitation. The process is not suited for the separation of copper from bismuth and antimony ; cadmium, arsenic, iron, and small amounts of tin do not interfere. I n the presence of zinc, the precipitate is sligbtly con- 30-5 0'.ANALYTICAL CHEMISTRY. 41 taminated with zinc oxalate. Numerous test analyses are given in By FRIEDRICH BULLNHEIMER (Chem. Centr., 1900, ii, 991-992 ; from Chem. Zeit., 24,870-871).- The following process is recommended for poor ores. 1-2 grams of the finely powdered sample are fus6d in a nickel crucible with 4 grams of sodium peroxide and 3 grams oE sodium hydroxide, first over a very small and then over a full flame until the bottom of the crucible begins to glow. When the mass has solidified, the crucible is placed while hot in a beaker containing some water and the solution is then transferred to a 250 C.C.flask. If the liquid is coloured green by manganese, it is decolorised by means of hydrogen peroxide. When cold, the liquid is diluted to the mark, and half of the filtrate is mixed with 20 grams of ammonium nitrate. After any silica or stannic acid has deposited, magnesium nitrate is added to precipitate any phosphoric or arsenic acid. Afher 6 hours, the liquid is filtered, the precipitate washed with ammoniacal water and the filtrate made faintly acid with nitric acid. 20-30 C.C. of mercury solution (200 grams of mercurous nitrate heated with 20 C.C. of strong nitric acid and alittle water and then diluted to 1 litre) are added, and after a few hours the liquid is nearly neutralised with ammonia.The precipitate is washed with water containing a little mercurous nitrate and then ignited and weighed. In the presence of much molybdenum, the precipitate ahould be mixed with ammonium chloride and then be reignited. the paper. L. DE K. Estimation of Tungsten in Ores. L. DE K. AnaJysis of Tin and Tin-plated Wares. By V. MAINSBRECQ, (Chem. Centr., 1900, ii, 743-744; from Rev. intern. falsijc, 13, 113--115).-Ten grams of the finely cut up sample are treated in an Erlenmeyer flask with strong hydrochloric acid and left overnight. Any undissolved matter, which may contain arsenic, antimony, and the bulk of the copper, is filtered off and treated with nitric acid, which converts the antimony into trioxide and dissolves the arsenic, which may be precipitated as ammonium magnesium arsenate. The hydro- chloric acid solution is oxidised with nitric acid, neutralised with ammonia, again slightly acidified with hydrochloric acid, and pre- cipitated with a strong solution of ammonium nitrate; the filtrate from the stannic hydroxide contains the lead and the remainder of the copper, which are precipitated as sulphides and then separated by the thiocyanate method.The filtrate contains zinc and iron, which are separated by means of ammonia, the zinc being finally weighed as sulphide. Tin and copper are, however, best determined in a separate portion. One gram is dissolved in nitro-hydrochloric acid, the tin precipitated by means of ammonium nitrate as directed, and the copper estimated in the filtrate either by potassium thiocyanate or colori- metrically. Tin-plate is simply dissolved in hydrochloric acid ; copper Sources of Lo8a in the Estimation of Gold and Silver in Copper Bars, and a Method for its Avoidance. By RANDOLPH VAN LIEW (Chm.Ccntr., 1900, ii, 992; from Eng. and Min. J., 69, 469, 498).--In order t o minimise the loss of gold and silver in is not likely to be present, L. DE K.42 ABSTRACTS OF CHEMICAL PAPERS, the assay of copper bars, the following method is proposed. One or two assay tons of the turnings is dissolved in the cold in 350 C.C. of water and 100 C.C. of nitric acid of sp. gr. 1.42. After 18-20 hours, a further addition of a t most 30 C.C. of acid must be made, and within 24-26 hours the solution will be completed. Any nitrous fumes are expelled by means o f a current of compressed air in about 20-30 minutes; sodium chloride is then added.The silver chloride, mixed with the insoluble matter containing the gold, is collected on a filter, care being taken to wash i t all down to the bottom. The cone of silver chloride is covered with 4-5 grams of assay-lead and with the filter heated at a temperature low enough to cause reduction of the silver chloride without any volatilisation of the chloride ; cupellation &c., is then carried on as usual. L. DE K. Assay of Cupriferous Materials for Gold and Silver, By L. D. GODSHALL (Chem. Centr., 1900, ii, 992-993; from Trams. Amer. Inst. Mirn. Engin., Feb., 1900 ; Berg.-Huttenm. Zeit., 59, -454).