THE QUANTITATIVE DETERMINATION OF ALKALOIDS BY MEANS OF IMMISCIBLE SOLVENTS BY MIRIAM GERTRUDE BUCK B. A. University of Illinois 1920 THESIS Submitted in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE IN CHEMISTRY IN THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS 1921 ' \H^\ ft UNIVERSITY OF ILLINOIS THE GRADUATE SCHOOL Se ptember 26 1Q9 1 f i I I ■4 ft ? ft n I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY- G ERTRUDE BUCK ENTITLED T HE QUANTITATIVE DETERM I NAT 1 ON ON ALKALOI DS BY MEANS 0? IMmISCIELE SOLVENTS BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF ...mASTE R . 0 E_ SCI EN CE Recommendation concurred in* Committee on Final Examination* *Required for doctor’s degree but not for master’s . ■ ■ I wish to express ray sincere thanks to Professor G-.D.Beal for the suggestion of this problem and the interest he has shown during the experimental work. Digitized by the Internet Archive in 2015 https://archive.org/details/quantitativedeteOObuck TABLE OP CONTENTS. I. INTRODUCTION: DISCUSSION OP THE PROBLEM p. 1 I I. HISTORICAL. I I I. EXPERIMENTAL. 1. Preparation of the tartrates of nicotine. 2. Preparation of nicotine hydrochloride. 3. Preparation of pure nicotine. 4. Determination of the equilibrium conditions for the partition of nicotine between aqueous neutral and acid solutions and an immiscible solvent ( ether ) : (at) Extraction of the neutral or acid aqueous solu- tion with ether. (b) Conditions of equilibrium in systems in which the alkaloid is being removed from its ether solution by an acid. (d) Extraction, with ether, of nicotine salt solu- tions to which sodium hydroxide has been added in amount equivalent to the acid in the salt. IV. SUMMARY. 12 BIBLIOGRAPHY. 13 OJ vo . •< t . THE QUANTITATIVE DETERMINATION OF ALKALOIDS BY MEANS OF IMMISCIBLE SOLVENTS. I. INTRODUCTION: DISCUSSION OF THE PROBLEM. One of the oldest and most widely used methods of alkaloL - al assay is based upon the principle that the majority of alka- loids are insoluble or only slightly soluble in water while their salts are insoluble in the organic solvents which are immiscible with water. This method is known as the "shaking out" process, made familiar by the " Dragendorff " ( 1 ) method of plant analy- sis and the "Stas-Otto" ( 2) method for no i son assay. The accur- acy of the method depends on several assumptions : first , that the alkaloidal salt is insoluble in organic solvents; second, that the free alkaloid is insoluble in neutral aqueous solutions; third, that there is no hydrolysis of the alkaloidal salt by the water present; and fourth, that there is no decomposition of the alkaloid by the organic solvent. Dra.gendorff and Otto must have recognized that the principle was not absolute as they make men- tion of individual exceptions. In most alkaloidal assays of this type, the sample is first extracted with dilute acid. This dilute acid solution is then shaken with an organic solvent to remove such substances as, col- oring matter, essential oils, certain organic acids, bitter prin- ciples, tannins, etc., which would later interfere with the quan- titative determination of the alkaloid. The solution is then made alkaline and shaken with an immiscible solvent. Purifica- tion is carried out by shaking the organic solvent layer with can acid, making alkaline, and again shaking with the immiscible sol- vent. This is repeated several times. The organic solvent is then removed by evaporation and the residue determined by direct weight, by dissolving in standard acid and titrating the excess acid with standard alkali, or by the uie of special precipitation reagents for individual alkaloids. Beal and Lewis ( 3 ) eall attention to several sources of er- ror which are introduced during the process. In the first place, the alkaloidal salt may be slightly soluble in the organic sol- vent. Then, too, there is the possibility of the alkaloidal salt being hydrolyzed, by the water present, into free alkaloid and acid, and this free alkaloid would be readily soluble in the or- ganic solvent. There are a few cases where the organic solvent decomposes the alkaloid or combines with it. These factors cause a decrease in the amount of alkaloid in the acid solution and a corresponding decrease in total alkaloid at the end of the assay. Finally, the shaking out of the alkaloidal solution in organic solvent introduces the possibility of the insolubility or slow solubility of the salt formed in the acid solution. ' - ..... ; • v . . ' . , . < .. 