244 ABSTRACTS OF CHEMICAL PAPEES ORGANIC ANALYSIS. Estimation of Carbon Monoxide by the Iodine Pentoxide Method. V. Froboese. (Zeitsch. anal Chem., 1915, 54, 1-11.)-In this method the U-tube containing the iodine pentoxide should be heated to 100' C. in a beaker of boiling water ; there is no need, to make the temperature higher by means of a paraffin bath (cf. ANALYST, 1914,39, 445). The amounts of iodine liberated or of carbon dioxide formed are both measures of the quantity of carbon monoxide present, but the more trustworthy results are obtained by estimating the carbon dioxide ; the latter may be estimated gasometrically, or it may be absorbed in barium hydroxide solution and the excess of this titrated, or the resulting barium carbonate may be collected, converted into barium sulphrtte, and weighed.I n this way carbon monoxide can be estimated in the presence of hydrogen or coal-gas, but the gas under examination must be freed previously from carbon dioxide, acetylene, and ethylene. Hydrogen liberates iodine from iodine pentoxide; at 70" C. the action is but slight, but at 160' C. a consider- able quantity of free iodine is formed, 80 that the amount of carbon monoxide cannot be calculated from the quantity of iodine set free.The iodine titration method may, however, be used for the estimation of small quantities of carbon monoxide in air, provided that a current of air free from carbon monoxide and carbon dioxide isORGANIC ANALYSIS 245 passed through the heated iodine pentoxide tube for one hour before and after the air under examination is conducted through the tube.The quantity of carbon monoxide in a gas is without influence on the results obtained by the iodine pentoxide method if the gas be passed through the apparatus at a rate not exceeding 1 litre per hour. w. P. s. Detection and Estimation of Small Quantities of Carbon Oxysulphide in Carbon Dioxide [Mineral Waters] and Well-Gases.L. Dede. (Chem. Zeit., 1914, 38, 1073-1075; through J. SOC. Chm. Ind., 1915, 34, 298.)-Certain mineral waters containing much carbon dioxide, although originally failing to respond to the usual tests for hydrogen sulphide, developed appreciable amounts of that substance upon keeping ; the gases from these waters, including one particular specimen of carbon dioxide used commercially for aerating purposes, behaved similarly.The presence of sulphur was established and its amount det'ermined in the original gases by fixation as lead sulphate, according to Dennstedt's combustion method. No sulphide was produced from dissolved sulphate either in the presence of ferrous bicarbonate or of cork fragments, even after keeping for long periods; its formation is ascribed to the presence of small amounts of carbon oxysulphide in the waters.The decomposition of this substance by water, being a reversible process, oocurs very slowly. in the presence of carbon dioxide : COS + H , 0 Q H 2 S + GO,. When water containing 0-05 per cent. of sodium bicarbonate was saturated with carbon dioxide containing 0.02 per cent. of carbon oxysulphide and preserved in stoppered bottles, the odour of hydrogen sulphide became perceptible after about four days.Carbon oxysulphide was most conveniently detected and determined by aspirating a slow stream of the gas through a Lunge bulbtube containing a dilute (1 : 1,000) solution of palladious chloride acidified with hydrochloric acid and kept at 50' C. : COS + P d Q + 4 0 = PdS.+ 2HCl+ GO,.In the presence of fairly large amounts of hydrocarbons and other reducing gases which cause the separation of metallic palladium, stronger solutions of palladium chloride (up to 4 : 1,000) were necessary. The precipitate was collected, washed, and dissolved in hot 10 to 15 per cent. hydrochloric acid containing a little potassium chlorate, the bulbs being washed out with the same reagent.The amount of sulphate in the diluted solution was estimated by precipitation with baxium chloride. Three successive quantities of 25 litres of the commercial carbon dioxide mentioned above yielded, respectively, 0.0057, 0.0059, and 0-0048 grm. BaSO,, corresponding to 0-0029 per cent. COS by weight. The gas obtained directly from a well by means of a partially immersed bell-jar and a 25-litre aspirator gave, similarly, 0.0037, 0.0031, 0.0038, 0.0035, and 0.0039 grm.BaSO, per 25 litres, corresponding to 0-0019 per cent. COS by weight. I t is suggested that the change in the physiological action of certain mineral waters upon storing may probably be connected with the gradual decomposition of small amounts of carbon oxysulphide. Estimation of Formic and Acetic Acids and the Separation of these Acids in very Dilute Solutions.E. Heuser. (Chem. Zeit,, 1915, 39, 57-59; through J. SOC. and Chem. Id., 1915, 34, 286.)