DETERMINATION OF THE CONSTITUTION OF FATTY ACIDS, 83 IX.-Determination of the Constitution of Fatty Acids. Part 11. By ARTHUR WILLIAM CROSSLEY and HENRY RONDEL LE SUEUR. IN the first part of this research (Trans., 1899, 75, 163), it was stated that among the acids then examined none contained alkyl groups in the a-position, and a s in such a case a new point of con- siderable interest is raised, we decided to prepare an acid of this type and submit it to the process there described, The acid selected was ethyl- isopropylacetic acid [a-isopropylbutyric acid], C,H,*CH(C,H,)*CO,H, which we have succeeded in preparing in large quantities from ethyl malonate. As the preparation is attended with many difficulties, we give full details of our experiments. Ethyl cyanoacetate has recently been used in synthetical work in- stead of ethyl malonate, with marked success ; we have also prepared ethylisopropylacetic acid from this substance a8 a starting point, but G 284 CROSSLEY AND LE SUEUR: DETERMINATION OF THE the method is not to be recommended, as the yield of pure acid is not nearly so good as when ethyl malonate is employed.Following the usual course of procedure when two alkyl groups have to be introduced into ethyl malonate, we first attempted to intro- duce the heavier isopropyl group and then the ethyl group, but after a, long series of preliminary experiments, of which it is not necessary t o give the details, we found that the yield of ethyl ethylisopropyl- malonate so obtained was, relatively speaking, very small ; it could, however, be increased to 75 per cent.of the theoretical by reversing the order of introduction of the alkyl groups, but even then the introduction of the isopropyl group is not an easy matter, the most satisfactory conditions being the following. After treating ethyl ethylmalonate with the calculated quantities of sodium and isopropyl iodide, the portion of the resulting liquid 'boiling below 230' (ethyl ethylisopropylmalonate boils at 230-235') was again treated with sodium and isopropyl iodide (see page go), by which means the yield of pure disubstituted ethyl malonate was increased from 46 to 75 per cent. of the theoretical. The liquid resulting from the hydrolysis of ethyl ethylisopropyl- malonate separates on distillation into two main fractions, one boil- ing at 202-205' (consisting of pure ethylisopropylacetic acid) and the other at about 165-175'.We were at first much puzzled to account for the presence of this liquid of lower boiling point, which had not the characteristic odour of a fatty acid, but smelt more like an ester. On consulting the literature of the subject, we found that Paal and Hoffmann (Ber., 1890, 23, 1497) noticed that when diethyl isoamylmalonate is hydrolysed with alcoholic potash, there is formed, besides the corresponding malonic acid, a considerable quantity of ethyl hydrogen isoamylmalonate, C , H I l * C H < ~ ~ ~ H , which on distillation loses carbon dioxide with production of ethyl isoamylacetate, C,H,1*CH2-C02C,H,. A similar series of reactions appears t o take place in the above instance ; the hydrolysis of diethyl ethylisopropyl- malonate gives rise to ethylisopropylmalonic acid,C,H,* C(C,H7)(C0,H),, and ethyl hydrogen ethylisopropylmalonate.On distillation, both these substances lose carbon dioxide, with the production of ethylisopropylacetic acid and its ethyl ester. The latter constitutes the fraction boiling at 165-175', as was proved both by preparing a specimen of pure ethyl ethylisopropylacetate, which was found t o boil at 164-165', and by hydrolysing the fraction boiling at 165-175', when it yielded pure ethylisopropylacetic acid. We were in all cases obliged to resort to this process of double hydrolysis, for no matter whether we employed 4 or 6 mols. of potass- ium hydroxide, we always obtained considerable quantities of the 2 2 5CONSTITUTION OF FATTY ACIDS.PART XI, 85 fraction boiling at 165--175O, even when the heating was continued for 16 hours. The derivatives of ethylisopropylacetic acid, prepared for the pur- pose of characterisation, are described in detail in the experimental part of this paper. It is interesting to note that the melting point of the amide is as high as 1 3 4 O , for Burrows and Rentley (Trans., 1895, 67, 511) found that the amide OF a similarly constituted and isomeric fatty acid (methylisobutylacetic acid) melted a t 90°, which, as they remark, is a very low meltiug point for an amide of a higher fatty acid. On comparing the physical constants and chemical properties of ethylisopropylacetic and methylisobutylacetic acids, the need of some accurate method for the determination of the constitution of such compounds is well brought out.Methylisobutylace tic acid, CH,. CH( C,H9)*C0,H. Acid ............ B. p. 204-205" Ethyl ester.. .... 