4 . . . X • t , ; . « . X ■ . ' . ?• as* 2 . It is the nurpose of this work to obtain results showing the partition Of the alkaloid, nicotine in particular, between the acid layer and the immiscible solvent using hydrochloric and tartaric acids under different concentration conditions. Equilibrium conditions were determined by shaking the acid solu- tion of the alkaloid with the organic solvent and also by shak- ing the solution of the alkaloid in the organic solvent with an acid . Results are expressed in terms of the " extraction factor", a term introduced by Beal and Lewis (4), in their work of the same nature. By "extraction factor" is meant the ratio of the amount of alkaloid in the layer of added solvent to the amount originally present in the first solution. The extraction factor shows at a glance the completeness of the extraction, an indicar- tion of the value of the extraction under those conditions. The partition ratio, the term used in the past, tells only the partition of one molecular species between two layers of equal volume, by definition of the term, partition ratio. II. HISTORICAL. Dr. C. Kippenberger (5)> in 1897, first called attention to the quantitative solution of the problem. He states clearly the possibility of error due to the hydrolysis of the salt. The use of Chloroform or a mixture of chloroform and a little alco- hol is suggested for a solvent. Three years later, Kippenberger (6), published a second pa- per in which he endeavored to establish the question on a firm- er basis. He worked with the alkaloids, strychnine, brucine, atropine, morphine, aconitine, veratrine, papaverine, narceine, codeine, emetine, pelletierine, cocaine, quinine, narcotine, sparteine, thebaine, hyoscyamine, daturine, scopolamine, and the base caffeine. He used chloroform and ether for shaking out. The action of the salts of the following acids was studied: hydrochloric acid, 21.9$ HC1; sulphuric acid, 40.1$ H^SCL; tar- taric acid, and oxalic acid. The effect of the addition of so- dium chloride to the acid solution was observed in some cases. To the acid solution, he added the immiscible solvent. Af- ter extraction, he separated the two layers, evaporated off the organic solvent and determined the amount of alkaloid and alka- loidal salt extracted. This was done by dissolving the residue in N/50 standard acid and titrating the excess acid with N/50 standard alkali. The value for acid neutralized by alkalbid in- dicated the amount of free alkaloid present in the residue. , . * . . 4 , . . • . • * 3. The solution was then made alkaline with an excess of sodium hydroxide and again extracted with chloroform. The amount of to- tal alkaloid was obtained by evaporation of the solvent and treat- ment with standard acid and alkali as before. Subtraction of tte first value, or free alkaloid present, from the second value, or total alkaloid, gave the amount of alkaloid present as salt. In 1901, Hans Proelss (7) gave a short description of the behavior of alkaloidal solution toward different solvents. He endeavored to determine first, the best solvent for alkaloids as a class, and second, the best solvent for individual alka- loids. He compared the relative extractive powers of ether, chloroform, benzene, and also mixtures of ether and chloroform, and alcohol and chloroform for the alkaloids, picro toxin, ber- atrine, strychnine, atropine, codeine, and morphine. He states that constant results could not be obtained of sufficient accur- acy to be anything more than comparative. His conclusion is that the best shaking out liquid for alkaloids in general is chloroform because of the solubility of most alkaloids in that solvent. Ed. Springer (8), in 1901, studied the effect of the sol- vent chloroform on the extraction of the following alkaloids: morphine, coniine, narcotine, strychnine, quinine, codeine, vereratrine, and cocaine from solutions made acid with sul- phuric, phosphoric, hydrochloric, tartaric, acetic, oxalic, and citric acids. The amount of alkaloid in the residue, after evaporation of the chloroform, was determined by titration in the same way as by Kippenberger . His work is of no value from a quantita- tive stand-point as he was unable to obtain checked results. In 1906, Simmer (9) published an important paper divided in three parts: 1 . The behavior of the salts of the common alkaloids to- ward extraction by chloroform and other important solvents . 