THE ANALYST. 193 ORGANIC ANALYSIS. Further Notes on the Simplified Method of Ultimate Combustion Analysis. M. Dennstedt. (Bel-. deut. Chem. Ges., 1908, 41, 600-604.)-The author gives a minute description of the manner in which his improved combustion apparatus should be worked when dealing with any substance of unknown properties. This general method is not so rapid as the special methods applied to known substances, but it is often necessary to use it. When the substance in the insertion tube has been placed in the combustion tube, the stream of oxygen is adjusted, and the194 THE ANALYST. oombustion flame is gradually advanced under the contact material until the back portion of the platinum next to the insertion tube begins to glow. The roof of the furnace is placed so that it covers not more than 1 cm.of the insertion tube. The volatilisation flame is then turned up as far behind the substance in the boat as possible ; only very volatile substances are so far affected, most substances remain unchanged. When the contact material is glowing brightly, the volatilising flame is advanced about 1 cm. every three or four minutes, until the substance melts or deoomposes; it is then allowed to remain stationary for fifteen minutes, provided signs of combustion are visible. If not, the combustion flame, with the cover, is advanced 1 or 2 mm. at a time towards the boat, until combustion begins. When combustion slackens down the flame is advanced a little more, until the substance is either volatilised or oarbonised. The adjustment of the combustion flame and cover must be such that the platinum in front of the insertion tube is maintained at a bright red heat.The volatilising flame is only brought further forward when no more combustion can be obtained with the combustion flame. Finally the cover is placed over the volatilising flame, and the stream of oxygen is increased in order to complete the combustion of the carbon. The calcium chloride used for absorbing the water of combustion should be gently ignited before filling the absorption tubes, and the latter should then be saturated with carbon dioxide. The soda-lime used for absorbing the carbon dioxide of combustion may often be too dry for its purpose. When very high temperatures are necessary for the combustion, the fusion of the end of the insertion tube to the walls of the combustion tube may be prevented by winding a few turns of platinum wire round the former.In the determination of sulphur care must be taken to wash out not only the insertion and combustion tubes, but also the boat after the combustion, since small quantities of sulphuric acid are retained by them. The boat should also be extracted with hydro- chloric acid before commencing an estimation of sulphur. (Cf. ANALYST, 1908,146.) J. F. B. New Volumetric Method for the Estimation of Albumin by Phospho- Tungstic Acid. Tsuchiya. (Zentralbl. inn. Med., 1908, 29, 105 ; through Chenz. Zeit. Rep., 1908, 32, 149.)-The reagent used consists of alcohol (96 per cent.) containing 1.5 per cent. of phospho-tungstic acid and 5 per cent.of concentrated hydrochloric acid. I t is stated to give no precipitate with normalurine, to give more accurate results than Esbach’s reagent, to give approximately correct results even with large quantities of albumin, and to give a precipitate containing approximately the same percentage of nitrogen 8s that contained in the albumin present in urine. A. G. L. The Quantitative Conversion of Aromatic Hydrazines into Diazonium Salts. F. D. Chattaway. (Proc. Chem. SOC., 1908, 24, 74-75.)-AAll primary aromatic hydrazines can be quantitatively converted into the corresponding diazonium salts either by chlorine or by bromine. If either halogen be allowed to act at a low temperature on the hydrazine dissolved in alcohol, the diazonium salt separates out in a pure state.Unless, however, the solid salt is required, theTHE ANALYST, 195 operation can be most easily carried out by dissolving the hydrazine in glacial acetic acid, cooling the solution to about -15" C. by the addition of crushed ice, and either passing in a rapid stream of chlorine or adding the calculated quantity of bromine dissolved in acetio acid and similarly cooled by ice. The formation of the diazonium salt is without doubt of two atoms of hydrogen in the hydrazino group by elimination of halogen acid, thus : C8H, .N.H C,H,.N.CI H.N.Cl -3 I -+ I H.N.H effected. by a substitution halogen, followed by the C,H,.N.CI Ill N As in other cases where hydrogen attached to nitrogen is replaced by halogen, an additive product is probably first formed, in this instance by the addition of four halogen atoms, each nitrogen atom becoming quinquevalent ; the elimination of halogen acid then gives rise to the diazonium salt, which, being stable under the conditions of the experiment, can be isolated.Determination of Benzene in Illuminating Gas. L. M. Dennis and E. S. Mecarthy. (Jozirn. Amer. Chem. Soc., 1908, 30, 233-247.)