THE ANALYST. 313 ORGANIC ANALYSIS. The Determination of Methyl Alcohol by the Iodide method. M. J. Stritar and H. Zeidler. (Zcit. anal Chem., 1904, xliii., 387-400.)-The method employed by the authors is practically identical with that devised by Zeisel and Fanto for the determination of glycerin (ANALYST, xxvii., 300; xxviii., 222). In test experiments 5 C.C. of an aqueous solution containing 0.0385 gramme of methyl alcohol were mixed in the small flask of the apparatus with 13.5 C.C. of pure hydriodic acid of specific gravity 1-96, and the following amounts of silver iodide obtained : (I) 0-2800 gramme, and (2) 0.2797 gramme ( = 0.0382 and 0.0381 gramme of methyl alcohol). In no case did the operation take longer than one hour from the time boiling commenced. It is essential that the concentration of the methyl alcohol solution should not exceed 1 per cent., and if possible not be less than 0.5 per cent.Obviously the absence of other substances that yield silver iodide or interfere with the process must be ascertained, such as, c.g., sulphur compounds, large amounts of chlorides, nitrates, etc. Dcterirnhtation of Methyl Alcohol in the Products of the Distillation of Wood.- Experiments are described in detail to show the influence of the different products on the results of the iodide method. The only substances that it is necessary to take into account are alcohols (allyl alcohol) and esters (methyl acetate). To obtain the correction for allyl alcohol, an aliquot portion (25 c.c.) of the solution is acidified with 5 C.C. of dilute sulphuric acid, and titrated with a solution containing 1.186 grammes of potassium bromate and 4.224 grammes of potassium bromide per litre, which is introduced with continual shaking, 0.2 C.C.at a time, at intervals of a minute, until the liquid becomes yellow. The number of C.C. of the bromide-bromate solution used gives the number of niilligrammes to be deducted from the weight of silver iodide in the methyl alcohol determination. The necessary correction for methyl acetate is obtained by determining the ester value of the solution in the usual manner. Crude wood spirit is prepared for the determination by diluting 5 C.C. of known specific gravity to 500 C.C. and using 5 C.C. of this solution. The weight of silver iodide obtained, multiplied by the factor 272.94, gives the amount of methyl alcohol in grammes per 100 C.C.Crude aqueous zooocl! spirit (Rohholzgeistwasser).-From 5 to 10 c.c., according t o the strength, are diluted with 40 C.C. of water, shaken for fifteen minutes with animal charcoal, filtered, and the filtrate and washings made up to 100 C.C. If an error of 2 per cent. is permissible, the treatment with animal cliarcoal may be omitted. Crzide pyroligwous acid.-Fifty C.C. are neutralized with sodium carbonate, and the liquid diluted to 250 C.C. and filtered, 100 C.C. of the still dark filtrate are rendered strongly alkaline with sodium hydroxide and distilled until 50 C.C. of the distillate have been collected, this is then treated with animal charcoal as above described. C. A. M.314 THE ANALYST.The Determination of Methyl Alcohol in Commercial Formaldehyde. M. J. Stritar. (Zeit. anal. Chenz., 1904, xliii., 401-403.)--Five C.C. of the formalin are diluted with 100 C.C. of water, treated with ammonium hydroxide in excess (about 10 C.C. of a 12 per cent. solution of ammonia are generally sufficient) and distilled, the first 50 C.C. of the distillate are rendered slightly acid with strong acetic acid and made up to 100 C.C. Five C.C. of this solution are then analysed by the iodide method (see preceding abstract), and the amount of silver iodide multiplied by the factor 54.57 to obtain the grammcs of methyl alcohol in 100 C.C. of the formalin. Formaldehyde itself acts upon hydriodic acid, yielding silver iodide as a, final product ; but the reaction takes place very slowly, and as only traces of formalde- hyde are left in the distillate obtained as described above, the results of the iodide method are not affected.Two samples of commercial formalin thus examined contained 16.5 and 15.14 grammes of methyl alcohol per 100 C.C. respectively. C. A. M. On the Quantitative Determination of Methoxyl. Wilhelm Kropatschek. (Mo?zatshefte f. Chew., xxv., 583.)