-When applying the usual dry assay, care should be taken to examine the slags and the used cupel for traces of gold and silver.When the material is very rich in copper, it will be necessary to test for and estimate it in the weighed silver buttons. The sample is dissolved in nitric acid and after being diluted with water, lead acetate and a corresponding quantity of sulphuric acid are added ; the precipitated lead sulphate carries down any metallic gold, and after being dried, the precipitate is cupelled as usual. The filtrate is precipitated with a chloride or a bromide, and lead sulphate or bromide precipitated in the solution as before, and the resulting argentiferous deposit also cupelled in the usual manner. The material may also be dissolved in nitric acid, evaporated with addition of sulphuric acid, and the resulting solution treated for a very short time with hydrogen sulphide. The precipitated copper sulphide carries down any gold and this is then recovered like the The wet assay is best conducted as follows.silver. L. DE K. Waters Contaminated by Cystine. By MARCEL MOLINI~ (Compt. rend., 1900, 131, 720--721).-The development of a yellow coloration by the action of the mercurichloride of sodium p-diazobenzene- sulphonate which is not destroyed by sulphurous acid cannot be taken as an indication of the presence of cystine in natural waters, because the same result is obtained with slightly acidified distilled water ; on the other hand, the coloration is never produced in neutral solutions (com- pare Causse, Abstr., 1900, ii, 457, 458).G. T. M. Estimation of Iodoform in Dressing Materials, By G. FRERICHS (Chem. Centr., 1900, ii, 785; from Apoth. Zeit., 15, 544. Compare Lehmann, Abstr., 1900, ii, 372, 767).-0ne or two grams of the material are put into a 250 C.C. Erlenmeyer flask and 10-20 C.C. of alcohol or “spiritus aethereus” are added. Excess of N/10 silver nitrate and 10 C.C. of dilute nitric acid are then added, and after 15 minutes the egcess of silver is titrated with N/10 potassium thiocyanate. SeveralANALYTICAL CHEMISTRY. 43 determinations should be made to see whether the iodoforrn is eq.ually Estimation of Prussian Blue in Spent Gas Purifying Material. By A. 0. NAUSS (Chm. Centr., 1900, ii, 785-786; from J. Gasbel., 43, 696--697).-Ten grams of the sample are repeatedly shaken in a half-litre flask with 50 C.C.of 10 per cent. sodium hydr- oxide. Aftor 15 hours, the mixture is diluted to 605 C.C. and filtered. Fifty C.C. of the filtrate are added to 10 or 15 C.C. of a hot solution of iron alum (200 grams to 1 litre of water and 100 grams of sulphuric acid) and the whole is heated until the peculiar sweet odour has dis- appeared; the precipitate is filtered off and washed in a hot water funnel and then heated for a short time together with the filter with a definite volume of N/50 sodium hydroxide. The excess of alkali is then titrated with N/50 acid until a permanent pale greenish-yellow coloration is observed. To estimate cyanogen in gases, 100 litres of the gas are passed through an absorption Bask containing 20 C.C. of ferrous sulphate (1 : 10) and 20 C.C.of potassium hydroxide (1 :3) and then through a second flask containing 10 C.C. of the mixture. The contents of the flasks are emptied into a half-litre flask, a little more potassium hydroxide and ferrous sulphate, and 1 gram of lead carbonate, to remove hydrogen sulphide, are added, and the whole is boiled for a few minutes. When cold, it is made up to 505 C.C. and 50 C.C. are titrated as directed. One c.e. of N/50 alkali =0*0007794 gram of cyanogen. Detection of Methyl Alcohol in Mixtures. By S. P. MULLIKEN and HEYWARD SCUDDER (Amer. Chem. J., 1900, 24, 444-452. Compare Abstr., 1899, ii, 3S8).-Jandrier (Abstr., 1900, ii, 52) has recommended the use of gallic acid instead of resorcinol for the detec- tion of methyl alcohol, since he found that the colour reaction of the former reagent with formaldehyde is much more delicate than that of the latter, and is less likely to be obscured by coloration occasioned by the simultaneous presence of other aldehydes.The authors find that this modification of their test for methyl alcohol is inadmissible, since many other organic substances yield traces of formaldehyde on oxidation, which would not be detected by resorcinol but give the colour reaction with gallic acid owing t o its greater delicacy. A number of precautions are mentioned which should be adopted before applying the resorcinol test, in order to remove substances from the solution which could prevent or obscure the appearance of the characteristic rose-red ring, and a method of applying the test in the presence of much ethyl alcohol is described in detail. E.G. By R. GRASSINI (L’Orosi, 1900, 23, 224-225 and 297-298).