after standing one day. Care was taken to have all the solutions of the same temperature. The results of polarization are contained in the following table:

Determinations of the effect of sodium hydrosulphite B. A. S. F. and rongalite C. on the polarization of dextrose, levulose, and sucrose.

Hydrosulphite has more effect on the polarization of the sugars than rongalite, and lowers the polarization of dextrose more than that of the other two sugars. With sucrose there is a change in the immediate polarization on the addition of 0.5 gram, but on standing one day, 0.25 gram reduces the rotation. Whether this action is due to an inversion of the sucrose, it is impossible to say at present. With levulose there seems to be a slight increase in the polarization, but in other experiments with this sugar no change in the rotation was noted. Rongalite has a slight action on the rotation of dextrose, but on sucrose and levulose there is apparently no action.

In regard to bleaching, hydrosulphite is much faster than rongalite, but it has the disadvantage that its decolorizations are not altogether permanent, and on standing a very short time sulphur is precipitated, making the liquid too cloudy for a reading. More work along this line is being done at the Bureau of Chemistry.

THE DETERMINATION OF SULPHUROUS ACID IN MOLASSES.

By FRITZ ZERBAN and W. P. NAQUIN.

At the last meeting of the association, in 1906, W. D. Horne called attention to several sources of errors in the provisional official method for the determination of sulphurous acid in foodstuffs when applied to sugar products. In the meantime, the food law has been enacted, and therefore an investigation into this question could no longer be deferred. The different methods that have heretofore been used were studied and compared. The work has not yet been completed, but it was deemed advisable to present the data collected so far mainly for the information of those who have to make determinations of sulphurous acid in sugar products. The work will be published later in full.

The following questions had to be studied: (1) Whether molasses contains any hydrogen sulphid or gives rise to it when boiled in the presence of free acid; and whether hydrogen sulphid and sulphurous acid can be given off

simultaneously without acting upon each other. (2) Whether there is any error due to outside sources (use of gas for heating, rubber for connections, tin for condensers). (3) Whether any volatile substances are given off during the distillation which are apt to be acted upon by iodin; and whether or not iodin is lost by volatilization. (4) Whether the use of a current of carbon dioxid is necessary. (5) Whether the total amount of sulphurous acid dis

tils over with the first half of the distillate.

According to our present knowledge of the composition of the sugar cane and its products most, if not all, of the organic sulphur is in the form of proteid sulphur. By boiling in an acid solution the proteids are decomposed, and so far the following cleavage products containing sulphur have been isolated: Cystin, cystein, a- and B- thio-lactic acid, thio-glycolic acid, ethylsulphid, methylmercaptan, and hydrogen sulphid. The thio acids mentioned decompose further upon prolonged heating, yielding hydrogen sulphid. The presence of hydrogen sulphid in the distillation products of molasses is therefore to be expected. This was suggested by W. D. Horne and also by Jerome Alexander in a paper on the determination of sulphurous acid in gelatin. The following experiments were performed to decide this question: Fresh cane juice was boiled down to sirup without the use of any chemicals; 100 grams were diluted to 400 cc, 5 cc of glacial phosphoric acid added, and distilled. The distillate was collected in a flask containing de Koninck's reagent for hydrogen sulphid (mercuric cyanid and ammonium chlorid in water). A black precipitate was obtained which after filtration was dissolved in bromin water. The solution gave a precipitate of barium sulphate when tested with barium chlorid. Experiments carried out in a similar manner with silver nitrate and lead acetate solutions gave the same results. After receiving a copy of Doctor Horne's paper, cadmium chlorid was used and our observations could be confirmed. The filtrate from the cadmium sulphid was also tested for sulphur, with positive results. The volatile sulphur which is not in the form of hydrogen sulphid may be due to the volatilization of substances of a mercaptan-like character. We have not been able so far, on account of the small quantities obtained, to identify them or to prevent their determination together with sulphurous acid. Similar experiments were performed with a molasses obtained by the sulphitation process as practiced in Louisiana. Hydrogen sulphid was again obtained, which proves that under the conditions of procedure sulphurous acid and hydrogen sulphid may coexist. It is, therefore, advisable to use a solution of cadmium-chlorid in the determinations, as recommended by Horne.