2. The appearance of decomposition through treatment with chloroform. 3. The reducing action of alkaloids. He found that many neutral salts are extracted by both chloroform and benzene; this is especially true of the salts of nitric and the halogen acids. Many authors had stated that the extraction of alkaloids by means of chloroform is attended with a decomposition of the chloroform giving rise to free hydrochloric acid. Experimenting with the following alkaloids: atropine, brucine, quinine, cincho- nidine, cinchonine, cocaine, codeine, morphine, narcotine, nico- tine, strychnine, veratrine, Simmer found that the action of the alkaloid on the chloroform is negligible except in the cases of < I * . - t . , ' ' , . . * 4. brucine and veratrine. In 1914, Harden and Elliott(IO) published a paper on the methods of extraction with immiscible solvents from the point of view of distribution ratios. They worked with the alkaloids, aconitine, atropine, codeine, morphine, quinine, and strychninq, using chloroform and ether as solvents. Ammonium hydroxide was used to make the solutions alkaline. Prom the distribution coefficient and a certain subsequent algebraic calculation, they could determine the number of ex- tractions necessary to remove 99.9 per cent of the alkaloid. The distribution ratio (d) is indicated by the expression Concentration in 10 cc . of water _ C _ , » Concentration in lOce. of non-aqueous solvent C ~ c The algebraic expression for the calculation of the num- ber of shakings necessary for an extraction indicated by btnrr ( — — Y* where X 0 \ e- da ) a = volume of the aqueous solvent. e = volume of non-aqueous solvent. d = distribution ratio. X 0 = original amount of material to be extracted in the aqueous layer. Xjj= amount of material in the water layer after"n" extractions . Beal and. Lewis (11) established equilibrium conditions for the following systems in the case of the alkaloids, aconitine, atropine, brucine, cinchonidine, cinchonine, cocaine, codeine, morphine, quinine, strychnine, and. veratrine :- (a) The alkaloidal tartrates, tartaric acid., water, and chloroform. (b) The alkaloidal tartseates, tartaric ax id, water, and ether . (c) Certain alkaloidal sulphates, sulphuric sxid, water, and chloroform. (d) Certain alkaloidal sulphates, sulphuric acid., water, and ether. (e) Certain alkaloidal hydrochlorides, hydrochloric acid, water and chloroform. They have determined the extraction factors for all of these systems, as well as those described in the literature, and have calculated the most favorable conditions for extraction. They state that the neutral tartrates were prepared by dis- . . . . . , 4 • - 5. solving the alkaloids in an aqueous acid solution, containing equivalent amounts of tartaric acid in a large excess of water at the boiling temperature. The acid salt which forms first stays in solution and the remainder of the alkaloid completely neutralizes it. On cooling the solution slowly, the neutral salt comes out in beautiful crystals. In one or two cases, it was necessary to evaporate some of the solvent water in order to get the right concentration for crystalization . They also state that the mon-acid may be prepared by dis- solving the alkaloid in a slight excess of acid in a small quantity of hot water. On cooling, the crystals of the acid salt will come out. Allen ( 12) states that nicotine forms two classes of salts, the mon-acid and di-acid salts. He further states that most of salts of nicotine crystalize with difficulty, the acid tartrate, to which he gives the formula C ## H /f N a ( C^H fc O^ ), -f- 2 H^O , being an exception, it forming in handsome tufts when ether is added to it*s alcoholic solution. Hudson (13) has investigated the solution of nicotine in water and found that five per-cent of nicotine is soluble at any temperature, and fifteen per-cent of water in nicotine. Until 1911 , the method for the determination of nicotine, as adopted by the Association of Official Agriculture Chemists was the Kissling Method (14). This method involves two series of operations, ( 1 ) extraction with ether, ( 2 ) distillation with steam and titration. This method in experienced hands and in the absence of certain interfering substances, unquestionably yields good results. The final determination is a volumetric one and the end point is rather indefinite especially in the titra- tion of distillates from eo imereial nicotine preparations. In 1909, Bertrand and Javillier (Ip) published a paper on the use of