-1f a dilute solution of cobalt chloride, the concentration of which is not less than 0.5 per cent. and is best about 5 per cent., is mixed with potassium thiocyanate solution and a layer of alcohol is poured on to the liquid, which is then gently shaken so that the two strata do not mix, the alcohol assumes a beautiful agure-blue colour which i s most intense at the surface of separation. distributed over the whole of the material. L. DE E(. L. DE K. Colour Reaction of Alcohol.44 ABSTRACTS OF CHEMICAL PAPERS. This eeaction is not disturbed by the presence of nickel salts except when the ratio of cobalt to nickel is extremely small; the coloration, is, however, destroyed by hydrogen peroxide.If ether is used in place of alcohol, no coloration is obtained, so that the reaction may be used as a means of detecting the presence of alcohol in ether, which, if it contains 5 per cent. of added alcohol gives a marked coloration. T. H. P. Estimation of m-Cresol in Cresol Mixtures. By HUGO DITZ (Zeit. angew. Chem., 1900, 1050--1052).-Raschig’s process (Abstr., 1900, ii, 694) is not applicable when the mixture contains more than 10 per cent. of phenol. The sample should be suitably diluted with a mixture containing no phenol but a known quantity of m-cresol. In Rascbig’s process, the o-cresol and p-cresol are oxidised to oxalic acid ; should this reaction be quantitative, it might prove of great value.If, however, phenol is entirely absent, t,he author thinks that the assay may be more conveniently effected by the bromine process proposed by himself and Cedivoda (Abstr., 1900, ii, 54). L. DE I(. Detection of Starch-Sugar in Wine. By ED. DELLE (Chern. Centr., 1900, ii, 744; from Rev. imttwm. faZs$c., 13, 131-132).- Tony-Garcin’s process is recommended. One hundred C.C. of the sample are shaken with 30 grams of animal charcoal and, after 2 hours, fihered. Normally fermented wine then shows a rotation of about + 13’. If the decolorised wine contains less than 2 grams of reducing sugar per Iitre and has a rotation of more than + 13’ before and after in- version, the presence of starch-sugar is proved. I f more sugar is Estimation of the true percentage of Starch in Potatoes.By GEORG BAUMERT and H. BODE [with A. FEST] (Zeit. angeur. Chern., 1900, 1074-1078, 11 11-11 13).-After reviewing earlier processes, the authors have finally adopted the following method : 3 grams of the finely ground, air-dried potato are stirred at intervals in a porce- lain beaker with 50 C.C. of cold water for an hour, the liquid is poured off as completely as possible through a small asbestos filter, and the latter returned to the beaker. Fifty C.C. of water are added, the beaker is closed with a lid and heated for three hours and a half in a Soxhlet digester a t a pressure of 3 atmospheres. After cooling, the contents of the beaker are rinsed with 150-200 C.C. of boiling water into a 250 C.C. flask and boiled for 10 minutes ; when cold, the liquid is made up to the mark and filtered.100-200 C.C. of the filtrate are mixed with 10-20 C.C. of aqueous sodium hydroxide (100 grams per litre) and from the clear and stable solution, 25 C.C. are taken and mixed with 100 C.C. of commercial absolute alcohol, about 1 gram of fine asbestos being added in order to facilitate the separation of the starch. The precipitate is now collected on a Soxhlet asbestos filter- tube and washed by means of the suction-pump with 80 per cent. alcohol. The starch is then dissolved by means of about 3-5 C.C. of 5 per cent. hydrochloric acid, reprecipltated with 25-30 C.C. of com- mercial absolute alcohol and again collected on the filter ; it is then present, the test becomes uncertain. L. DE I(.ANALYTICAL CHEMISTRY.45 washed first with 80 per cent. alcohol, then with absolute alcohol, and finally with ether. After drying to constant weight a t 120-130’ in a current of air, the tube is weighed; it is then ignited in a current of air or oxygen and reweighed, when the loss represents the amount of starch. On account of the dilution caused by the alkali, it should be multiplied by 1.1. I n order t o obtain the percentage of starch in the raw potatoes, the following formula is applied, 8 = lOO(v x 1*1)/Om3 x Zp, in whichf repre- sents the weight of the raw potatoes, I the weight of the air-dried mass, and w the loss on ignition, Further experiments have con- clusively shown that the loss on ignition really represents pure starch. L. DE K. New Process for the Estimation of Glycogen.By GEORU LEBBIN (Chern. Centr., 1900, ii, 880-881 ; from Zeit. ofemtl. Chem., 6, 325--327).