Gas can be used for heating the distilling flask without any risk. But if it be found necessary to concentrate the distillate, this should be done in flasks with narrow openings. Rubber connections should be avoided as much as possible, especially at that side of the condenser which is connected with the receiver containing iodin or bromin. It was found that the use of rubber may introduce considerable error. When two glass tubes are to be connected with rubber hose the two ends of the glass tubes should touch. In a number of distillations a Kjeldahl condenser with tin serpentines was used, and in this case a slight deposit of sulphur was invariably noticed in the glass tube extending into the receiver. When glass condensers were substituted for those of tin that deposit could not be observed. The deposit is probably due to the interaction of hydrogen sulphid and sulphur dioxid, but the tin seems to play some part in this reaction.

Doctor Horne called attention to the possibility that sugar products like molasses when distilled in acid solution might give off volatile organic substances which are acted upon by iodin. In order to test this question, a sample

of molasses was secured from a house where no sulphur is used. In one experiment the distillate was collected in bromin water after washing with cadmiumchlorid solution, and the sulphur determined gravimetrically in the cadmium sulphid as well as in the filtrate from it combined with the bromin water. In a second determination the official method was used, the dilution in the distilling flask and the quantity of liquid distilled over being the same as in the first experiment. It was found that the official method yielded considerably more "sulphur" than was obtained by the gravimetric determination. However, this discrepancy is partly due to the fact that in spite of the U tube provided for in the official method some iodin volatilizes during the distillation. This could easily be shown by connecting the U tube with a glass tube extending into a small flask which contained a solution of potassium iodid. When the distillation was finished the potassium iodid had taken up a considerable quantity of iodin, not enough, however, to make up for the difference from the gravimetric method.

Several experiments conducted with and without the use of a current of carbon dioxid proved the necessity of using this precautionary measure, although the influence was not so great as we had anticipated. The inlet tube for the carbonic acid gas must extend into the liquid to prevent it from distilling back into this same tube.

After these preliminary experiments the following mode of procedure was adopted, avoiding the errors recognized so far: Transfer 50 grams of molasses to a flask of sufficient capacity and make up to 400 cc with air-free water. Close the flask with a rubber stopper (rubber can safely be used at this point) having three perforations. One of these is for the inlet tube for the carbon dioxid, the second for a small separatory funnel filled with glacial phosphoric acid, the third for a catch-all connected by means of a good cork with a Liebig condenser. Connect the other end of the condenser with a glass tube leading into a filtering flask containing a small quantity of a 2 per cent solution of cadmium chlorid. The glass tube must terminate below the surface of the liquid. After the filtering flask follows a Ụ tube containing a little water and then a glass tube leading into bromin water in a little Erlenmeyer flask. The U tube prevents sucking back of the bromin water into the cadmium solution. After the receivers have been put in place, carbon dioxid is passed through the entire apparatus for about ten minutes. Then about 5 ce of phosphoric acid are added from the separatory funnel and 200 cc of the liquid distilled over. When the distillation is finished, the bromin water in the Erlenmeyer and the contents of the U tube are transferred to a beaker, and the cadmium solution containing the distillate is filtered into the bromin water. The excess of bromin is boiled off and the sulphuric acid is precipitated as barium sulphate and weighed. If the hydrogen sulphid is to be determined also, the cadmium sulphid is dissolved in bromin water and the solution treated like the others.

It was deemed inadvisable to distil to a smaller volume than 200 cc in order to avoid charring of the mass at certain points and subsequent reduction of sulphuric acid.