silico-tungstic acid as an aid m the quantitative determination of nicotine. In their method, the nicotine was precipitated with silico-tungstic acid, the nicotine liberated from the nicotine silico- tungstate by the addition®*ealcined . magnesia, the mixture distilled with steam and the nicotine determined in the distillate by titration with standard acid. In conclusion, Bertrand afcd. Javillier suggested that the nicotine might again be precipitated from the distillate with silico-tungstic acid and determined gravimetrically by ignition to Si-C^ and WC^ . They quote no experimental work on this theoret- ical consideration. Chapin (16) reports the development of this method as car- ried out by the United States Bureau of Animal Industry. It gives results which agree favorably with those obtained by the 6 . "i Kisslin lethod. They give 0.114 as the factor for obtaining the amount of nicotine from the residue of silico-tungstic oxide III. EXPERIMENTAL. 1. Preparation of the tartrates of nicotine. The attempt was made to prepare the tartrates of nicotine as suggested by Beal and Lewis. For the neutral salt the alkaloid was dissolved in an aqueous acid solution containing an equiva- lent amount of tartartic acid in a large excess of water at the boiling temperature. For the acid salt, the alkaloid was dis- solved in a slight excess of acid in a small quantity of hot water. Neither salt would crystallize out either upon cooling, concentrating and cooling, or placing in a freezing mixture. Salting out with Rochelle salts was then tried. A salt crys- tallized but upon analysis and microscopiG.nl examination was proved to be only impure potassium bitartrate. The next procedure was to dissolve the tartaric acid in alcohol and add the nicotine. The solution was then concentrated on a steam bath. The salt would not crystallize until ether was added to the alcoholic solution. No matter what proportions of nicotine and tartaric acid were used, the analysis of the salt showed it to be the impure acid tartrate. The salt did not form in handsome tufts as stated by Allen but formed in a pink gum- my mass which crystallized upon standing in the refrigerator. This salt, as prepared using .an excess of tartatic acid, con- tained 32.68 per cent nicotine. The theoretical per cent of nic- otine, if C jo H I4 N- (CuH^O. ) x + 2 HjjO is the formula, is 32.54 per cent, if O lt ' ' C H H fc i s the formula, is 35 . 07pereent . Svi- dentally the crystals contain two molecules of crystallization. 2. Preparation of nicotine hydrochloride. Nicotine was disseized in anhydrous ether and dry hydrochlo- ric acid gas passed into the solution. The hydrochloride separa- ted in a gummy mass which became crystalline upon drying in a vacuum dessicator. This salt upon analysis showed 30.62 per cent hydrochloric acid, showing that the acid hydrochloride was the one which formed. Theoretic.ally , the anhydrous salt should be 31.04 percent hydrochloric acid and the salt with one molecule of v/ater of crystallization should be 28.83 per cent hydrochloric ac i d . 7. 3. Preparation of pure Nicotine. As the supply of pure nicotine was limited, it v/as neces- sary to prepare some from Black Leaf Forty, an insecticide con- sisting of a forty per cent solution of nicotine sulphate. A perforator was designed, for this purpose. The solution of Black Leaf Forty, made decidedly alkal ne with a concentrated solution of sodium hydroxide, was placed in flask A, Fig. 1. In flask B, which was placed upon a steam bath, was an almost saturated solution of tartaric acid., Ethei poured into flask A through the condenser until it flov/ed out of the arm '. As the ether passed upward through the tartaric acid, the latter ex- tracted the nicotine leaving the fats and other substances in the ether. As flask B was heated, the ether distilled through arm D and ran back into flask A where it extracted more nico- tine and again ran into flask B. When the solution of tartaric 3,cid became saturated with nicotine, it was renewed. The acid solution was then made alkaline with sodium hydroxide and shaken several times in a separatory funnel with ether to remove the nicotine. The ether was evapora- ted off in a vacuum dessicator and the nicotine distilled under vacuum. Two samples of nicotine, thus prepared, analysed 98.81 per cent and 97.23 per cent pure nicotine. 