in fruits, pure fruit jellies, jams, and honies in which large amounts of invert sugar are present. The action of acid on invert sugar mixed with glucose is the same as it is on the same amount of invert sugar alone, as the following results show: 13.028 grams invert sugar + 25 cc. glucose solution This shows that when glucose which contains invert sugar is polarized before and after inversion, there will be a change of polarization even if there is no sucrose present. This condition would be almost always found in honey adulterated with glucose. This obviously can not be corrected by the formula given, and the writer is working on the use of inverting agents, such as citric acid, which have practically no effect on the opticity of invert sugar. These results show that if you eliminate the effect of the acid used in inversion, the Clerget formula will become a constant for any given temperature regardless of the concentration of the sugar solution. This can be seen by noting from Fig. I that the increase of invert reading caused by the hydrochloric acid is 0.05 of the invert reading, so that Herzfeld's formula would be if the effect of the acid was removed, or a constant at any temperature. If an inverting agent could be obtained which had no effect on the invert sugar, the question of concentration would practically be eliminated. But the elimination of this error would necessitate the determination of a new factor which would in all probability be 141.79 or perhaps a little less. From this work it will be seen that the smaller the amount of hydrochloric acid used, the less is the effect on the negative reading, and therefore the smaller the error due to concentration. It will be seen from Table II that the effect of N/4 hydrochloric acid is comparatively small, and the factor for any given temperature approaches a constant for all concentrations. All will agree that this is a most desirable condition for attaining accuracy. The German method of using 1⁄2 normal weight (13.024 grams) and 5 cc. of hydrochloric acid reduces the error, as compared with the A. O. A. C. method, due to concentration, to a considerable extent, for two reasons. The smaller amount of acid has less effect on the negative reading and it has been shown from Table II that the smaller the amount of sugar used the less is the variation. As, for instance, the factor varies from 141.85 to 142.66, for I gram to 13 grams of sugar, a variation of 0.81°, and with the A. O. A. C. method it varies from 141.85 to 144.00, 2.1°. The use of less acid for the inversion is desirable and it is in this direction that investigation should tend. The smaller the amount of the acid the less is the danger of destruction of the sugar and the smaller the error due to concentration, but it must be remembered that every change in the strength or amount of acid used requires the most careful determination of a new factor. The work can be summed up in this way: 1. Hydrochloric acid increases levo-rotation of an invert sugar solution. 2. This increase, other things being equal, is proportional to the quantity of hydrochloric acid used. 3. Other things being equal and temperature varying, hydrochloric acid increases levo-rotation by a definite per cent. of the polarization. 4. In order to correctly calculate the percentage of cane sugar in invert sugar by Clerget's formula a correction depending on the amount of hydrochloric acid used must be made, which can be calculated from Fig. 1 or 2. All readings of the polariscope should be made at or about 20°. THE SUBSTITUTION OF HYDROGEN FOR CHLORINE IN TRICHLORMETHYLPARACONIC ACID. (SECOND PAPER.) BY HENRY C. MYERS. Received February 22, 1902. IN the Journal of the Chemical Society for June, 1897, I called attention to some extremely unstable condensation products resulting from the reduction of dichlormethylparaconic acid in attempting to eliminate the remaining chlorine atoms. One compound being monochlordiparaconic acid, C,H,CIO,, and another, also an acid, having the formula C,H,,O2. Secondary to other published work, I have continued these investigations. Finding that the reduction of the dichlor acid by various agents was not in the direction of methylparaconic acid, which substance I hoped to reach by this method, I decided to treat the trichlor acid with sodium amalgam, the trichlor acid being readily prepared on condensation of chloral with sodium succinate in the presence of a suitable dehydrating agent.1 Trichlormethylparaconic acid was dissolved in water, treated with