Finally the influence of concentration was studied and very surprising results obtained. When the same quantity of molasses was diluted in one case to 600 cc and in the other to 400 cc, both being distilled down to 200 cc, the first experiment gave a much higher result than the second. This experience led us to investigate how much of the liquid must be distilled to obtain the entire quantity of SO2. It was found that by adding again 200 cc of water after having distilled from 400 to 200 cc new quantities of sulphurous acid were obtained, and it was necessary to repeat this procedure four times to distil

over practically all of the SO2. In this case the last distillate contained only traces of sulphur. It was thus shown that by distilling over 800 cc the SO2 could be satisfactorily determined. The method would be facilitated by making up 50 grams of molasses to 1,000 cc and distilling down to 200 cc. In this case the receivers had naturally to be changed to some extent. Results agreed practically with those obtained by separate distillations. After the true amount of SO2 in a certain molasses had been determined in this way, the official method was tested for its accuracy. By distilling in the same way from 1,000 cc to 200 cc of a N/10 iodin solution for the collection of the distillate and titrating with thiosulphate, results considerably higher than those by the above method were obtained.

The necessity of distilling off 800 cc of liquid renders the method very tedious, and it would be desirable to devise one which requires less time and labor. It is possible that the sulphurous acid may be determined in some other way than distillation. Direct titration is impossible on account of the dark color of the products under examination, but some method of precipitation may directly or indirectly accomplish the purpose. It would also be desirable to find some method by which the different forms of sulphur dioxid in molasses could be distinguished.

Mr. Wiley called attention to the value and timeliness of Mr. Zerban's paper, inasmuch as both State and Federal food officials are concerned with the difficulties in distinguishing between natural sulphur and added sulphites.

Upon motion by Mr. Wiley, election of officers was made special order for 10 o'clock Friday; and on motion by Mr. Van Slyke reports of all special committees, except the committees on recommendations of referees, were made special order for 11 o'clock.

The association adjourned for the day.

THIRD DAY.

FRIDAY-MORNING SESSION.

REPORT ON MEDICINAL PLANTS AND DRUGS.

By L. F. KEBLER, Referee.

The burden of other work imposed upon the staff of the drug laboratory and other pharmaceutical chemists of the country in consequence of recent food and drug legislation, both National and State, has interfered somewhat with the prosecution of the cooperative work on drug assaying during the past year. It is hoped that when the most urgent problems have been solved the study of methods will receive an amount of attention commensurate with its fundamental importance. In the meantime it has been deemed advisable not to depend upon the cooperation of chemists generally in the work, but to continue the same within the drug laboratory as circumstances permit until the time is more propitious for general participation.

In view of the legal status conferred upon the United States Pharmacopœia, Eighth Revision, it seemed desirable to study all of the plant drug assays official therein, comparing them with some of the best available nonofficial methods. After this work was begun, the committee of revision of the Pharmacopoeia promulgated two lists of additions and corrections under date of May 1 and June 1, 1907, which introduced some minor changes in the details of certain methods. These changes were recognized in the work subsequently done.

The work this year covered methods for the assay of aconite leaves, aconite root, belladonna leaves, belladonna root, cinchona bark (yellow and red), coca leaves, colchicum corm and colchicum seeds. Samples of these drugs, delivered as being of United States Pharmacopœia quality, were made the basis of study, and the following programme of work was pursued by three collaborators in the drug laboratory. All calculations and solutions, except as otherwise specified, were based on the data of the United States Pharmacopoeia, Eighth Revision.

DIRECTIONS FOR THE WORK.

VOLUMETRIC SOLUTIONS.

Prepare a standard normal sulphuric acid, determining the factor by two or more of the following methods:

(1) The United States Pharmacopoeia method by titration with normal potassium hydroxid solution which has been standardized against purified potassium bitartrate.

(2) By titration of the ignition residue from a weighed portion of potassium bitartrate.

(3) By titration of a weighed amount of freshly ignited sodium carbonate. (4) By precipitation as barium sulphate and gravimetric determination. (5) By evaporation with slight excess of ammonium hydroxid, and determination after drying at 125° C. as ammonium sulphate.

From this normal acid prepare a decinormal sulphuric acid by dilution. Prepare a fiftieth normal solution of potassium hydroxid and determine the factor by titration with the decinormal sulphuric acid prepared as just directed.

DETERMINATION OF ALKALOID.

Total extraction method.

Into a 200 cc flask weigh 10 grams of the powdered drug, add about 75 ce of ether-chloroform mixture (5 to 1 by volume), rotate and add 5 ce of a mixture