-The estimation of formic and acetic246 AESTRA CTS OF CHEMICAL PAPERS acids by acidifying with sulphuric acid and distilling with steam requires too much time for ordinary use, and if the sulphuric acid be replaced by phosphoric acid, the results are vitiated hy the latter passing over into the distillate.Good results are obtained if the vapours are passed through a flask filled with glass beads and heated by a water-bath on their way to the condenser (cj. Wenzel, Monats. Chem., 1897, 659). The sample is mixed with 50 C.C. of water and 50 C.C. of phosphoric acid of sp. gr.1.2 and distilled at 44" C. (56 mtn. pressure) until the volume is reduced to 50 C.C. A further 50 C.C. of water is then added, and the mixture again distilled to a volume of 50 C.C. A current of air free from carbon dioxide is drawn through the apparatus during distillation. The use of a flask filled with glass beads is also necessary in determining acetic acid by distillation in a mixture of formic and acetic acids after destroying the former by oxidation with bichromate and sulphuric acid.Separation of Gases by Fraotional Distillation in a Vaeuum at Low Temperatures. G. A. Burrell and I. W. Robertson. (J. Ind. and Eng. Chem., 1915, 7,209-210.)-The authors summarise their previous work (ANALYST, 1914, 39, 414; 1915, 162, 163), and give particulars of further separations that they have effected by similar methods.From a mixture of gaseous olefines, ethylene can be separated by cooling to -140' C., and then exhausting the containing vessel by means of the vacuum-pump. From the residual liquid propylene can be removed by allowing the temperature to rise to - 120" C. and again exhausting. Mixtures of gaseous paraffins and olefines can be analysed as follows: The methane is removed by cooling to the temperature of liquid air and exhausting. Ethane and ethylene are removed at -140" C., and propane and propylene at -120" C., leaving the butanes and butylene as a residue.The pairs of constituents are then analysed by combustion with oxygen, Like fractional distillation of liquids at temperatures above room temperature, the above methods do not give sharp separations at one operation; but methane can be sharply separated in two operations, three or four being necessary to recover in a pure state the whole of the ethane or propane from a, mixture of gaseous paraffins.I t is even possible to separate ethane from ethylene, but the process is so tedious that it is not reoommended. G. C.J. Rapid Method of Fractionating Gases at Low Temperatures. G. A. Burrell and I. W. Robertson. (J. Ind. and Eng. Chem., 1915, 7, 210-211.)-The method is essentially identical with that already described (see preceding abstract), the novelty consisting in detailed instructions for securing sharp separations with a minimum expenditure of trouble and time. The following details show the best means of making a complete analysis of illuminating gas similar to that supplied to New York City, the composition of which is given below.About 1,000 C.C. is taken for analysis, freed from carbon dioxide by means of potash, and liquefied by means of liquid air. At the temperature of liquid air as much gas as possible is pumped off, constituting fraction A, which is preserved in a separate container.The residual liquid is next allowed to attain a temperature approximating, but not higher thanORGANIC ANALYSIS 247 Per Cent. Carbon dioxide ... ... 4.7 Carbon monoxide ... ... 29.8 Hydrogen ... ... ... 32.0 Methane ... ... ... 13.1 Ethane ... ... ... 2.8 Propane ... ... ... 0.3 Per Cent. ... Ethylene ... ... 9.8 Propylene ... ... ... 2.7 Butylenes ...... ... 1.6 Nitrogen ... ... ... 1.7 Benzene ... ... ... 1.5 Estimation of Unsaponiflable Matter in Greases. E. Twitchell. (J. Ind. and Eng. Chem., 1915, 7, 217-218.)-Five grms. of the sample (or, preferably, of the fatty acid prepared for 6 L titer " test, as this is cleaner) is saponified with alcoholic potash in a dish and evaporated nearly to dryness. A little alcohol is added, and the water and the solution obtained is washed into a separator. The alcohol-water ratio should be about 1 : 4, and the total volume 150 to 200 C.C.The soap solution is shaken twice with ether, using 50 C.C. each time. A third treatment with ether extracts so little that for ordinary analytical work it may be omitted. Owing to the comparatively high concentration of alcohol, the separation of the liquid into clear and sharply defined layers is almost instantaneous.More soap passes into the ethereal layer than is the case when less alcohol is used, but this is readily corrected for. The ethereal solution is washed once with water, then with hydrochloric acid, again with water, transferred to a weighed dish, and evaporated. The residue is dried at 110" C., weighed, dissolved in neutral alcohol, and the solution titrated with standard alkali.The fatty acids thus found, derived from soap dissolved by the ether and subsequently decomposed by hydrochloric acid, are calculated as oleic acid and deducted from the weight before obtained. The method, which has been in use five years, never gives lower results than other methods.Occasionally the results248 ABSTRACTS OF CHEMICAL PAPERS are higher, but are believed to be more exact, examination proving absence of neutral fats and of ethyl esters, which might conceivably be formed when the alcoholic ethereal extract is shaken with hydrochloric acid. G. C. J. Estimation of Small Quantities of Hydrocyanic Acid. A. Viehoever and C. 0. Johns. (J. Amer.Chem. Soc., 1915, 37, 601-607.)