sodium hydroxide to retard the action of free acid, small pieces of solid sodium amalgam added occasionally and the cylinder surrounded with ice. Frequently during a two-days' treatment, sulphuric acid was added to prevent the solution becoming more than weakly alkaline. Eventually sulphuric acid was added in large excess and the mixture extracted with ether; on recrystallization from water the result was found to be an almost quantitative yield of the dichlor acid. This method, as regards the preparation of dichlormethylparaconic acid, is much simpler than the laborious process of reducing with zinc dust and acetic acid and finally precipitating with hydrogen sulphide as described by Miller (Loc. cit.), large amounts of zinc being precipitated completely only on repeated treatment with hydrogen sulphide, and filtration being extremely slow and unsatisfactory. 1 Miller: Ber. d. chem. Ges., 23, R. 92 (1890); Ann. Chem. (Liebig), 255, 43. LONG-CONTINUED REDUCTION. The above treatment with sodium amalgam when continued for a week or ten days resulted in three definite compounds, one of which has never been previously observed; (a) the dichlor acid with its characteristic properties and melting-point of 142°; (b) monochlordiparaconic acid of orange color and melting-point of 220°, obtained by recrystallization from alcohol; (c) a new acid in relatively small amounts with melting-point of 126°-127° and found in the final ether extraction as follows: On completion of the treatment with sodium amalgam the solution was acidified with hydrochloric acid. The monochlordiparaconic acid being insoluble in water formed an extremely voluminous precipitate which was filtered off and recrystallized from alcohol. The filtrate on extraction with ether yields the white crystalline dichlor acid, readily purified by crystallization from water. Repeated extraction with ether gives a yellowish oil which on long standing over sulphuric acid forms well defined crystals. These on being recrystallized from water and washed with ether have a constant melting-point of 126°-127°, resemble benzoic acid in appearance, do not decompose on standing for a year, are soluble in sodium. carbonate solution, acid to litmus, and do not decompose on repeated melting. BEHAVIOR OF MONOCHLORDIPARACONIC ACID TOWARDS HEAT. In making melting-point determinations of this acid in the usual way, with capillary tube and paraffin-bath, it was noticed that the acid on melting underwent decomposition giving off a gas in small bubbles. In order to observe the change more closely, a long slender test-tube was substituted and considerable material used. In this way it could be observed that the substance really began its decomposition at about 190°, giving off fumes of hydrochloric acid and depositing white needle-like crystals on the tube just above the paraffin. I doubt if the true conditions of meltingpoints so called, can be observed in minute capillary tubes. These observations resulted in my placing about 3 grams of material in a U-tube heating the same as high as 226°, meanwhile drawing air through the apparatus. The air was previously freed of moisture and carbon dioxide; the products of the change were drawn through calcium chloride, silver nitrate and caustic potash. Sublimation began at 181° followed by an effervescence of gas at 190°. The following figures show the total loss in weight in the U-tube and the corresponding increase in weight in each of the receiving tubes: CALCULATED Loss CORRESPONDING TO I MOLECULE OF EACH GAS USING As the weights of water and carbon dioxide are too low for the loss of a single molecule and the total gain corresponds almost exactly with the loss of a single molecule of hydrochloric acid (calculated 0.5830, found 0.5885) I assume that the change is represented by the following reaction: and that the increase in weight in the calcium chloride and caustic potash tubes were due to hydrochloric acid alone. CONCLUSION. Trichlormethylparaconic acid on reduction with either zinc dust or sodium amalgam in the presence of acids produces the dichlor acid in almost quantitative amounts, except where treatment is long continued with the amalgam. In the latter case, three distinct and well defined acids result; viz., Dichlormethylparaconic acid, monochlordiparaconic acid, and a new acid with melting-point at 126°-127°. The dichlor acid on reduction produces monochlordiparaconic acid, according to the following equation :1 2С ̧H ̧O ̧Cl2 = C„H„O2C1 + 3HCl + 3CO2. 1J. Chem. Soc., June, 1897. |