17 1 65 -- 1 6 6 Chloride 9 , 152-153 Anilide ......... M.P. 110-111 9 9 90 Amide.. .......... Paratoluidide 1 ) S6 ......... ... Ethylisopropylacetic acid, C,H,*;CH (C, H 7) CO@. B. p. 202-203O 9 , 150-153 M.p. 114-115 ?, 164-165 9, 134-1 34 -5 Y 9 122-123 It will be noticed that the boiling points or melting points of the isomeric derivatives are almost identical, except in the case of the amides and paratoluidides, two derivatives which are not, as a rule, employed for the characterisation of fatty acids, but which, in view of the above, would appear to be worth more general investigation and use.In applying our method to ethylisopropylacetic acid, we first con- verted it into the ethyl ester of the corresponding a-bromo-acid, which was then treated with diethylaniline (see p. 95). From a glance at the formula of ethyl a-bromo-ethylisopropylacetate, it is evident that the elimination of hydrogen bromide may take place in two ways : (CH,),CH* $7Br*CO,Et (CH,),C:y.CO,Et (CH,),CH*fi*CO,Et 7H2 or VH 7% 3 CH3 CH3 Ethyl BB-dimethyl-a- Ethyl B-methyl-a-iso- e thy lacrylate. propylacrylate. and from oxidation experiments we conclude that both reactions occur. It appears, however, that the greater portion of the resulting unsaturated ester is ethyl pp-dimethyl-a-ethylacrylate, that is to say, the hydrogen atom of a CH group is more easily removed, along CH,86 CROSSLEY AND LE SUEUR: DETERMINATION OF THE with the bromine atom, than one of the hydrogen atoms of a CH, group.Diethylaniline therefore seems to exert a certain selective action in such cases, a point on which we hope shortly to furnish more definite information. That the course of the reaction would be twofold was t o be expected from experiments which have been described by W. H, Perkin, jun. (Trans., 1896, 69, 1466, 1490), who has shown that when ethyl a-bromomethylisopropylacetate is treated with quinoline or alcoholic potash a mixture of ethyl trimethylacrylate and ethyl isopropylacryl- ate is obtained. (CHJ,CH*FBr C0,Et (CH,),C:$!*CO,Et - and CH3 CH, Unfortunately, no definite information is given of the ethyl esters produced, but it is shown that ate does not react with the sodium compound (CH,),CH*fl*CO,Et as to the amounts ethyl trimethacryl- of ethyl malonate, CH,.wherea8 ethyl isopropylacrylate does, and calculating from the experi- mental data given, it would appear as if the original mixture of unsaturated ethyl salts consisted of approximately 70 per cent. of the former with 30 per cent. of the latter, thus pointing to the fact that in this case also the hydrogen atom of the CH group is the more t-eadily eliminated. When the mixture of acids resulting from the hydrolysis of the unsaturated ethyl esters is oxidised first with potassium permanganate and then with a mixture of potassium dichromate and dilute sulphuric acid, acetone, acetic acid, propionic acid, and iaobutyric acid are obtained, as was to be expected, P/3-dimethyl-a-ethylacrylic acid giving rise t o acetone and propionic acid, and /3-methyl-a-isopropylacrylic acid to acetic and isobutyric acids by rupture of the double linking. When 25 grams of the acids were oxidised, we obtained, in separate experiments, 1.136, 1.84, and 2.07 grams of acetone.I n the first two instances, the acetone was estimated by the iodoform method, and in the third by precipitation of the p-bromophenylhydrazone. We thought at first that by determining the amount of acetone produced we should get some idea of the proportion of the two unsaturated acids present, but from what follows it will be evident that no such information can be gathered from this experiment. Unfortunately, it is also impossible to accurately estimate the amounts of the various fatty acids produced, but acetic acid is obtained in largest quantity, propionic acid next, and but very little isobutyric acid, Had no side issues to be considered, the amount of propionic acid should be propor- tional t o the amount of acetone, and the amount of acetic acid pro- portional to that of isobutyric acid, but in neither case does this hold good, there being much less acetone than would correspond toCONSTITUTION OF FATTY ACIDS.PART 11. 