-The ammoniacal nickel nitrate reagent proposed by Dennis and O'Neill was found to be inefficient as an absorbent of benzene, except in those cases where the gas also contained cyanogen, and this led to the preparation of ammoniacal nickel cyanide. This reagent quantitatively removes benzene from its admixture with air, and completely removes it from coal-gas in two minutes, while it does not absorb ethylene or other constituent of coal-gas, with the exception of those gases soluble in caustic potash.These should, of course, be first removed before treating with the nickel cyanide reagent. The reagent, which is capable of absorbing four times its volume of benzene, is prepared by adding 25 grams of potassium cyanide, dissolved in 40 C.C. of water, to 50 grams of cryatallised nickel sulphate, dissolved in 75 C.C. of water; 125 C.C. of ammonia (specific gravity 0.910) is then added to the mixture, which is shaken until the nickel cyanide is dissolved, and then cooled to 0" C. After twenty minutes the separated crystals of potassium sulphate are removed, and 18 grams of citric acid, dissolved in 10 C.C. of water, added to the clear solution. After ten minutes at 0" C.the greenish-blue supernatant liquid is decanted, and shaken in a Hempel gas pipette with 2 drops of liquid benzene for two minutes. This addition of benzene is advisable, since it was found that the freshly prepared reagent does not absorb well until it has been used a few times. The absorptions should be carried out in the Hempel pipette used for treating heavy hydrocarbons with fuming sulphuric acid, but it is better to have the upper bulb, which is filled with the broken glass, of a diameter of 4.6 cm. The gas to be examined should be repeatedly passed into this pipette containing the reagent, and back again, for two minutes, when the ammonia vapour left in the residual gas is absorbed by similar treatment with 5 per cent. sulphuric acid. The insoluble compound formed in the reaction, Ni(CN),.NH,.C,H,, should be filtered out after the reagent has been in use for some time, to prevent clogging of the pipette.The authors state that Morton's method for the estimation of benzene by196 THE ANALYST. means of strong sulphuric acid (ANALYST, 1907, 32, 61) gives inaccurate results, frequently failing to completely remove benzene, while absorbing a variable but considerable amount of ethylene. A. R. T. The Estimation of Phenols in Gas Liquors. F. W. Skirrow. (Jouriz. Gaslight, 1908, 100, 357-360.)-Cf. ANALYST, 1908, 134. For the colorimetric estimation of phenol in the gas itself, 50 to 60 litres of the sample are filtered through cotton-wool to remove tar-fog, and passed through two flasks containing 500 C.C. of 20 per cent.sodium hydroxide solution. The liquid in the flasks (after removal of cyanogen as iron ferrocyanide) is titrated with iodine and thiosulphate, and diluted until the red colour of the tri-iodophenol is only just visible, the solution then containing 0.00025 per cent. of phenol. The tar-acids in gas liquors were separated into (a) phenol, ( b ) cresols and crystal drainings, and (c) residues of high boiling-point. The original tar contains more cresol than phenol, but in the washing out with arumoniacal liquor, more of the phenol is dissolved. The presence of homologues of phenol introduces an error in the estimation of the phenol, the results being about 5 per cent. too low. Deter- minations of the oxygen absorption by means of permanganate, as in the method used in sewage works analyses, showed that 1 gram of phenol absorbed 1,401 grams of oxygen, 1 gram of cresol 1,245 grams of oxygen, and 1 gram of the average tar- acids from gas liquor 1,343 grams of oxygen.C. A. M. Determination of the Solidifying - Point of Commercial Paraffin. R. Kissling. (Chenz. Rev. Fett- 21. Hnrx-Ind., 1908, 15, 46-49.)-1t is here shown that in the determination of the solidifying-point by Shukoff's method considerable C T ~ O ~ S may result unless the conditions are invariably identical. Thus the results are influenced by such factors as the method and duration of shaking. If, for example, the tube be shaken and then allowed to stand, the thermometer becomes constant after a certain period of initial shaking, but the solidifying-point differs from that obtained by alternately shaking the tube and allowing it to stand until the paraffin froths and the thermometer becomes constant.To obviate such sources of error, the author has devised the following modification of the method : The Dewar's vacuum tube containing the melted paraffin is immersed to within about 10 mm. of the top in water a t a temperature about 5" C. above that of the solidifying-point expected, and as soon as the temperature of the paraffin is the same as that of the water the tube is vigorously shaken. When frothing of the paraffin takes place, the tube is allowed to stand, and the temperature taken from minute to minute, until three or more reading8 are identical, the mean solidifying-point being thus obtained. Owing to the influence of the varying proportions of the different constituents, the temperature curves of paraffins having the game mean solidifying-points may vary considerably. C.A. M. The Estimation of Tarry Substances in Mineral Oils. E. Lecocq. (Bull. SOC. Clzirn. Belg., 1908, 22, 81-87.)