-Twenty grams of horse flesh or liver are treated in a porcelain dish with 90 C.C. of water and 10 C.C. of 15 per cent. potassium hydroxide and heated over wire gauze until completely dissolved, when the whole is evaporated down to 30-35 C.C. After diluting to exactly 50 or 100 C.C. and filtering through glass wool, 25 C.C. are mixed with 50 C.C. of a mixture of 90 parts of alcohol and 10 parts of 40 per cent. potassium hydroxide. The precipitated crude glycogen is left over- night, collected on a filter and washed with alkaline alcohol. The filter is perforated and the contents are washed into a 100 C.C. flask with 80 C.C. of boiling water. The flask is well shaken, and when cold the contents are neutralised with 10 per cent, hydrochloric acid using tincture of litmus as indicator. After adding 3-4 more drops of the acid, 5-10 C.C.of Briicke’s reagent are added, the liquid diluted to the mark and filtered. Fifty c.c.of the filtrate are mixed with 75 C.C. of 95 per cent. alcohol, the precipitate is collected the next morning on a tared filter, and washed first with alcohol and then with ether, After being dried and weighed, the filter is burnt and any ash allowed for. L. DE K. Estimation of Acetic Acid. By V. DELFINO and M. MIRANDA (Chem. Centr., 1900, ii, 1038; from Mon. Xci., [iv] 14, ii, 696).- The solution of the acetic acid (? acetate) is boiled with an excess of ferric chloride, The resulting ferric acetate is decomposed into acetic acid and ferric hydroxide which then becomes anhydrous and ad- heres firmly to the sides of the beaker; it is dissolved in warin sul- phuric acid, and reduced by means of a strip of silver foil.After removing the silver from the solution by means of hydrochloric acid, the iron is titrated as usual. L. DE K. Acidimetric Estimation of Protocatechuic Acid. By HENRI IMBERT (Bull. Xoc. Chim., 1900, [iii], 23, 832--834).-According to Massol (Abstr., 1900, i, 600), protocatechuic acid crystallises in yellow needles melting at 199’ and is neutralised by 1.5 mols. of alkali when phenolphthalein is used as the indicator. The author now describes experiments to confirm his former statement (Abstr., 1900, i, 226), that protocatechuic acid is monobasic towards phenolphthalein, although46 ABSTRACTS OF CEEMICAL PAPERS.the end point is somewhat obscure. colourless, as it is usually stated to be, and melted at 194-195’. Moreover the acid employed was N. L. New Method for the Detection of ‘‘ Saccharin ” and Salicylic Acid alone or in presence of each other, By E. RIEGLER (Chem. Centr., 1900, ii, 880 ; from Pharm. Centr.-Halle, 41,563-564). -0-01-0°02 gram of “saccharin” is dissolved in 10 C.C. of water and 2 drops of 10 per cent. sodium hydroxide and put into a separ- ating funnel. Solution of p-diazonitroaniline is added drop by drop with constant shaking until the greenish-yellow colour of the liquid has disappeared. After shaking with 10 C.C. of ether, the aqueous layer is removed, and 20-30 drops of 10 per cent. sodium hydroxide are added.At the place of contact between the ether and the soda, a fine green ring is formed; on shaking, the ether turns green and the alkali yellowish-brown. On removing the aqueous layer and adding 5 C.C. of strong ammonia, this turns a beautiful green, whilst the ether is decolorised. Under the same circumstances, salicylic acid first shows an intensely red ring, then a red aqueous but colourless ethereal layer which remains colourless when treated with ammonia, the latter turning red. If 0*02-0.03 gram of a mixture of I‘ saccharin ” and salicylic acid is treated in the same manner, there is obtained first a red, somewhat brownish, ring, then a green ethereal, and a red aqueous solution; on addition OF ammonia, a colourless ethereal and a violet ammoniacal layer.The colour of the latter varies i n shade according to the re- spective amounts of the two substances. The reagent is prepared by dissolving 2.5 grams of p-nitroaniline in 25 C.C. of water and 5 C.C. of sulphuric acid. Twenty-five C.C. of water and a solution of 1.5 grams of sodium nitrite are added, and the Estimation of 6cSaccharin” in Beverages. By ED. DELLE (Chem. Centr., 1900, ii, 744-745; from Rev. intern. falsijic, 13, 121--122).-1n the absence of salicylic acid, 50-100 C.C. of tbe liquid free from alcohol are acidified with phosphoric acid and shaken with ether, or better with a mixture of ether and light petroleum. The ethereal liquid is evaporated, and the residue carefully fused with potassium hydroxide. The fused mass is dissolved in water, the liquid acidified with acetic acid, and the salicylic acid formed extracted with benzene.The weight of the salicylic acid multiplied by 0.77 equals the (‘ saccharin.’’ It may also be estimated colorimetrically by means By PAUL VIETH and M. SIEGFELD (Chem. Centr., 1900, ii, 922; from Milch Zeit., 29, 593--597).-As the result of a long series of experiments, the authors state that the natural acidity of milk as found by titration with barium or sodium hydroxide with phenolphthalein as indicator is subject to serious variation. Determination of the Specific Heat of Fats. By VANDEVYVER- GRAU (Chem. Centr., 1900, ii, 923-924 ; from Ann. Chim. anal. up#, 5, 321--323).