87 the amount of propionic acid, and much more acetic acid than would correspond to that of isobutyric acid, The explanation of this is not difficult to find, for we have satisfied ourselves that, under the conditions employed, both acetone and isobutyric acid are further oxidised, and thus confirm the statements of Schmidt (Bey., 1874,7, 1363) and Hercz (AnnuZen, 1877, 186,258).Supposing, then, that acetone and isobutyric acid are formed in our experiments as primary oxidation products, they would immediately encounter an excess of oxidising agent, and consequently mould both be largely converted into acetic acid, a fact which would account for the relatively large amount of this acid and for the small amounts of acetone and isobutyric acid which we find. In order to make our experiments more complete and also to supply supplementary evidence, we are a t present investigating more closely the question of the oxidation of the lower fatty acids. Applying the results of our experiments to the determination of the constitution of the acid in question, which for the moment we may presume t o be of unknown structure, the argument would be as follows, Analysis gives the molecular formula C,HI4O2, and as the acid is capable of bromination by Volhard's method, the bromine atom must occupy the a-position.By eliminating this bromine atom together with a neighbouring hydrogen atom, an unsaturated substance is formed, whence it follows that, in the latter, the double bond is between the a- and P-carbon atoms, and that the carbon atom of the 00 or C02H group of any ketone or acid produced from this un- saturated substance by oxidation at the double bond must also have occupied either the a- or P-position in the original acid, I n the case under investigation, we find acetone as one oxidation product, therefore the carbon atom of the ketonic group must have occupied either the a- or P-position in the unsaturated acid, A con- sideration of these alternatives leads to the following results : (1) Occurrence in the p-position.-It follows from what has just been said that the correct grouping of five of the seven carbon atoms is (CH,),C:C*CO,H.Only two carbon atoms are not accounted for, and as there *is but one unsatisfied bond, these must be attached as an ethyl group, so we get as the formulae of the unsaturated and saturated (ethylisopropylacetic) acids. The unsaturated acid, on oxidation, would give acetone and propionic acid, both of which we have been able to identify. ( 2 ) Occurrence in the a-position.-We should then have the following88 CROSSLEY AND LE SUEUR: DETERMINATION OF THE CH grouping, CHS>OCO,H, which, as it necessitates the acceptance of a 3 '* quinquevalent carbon atom, may be a t once dismissed. It might, however, be urged that all the acetone comes from the further oxidation of isobutyric acid.Then, arguing on the same lines as above, we may first place the carbon atom of the UO,H group in isobutyric acid in the P-position, and obtain the following grouping for six out of the seven carbon atoms : (CH,),CH* CH:G*CO,H leading to (CH,),CH* CH:v*CO,H QH3 for the unsaturated acid : an acid which would give acetic and isobutyric acids, together with acetone from the latter on oxidation, but from which it would be impossible to obtain propionic acid. If the carbon atom of the CO,H group is placed in the a-position, we obtain the formula for the unsaturated and the saturated acids.The former would give acetic and isobutyric acids on oxidation, whilst the latter is ethyliso- propylacetic acid, from which acetone and propionic acid can be ob- tained as already shown (p. 8'7). We ccnclude therefore that : (1) With an acid of the type of ethylisopropylacetic acid, our method works just as well as in the cases previously examined (Zoc. cit.). (2) The elimination of hydrogen bromide from the ester of the corresponding bromo-acid takes place in two ways (p. 85). (3) Acetone and propionic acid result from the oxidation of Pp-di- methyl-u-ethylacrylic acid, and acetic and isobutyrio acids from the oxidation of P-methyl-a-isopropylacrylic acid. For the general application of this method, all fatty acids map be divided into four groups : Group I.-Those containing two hydrogen atoms in the a-position, and either one ortwo in the P-position. The theoretical considerations connected with acids of this type have already been given (Crossley and Le Sueur, Trans., 1899, 75, 161), and the process has been carried out with valeric, isovaleric, and isobutylacetic acids.Group 11.-Those containing only one hydrogen atom in the a-position, and either one or two in the P-position. The present communication gives the theoretical considerations and a practical illustration of the method as applied to an acid of this type. Ethylisopropylacetic acid is a particularly good example, be- I n all three cases, the method works well.CONSTlTUTION OF FATTY ACIDS. PART 11.89 Acid. Qroiip I. Group 11. Group 111. A Butyric acids .. ,,. . .. . , . . , . . Valeric acids . . , .. . . . . . . , , . . 1 Hexoic acids . . # ... , . . . . . , . 3 3 1 Heptoic acids ... ... .. . ... .. . 7 6 3 1 2 1 1 cause it contains only one a-hydrogen atom, but two P-carbon atoms, t o one of which there is attached one hydrogen atom, and to the other two hydrogen atoms. Group 111.