-Holde's method of extracting tarry substances from mineral oils by means of petroleum spirit or a mixture of 4 partsTHE ANALYST. 197 of alcohol with 3 parts of ether (cf. ANALYST, 1903, 28, 154, 299) has given good results in the author's hands, but in the case of cylinder oils containing paraffins the residual tarry matter frequently contains paraffin in spite of prolonged washing. The author therefore recommends distillation of the oil in a current of superheated steam before applying Holdds method.By this means both the solid and liquid hydrocarbons are removed, and the residue will finally consist only of tarry substances mixed with traces of oil adhering to the sides of the flask. A few glass b e d s are introduced to break up the residue, which is then treated with, e.g., 50 C.C. of the mixture of alcohol and ether. The closed flask. is vigorously shaken for some minutes, and allowed to stand for several hours, the liquid being then decanted and the deposit washed with the mixture of alcohol and ether until the filtrate ceases to be fluorescent. The residue may now be dissolved off the filter by means of a few C.C. of hot benzene, the solution evaporated in a tared basin, and the residue dried at about looo C. until constant in weight.Used with this modification, Holde's method gives exact results for all kinds of oil. With regard to the permissible limits for tarry matters in cylinder oils, the author considers that such oils should fulfil the following requirements : For Temperatures Corresponding to about 1 atmosphere Coinsponding to about 4 crtmos heres Corresponh'ing to about 5 to 7 aniiosphcres Corresponding to nhout 9 to 12 atmospheres Corresponding to higher pressures and for superheated steam Specific Viscosity a t 50'' C. 8 to10 About 15 About 20 About 25 30 t o 50, according to temperature of usage Distillation in Steam. About 75 per cent. should distil botweeii 150" and 130" C. About 7 5 per cent. between 150" arid 200" C. 7 5 per cent. between 160" and 220" c. 7 5 per cent. between 170" and 230" c.Variable ; highest observed, 75 per cent. between 2-10" iilld 260" c. Limit for Tarry Matter. None None 2.00 3.0 3.0 to 4.0, xcording to teiii 1)rrature o t usage. Limit for Pnrilffins. Per Cent. 3 '50 3.50 3.50 1.0 (locomotives), 3-50 (0 the I' machinery) 4.0 I I C. A. M. The Keeping Power of Fehling's Solution, and the Volumetric Process of Estimating Sugars with It. F. Watts and H. A. Tempany. (Jount. SOC. Chem. Ind., 1908, 27, 191-193.)-1t is stated that Violet's modification of Fehling's solution may be kept ready mixed for many months without fear of deterioration if light and air be excluded from it. For determining the end-point of the titration in the volumetric estimation of reducing sugars, the authors use potassium ferrocyanide solution acidified with acetic acid.A drop of the titration liquid is placed on a thrice-folded filter-paper, and when the liquid has soaked through the three layersI98 THE ANALYST. the lowest one is tested with a drop of the indicator. The reducing sugar content of the solution under examination should be confined within narrow limits (0.3 to 0.8 gram of sugar per 100 c.c.), in order to prevent the results obtained from being influenced by continual alterations of the bulk of the solutions during the titration ; the error introduced then becomes approximately constant. Allowance is also made for it if the Fehling’s solution be standardised under similar conditions. In the estimation of reducing sugar in raw sugars, a correction must be made for the reducing action of the sucrose itself.As the reeult of their experiments, the authors state that 1 gram of sucrose dissolved in 100 C.C. of water possesses a reducing power equal to 0.0033 gram of invert sugar dissolved in the same volume of water. w. P. s. The Chloramine Reactions of the Proteins and Industrial Applications. C. F. Cross, E. J. Bevan, and J. F. Briggs (Jounz. SOC. Chem. I d , 1908, 27, 263-264).