-Pure tristearin has, a t - 6*, a fairly constant specific whole is diluted to the mark. L.DE K. of ferric chloride. L. DE K. Acidity of Milk. L. DE K.AN ALP TICAL CHEMISTRY. 47 heat of 0.336. This increases at lo", ZOO, 304 and 40' to 0.397, 0.409, 0,449, and 0.501, and is at 50' 0.510. A t 50" to 55', the edges of small particles of tristearin begin to soften, and the specific heat quickly rises t o 1.3-1.4. Palmitin shows a similar curve, the specific heat varying from 0.330 at - 7" to 0.478 at + 60°, and more than 1 at 3-4' below the melting point, 66.5'. Other fats behave in the same manner. The author considers the determination of the specific heat of fats, if not impossible, at least extraordinarily difficult, as the heat Estimation of Fat in Finely Powdered Substances, parti- cularly in Cocoa and Cocoa Mixtures. By P. WELMANS (Chem.Centr., 1900, ii, 786-787; from Zed. ofentl. Chem., 6, 304-314).- Five grams of the material are put into a separating funnel and shaken with 100 C.C. of ether saturated with water for a few minutes ; 100 C.C. of water saturated with ether are then added, and the whole is again vigorously shaken until a complete emulsion has formed. When ths ether has sufficiently separated, 50 or 25 C.C. are pipetted off, the ether is distilled off, and the residual fat weighed. I n calculating, it must be remembered that the total volume of the ethereal layer is not 100 c.c., but 100 C.C. plus the volume of the f a t ; the sp. gr. of cocoa fat may here be taken as 1. Bonnema's tragacanth method (Abstr., 1899, ii, 822) is also service- able if due correction is made for the volume occupied by the fat.of fusion causes an elevation of temperature. L. DE K. L. DE I(. Absolute Iodine Absorption Number of Fats. By MASSJMO TORTELLI and R. RUUGERI (Annali del Lab. Chim. Centr. delle Gabelle, 1900,4, 191-204; L'Orosi, 23, 109--122).-The authors ascribe the varying values obtained for the absolute iodine absorption number of one and the same fat by Wallenstein and Finck's method (Chem. Zeit., 1894, 18, 1188-1191), to (1) the use of too small a quantity of mate- rial, and (2) a slight oxidation of the liquid fatty acids. These dis- turbing influences are avoided in the following process. Twenty grams of the oil or fat are saponified with alcoholic potassium hydroxide and afterwards converted into the lead soap, which is washed with 200 C.C. of hot (not boiling) water three times in succession, allowed to cool, dried between filter paper and heated in a reflux apparatus on a water- bath with 220c.c.of ether for 20 minutes. The flask is then cooled in a mixture of ice and water, and maintained at a temperature of 8-10' for 2 hours, after which the liquid is filtered through a pleated filter into a 200 C.C. flask, made up to volume with pure ether, and the stoppered flask left all night in flowing water. The clear ethereal solution, filtered if necessary, is next removed t o a separating funnel in which the soluble lead salt is decomposed by 150 C.C. of 20 per cent. hydrochloric acid; after vigorous shaking, the two layers are allowed to separate, and the lower aqueous liquor, together with the lead chloride, run off from the funnel. This is repeated with another 100 C.C.of the acid, after which the ethereal solution is washed twice with 150 C.C. of distilled water, filtered, and distilled u r t i l 40-60 C.C. remain. The residue is then placed in a 100 C.C. flask immersed to its neck in a water bnth which is heated to boiling while48 ABSTRACTS OF CHENICAL PAPERS. n current of carbon dioxide is passed through a narrow tube bent at a right angle and drawn out nearly to a point, and reaching almost to the bottom of the flask. After an hour’s heating the flask is removed from the bath, and 8-15 drops (according as the oil contains more or less cotton seed oil or other drying oil), neglecting the first few, aspirated into a tared flask, and tho iodine number determined in the usual way.A large number of fats and oils have been examined by this method, the values obtained being greater, and nearer to the theoretical values, than those given by any process previously proposed. T. H. P. Solidiflcation Point of Fatty Acids. By R. MORESCHIN~ (Anmli delLa6. Centr. delle Gabelle, 1900,4,293-302. Compare Abstr., 1900, ii, 465).-The true solidifying point of a mixture of fatty acids obtained from soap may be found as follows. A test-tube 22-25 mm. wide is fixed by means of a stopper in a cylindrical glass vessel containing air having an initial temperature about 1 5 O lower than the required solidi- fying point, About 20 grams of the mixed acids are fused and poured into the test-tube, the change of temperature being noted. As soon as the rate of cooling shows signs of slackening, the mass is subjected to vigorous stirring, this being kept up as long as the temperature con- tinues to fall.After a time the mercury begins to rise and reaches a maximum a t which it remains motionless for some two minutes ; this maximum represents the true solidification point. By THEODOR GEUTHER (Chem. Centr., 1900, ii, 879-880; from Zeit. ofentl. Chem., 6, 328--329).