--Tn Gregor’s modification (Monatshefte f. Chem., xix., 116) of the Zeisel method a solution of potassium arsenite in potassium carbonate is substituted for the mixture of amorphous phosphorus and water generally used as washing liquid, in order to prevent the formation of silver phosphide, supposed to be due to mechanically carried over phosphorus. It has since been shown that this may lead to serious errors, a crystalline precipitate containing iodine being deposited from the arsenite solution.The author now finds that a deposit of the formula KI+As,O, is indeed produced if the concentration of the washing liquid recom- mended by Gregor in his paper, but not actually used by him, is employed, but that exact results are obtained if the concentration does not exceed 1 to 2 per cent. By means of experiments on pure hydrogen iodide, prepared by the action of iodine OD formic acid-ie., from phosphorus-free materials-he also shows that the formation of the silver phosphide is due not to the mechanical carrying over of amorphous phosphorus, but to the presence of hydrogen phosphide in most ordinary samples of “ pure ” hydrogen iodide made from iodine and phosphorus. This hydrogen phosphide is completely absorbed by Gregor’s washing liquid. A.G. L. The Hydration of Alcohols by Calcium and Barium Oxides. L. Crismer (BUZZ. Xoc. Chiim. BcZg., 1904, xviii., 128, 129.)-If anhydrous methyl alcohol be heated with calcium oxide freshly prepared from marble and then distilled, the dis- tillates show critical temperatures of solution (in a suitable solvent, such as paraffin oil or petroleum spirit), corresponding to amounts of water ranging from 0-12 to 2 per cent. Hence the author concludes that the ordinary method of distilling methyl alcohol with lime ought to be abandoned. I n like manner ethyl alcohol is hydrated by barium oxide, though not by calcium oxide. The author has prepared pure methyl, ethyl, and propyl alcohols by utilizing their different critical temperatures of solution.He shows that the difference between the three values is a constant. Thus, the fol-THE ANALYST. 315 lowing critical temperatures were obtained with paraffin oil as the solvent : Methyl alcohol, 166" C. ; ethyl alcohol, 89.7"C. ; and propyl alcohol, 13.5" C, ; difference, 76.3" and 76.2O C. C. A. M. Notes on the Determination of the 100" Point on Ventzke'rJ Saccharimeter Scale. Schoenrock. (Zeits. Zt&e&td. ; through Chenz. Zeit. Rep., 1904, xii., 177.)- The author, in a very exhaustive paper, comes to the following conclusions : (1) To obviate the effect of various sources of light on different observers, the light should be purified by passing through a 1-5-centimetre film of 6 per cent.potassium bichromate solution. (2) According to Herzfeld's observations, a plate of quartz of 100" Vent& gives a rotation, at 20" C., of 34.657 scale degrees with pure sodium light. substance giving different values for different lights has not the same dispersion as quartz. (4) Increase of rotation value z1 is causgd by risihg temperature, according to the formula : (5) In the case of quartz, v is independent of t when the temperature of both quartz- plate and apparatus coincide. (6) If the rotation of an approximately normal sugar solution (this and the apparatus at 2Oo)=S,,,Ventzke, and at to=&, then s20 = S, + S,O*O00609 ( t - 20). (7) The rotation of a normal sugar solution at 20" with spectrally purified sodium light is 0.010 scale degrees greater than the corresponding angle of the normal quartz-plate of 100" Ventzke.(8) For accurate analyses observn- tion tubes, which can be filled after both cover-plates are in position, should be used. ( 3 ) vt = ~~~o + ~,,0*000148 ( t - 20). H. A. T. On the Reaction of Lard from Cotton-seed Meal-fed Hogs with Halphen's Reagent. Elton Fulmer. (JOZLY72. Amcr. ChCY72. Soc., sxvi., 837.)