-Those containing no hydrogen atoms in the a-position. With an acid of the constitution CR,*CO,H, our method cannot be employed, for it is well known that, unless an acid contains an a-hydrogen atom, it cannot be brorninated by Volhard’s method, as tohe molecule breaks up under these conditions (compsre Reformatzky, Be?*., lS90, 23, 1594; Aumers and Bernhardi, Be?*., 1891, 24, 2210).There are, however, so,few acids of this ty1;e, that, given a fatty acid incapable of being brominated by Volhsrd’s method, without decom- position, this in itself would be strong evidence as to its constitution. Group 1V.-Those containing two hydrogen atoms in the a-position and none in the P-position. We cannot ascertain t h a t an acid of this type has been described. It seems probable that such an acid should result from the product of the interaction of ethyl sodiomalonate and dimethylethylcarbinyl bromide, Group IV. A 1 1 E x P E R I M E N T A L. 1, E t IL y li s o p Y o p y l u c e t i c A c i cl. Prepavation of 23th y lisopropy lucetic Acid from Ethyl Eth ylmalonat e. After making a number of preliminary experiments, large quantities of ethyl ethylmalonate were worked up in the following manner.24 -8 grams of sodium (I mol.), dissolved in 300 C.C. absolute alcohol,90 CROSSLEY AND LE SUEUR: DETERMINATION OF THE were mixed with 200 grams of ethyl ethylmalonate (1 mol.), 190 grams of isopropyl iodide (1 mol.) were gradually added, and the whole heated on a water-bath for 15 hours. The greater portion of the alcohol mas then distilled off, and after adding water, the whole was extracted with ether, the ethereal solution dried over calcium chloride, and the ether evaporated. After distilling the residue three times in air, using a fractionating column, the following fractions were collected : Below 210 "... ...... 9 grams 225-230° ......... 28 grams. 210-225 .........66 ,, 230-235 ......... 113 ,, An analysis of the fraction 230-235' gAve the following numbers : C = 62-57 ; H = 9.55. 0.1182 gave 0.2712 CO, and 0.1016 H,O. Ethyl ethyZiso~ro~yZma7onate, C,H5* C( C,H7)(C0,C,H,),, is a clear, colourless, mobile liquid possessing a pungent, rather unpleasant smell, and boiling a t 232-233'. The yield of ester obtained is only 46 per cent. of the theoretical, but this can be materially in- creased by again treating the lower fractions with sodium and isopropyl iodide, in proportions calculated on the supposition that the fraction 225-230' contains 25 per cent. of unchanged ethyl ethylmalonate C,,H,,O, requires C = 62.60 ; H = 9.56 per cent. 210-225 ,, 50 ,, 9 , ?, 9 9 below 210 ,, 100 ,, >, 9 , ,? I n this way, 70 grams of liquid boiling between 230-235' were obtained, thus bringing the total yield of ethyl ethylisopropylmalonate up t o 75 per cent.of that theoretically obtainable from the ethyl ethylmalonnte employed. EthyZisoprop~ZmccZonic m i d , C2H5*C(C3H7)(C02H)2, was obtained by hydrolysing the ester with alcoholic potash, and after recrystallisation from benzene was analysed. 0.1670 gave 0.3370 CO, arid 0.1194 H,O. This acid is insoluble in cold light petroleum (b. p. SO-loo'), readily soluble in acetone, alcohol, or water, and crystallises from benzene in beautiful, glistening needles melting at 131-131*5'. At a higher temperature, it evolves carbon dioxide, giving rise to the corresponding fatty acid. The silver salt, prepared in the ordinary manner, is a white, curdy precipitate.0,2032 gave on ignition 0.1134 Ag. C= 55.04 ; H=7*94. C8H,,0, requires C = 55-17 ; H = 8.05 per cent, Ag=56°80. C,€l,,O;,Ag, requires Ag = 55.67 per cent.CONSTITUTION OF FATTY ACIDS. PART 11. 91 EthyZisopropyZcccetic acid, C,H,*CH(C,H7)*C02H.-422 grams of ethyl ethylisopropylmalonate were heated for 10 hours on a water-bath in quantities of 100 grams a t a time with 150 grams (6 mols.) of potass- ium hydroxide dissolved in alcohol. Water was then added, the alcohol evaporated, and after acidification with sulphuric acid the whole was extracted with ether, the ethereal solution carefully dried over calcium chloride, and the solvent evaporated. The residue, after heating to eliminate carbon dioxide, was repeatedly distilled, using a fractionating column, when the following fractions were collected : Below 190 O.........69.5 grams. 196-200' ...... 12.0 grams, On further distillation of the portion boiling between 200-205°, pure ethylisopropylacotic acid boiling constantly at 202-203' was obtained. It is a clear colourless, oily liquid, with a disagreeable and penetrating odour, similar to that of other acids of the fatty series. It is readily attacked by potassium permanganate in alkaline solution. 190-196 ......... 18.0 I , 200-205 ...... 