ANALYTICAL EDITION
200
column of the table are computed from the measured potentials. The two figures for the pH of each solution indicate the reproducibility obtained on refilling the spiral. It will be seen that there is agreement within about 0.01 p H unit between these measured values and the published “accepted” values given in column 4 within the pH range 1 to 8. Above the latter value, however, there is the familiar deviation of the glass electrode. The accepted values are all on the basis that the potential of the tenth normal calomel electrode is 0.3376, The value for the pH of 0.1 N hydrochloric acid is from Clark ( I ) , that of the phthalate buffer from work and the pH values of the diethylbarbituric of the authors acid buffers from Michaelis (6). The pH of the 0.01168 N hydrochloric acid has been found by interpolating a value of the activity coefficient from the values given by Scatchard ( 7 ) . It seenis probable that within the range in which the glass electrode is effective the new measured values are as accurate‘as the published ones. The potential measurements recorded in Table I were obtained in a constant-temperature room a t 25” * 0.1” with the potentiometer and Compton electrometer described by MacInnes and Belcher In addition many routine measurements have been made in this laboratory on biological solutions using for the potential determinations the vacuum tube apparatus described by Hill ( 2 ) . With both these potential-measuring equipments insulation and electrostatic screening have been carefully studied. Although the authors do not suggest any relaxation of care with respect t o these factors, their importance is not as great as it sometimes is in work of this kind, because the glass elec-
(e),
(e).
Vol. 5, No. 3
trodes described in this article have comparatively low resistances. TABLEI. COMPARISON OF ACCEPTED AND MEASURED PH VALUES O F BUFFERSOLUTIONS MEASURED E. M. F. PH VALUE Measured Accepted -0,1360 1.06 1.06 -0,1362 1.06
SOLUTION 0.10 N Hydrochloric acid 0.01168 N Hydrochloric acid
-0,0807
-0.os12
1.99 1.99
1.9s
0.05 N Potassium acid phthalate
0.0370 0.0367
(3.98)
3.98
Diethylbarbiturate buffer
0.2143 0.2137
6.98
6.98
7.00
Diethylbarbiturate buffer
0.2753 0.2743
8.00
8.01
8.00
Diethylbarbiturate buffer
0.3297 0.3289
8.93 8.92
9.00
Diethylbarbiturate buffer
0.3601 0.3614
9.44 9.45
9.60
LITER.4TURE
CITED
(1) Clark, W. M., ”Determination of Hydrogen Ions,” 3rd ed., Williams and Wilkins. 1928.
( 2 ) Hill, S. E., Science, 73, 529 (1931). (3) Kahler, H., and De Eds, F., J . Am. Chem. Soc., 53, 2998 (1931). (4) MacInnes, D. A., and Belcher, D., Ibid., 53, 3315 (1931). ( 5 ) MacInnes, D. A., and Dole, M . , IND.ENG.CWEM., Anal. Ed., 1, 57 (1929); J. Gen. Physiol., 12, 805 (1928-29); J . Am. Chem. Soe., 52, 29 (1930). (6) Michaelis, L., J. Biol. Chem., 87,33 (1930). (7) Scatchard, G., J . Am. Chem. Soc., 47,648 (1925). (8) Thompson, M. R., BUT.Standards J . Research, 9, 833 (1932). RECEIVEDFebruary 11, 1933. ~
Estimation of Dextrin in the Presence of Glue JEROME ALEXANDER,50 East 41st St., New York, N. Y.
A
DMIXTURE of dextrin powder with glue may be readily
detected by microscopic examination, using mineral oil or some high-boiling alcohol as a mounting medium. However, if the dextrin has been dissolved in the glue liquor prior to drying the glue, the addition is not so obvious. Some dextrins will not respond to the iodine test, especially if the glue is highly colored; but any dextrin may with certainty be detected by hydrolyzing the mixed glue with dilute hydrochloric acid and then testing the liquid for dextrose with Fehling’s or Benedict’s qualitative solution. The protective colloid action of the protein and its hydrolysis products tends to make the copper oxide precipitate so fine that it appears yellow, and if all the copper is not reduced, the solution may appear to be green. For the precise estimation of dextrin in glue, the follow. ing method proved satisfactory: Soften 0.6 gram of the glue under test by soaking in 10 cc. of dilute hydrochloric acid (5 parts of concentrated hydrochloric acid t o 100 parts of water). Hydrolyze the mixture slowly at low tem erature (overnight a t about 60” C.). After neutralization mafe up the solution to exactly 50 cc., and determine the dextrose content by Benedict’s quantitative method. BENEDICT’S QUANTITATIVE SOLUTION (2). Dissolve 200 grams of sodium citrate, 200 grams of sodium carbonate, crystallized (equivalent to 75 grams of anhydrous sodium carbonate), and 125 grams of potassium thiocyanate (sulfocyanide) in enough hot distilled water to make up t o about 800 cc. Dissolve separately 18 grams of purest copper sulfate crystals (air dried) in about 100 cc. of distilled water, and pour into the first solution with constant stirring. Add 5 cc. of a 5 per cent solution of potassium ferrocyanide, and then make up to 1000 cc. with distilled water. The solution is stable, and 25 cc. are reduced by 0.05 gram of dextrose.
To make the analysis, a 150-cc. flask is held in a stand over a Bunsen burner, a t such a height that the reagent solution can be kept briskly boiling by a small flame. In this flask are placed 3 to 4 grams of anhydrous sodium carbonate and 25 cc. of the reagent, the mixture being heated until most of the carbonate is in solution. The glue solution is now run in slowly, until a chalk-white precipitate is formed and the blue color lessens perceptibly in intensity. The glue solution is now run in still more cautiously, with constant boiling, until disappearance of the last trace of blue indicates the end point. If less than 5 cc. of glue solution has been used, the solution is accurately diluted so that about 10 cc. will be necessary, and the titration repeated with this diluted solution. Should the mixture in the flask become too thick, boiled distilled water may be added to replace that lost by evaporation. In using this method, no trace of reduction was shown by a glue known to be pure, which was run as a check or blank in the quantitative test, after having been proved free from reducing substances by qualitative test. Mixtures of this glue with several commercial dextrins gave concordant results. In converting the dextrose figure to dextrin, it should be remembered that commercial dextrins contain variable percentages of water, dextrose, etc. The dextrose figure multiplied by the factor 0.9 gives a close approximation to the amount of dextrin present (1). (1) Allen, “Commercial Organic Analysis,” 5th ed., Vol. I, p. 530, Blakiston, 1923. (2) Cole, S. W., “Practical Physiological Chemistry,” 6th ed., Simpkin, Marshall, Hamilton, Kent and Co., London, 1920. RBCEXYEDFebruary 10, 1933.