-The author finds that lard from hogs fed even for short periods on cotton-seed meal will give a coloration with Halphen's reagent, equivalent to an apparent content of cotton-seed oil of 0.4 to 15 per cent. The colour is given by lard rendered from all parts of the body of the hog, the strongest colour being given by kidney fat and the faintest by intestinal fat.The colour-producing substance is very persistent when once deposited, a reaction being given by fat from animals killed five months after receiving their last ration of cotton-seed meal. A, G. L. A Comparison of the Halogen Absorption of Oils by the Hubl, Wijs, Hanus, and McIlhiney Methods. L. M. Tolman. (Jozmz. Amer. Chent. xoc., xxvi., 896.)-Froni a. number of comparative determinations, the author comes to the conclusion that much better results are obtained by the Wijs and Hanus solutions than by the Hiibl. With the Hanus solution an excess of 60 to 70 per cent. is necessary to ensure rapid absorption ; the Wijs solution works well with an excess of 35 per cent., but gives higher results. I n either case thirty minutes is sufficient time for the absorption.For ordinary work, a solution of iodine chloride in carbon tetrachloride is best, when a determination of substitution is to be made ; otherwise acetic acid is the better solvent. Bromine solutions are not generally satisfactory. A. G. L.316 THE ANALYST, Contributions to the Analysis of Rose Oil. P. Jeancard and C. Satie. (BzdZ. Xoc. Chim., 1904, xxxi., 934-937.)-A green oil distilled from rose flowers after removal of the petals (Le., from the calices, pistils, stamens, etc.) had the following characteristics : Freezing-point, +So C. ; specific gravity at 15" C., 0.8704 ; optical rotation at 15" C., - 41" ; acid value, 6.12 ; saponification value, 22.4 ; stearoptene, 51.13 per cent. ; total alcohols, 13.99 per cent. ; and citronellal 13.56 per cent, The stearoptene consisted, in the main, of a substance melting at about 14" C.A specimen of oil obtained horn the tea-rose had a characteristic odour, and contained 72 to 74 per cent. of stearoptene, although its freezing-point was only 23.5" C. This stearoptene consisted of two bodies, one of which melted at about 14" C., and the other at 40" C., and apparently contained the same constituents as the oil from the calices. The authors point out that these facts show that it is not rational to base a judgment as to the purity of rose oil solely on the freezing-point, since the latter does not necessarily correspond with the proportion of stearoptene. In their opinion, it is best to make a determination of stearoptene and citronellal in the following manner : Ten grammes of the rose oil are mixed with 50 C.C.of acetone and cooled to about -loo, without shaking. The stearoptene is collected on a weighed filter, which is kept cool by a freezing mixture, and after being washed several times with acetone, is dried in vacuo and weighed. The filtrate, which should yield no further deposit on chilling, is distilled in a weighed flask, the distillation being finished under diminished pressure, and the residue of eleoptene weighed and submitted to further tests. Ten samples of Eastern oil of known purity thus examined gave results ranging between the following limits : Oil.-Freezing-point, 19" to 21" C. ; stearoptene, 18 to 23 per cent. EZeoptene.-Specific gravity at 15" C., 0.886 to 0.888; optical rotation at 15" C., - 1" to -3" ; solubility in 70 per cent.alcohol, 1.5; acid value, 1 to 2 ; saponification value, 10 to 12 ; total alcohols, 84 to 88 per cent. ; and citronellal, 30 to 40 per cent. In practice the authors have adopted provisionally the following limits : Provence Oil. Bulgarian Oil. Stearoptene, per cent. ... ... ... 30 to 35 18 to 23 Citronella1 in the eleoptene, per cent. . . 20 to 23 30 to 40 The limits for Provence rose oil are thus v e y similar to those given by Schimmel and Co. for Saxony rose oil. C. A. M. Testing Colloids. Edmund J. Mills and Archibald Gray. (Jourqz. SOC. Chenz. Iqzcl., xxiii. 526.)