128.0 9 , 0.1466 gave 0.3476 CO, and 0.1412 H,O. C7HI4O2 requires C = 64.61 ; H = 10.77 per cent. The siZvei* salt, prepared in the usual manner, is a white caseous 0,1988 gave on ignition 0,0902 Ag. C7H,,0,Ag requires Ag = 45.57 per cent. We were for some time unable to account for the large fraction boiling below 190°, of which the major portion passes over between 165' and 170'; fhrther experiments showed, however, that the latter temperature is about the boiling point of ethyl ethylisopropylacetate (164-165'), and the course which the reaction is supposed to follow has already been explained in the introduction (p.84). The whole of the liquid boiling below 200' was therefore again hydrolysed with alcoholic potash, when a further 60 grams of pure ethylisopropyl- acetic acid were obtained, thus increasing the yield to 80 per cent. of that theoretically obtainable from the corresponding disubstituted malonic ester. C=64*66 ; H=10-71. precipitate. Ag= 45.39. Preparation of EthyZisopropyZacetic Acid from Ethyl a-Cyanoacetate.Ethyl a-cyanoacetate was first converted into ethyl a-cyanobutyrate, C2H5* CH(CN)*CO,C,H, (b. p, 208--209O), by heating with sodium and ethyl iodide in alcoholic solution. Three soda-water bottles, each containing 6 grams of sodium dis- solved in 60 c . ~ , absolute alcohol, 31 grams of ethyl a-cyanobutyrate,92 CROSSLEY AND LE SUEUR: DETERMTNAT'ION OF THE and 40 grams of isopropyl iodide, were securely corked, and heated in a water-bath for 20 hours, Water was then added to the contents of the bottles, and the whole extracted with ether, &c. ; the resulting liquid was then fractionated, using a column, with the following results : Below 205' ......... 11.5 grams. 230-225' ......... 35-5 grams. 205-215 ......... 3.5 ,, 225-230 ......... 23.0 ,, 215-220 .........10.5 ,, A nitrogen determination in the Fraction 225-23OOgave the follow- 0.1416 gave 10 C.C. moist nitrogen at 23' and 766 mm. Ethyl a-cy(moet?$isopropyZacetate, C,H,* C( C,H,)( CN) C02C2H5, is a clear, colourless liquid having a faint odour of peppermint, and boils a t 226-227' under 756 mm. pressure. The yield is poor (20 per cent. OF the theoretical), but may be increased to 50 per cent. of the amount theoretically obtainable from the ethyl cyanobutyrate used, by again treating the lower fractions with sodium and isopropyl iodide, on the same principle as described in the preparation of ethyl ethylisopropyl- malonate (p. 90). Dilute sulphuric acid seems to be the best agent for hydrolysing this ester, but the results are not very satisfactory, and the yield of fatty acid is small. The ester was boiled for 30 hours with four times its weight of 55 per cent. sulphuric acid, and after dilutiug with water the whole was extracted with ether, the ethereal solution dried over calcium chloride, and the ether evaporated. On fractionating the resulting liquid, the major portion distilled between 200' and 205", and by repeating the process an acid liquid was obtained boiling at 202-203' and having all the properties of ethylisopropylacetic acid.It gave a white, insoluble silver salt, which was analysed : ing numbers : N = 8.02. C1,H1702N requires N = 7.65 per cent, 0.1532 gave on ignition 0.0702 Ag. Ag = 45.82. C7H,,02Ag requires Ag = 45-57 per cent. The p-toluidide crystallised from light petroleum (b. p. 60-SOo) in slender, glistening, silky needles melting a t 121-122' (compare I f ethyl cyanoethylisopropylacetate is heated with dilute sulphuric acid only for 15 hours, the greater part of the resulting liquid was found to boil between 160-170°, a small amount passing over between 230' and 240°, which solidified on cooling.The fraction 160-170" consisted probably of the nitrile of ethylisopropylacetic p. 94).CONSTITUTION OF FATTY ACIDS. PART IT. 03 acid, which, from analogy, would be expected to boil a t about this temperature ; moreover, when heated with dilute sulphuric acid, this fraction is converted into ethylisopropylacetic acid. The solid fraction (b. p. 230-240") crystallised from light petroleum in white, silky needles melting a t 133*5-134" (see p. 94) and consisted of ethyliso- propylacetamide.0,1338 gave 12.8 C.C. moist nitrogen a t 17" and 752 mm. N = 10.98. C,H,,ON requires N = 10.85 per cent. I n one of the first experiments me made on the hydrolysis of the above ester, a portion was heated for 12 hours with 5 times its volume of concentrated hydrochloric acid, but, as on extraction with ether, &c., the boiling point of the resulting liquid mas found to be that of the original -ester, the whole was heated with concentrated aqueous potassium hydroxide for 15 hours, during which process considerable quantities of ammonia were evolved. The residue extracted from the acidified liquid was found to boil for the most part below 200°, and n small portion boiling between 230' and 240' solidified on cooling. This substance crystallises from water, or, better, from light petroleum (b.p .80-100°) in beautiful, silky needles melting a t 122--122*5O, and does not sublime when heated in a dry test-tube. 0.1416 gave 13.8 C.C. moist nitrogen a t 23.5' and 760 mm. N = 10.95. C,H,,ON requires N = 10.85 per cent. Although resembling ethylisopropjlacetamide, it melts at a tempera- ture 13 degrees lower, and we have, so far, been unable to satisfactorily prove its exact nature, the amount obtained being very small, Derivatives of Ethglisopopylacetic Acid. Ethyl ethylisopropylncetute, C2H,* CII( C,H,) * CO,C,H,.