-The author prepares the reagent by pouring 25 C.C. of water on to 5 grams of powdered sodium phosphotungstate and then adding a t once 30 C.C. of pure nitric acid of sp. gr. 1.39. This reagent keeps for about a year. To apply the test, 5 grams of the filtered lard are put into a tared test-tube, 3 grams of pure chloroform are introduced, and then 20 drops of the reagent.After thoroughly shaking, the tube is put aside and the coloration observed within two minutes ; colora- tions forming after two minutes should be ignored. The presence of only 5 per cent. of vegetable oil betrays itself by a plain, dark-green coloration. A yellowish-green coloration may be due to stale or rancid By ALB. J. J. VANDEVELDE (Chem. Centr., 1900, ii, 783 ; from Bull. Assoc. Belg. des Chimistes).-A criticism of Breinl’s process (Abstr., 1899, ii, 824). Light petroleum should on no account be used in this test as it also gives the reaction, particularly if it has a high boiling point. It follows that the reaction is not due to nitrogenous matters as suggested by Breinl. T. H. P. Modiflcation of Welman’s Reaction. lard.L. DE I(. Breinl’s Reaction for Sesame Oil. L. DE K. Modification of Hiibl’s Method of Estimating the Acid and Saponification Numbers of Wax, By OTTO EICHRORN (Zeit. anal. Chem., 1900, 39, 640--645).-The difficulty of saponifying wax, especially Russian wax, may be obviated by dissolving it in amyl alcohol. For estimating the free acids, 6 grams of wax are coveredANALYTICAL CHEMISTRY. 49 with 60 C.C. of pure amyl alcohol in a conical flask, and heated on an asbestos plate to boiling. Phenolphthalein is then added, and the titration made with N/10 alcoholic potash, which is added until the red colour disappears slowly. The liquid should be kept warmienough to prevent separation of the wax, and the titration should be finished at about 60°, higher temperatures being avoided near the end to prevent risk of saponification, The acid values so obtained are lower than those hitherto accepted.For the saponification, 5 grams are dis- solved in 60 c . ~ . of amyl alcohol, mixed with 25 C.C. of N alcoholic potash, the mixture gently boiled for half an hour, and titrated back with N/2 hydrochloric acid, after adding phenolphthalein. As soon as the solution has become colourless, the flask is heated until the liquid separates into two layers, and tbe addition of acid is continued until the lower, aqueous layer remains yellow even on long heating. A little water (10-25 c.c.) may be added to render the change more distinct. A blank experiment, conducted in the same manner, is indispensable. The presence of ceresin or paraffin does not interfere in any way.Analysis of Fruit Essences. By GUIDO FA~RIS (Anwli del Lab. Chirn. Centr. dells Gabelle, 1900, 4, 41--141).-A large number of experiments have beenmade on the methods of detecting the com- mon adulterants of oil of lemons. The results show that there is no absolutely certain way of determining the presence of extraneous sub- stances, although much can be learnt by the methods proposed by Schimmel & Co. and by Soldaini and Bert& which make use of frac- tional distillation and measurements of specific rotation. Other tests such as colour reactions and determination of physical constants, which are useful in some cases, are also considered. The addition of oils, resins, or balsams to oil of bergamot increases both its density and the residue left on evaporation, whilst oil of turpentine produces a diminution in the rotatory power and in the amount of esters.The admixture of lemon or orange oil with oil of bergamot increases its specific rotation and lowers its density, the amount of esters and the residue left on evaporation. Methods for examining sweet and bitter orange oils and oil of man- darins are also referred to. Valuation of Lemon Oil. By J. WALTHER (Chem. Centr., 1900, ii, 903; from Pharm. Centr.-H,, 41, 585-588. Compare Abstr., 1900, ii, 173, and Schimmel & Co.’s beschaytsber., April, 1900).- I n the author’s method of testing lemon oil, the sodium hydrogen carbonate should be added to the mixture of lemon oil and hydroxyl- amine hydrochloride in fine powder, so that it is gradually dissolved during the digestion and decomposes the hydrochloride; 80 parts of the carbonate are required for about 70 of hydroxylamine hydrochlor- ide. The end reaction with methyl-orange is described in the original By J.WALTHER (Chem. Centr., 1900, ii, 970 ; from Pharm. Centr.