-This paper contains a preliminary account of a physical method for testing colloids-namely, the determination of the modulus of elasticity- which, while capable of giving comparative results of practical value, is not claimed to be entirely successful, The method consists in coating a standard cotton thread with a given colloid, drying the thread, and determining the modulus of elasticity of the coating, which is directly related to the stiffening power of the colloid.Brook's No. 12 white cotton thread is suitable on account of its freedom from finishing ingredients. The thread is first slightly singed to burn away the projectingTHE ANALYST. 317 ends surrounding it, and is then immersed in a standard solution or paste of the colloid, such as gum arabic, potato starch, etc. About a metre of this coated thread is then drawn repeatedly through a german-silver nozzle (Fig. l), the narrow bore of which is 0-5 millimetre in diameter.The thread is finally hung up to dry in the air with a weight of 18 grammes attached to its lower end. The dried threads are cut into about 0.2 metre lengths, and kept in closed tubes until required for testing. For the determination of the modulus the piece of thread is placed with its ends in the glass tubes carried by the two blocks (Fig. 2), which can be moved nearer to or further away f r o m e a c h other by means of t h e r i g h t FIG. 1, and left hand screw working in a bed beneath them. FIG. 2. The exposed lengthof the thread is next measured by a millimetre scale. The lowering of the thread, produced by hanging on it small S-shaped copper wires weighing 0.03 gramme, is determined by measurement with a cathetometer, or by using a low-power microscope with a micrometer eye- piece, the measurements being made through a screen of thin glass.Then, if E is the modulus, P the weight in kilogrammes, 1 the exposed Iength (rnillimetres) of thread, s the deflection produced by the weight, and r the radius of the thread, 13 E = x x --( = kilogrammes per square millimetre of section). 12 s rr4 The value of .the modulus for the thread alone was found by the authors to be 86.9, a quantity to be deducted from all the observations. Two independent effects take place in the coating of the thread with these colloids. The cotton in contact with the gum arabic, for instance, takes up the lime or gum, which becomes com- bined, and the modulus comes out too low. When this has ceased, only a stiffening effect takes place, and thus the results obtained on the stronger solutions are more nearly normal.The authors found the comparative values of potato starch, maize starch, and gum arabic by the modulus of elasticity to be as 3 : 4 : 7 (35.7 : 48.4 : SZ-S), and the prices of the two first-named were strictly in this ratio, Maize starch is too variable in price to admit of m y comparison. A. R. T. The Use of Sodium Peroxide i n the Qualitative Analysis of Organic Sub- stances. (D. Chenz. Ges. Bey.: 1904, xxxvii., 2155; through Chem. Zeit. Rep., 1904, xviii., 206.)-Advantage is taken of the oxidizing action of sodium peroxide for the identification of S, C1, Br, I, P, As, in organic compounds. The peroxide is mixed with one-twenty-fifth of its weight of a substance con- taining much C and H (naphthalene, cinnamic acid, etc.) and kept in well-stoppered bottles. H. H. Pringsheim.318 THE ANALYST. A small quantity of this oxidizing mixture with a few drops or grains of the substance to be tested is heated in a Bunsan flame. The cooled mass is dissolved in water and divided into two portions. One, rendered acid with HNO,, serves for the determination of halogens and phosphorus, in the other ; acid with HCI, sulphur and arsenic are determined. H. A. T. The Use of Pumice-Stone in the Incineration of Organic Matter. M. Duyk. ( A m . de Chim. anal., 1904, ix., 252-254.).-It is possible to obtain rapid and complete incineration of even the most refractory substances, such as albumin, yeast, wool, etc., by mixing them with an equal weight of coarsely powdered, recently ignited pumice-stone, and heating the mixture at it moderate heat in a muffle or over a gas flame. C. A. M.