-Pure ethyl- isopropylacetic acid was dissolved in absolute alcohol, dry hydrogen chloride passed in to saturation, and the whole left for 24 hours. After pouring into water and extracting with ether, &c., it, was found that very little of the ethyl salt had been formed, as nearly the whole of the residue boiled between 200' and 2 0 5 O , and consisted therefore of un- changed acid. This observation appears t o be in accord with those of Menschutkin and others, who have shown that acids of this type are not readily esterified in this manner.After several further fruitless experiments, the following method was adopted. 10 grams of the pure acid were dissolved in 25 C.C. of absolute alcohol, and after adding 10 C.C. of concentrated sulphuric acid, the vhole was heated on a water-bath for 8 hours; water was then added, and the oil which separated was extracted with ether. The ethereal94 CROSSLEY AND LE SUEUR: DETERMINATION OF THE solution was washed with dilute sodium carbonate, then with water, carefully dried over calcium chloride, and the ether evaporated, On fractionating the residue, 11 grams (91 per cent.) of a liquid boiling constantly a t 164-165" under 765 mm.pressure were obtained, 0.1486 gave 0.3708 CO, and 0.1510, H,O. The ester is a clear, colourless, mobile liquid with a characteristio, penetrating smell. EthyZisopropp?acetumide, C,H,* CH(C3H7) COwNH2.-The ester was heated in a sealed tube with concentrated aqueous ammonia for 12 hours at 180°, but was recovered unchanged. The same result was obtained when alcoholic ammonia was employed, Some of the pure acid was then heated for 10 minutes in a reflux apparatus with the calculated quantity of phosphorus pentaohloride. After distillation, the crude chloride boiling between 150° and 153O was slowly poured into concentrated aqueous ammonia, when a white solid separated, which was collected, spread on a porous plate, crystal- lised from light petroleum, and analysed, 0.1166 gave 11.0 C.C.moist nitrogen at 21Oand 764 mm. N = 10.80. C,HI,ON requires N = 10.85 per cent. The amide is but slightly soluble in chloroform, benzene, or acetone, but readily in alcohol or water, and crystallises well from either light petroleum or water in long, white, silky needles melting at 134-134.5O. When heated in a test tube, it sublimes in needle-shaped crystals. EthyZisopropykccetarLilide, C,H,* CH(C,H,) *CO*NH* C,H,, prepared by heating the pure acid with twice its weight of pure aniline for 24 hours, was purified in the usual manner, and finally by recrystal- lisation from light petroleum (b.p. 100-120°), from which it separated in clusters of small, glistening needles melting a t 114-1 15O. It is insoluble in water, but readily soluble in alcohol, chloroform, acetone, or ether. C = 68.06 ; a= 11.33. C,H,,O, requires C = 68.35 ; H = 11 -39 per cent, 0.1580 gave 9.6 C.C. moist nitrogen at 20' and 762 mm. C,,H,,ON requires N = 6.83 per cent. EthyZiso~opyktcetpara~oZ~~~~de, C2H5* CH(C,H7)* CO*NH* C,H7, pre- pared and purified in a similar way to the anilide, crystallises from light petroleum (b. p. 60--80°) in glistening, feathery needles melting at 122.5-123O. It is insoluble in water, but readily soluble in the ordinary organic solvents, even in the cold. N=6*36, N = 6-96.0.1994 gave 11.0 C.C. moist nitrogen at 19' and 762 mm. C,,H,ION requires N = 6.39 per cent.CONSTlTUTION OF FATTY ACIDS, PART 11, 95 11. E t h y I a- B r o m o e t h y I i s op r op y l a c e t a t 8. Ethyt a-bromoethylisopropyhcetate, C,H,* CBr(C,H,) * CO,C,H,.- Ethylisopropylacetic acid was treated in quantities of 50 grams at a time with 4 grams of amorphous phosphorus and 120 grams of dry bromine (Volhard, dnnalen, 1887, 242,61). On pouring the resulting bromide into alcohol, very little heat is developed at first, but on standing the temperature gradually rises, and eventually it is necessary to moderate the reaction. The resulting ester of the bromo-acid is a colourless, mobile liquid, possessing a pungent, characteristic smell somewhat resembling that of peppermint.It boils constantly a t 135-136' under 59 mm. pressure, and is obtained in almost theoretical amount (92-93 per cent.). 