-The authors have extended the work of Raschig (Chem. Zeit., 1907, 31, 926), who obtained monochloramine by the action of hypochlorites on ammonia, and have shown that the proteins and their derivatives, in virtue of their amino groups, form similar compounds. ,411 these chloramines possess the property of liberating iodine from potassium iodide and oxidising sodium arsenite, and thus resemble hypochlorites, but they differ from the latter in the fact that fhey are not destroyed by hydrogen peroxide and possess no apparent bleaching action.Soluble chloramines may exist in bleaching liquors which have been used on substances like flax, which contain residues of protein matters, and in the ordinary assay of such liquors they wopld wrongly be estimated as ‘‘ bleaching chlorine.” The amount of chloramine in these liquors may be estimated approxi- mately by destroying the hypochlorite with an excess of hydrogen peroxide, discharging this excess by acidified permanganate from a burette, adding potassium iodide and starch, and titrating the iodine liberated by the chloramine with & thio- sulphate. The chloramines of any insoluble protein substance may be prepared by means of hypochlorites, hypochlorous acid, or moist chlorine gas ; under this treatment gelatin becomes insoluble in hot water.The action tends to be superficial, owing to the fact that the chloramine produced constitutes a semi-permeable membrane through which further quantities of the chlorinating agent cannot penetrate. The active chlorine fixed in the form of chloramine is estimated, after thorough washing, by steeping the substance in a known excess of sodium arsenite for a sufficient length of time to ensure complete osmosis (about one hour) and titrating back the excess of arsenite by iodine. A dissolved protein-e.g., gelatin- may be estimated by means of the chloramine reaction by distributing a known weight of it on a porous material, such as cotton yarn, and exposing it, in the moist state, for about one hour in an atmosphere of chlorine gas. The excess of chlorine is then removed most conveniently, not by washing, but by suspending the yarn in a strong current of air from an electric fan for at least an hour and a half.The estimation of the fixed chlorine is then carried out, by means of standard arsenite. The value for air-dry gelatin determined in this way was 15.4 per cent. of active chlorine. This empirical factor may be used for the estimation of gelatin inTHE ANALYST. 199 paper, etc. ; previous combination of the gelatin with formaldehyde does not affect the result. Proteins in solution may be treated with a basic solution of hypochlorite, and estimated by the peroxide method mentioned above.A rapid qualitative test for the general or localised presence of proteins in natural or manufactured products is carried out as follows: The material to be tested is either damped with water and exposed to an atmosphere of chlorine gas for half an hour, or it is steeped directly in a dilute acidified solution of bleaching-powder. The chlorinated substance is rinsed in water, and the excess of chlorine is removed by immersing the substance for five minutes in a 2 per cent. solution of sodium phosphate previously heated to 45' C. The material is then treated with a solution of potassium iodide and starch, a blue stain being developed in those places where protein is present. The sodium phosphate also precipitates traces of ferric salts which would liberate iodine in the acid state.J. F. B. Quantitative Estimation of Phenolic Hydroxyl Groups. J. Herzog and V. Hancu. (Ber. deut. Chem. Ges., 1908, 41, 638-639.)-Herzog (Bcr. deut. Chena. Ges., 1907, 40, 1831) has shown that diphenylurea chloride yields well crystallised diphenylurethanes with phenolic substances, and that these compounds are very readily saponified by potassium hydroxide yielding diphenylamine, which is volatile with steam, and the potassium salt of the phenol, which is not volatile. The saponi- fication of these urethanes affords a convenient means for the estimation of the percentage of hydroxyl groups in the original phenol. For this purpose about 1 gram of the urethane and 8 C.C. of alcohol are heated with an excess of potassium hydroxide in 8 pressure-bottle for one hour in a boiling water-bath. The product is transferred to a distillation-flask, the bottle being rinsed out twice with 2 C.C.of alcohol. The distillation with steam, which follows, should be conducted slowly, so that the distillate containing the diphenylamine comes over drop by drop. Any solid deposit in the condenser is removed at the end of the distillation by allowing the condenser to become warm. After standing for twenty-four to forty-eight hours the distillate becomes clear, and the diphenylamine is collected on a tared filter and dried at 30' C. J. F. B. Quantitative Estimation of Methoxyl and Methylimino Groups. A. Kippal. (Ber. dezd. Chem. Ges., 1908, 41, 820-821.)