4. Determination of the equilibrium conditions for the partition of nicotine between aqueous neutral and acid solutions and an immiscible solvent (ether ) . (a) Extraction of the neutral or acid aqueous solution with ether. A weighed amount of nicotine v/as washed into a two hundred and fifty cubic centimeter volumetric flask, a calculated amount of normal acid was added and the solution made up to the mark with distilled water. Twenty-five cubic centimeter portions were m e a s ured into flasks, twenty-five cubic centimeters of ether added, the flasks stoppered and placed in a water thermostat and shaken for about two and one half hours. The thermostat, which was accurate to within a tenth of a degree, v/as maintained at a temperature of twenty-five degrees. After the time of shaking had elapsed the flasks were re- moved, and ten cubic centimeters of the ether layer pipetted into a beaker containing a small amount of water and about five cubic centimeters of dilute hydrochloric acid. After stirring, the ether was evaporated off on the steam bath. The solution was then diluted to about seventy-five cubic centimeters and an excess of silico-tungstic acid v/as added. It was then stirred . . . 8 . 1 until the precipitate which was at first amorphous became crys- talline. The beakers were then removed from the hot plate, al- lowed to stand over night, the contentc^’iltered on quantita- tive filter paper, and the precipitate* with cold water contain- ing one cubic centimeter of concentrated hydrochloric acid per liter. The filters were then ignited in an electric muffle at bout 850 ° to constant weight. The amount of nicotine extract- ed was determined by use of a factor 0.114 and the extraction factor calculated. TABLE I. Solution containing one mole of tartaric acid to four moles of nicotine. Wt. of nicotine in 10 cc. solution Length of extraction. Wt.of nicotine extracted. “ T Extraction factor. 0. 042476 gm. 2- hr. 45niin. 0,016006 gm. 0.3768 0.042476 2hr . 45min. 0.016006 0.3768 0.042476 2 hr. 45ciin. 0.016006 0.3768 0.042476 2 hr. 4 pmin . 0.016051 0.3779 0.042476 2 hr. 50min . 0.016300 0.3838 0.042476 2 hr. 50m in . 0. Cl 6200 0.3814 0.042476 2 hr. 50min. 0.016210 0.3816 0.070300 2 hr. 30min . 0.026220 0.3735 0.070300 2 hr. 30min. 0.026500 0.3770 0.070300 2 hr. 30min. 0.025450 0.3762 0.070300 2 hr. 30min . 0.0264-0 0.3762 Average 0.3780 TABLE II. Solution containing moles ; one mole 0 of nicotine f hydrochloric acid to two • 7t. of nicotine Length of in 10 cc. solution. extraction. Wt. of nicotine extracted . Extraction factor. 0.05692 gm. 0.05692 0.05692 2 hr. 2 hr. 2, hr. 50 min. 50 .min. 50 min. 0.02231 0.02247 0.02256 Average 0.3917 0.3949 0.3963 0.3943 / ) 9. There was no extraction from solutions containing as much or more than one half mole of tartaric acid to one mole of nico- tine. The extraction factors obtained for a solution containing one mole of tartaric acid to four moles of nicotine are fouftd in TABLE I. There was no extraction from a solution containing one mole of hydrochloric acid to one mole of nicotine. The factors obtain- ed for a solution containing one mole of hydrochloric acid to two moles of nicotine are found in TABLE II. They are slightly higher than those obtained from the tartaric acid solution of equivalent strength which are found in ’EableJ.. (b) Conditions at equilibrium in systems in which the alka- loid is being removed from its ether solution by an acid • The quantity of nicotine was weighed into the shaking flasks;, twenty-five cubic centimeters of ether added and the flasks shaken to dis&dlve the nicotine in the ether. Twenty-five cubic centimeters of normal acid were then added and the flasks shaken in the thermostat. Runs were made to see if there was more extraction by the acid at the end of an hour than at the end. of a half hour but no appreciable difference was observed. The results of the ex- tractions with hydrochloric acid are given in TABLE III and those with tartaric acid in TABLE IV. TABLE III. Solution of nicotine in ether shaken with N/ 1 HC1. Wt. of nicotine Wt.of nicotine Extraction in 25 cc. ether. extracted . factor. 0 0. 1494 gm. 0 . 133465gni. 0.8933 0. 1327 0. 1 18202 0.8952 0. 1442 0. 129076 0.8951 0. 1440 0. 123889 0.8849 0. 1518 0.136372 0.8983 Average 0.8934 i . . ■ . , ' 10 . TABLE IV. Solution of nicotine in ether shaken with N/1 HTar. Wt.of nicotine in 25 cc. ether. 1 - - - - Wt.of nicotine extracted . Extraction factor. 0. 1080 grn. 0.09670 gm. 0.8954 0. 1 167 0. 10275 0.8799 0.0824 0.07097 0.8650 0. 1 177 0. 10887 0.9252^ 0.2014 0. 18440 0.9155 from Black Leaf Forty 0. 