-1n the colorimetric estimation of cyanides by the Prussian blue reaction, the maximum formation of blue is only obtained when the volume of the solution tested is sufficiently small. The solution of hydrocyanic acid must therefore be concentrated, which may be done without appreciable loss by evaporating under reduced pressure in presence of a small excess (0.02 to 0.1 grm.) of sodium hydroxide.The evapora- tion is performed in a distillation flask, which is heated in a water-bath at 70" C., and loss of cyanide by entrainment is avoided by interposing an adapter, like that employed in distillations of ammonia by the Kjeldahl method, between the flask and the condenser. With quantities of about 0.5 mgrm. of hydrocyanic acid or 1 mgrm.of potassium cyanide, it is essential that the solution be evaporated to a volume not exceeding 1 C.C. ; 0-2 to 0.5 C.C. of a freshly prepared 3 per cent. solution of ferrous sulphate and 0.05 grrn. of potassium fluoride are then added, the flask is again connected with the vacuum pump, and the contents are mixed by agitation, Quick exhaustion of the air is necessary, because it ie desirable that the reaction take place without the formation of any considerable quantity of ferric salt by oxidation.After five to ten minutes a t the ordinary temperature the flask is detached from the pump, and the mixture acidified with 30 per cent. nitric acid. Sulphuric acid may be used, but hydrochloric acid in excess tends to produce a green colour. When only traces of hydrocyanic acid are present, it is sometimes necessary to warm to about 50" C.before the colour appears. It is noted that the presence of the potassium fluoride has a most favourable effect on the formation and clearness of shade of the blue, The results are obtained by colorimetric comparison with a standard solution of potassium cyanide similarly treated. When less than 1 mgrm.of potassium cyanide is present, the quantities of the reagents should be reduced, as an excess is to be avoided. As a qualitative test, the reaction is sensitive to 0-00002 grm. of potassium cyanide. J. F. B. Analysis of Vegetable Parchments. R. W. Sindall and W. Bacon. (Chem. Engineer, 1915, 21, 75.) - The preparation of slides for the microscopic analysis of vegetable parchments is difficult on account of the resistance of the paper to disintegration into its constituent fibres.Chemical methods of disintegration- e.g., by the use of cuprammonium solution (Vidal) or of moderately diluted sul- phuric acid (Bartach)-are open to the objection that some of the fibres are more profoundly attacked than the others, so that in papers composed of a mixture of rag and wood cellulose fibres, a portion of the latter may be lost.The authors therefore prefer a mechanical method of disintegration, such as is readily performed by a small vertical coffee-mill. About 0% grm. of the sample is crumpled up with hot water, then torn into small piecee, and passed through the mill several times with a little water. The friction of the particles of paper upon each other andORGANIC ANALYSIS 249 between the closely set surfaces of the mill separates the fibres fairly completely, and the resulting pulp can then be mounted for microscopic examination in the usual.manner. In staining with the iodine-zinc chloride reagent, the solution should be diluted before applying it to the preparation, as the chemically modified fibres of vegeta.ble parchment react so strongly with the ordinary stain that all distinguishing tints are obscured.J. F. B. Analytical Distillation of Petroleum. W. F. Rittmann and E. W. Dean. (J. Ind. and Eng. Chem., 1915, 7, 185-195.)-A comparative study of three types of apparatus-(1) Those designed to effect fractionation in a high degree, such as that of Hempel; (2) those designed to prevent any cooling of vapour between the surface of the boiling liquid and the outlet to the condenser, of which Allen and Jacobs’ apparatus (U.S.Bweau of Mines, BzlZZ. 