470 ABSTRACTS OF CHEMICAL PAPERS ORGANIC ANALYSIS. Salting Out of Amino-Acids, and their Separation by Means of Neutral Salts. P. Pfeiffer and F. Wittka. (Bw., 1915, 48, 1041-48; through J. &c- Chem. I d . , 1915,34,976.)-0n addition of saturated solutions of neutral salts to satu- rated solutions of amino-acids, glycocoll, tyrosine, and aspartic acid gave no precipi- tate ; alanine was only precipitated by ammonium sulphate, whereas d,Z-leucine and d,Z-phenylalanine were precipitated by sodium and potassium chlorides, potassium oxalate, magnesium and sodium sulphates, sodium and potassium acetates, but best by ammonium sulphate.In these precipitates, the amino-acids are not combined with the salts, but are in the free, crystalline condition. Phenylalanine did not give a precipitate with potassium or aluminium sulphete, ammonium, calcium, barium, magnesium or aluminium chloride, or with barium bromide.The results afford further confirmation of the view that proteins are com- plexes of amino-acids, the salting out of which can no longer be accepted as a criterion of their colloidal character. Phenylalanine and I-leucine may be separated from glycocoll by means of ammonium sulphate.The calcium chloride compounds of glycocoll and d-alanine have been separated by the difference in their solubility in aqueous alcohol. Disturbing Factor in Barfoed's Test. W. H. Welker. (J. Amer. Chm. SOC., 1915, 37, 2227'-2230.)-Barfoed's cupric acetate solution has been used exten- sively in bio-chemistry for the detection of reducing monosaccharides, and since it does not reduce disaccharides such as maltose, it has been employed to show the difference between the products obtained by the action of diastase and of dilute acids respectively on starch pastes.The author finds that when starch is boiled with hydrochloric acid the solution at that stage, when the iodine reaction vanishes, does not reduce Barfoed's reagent, but that a greenish-white precipitate is formed.The cause was traced to the presence of sodium chloride (the solutions had been neutralised with sodium hydroxide and acidified with acetic acid), a very small quantity of which interferes with the reaction. The precipitate contains copper, sodium, chlorine, and acetic wid, but its exact composition has not been determined.w. P. s.ORGANIC ANALYSIS 471 Estimation of Formic Acid in the Presence of Acetic Acid. R. Lauff- mann. (Chem. Zeit., 1915, 39, 575; through J. SOC. Chem. Ind., 1915, 34, 1008.) -A method described recently by Heermann (Chem. Zeit., 1915, 39, 124) depends on the fact that formic acid is capable of expelling acetic acid from acetates. A mixture of sodium formate and sodium acetate is dried at 125' C., and weighed, then treated with formic acid, evaporated, the residue again dried at 125' C., and weighed, the quantity of acatic acid being calculated from the loss in weight, and the amount of formic acid present originally being ascertained indirectly.Investigation of this method showed that it yields results which are only approximately correct.Solu- tions containing about 1 per cent. of formic and acetic acids together, the quantity of formic acid varying from 0.3 to 0.7 per cent,, were neutralised, evaporated, and the residues dried and weighed. The latter were then twice evaporated with the addition of formic acid, followed by four evaporations with the addition of water. The quantity of formic acid found was about 0.02 per cent.too high, whilst that of the acetic acid was too low. Analytical Distillation of Petroleum.-11. W. F. Rittman and E. W. Dean. (J. Ind. and Eng. Chern., 1915, 7, 754-760.)-1n continuation of their work (ANALYST, 1915, 249), the authors have now examined the merits for their particular purpose of every type of fractionating column hitherto described. A Hempel column is finally recommended.The only columns giving more perfect fractionation were the evaporator still-head of Young and an 18-inch rod and disc column, but the latter is unsuited for work above 225" C., as it floods owing to excessive reflux condensation, whilst the evaporator still-head is fragile, complicated, and only very slightly more efficient than a simple Hempel column. The sole disadvantage of the latter, on which Young laid stress, is the comparatively large amount of liquid trapped by it at the end of a, distillation, but with quantities of 200 c.c.under examination this is less important than when smaller quantities must be taken.An advantage of the Hempel column is that it can be constructed in one piece with the distilling flask. The form of apparatus finally recommended is a distilling flask with a bulb of 300 C.C. capacity, and a neck 1 inch in diameter and 6 inches long from the bulb to the side tube.A copper wire with a hook at the top, to facilitate withdrawal, and bent into a spiral near the bottom of the neck of the flask, Berves to support 5 inches of aluminium beads, which fit together less closely than glass beads, and are therefore less liable to be flooded at high temperatures by excessive reflux condensation.If the neck of the flask is as narrow as Q inch, the results are not materially different. The length, however, should closely approximate 6 inches and the height of the column of beads 5 inches to obtain strictly comparable results. A longer column gives slightly better fractionation, but is unwieldy and leads to excessive condensation at high temperatures.The reduction of cracking is desirable in the analytical distillation of petro- leum, but it is shown that the importance of this factor has been overestimated in the past. The simpler hydrocarbons are stable at much higher temperatures than those of higher molecular weight, and the tendency for the products of reflux action to be cracked by superheating is negligibly small.It is the residual hydrocarbons in472 ABSTRACTS OF ,CHEMICAL PAPER,R the still that tend to become cracked, and the extent of the action is a function of time and temperature. I t is shown that when vapour at 300" C. was passing from a Hempel column, as described, to the condenser, the temperature of the liquid in the still approximated 340" C., against 310" C.when no fractionating column was used, so that simple distillation can only be carried about 30" C. higher than distillation in- volving efficient fractionation, with the same possibility of cracking in each case. In the apparatus described, the authors distil 200 C.C. of petroleum at the rate of 2 to 2.5 C.C. per minute.G. C. J. Estimation of Solid Matter in Smoke. A. Gautier. (Bull. SOC. Chim., 1915, 17, 273-278.)--For the estimation of soot, tarry matters, etc., in smoke, an iron tube closed at one end is inserted through a hole in the chimney, where it reaches to the opposite wall. In this tube are two slits directed downwards, and a second narrower tube of nickel with open end is fitted into it.The other end of the narrow tube is connected with an absorption vessel consisting of two superposed bulbs cooled in a tank of water. The smoke is drawn into the space between the iron and nickel tubes by means of a suitable aspirator, and then passes upwards through the bulbs, which contain water and small glass balls to retain the bulk of the soot, etc. It is next drawn through a vertical filter composed of three pieces of filter-paper fixed between platinum cones, having a lateral tube which is heated to 100" C. to prevent condensation of steam. The fine particles of solid matter are retained by the filter, and the gas leaving it is finally drawn through a gas meter. The filter- papers, which have previously been dried stti 100" C. and weighed, are then trans- ferred to a funnel, the contents of the bulbs collected upon them, and the papers washed with ether and hot alcohol, to remove tar, etc., dried at 100" C., and weighed. The amount of solid matter i n coal smoke usually averages 0.003 to 0*008 grm., but may reach 0.890 grm. or more per cubic metre. C. A. M.