-H., 41, 613-616). -The quantity of carvone in ethereal oils is estimated by a method similar to that used for lemon oil (Abstr., 1900, ii, 173) 2-5 grams M. J. S. T. H. P. paper and numerous analyses are quoted. E. w. w. Estimation of Carvone in Ethereal Oils. VOL. LXXX. ii. e(50 ABSTRACTS OF CHEMICAL PAPERS. of the oil are mixed with 10 grams of a freshly-prepared solution of hydroxylamine hydrochloride (2 : 3), 25 C.C. of alcohol free from alde- hyde, and 2 grams of sodium hydrogen carbonate, and the mixture gently boiled for half an hour.To the cold solution, 6 C.C. of hydro- chloric acid of sp. gr. 1.12 are added, the volume made up to 500 c.c., and the unused hydroxylamine determined in 25-50 C.C. of the fil- tered liquid. The standard solution of alkali should be free from carbon dioxide. Three samples of caraway-seed oil of sp. gr. 0.913 a t 1S0, 0.907 a t 19O, and 0.893 at 15' were found to contain 52*5,48*5, and 10.0 per cent. of carvone respectively; a curled-mint oil of sp. gr. 0.932 at 18' contained 51 per cent. and another of sp. gr. 0.887 at 20°, 16 per cent. of carvone. Acid and Saponiflcation Numbers of some Copals. By WALTHER LIPPERT and H. REISSIGER (Zeit. angew. Chem., 1900, 1047--1050).-The following varieties of copal have been tested ; Angola copal, Angostura copal, Benguela copal, Brazil copal, Cameroon copal, Kauri gum, Dammara resin, and Madagascar, Zanzibar, Manilla, and Sierra Leone copals.The results have been duly tabulated. One gram of the finely powdered copal is treated for a few hours with 10 C.C. of chloro- form, 25 C.C. of alcohol are added, and the whole is gently warmed; if no solution takes place, a mixture of equal parts of ether and alcohol should be tried. Dichlorohydrin, which has been proposed as a universal solvent for resins, does not answer for copals. When cold, the acidity is taken with N/2 alcoholic potash using phenolphthalein as indicator. The saponification number was estimated in the usual manner by heat- ing with excess of standard alcoholic potash ; the use of chloroform is, of course, excluded here.Henriques' cold saponification process does Detection a n d Estimation of '' Dulcin " (Phenetolecarb- amide) in Articles of Food. By J. BELLIER(Chem. Centr., 1900, ii, 822; from Ann. Chim. anal. appl., 5, 333-337).-Lemonades or syrups, after dilution with water, are rendered alkaline with ammonia and shaken with ethyl acetate. In testing wine, 200 C.C. of the sample are treated with 2 grams of mercuric acetate and a little ammonia, and the filtrate is then shaken with 50 C.C. of ethyl acetate. When dealing with beer, 200 O.C. are mixed with 2-3 grams of sodium phosphotung- state and 10--201 drops of sulphuric acid, and the filtrate is shaken with 50 C.C. of ethyl acetate, after rendering alkaline with ammonia.I n either case, the solution in ethyl acetate is evaporated to dryness and the residue dissolved in a little sulphuric acid and mixed with a few drops of strong solution of formaldehyde. On adding water, a flocculent precipitate is formed which after 24 hours may be collected, washed, dried, and weighed ; its weight represents that of the '' dulcin." The Constituents of Digitalis and their Toxicological De- tection. By DIOSCORIDE VITALI (Chem. Centr., 1900, ii, 881 ; from Boll. Chim. Farm., 39, 597-602).-When testing meat which bad E. W. W. The acid number is estimated by direct titration. not work well with copals. L. DEK. It may be further identified by Jorissen's reaction. L. DE K.ANALYTICAL CHEMISTRY. 51 been purposely poisoned with digitalis, the author was only able to detect digitoxin.The meat is extracted with dilute alcohol, the alcohol evaporated, and the residual liquid treated first with lead acetste and then with sodium sulphate. The filtrate, after being ren- dered alkaline with ammonia, is shaken with chloroform, and from this the digitoxin is precipitated by adding 20 times its volume of a mixture of 1 part of ether and 7 parts of light petroleum. After a further purification with alcohol and ether, the digitoxin gives the re- action with Keller and Kiliani’s reagent very distinctly. L. DE K. Isolation and Estimation of Colchicine. By ALBERT B. PRES- COTT and HARRY M. GORDIN (Chem. Centr., 1900, ii, 784-785; from Apoth. Zeit., 15,521 -522).