-The author records a case- that of the methylic ether of @-hydroxypyridine betaine-in which the determination of the inethoxyl group by Zeisel's method gave results which were much too low owing to the fact that, under the action of boiling hydriodic acid, the methyl group is liable to migrate and attach itself to the nitrogen.On the other hand, the method of Herzig and Meyer, for the inclusion in the estimation of the methyl residue attached to nitrogen, gave too high results owing to the splitting up of the betaine, with the formation of methyl iodide from other portions of the molecule. J. F. B. Action of Ozone on Compounds with Triple Bonds. E. Molinari. (Ber. dcut. Chem. Ges., 1908, 41, 585-589.)-1n a recent paper (ANALYST, 1908, 22) the author proposed a method for the differentiatios of compounds with triple bonds200 THE ANALYST.from those with double bonds by means of ozone. The observations on which this method was based met with a direct contradiction on the part of Harries (ANALYST, 1908, 99). The author now reaffirms all his previous statements, and attributes the difference between his observations and those of Harries to the fact that he used ozonised air, whilst Harries always employed ozonised oxygen. J. F. B. Physical Constants of Mixtures of Water and Glycerin. P. M. Strong. (Clwm. Zeit., 1908,32, 334.)-The values given in the following table were determined at 18" C., the specific gravity being found by means of a Westphal's balance, the refractive index in Abb6's refractometer, the surface tension by means of a Duclaux's pipette of 5 C.C.capacity, delivering 100 drops of distilled water at 15" C., and the viscosity in Ostwald's viscosimeter : Glycerin. Per Cect. 0 10 20 30 40 50 60 70 80 90 100 Specific Gravity. 1.000 1 *0295 1.0681 1.1094 1,1340 1.1602 1.1860 1.2142 1.2330 1.2599 1.0800 (?) Refractive Index. 1.3277 1.3390 1.3528 1.3650 1.3787 1.3891 1.4020 1.4129 1.4386 1.4529 1.4650 Surface Tension. Number of' Drops. 102 106 110 114 119 124 130 136 147 158 172 Viscosity Coefficient. loO0O 1.3137 1.7197 2-5340 3.6451 5.4108 7.0716 14.2094 48.1632 81.0256 777 -5382 C. A. &I. The Goldenberg Method for the Estimation of Tartaric Acid in Argol, Wine Lees, and other Crude Materials containing Tartaric Acid. Chemische Fabrik, formerly Goldenberg, Geromomt and Co. (Zeit. Anal. Chem., 1908, 47, 57-69.)-The following manner of carrying out this process is put forward in the hope that it may form the basis of an international method for estimating tartaric acid in crude articles of commerce.Six grams of the sample containing more than 45 per cent. of total tartaric acid, or 12 grams if the sample contain less than 45 per cent. of the acid, are digested for ten minutes with 18 C.C. of hydrochloric acid (specific gravity 1-10), The whole is then transferred to a 200 C.C. flask and diluted to the mark with distilled water. After mixing, the contents of the flask are filtered through a dry filter. One hundred C.C. of the filtrate are placed in a 300 C.C. beaker, in which 10 C.C. of 66 per cent. potassium carbonate solution have been placed pre- viously, and the mixture is boiled for twenty minutes in order that any calcium car- bonate may be converted into the crystalline form.The solution, together with the precipitate, is then rinsed into a 200 C.C. flask, cooled, diluted with water to the mark, mixed and filtered. One hundred C.C. of the filtrate are evaporated to a volume ofTHE ANALYST. 201 15 c.c., and, while the solution -is still hot, 3.5 C.C. of glacial acetic acid are added slowly, with constant stirring, the mixture being then stirred for a further five minutes. After the lapse of ten minutes, 100 C.C. of 95 per cent. alcohol are added, and the stirring is continued for five minutes. The whole is next allowed to stand for ton minutes, then filtered, and the precipitate is washed with alcohol until free from acidity. The filter and precipitate are then washed into a porcelain basin with 200 C.C. of hot water, and the boiling solution is titrated with 2 sodium hydroxide solution, using litmus paper as indicator. The sodium hydroxide solution must be standardised by titration on a known quantity of pure potassium hydrogen tartrate. As no allowance is made for the volume of the insoluble matters when the sample is made up to the volume of 200 c.c., the results obtained are corrected as follows : in the caw of samples containing less than 45 per cent. of total tartaric acid 0.80 per cent. is subtracted from the quantity of acid found ; for samples containing from 45 to 60 per cent. the correction is 0-30 per cent. ; for those containing from 60 to 70 per cent., 0.20 per cent.; whilst no correction is applied to the result if the sample contain more than 70 per cent. of total tartaric acid. w. P. s.