0.2440 gave 0.1950 AgBr. Br = 33.97. C,HI7O,Br requires Br = 33.76 per cent, Treatment of Ethyl a-Bromethyli~opi~opyZ~1cetats with BietJhylaniline, 356 grams of the ester were treated in portions of 50 grams (1 mol.) a t a time with 63 grams (2 mols.) of freshly distilled diethyl- aniline, and worked up in the manner already described (Trans., 1899, 76, 166). As the residual liquid was found to contain traces of bromine, it was again heated with 1 mol. of diethylaniline for 8 hours, and then distilled in air. After five distillations, the follow- ing fractions were collected under 748 mm. pressure : 170-175' ......2 grams. 185--200' ...... 5 grams, 175--180 ...... 170 ,, Above 200 ...... 28 ,, 180-185 ...... 5 ,, The frccction boiling at 175-1SO'was found to be entirely free from 0.1094 gave 0.2774 CO, and Oa1008 H,O. C,H1,O, requires C = 69.23 ; H = 10.25 per cent. So far this is the only instance in which we have found it possible to entirely eliminate bromine by using diethylaniline ; as a rule, the analyses of such unsaturated esters do not lead to satisfactory results, owing t o the presence of small amounts of halogen. This fraction, amounting t o '73 per cent. of the product, is a clear, colourless, mobile liquid possessing a very pungent odour of peppermint ; although of fairly constant boiling point, it is not a simple substance, but as the oxidation experiments show, consists of a mixture of ethyl P/3-dimethyl-a-ethylacrylate and ethyl p-methyl-a-isopropylacrylate. bromine, and gave the following numbers on analysis : C = 69-15 ; H = 10.24.96 CROSSLEY AND LE SUEUR: DETE€iMINATION OF THE Hydrolpis of the Mixed Esters.170 grams of the mixed esters were hydrolysed in two portions with 120 grams of caustic potash dissolved in alcohol. The mixture of acids obtained from the product had the remarkable property of boiling quite constantly at 136O under 55 mm. or at 203-204' under 760 mm. pressure. It is a clear, colourless liquid, with a faint but sharp odour ; i t rapidly decolorises a chloroform solution of bromine, and also an alkaline solution of potassium permanganate. Oxidation of the mixed Unsaturated Acids. The mixed acids were oxidised (cornpare Trans., 1899, 75, 165) in quantities of 25 grams, first, with 23 grams of potassium permanganate and then with 60 grams of potassium dichromate dissolved in dilute sulphuric acid.During the second operation, an absorption apparatus was connected with the end of the condenser, in order to avoid loss of acetone. The oxidation took place very readily, in spite of the fact that the acids are insoluble in water, and was complete i n from 3-4 hours. The product wa8 distilled with steam until no more acid passed over, the distillate carefully neutralised with potassium hydroxide, again distilled with steam, and the second distillate treated with a n alcoholic solution of p-bromophenylhydrazine, when a copious precipitate separated.This was collected and recrystallised from light petroleum, when small, glistening leaflets of the p-bromophenyl- hydrazone derivative of acetone, (CH,),C:N*NH*C,H,Br, were ob- tained melting at 94-95" (compare Neufeld, A?znat!en, 15135,248, 96). C9HllX2Br requires Br = 35.24 per cent. The contents of the distillation flask were evaporated to dryness, and the potassium salts, rendered anhydrous by heating at 100" i n a n air oven, were distilled with concentrated sulphuric acid, when a good deal of charring occurred. The liberated fatty acids (13-14 grams from 25 grams of the unsaturated acids) were dried by standing over concentrated sulphuric acid and then distilled fractionally, when no substance of higher boiling point than 160" was obtained. I n subsequent experiments, the process was somewhat modified, inasmuch as the potassium salts were dissolved in dilute sulphuric acid, and the whole extracted 10 times with pure ether.The ethereal solution was dried over calcium chloride, the ether removed by evapora- tion, and the residual fatty acids distilled, the following fractions being collected : 0.1212 gave 0.0994 AgBr. Br = 34-90, 100-1 1 0" 125-1 35" 145-150" 110-1 25 135-145 150-160CONSTITUTION OF FATTY ACIDS. PART 11. 97 The thermometer rises very rapidly from 160' t o 200°, at about which temperature the residue passes over, and consists presumably of incompletely oxidised acids, The various fractions were further purified by drying over anhydrous sodium sulphate and repeated distillation. Fraction boiling at 100-1 25".-This contains aeetic acid, CH,*CO,H.A portion of this fraction boiling at 118' was converted into the silver salt, which after recrystallisation from water was analysed. 0.1288 gave on ignition 0.0834 Ag. C,H,O,Ag requires Ag = 64.67 per cent. A further quantity of the acid boiling between 117-119' was con- verted into the anilide, which crystallised from water in pearly scales melting at 112-112.5". 0.1024 gave 9.2 C.C. moist nitrogen at 20' and 764 mm. N= 10.33. Fraction boiling at 135-1 45".