-Twenty-five grams of powdered colchicum corm or seeds are extracted in a Soxhlet apparatus for 2 hours with 35 per cent.alcohol ; the alcohol is distilled off, the oily residue rinsed with about 10 C.C. of hot water in a small separating funnel, and then shaken with 2-3 C.C. of light petroleum. After 15-20 minutes, the oily layer floats on the aqueous liquid. Without filtering, the latter is drawn off into a 100 C.C. flask, the stem of the funnel is rinsed with a little water, and the oily matter is again shaken with 10 C.C. of water. The whole is now put into a small dish, the light petroleum is removed by gently warming, and the contents are again transferred to the funnel and once more shaken with 2-3 C.C. of light petroleum ; by this repeated treatment, the whole of the alkaloid is removed from the oil.The turbid aqueous solution is diluted to 100 c.c., 1-2 grams of talc are added, and after shaking it is passed through a dry filter. The alkaloid may be estimated gravimetrically by first removing the water by means of a current of air, and finally drying the residue in a vacuum or over sulphuric acid. Volumetrically, it may be estimated by boiling it for 2 hours in a reflux apparatus with N/40 potassium hydroxide, and titrating the excess of the latter with N/40 hydro- chloric acid. L. DE K. Analysis of Quinine Glycerophosphate. By LEON PRUNER (J. I’harm., 1900, [vi], 12, 309-312).-The phosphoric acid in quinine glycerophosphate cannot be estimated by calcination and subsequent oxidation with nitric acid, as the carbon produced during calcination reduces the phosphoric acid and a loss of phosphorus occurs.Quinine is not completely precipitated from quinine glycerophosphate by alkalis in the cold. The following method of analysis gives accurate results :-the quinine glycerophosphate is dissolved in dilute nitric acid (1 : lo), a solution of sodium hydroxide added and the precipitated quinine filtered off. The filtrate is boiled for some time, when the re- mainder of the quinine is precipitated and may be removed by extraction with ether or chloroform. The mother liquor from the ethereal or chloro- form extraction is then mixed with a large excess of nitric acid together with 2-3 grams of potassium nitrate, evaporated t o dryness and calcined to a white ash, care being taken that sufficient nitric acid is present to avoid the presence of any free carbon during calcination.The phosphoric acid present in the residue may then be estimated by any of the usual methods. H. R. LE S. 4-252 ABSTBACTS OF CHEMICAL PAPERS, Analytical Chemistry of the Alkaloids. IV. Action of Bromine on Strychnine and Brucine. By EABL KIPPENBERGER (Zeit. and. Chem., 1900, 39, 609-627).-By acting on an acid solu- tion of a strychnine salt with an excess of bromine water, a gradual substitution of bromine for hydrogen takes place ; the monobromo- derivative is formed almost instantly, in the course of an hour di- and tri-bromostrychnine are formed, whilst the substitution of more hydrogen atoms only takes place after several hours, and in the course of 12 hours 16-17 atoms of bromine are taken up, with formation of strychnine perbromide. The presence of sodium chloride greatly ac- celerates the reaction. When moderate amounts of potassium bromide are present, yellow, nearly insoluble, monobromostrychnine hydrobromide perbromide, C,1H,20,N,Br,HBr,Brz, is rapidly formed together with a little of the dibromo-compound ; the presence of larger amounts of potassium bromide restricts the internal substitution. I n no case could more than 2 atoms of bromine be introduced into the molecule. Brucine forms a similar hydrobromide perbromide, but in conse- quence of its solubility in water, especially in the presence of acids, being greater than that of the strychnine compound, there is a greater tendency for the exterior atoms of bromine to effect substitution in the brucine molecule. The presence of both sodium chloride and potassium bromide favours the additive and hinders the substitutive reaction. The bromobrucine compounds exhi bit a rose-red colour. How- ever large the excem of bromine used, no perbromide containing more than two atoms of additive bromine could be obtained. Brucine salts behave in a similar manner. M. J. S. New Test for Gelatin and Isinglass. By OTTOMAR HENZOLD (Chem. Centrr., 1900, ii, 744, from Zeit. ofentl. Chem., 6, 292-293).- Fruit jellies in which gelatin is suspected are boiled with water, the hot liquid is filtered, and a portion of the filtrate is mixed in a test- tube with an excess of solution of potassium dichromate (1 : 10). The mixture is again heated to boiling, then rapidly cooled and mixed with 2-3 drops of sulphuric acid. Gelatin gives a white, finely divided precipitate which becomes sticky after some time ; vegetable jellies such as agar-agar, Iceland moss, and carrageen give no pre- Artiflcial Silk and its Distinction from the Natural Product. By A. SOLARO (Annali del Lab. China. Certtr. delle Gabelle, 1900, 4, 503--527).--Chemical and microscopical methods are given for dis- tinguishing natural silk from the artificial preparations having a basis of nitrocellulose or gelatin. cipitat e. L. DE K. T. H. P. Estimation of Humus. By CHARLES RIMBACH (J. A w . Chew, Soc,, 1900, 22, 695-703),-See this vol., ii, 37.