19) was selected as example ; and (3) those in general use in the petroleum industry, which come midway between these types, and of which Ubbelohde’s was selected as example. As might be expected, the gradation in specific gravity of successive fractions showed that Hempel’s method effected the highest degree of separation, Allen and Jacobs’ the least.On redistillation of the separate fractions in the apparatus with a Hempel column, it was found that on the average only 14 per cent. of each of the fractions from Allen and Jacobs’ apparatus distilled between the original temperature limits, whilst the percentage in the case of fractions obtained by Ubbelohde’s method was about 25, and in the case of fractions originally distilled from Hempel’s appa- ratus it was 55.The comparative extent of fractionation effected by the three methods is brought out in several tables and curves, but most strikingly by a table obtained from one of the curves. The curve was constructed by plotting the specific gravities of each small fraction obtained by redistillation, in the Hempel apparatus, against the percentage of distillate collected when that small fraction was half collected.By interpolation, the specific gravity of the distillate at any stage of the redistillation could be obtained. I t was found that when crude petroleum was distilled in a Hempel apparatus, the receiver change! at each 25’ C.rise of tem- perature, and any one of the fractions redistilled in the same apparatus, the differ- ence in specific gravity of the liquid coming over when only 5 per cent. had distilled and that coming over when only 15 per cent. remained in the still was only 0.010 when the fraction submitted to redistillation was a high fraction, increasing to 0.024 for the lowest fraction.When a fraction obtained by Ubbelohde’s method was similarly distilled in the Hempel apparatus, the corresponding rise in specific gravity between 5 per cent. and 85 per cent. distillate (chosen to eliminate irregularities at the beginning and end of a, distillation) was from 0.025 for fractions boiling above 275’ C. to 0.047 for the lowest fraction.The corresponding range, starting with Allen and Jacobs’ method, was as great as 0.043 to 0.075. I t is sometimes assumed that any method of distillation will give useful information, provided it is somewhat rigidly defined. In support of this it might be alleged that when temperatures are plotted against percentage of distillate below that temperature, nearly straight lines are traced in the case of petroleum, and that these lines, though not identical if varying methods be adopted, are substantially250 ABSTRACTS OF CHEMICAL PAPERS parallel.This is approximately, but only approximately, true, and the divergence is most marked at the bottom of the curve-that is to say, the most important part of its range, having regard to the present high price of motor spirit compared with other petroleum distillates.No method is yet recommended; but the work now described is part of an attempt by the United States Bureau of Mines to improve on existing methods. The ideal method must minimise, or, if possible, eliminate ( ‘ cracking ”; the apparatus must be standard and easily obtainable; and the results must meet the needs of both the scientist and the oil technologist.They need not be identical with those obtained in the refinery-in fact, they cannot be, having regard to the variations in works procedure. G, C. J. Comparative Study of Methods for the Quantitative Estimation of Sulphur in Peptone. H. W. Redfield and C. Huckle. (J. Amer. Chem. SOL, 1915, 37, 607-611.)-0f the various methods tested, the Liebig-Koch method (diges- tion with nitric acid and fusion with potassium nitrate and hydroxide) gave the highest and most consistent results. A slight modification was introduced : the solution of 2 grms.of peptone, after digestion with nitric acid (sp. gr. 1-4) on the water-bath for two hours, was introduced into a nickel crucible containing 8 grms. of potassium hydroxide and 0.5 grm.of potassium nitrate and evaporated over an alcohol lamp, a jet of air being directed over the surface of the liquid ; in this way a perfectly homogeneous mixture for fusion was insured. I n transferring the liquid from the beaker to the crucible, the particles adhering to the sides of the beaker were wiped off with a piece of filter-paper wrapped round the end of a glass rod, the paper being added to the solution in the crucible.Tests were made which proved that no sulphur was lost in the form of volatile compounds during the preliminary digestion with nitric acid in open vessels. The time of the digestion beyond two hours had no influence on the results. The following results for total sulphur in Witte peptone, corrected for moisture and for sulphur introduced by the reagents, are recorded : Liebig method, 0.9902 ; Osborne method, 0,9479 ; A.O.A.C. method, 0.9664 ; Liebig-Koch method, 1.0061 per cent. For the determimtion of loosely bound sulphur, the Schulte method was approved, and for that of the easily oxidis- able sulphur, the method of digesting 2 grms. of peptone with 20 C.C. of a saturated solution of potassium chlorate in nitric acid for two hours on .the water-bath, evaporating to dryness, adding 5 C.C. of hydrochloric acid and 20 C.C. of water, evaporating to dryness, dehydrating and determining the sulphate, gave the most consistent and valuable results. The numbers recorded are-for loosely bound sulphur by the Schulte method, 0.3813 per cent.; for easily oxidisable sulphur, 0.5194 per cent. The portion of peptone soluble in alcohol showed total sulphur 0.7950 per cent. by the Liebig-Koch method, the residue insoluble in alcohol, 1.0743 per cent. J. F. B.