1173 0.10912 0.9315 r average 0.9229. 0.1302 0. 1 1950 0.9191 0.1277 0. 1 1780 0.9233J Average 0.9068 By use of the extraction factor, it is possible to deter- mine how many extractions are necessary for complete removal of an alkaloid. For instance, if there were one gram of nico- tine in the ether solution, the first extraction with N/1 Hy- drochloric acid would remove 89.34 per cent of one gram or 0.8934 grams leaving 0.1066 grams. The second extraction would remove 89.34 per cent of the remaining 0. 1066 grams leaving 0.0114 grams. The third extraction would remove 89.34 per cent of the remaining 0.0114 grams and leave only 0.0012 grams or 0,12 per cent. From this, it is evident that three extractions would give practically complete removal of the nicotine. (c) Extraction, with ether, of nicotine salt solutions to which sodium hydroxide has been added in amount equivalent to the acid in the salt. The nicotine tartrate was weighed into the flasks, standard sodium hydroxide added to neutralize the acid** J di stilled water added to bring the volume to twenty-five cubic centimeters. An equal volume of ether was then added and the whole shaken in the thermostat for two hours. In the case of the hydrochloride, a calculated amount of normal hydrochloric acid was s.dded to the nicotine and standard sodium hydroxide added to neutralize . it. For some reason, close checks were not obtained. The results are given in TABLE V and TABLE VI. 1 1 . TABLE V. Nicotine tartrate in solution . NaOH added to neutralize HTar. Wt.of salt in 25 cc. solution. Corresponding Wt.of nicotine v/t.of nicotine. extracted. Extraction factor. 0.2941 gin. 0.0961 gm. 0.06765 gin. 0.7038 0.3652 0.1153 0.08880 0.7450 0.4749 0.1552 0. 1 1697 0.7535 0.3297 0. 1077 0.08137 0.7558 0.2844 0.0929 0.07002 0.7568 0.4143 0. 1354 0. 1C4C0 0.7648 0.3544 0. 1158 0.09147 0.7900 0.3995 0.1305 0.10355 0.7935 0.4585 0. 1498 0. 1 1925 0.7970 0.3348 0.1094 0.08755 0.7985 : 0.3623 0. 1 184 0.09537 0.8050 Average 0.7694 Nicotine plus TABLE VI. N/1 HC1 to make C H N (HC1) . NaOH added to neutralize HC1. Wt.of nicotine in Wt.of nicotine Extraction 25 cc.of solution. extracted . factor . 0.08322 gm. 0.060977 gm. 0.7330 0.08322 0.061 161 0.7349 0.08322 0.064182 0.7712 0.08322 0.064210 0.7716 0.08322 0. 066063 0.7933 0.08322 0.066177 0.7952 Average 0.7665 I 12 . IV. Summary. 1 . Pure nicotine has been obtained from Black Leaf Forty by use of a perforator, subsequent extractions, and vacuum distillation. 2. It is impossible to obtain the tartrate of nicotine from a water solution. It is readily obtained from its alcohol- ic solution upon the addition of ether. 3. It seems evident that the acid salts of nicotine are the only ones which it is possible to obtain in crystalline form. In contrast to this conclusion is the striking Wal the neutral salts are sufficiently stable in solution to prevent any extraction of the nicotine by ether. In fact, acid of half the concentration necessa,ry to form the neutral salt is sufficient to prevent extraction of the nicotine from the sol- ution . 4. Equilibrium conditions have been determined For the partition of nicotine between aqueous and acid solutions ( tartaric and hydrochloric acids) and an immiscible solvent (ether) . 5. The most practical method for the determination of alkaloids involves the extraction of the alkaloid from an aqueous solution by means of an immiscible solvent, such as. chloroform or ether. . r •. UkJ r ' i . . • . . . 1 3. BIBLIOGRAPHY. 1. Dragendorff: Die ger^ich^tlich-chemische Ermittelung von Giften, 4 Aufl., p, 151. 2. Stas-Otto : Ausmittelung der Gifte., 7 Aufl., pp. 144& 280. 3. Beal and Lewis: J, Amer. Pharm. Assoc., 5, 812-837, 1916. 4. Beal and Lewis: Loc . cit. 5. KippeShberger, G. : Grundlagen ftir den Nachweis bon Gif tstof- fen bei gerichtlich-chemischen Uutersuch- ungen, p, 56. 6. Kippenberg*€rC . : Zts. f. Anal. Ch., 1900, 39, 290-314. 7. Proelss, Hans : Apoth. Ztg., 1900, 16, 289-493. 8. Springer, Ed . : Apoth. Ztg., 1901, 17, 225-226. 9. Simmer; Arch. d. Phar., 1906, 244, 672. 10. Harden and Elliott: J, Ind. Eng. Ch., 6, 928. 11. Beal and Lewis: Loc. cit. at 3. 12. Allen: Comm. Org. Anal. 4th Ed. VI, 238.' 1 3. Hudson, C . S. : Zeit. Physikal Chem. , 1904, 47,113. 14. U. S.D.A. Bur. of Chem., Bull. 107, rev. 1910. 1 5. Bertrand, G. ,& Javillier,M, : Bull, des Sciences Pharrnacol- ogiques, tome 166, no.1 pp. 7- 14, Jan. 1909. 16. Chapin, Robert M. : U. S. Dept, of Agr., Bur. of Animal Ind., Bull. 133.