-This contains propionic acid, The silver salt, prepared from a portion of the acid Ag = 64.74. CH,*CO*NH*C,H, requires N = 10.37 per cent. CH,*CH,*CO,H. boiling at 140°, gave the following numbers on analysis : 0.1520 gave on ignition 0*0908 Ag. C,H50,Ag requires Ag = 59.66 per cent.The anilide prepared from the acid boiling between 139-141' crystallised from light petroleum in glistening leaflets melting at 103-104' (compare Trans., 1898, 73, 34). Ag = 59.73. 0.1429 gave 12 C.C. moist nitrogen at 20" and 753 mm. C,H,*CO*NH*C,H, requires N = 9.39 per cent. Fraction boiling at 150-1 60°.-This contains isobutyric acid (CH,),CH*CO,H. The portion of this small fraction boiling between 153" and 157" was converted into the calcium salt by heating the acid dissolved in water with pure calcium carbonate, filtering, and evaporating the filtrate in a vacuum, when radiating clusters of silky needles were obtained, in which, after drying in air, the water of crystallisation was determined. N=9.51. 0.3536 lost 0.0702 at 155".0.2368 gave 0.1536 CaSO,. (C,H,O,),Ca requires Ca = 18-69 per cent. According to Chancel and Parmentier (Compt. red., 1887,104,477), calcium isobutyrate should contain 5H,O, requiring 29.6 per cent. H,O, but, as has repeatedly been shown (Trans., 1898, 73, 15, 35 ; also 1899,75, 185), the salt prepared i n the above manner contains H,O= 19.85. Ca = 19-07. VOL. LXXVII. H98 DETERMINATION OF THE CONSTITUTION OF FATTY ACIDS. water corresponding more nearly with 4H,O (25.17 per cent.), the amount depending to a large extent on the length of time during which the salt has been exposed to air (Trans., 1898, 73, 15). The unusually small proportion of water found above appears to be due to the facts that, first, as the calcium analysis shows, the salt was not quite pure, and, secondly, when the solution of the salt was evaporated sufficiently for it to crystallise, the surface became coated with a crust, from which, on account of the small amount of material, it was very difficult to completely detach the crystals.As mentioned on p. 95, when ethyl a-bromoethylisopropylacetate is treated with diethylaniline, 28 grams of a substance boiling above 200' were obtained, besides the unsaturated esters, On distilling this under 65 mm. pressure, 8 grams were found to boil below 200°, and 14 grams between 205-2109 As both fractions contained nitrogen, it was thought that an ethyl ester of an anilino-acid might have been produced, for, as Bischoff (Bey., 1898, 31, 3015) has shown, these substances are formed by the action of dimethylaniline on the ethyl esters of a-bromo-acids of the fatty series, but it appears not to be so in this case, The fraction boiling below 200' was treated with ;alcoholic potash, the alcohol evaporated, and a small amount of an insoluble nitrogen- ous substance removed by ether.On acidifying and extracting with ether, 7 grams of a liquid boiling between 207-209' were obtained, which exhibited all the properties of the mixture of unsaturated acids described on p. 96. A portion was converted into the silver salt and analysed. 0.1708 gave, on ignition, 0.0788 Ag. The fraction boiling at 205--210' was again distilled, and a portion 0.1558 gave 0,4264 CO, and 0.1288 H,O. 0.2261 N = 7.06. 0.2090 ,, 11.9 C.C. ,, > ? ,, 15' ,, 770 mm. N=6.76. C,,H,,O,N requires C = 73.64 ; H = 9.71 ; N = 5.05 per cent. These numbers do not agree with those required for the ethyl ester of the corresponding anilino-acid. Some of the liquid was then heated with excess of alcoholic potash for 7 hours, Water was added and the alcohol evaporated, when on extracting the alkaline liquid with ether, nearly the total weight of liquid originally taken was obtained. It boils constantly at 290-295O under 750 mm. pressure, is a thick, pale yellow liquid with a marked blue fluorescence, and gave C = '76.52, Ag=46.13. C7Hl10,Ag requires Ag = 45.96 per cent. boiling at 207-208° under 60 mm. pressure was analysed : C: = 74.64 ; H = 9.18. ,, 13.6 C.C. moist nitrogen at 16' and 766 mm.HEWITT : PREPARATION OF BENZENEAZO-0-NITROPHENOL. 99 H=8*99, N=7*31 per cent. as the mean of three closely agreeing analyses. It dissolves in concentrated sulphuric or hydrochloric acids, but is thrown out of solution on adding water. With concentrated sul- phuric acid and a crystal of potassium dichromate, it gives a rose-pink colour turning to greenish-brown. Concentrated nitric acid dissolves it, and on evaporation a rose-pink colour is developed. Potassium permanganate is readily decolorised by it, but no definite oxidation product could be isolated, and we have so far been unable to decide the nature of this substance. Our thanks are due to the Research Fund Committee of the Chemical Society for a grant defraying in part the cost of the materials used in this investigation. CHEMICAL LABORATORY, ST. THOMAS’S HOSPITAL.