known volume with 0·1 normal acid. It was afterwards found possible to titrate the copper salt solution directly, as the intense blue colour of the copper-ammonia ions disappears before the end point is reached, and the colour due to the copper ions in the very dilute solution which we used in the titration was much too weak to mask the end point in any way. The amount of ammonia found by these methods agreed perfectly. The total amount of ammonia found was usually a trifle less than that added, but the difference was nearly always less than 1 per cent. of the total, and never exceeded 1:5 per cent.

The aqueous solution was obtained by taking a known volume (a) of normal salt solution, adding a known volume (b) of approximately normal aqueous ammonia solution (sufficient to precipitate the metal hydroxide completely, then wholly redissolve it) and then (50-a-b) c.c. of water. The amount of chloroform used was either 100 c.c., 70 of which were taken for the titration, or 50 c.c., of which 40 were taken. The amount of 0.1 normal acid used in the titration varied between 4 and 16 c.c. for the chloroform, and between 10 and 40 c.c for the aqueous solution.

Distribution coefficient of Ammonia between Water and Chloroform. The results obtained at 20° with varying absolute concentrations of ammonia are given in the following table:

The deviations from the mean value shown by the above numbers are quite irregular, and no regular variation of the distribution-coefficient with the concentration is observable. The accuracy of the later determinations in the paper (after some experience had been gained) is probably much greater than that measured by the deviations of the single observations in the above series from the mean.

The coefficient decreases with rise of temperature. Thus, at 25° the following results were obtained:

Experiments with Copper Sulphate Solution.

The requisite volume of normal copper sulphate solution was introduced into the separating funnel, and varying volumes of approximately normal ammonia solution added; then a volume of water so as to make the total addition up to 50 c.c. The distributioncoefficient was determined at 20° with the following results:

* The ammonia was determined by distillation; in the other cases by direct titration. In the cases marked with an asterisk also 50 c.c. of chloroform were used, and 40 taken for the titration, otherwise 100 were used and 70 taken.

agent. With B-phenoxycinnamic acid, we find, however, that condensation to the phenyl derivative of benzo-y-pyrone, namely, flavone, does not occur, as the acid breaks up according to the equation,

C,H,•C(O•C,H,):CH•CO,H+H,O=C,H,*CO•CH,CO,H+CH•OH, yielding benzoylacetic acid and phenol, the latter, however, under the influence of the sulphuric acid, being transformed into its sulphonic acids. Phenoxyfumaric acid, on the other hand, being more stable, as has already been pointed out in the previous paper, partially undergoes the desired change and yields benzo-y-pyronemonocarboxylic acid. This acid, on distillation in a vacuum, loses carbon dioxide and -CO.CH benzo-y-pyrone, CH the hitherto unknown parent sub

stance of flavone and the yellow plant dyes such as chrysin, quercetin, and fisetin, is formed. The yield of the benzo-y-pyronecarboxylic acid is not very satisfactory, since phenoxyfumaric acid appears also to break up in a manner similar to B-phenoxycinnamic acid, with formation of the products of the decomposition of oxaloacetic acid, namely, oxalic and acetic acids.

EXPERIMENTAL.

Ethyl B-o-Methoxyphenoxycinnamate,

CH•C(O•CH¢•O•CH,):CH•CO,C,H.

The experiments previously described were exclusively directed to the study of the interaction between monohydric phenols and esters of the acetylene series, but we have now commenced to investigate the behaviour of dihydric phenols towards these esters. Up to the present we have examined the reaction of guaiacol with ethyl phenylpropiolate. Their union is effected as in the former cases by dissolving sodium (1 at.) in an excess of hot guaiacol, adding the ester (1 mol.), and then warming the mixture until the sodium derivative of the phenol has dissolved. The viscous product on cooling sets to a solid mass; this is agitated with dilute sulphuric acid and ether, the ethereal layer freed from the dissolved guaiacol by shaking it with caustic potash, and the oil which remains after removing the ether by evapora tion distilled in a vacuum. Ethyl B-o-methoxyphenoxycinnamate is a colourless, viscous oil with an aromatic odour, and boils at 230-231° under 15 mm. pressure. On analysis:

0.2038 gave 0.5437 CO, and 0.1122 H2O. C-72-75; H=6·11. C18 H1804 requires C=72·48; H=6.04 per cent.

In its properties, the ester resembles the other aryl ethers of ethyl B-hydroxycinnamate. On boiling it with alcoholic potash for

1 hour on the water-bath, hydrolysis to the corresponding acid takes place.

4

B-o-Methoxyphenoxycinnamic acid, C ̧H ̧·С(O·C ̧H ̧·O·CH2):CH·CO2H, is precipitated from the alkaline solution by dilute sulphuric acid as a sticky product which gradually sets to a hard mass. This dissolves freely in alcohol, and the alcoholic solution, on dilution with water, gives an emulsion which, on standing, changes into colourless prisms. These begin to soften at 120°, and melt at 138° with evolution of gas. On analysis:

0.1914 gave 0.4985 CO, and 0.0907 H2O. C-71-03; H=5.29.

C16H1404 requires C=71·11; H=5·18 per cent.

The solution of the acid in ammonia gives with silver nitrate a white precipitate of the silver salt which darkens at 100°.

o-Methoxyphenoxystyrene, CH, C(O·CH, O·CH):CH2, is obtained when the acid is heated, carbon dioxide being evolved. It distils at 183° under 15 mm. pressure as a colourless oil which solidifies after some time, and melts again at 43-44°. On analysis:

0.1642 gave 0-4800 CO2 and 0·0910 H2O. C-79-87; H=6·15. C15H1402 requires C-7964; H=6·19 per cent.

Condensation of Thiophenol with Ethyl Phenylpropiolate and Ethyl Acetylenedicarboxylate.

The action of thiophenol on the esters of the acetylene series resembles in every respect the behaviour of the phenols towards these compounds. The sodium thiophenolate required for the experiment is prepared by boiling a solution of thiophenol in toluene with sodium (1 at.) in a flask with a reflux condenser until the metal has disappeared and the precipitate is quite white. The required quantity of the ester (1 mol.) is then added, and the resulting product isolated in the same manner as are the corresponding phenoxy-compounds.

Ethyl B-thiophenylcinnamate, CH, C(S.C2H2):CH CO2CH, is in this way formed from thiophenol and ethyl phenylpropiolate. After removal of the excess of thiophenol by shaking the product with caustic potash and then boiling off the toluene used in the reaction, it remains behind as an oil which rapidly solidifies. The ester is fairly soluble in ether, readily so in boiling alcohol, and crystallises from the latter in slightly yellow spangles which melt at 95-96°. On analysis:

0.1577 gave 0.4156 CO2 and 0·0832 H2O. C=71·87; H = 5.86. 0.2324 01927 BaSO4. S=11·40.

C17H6O2S requires C71.83; H=5·63; $= 11.27 per cent.

VOL. LXXVII.

4 L

B-Thiophenylcinnamic acid, CH ̧·C(S•C2H2):CH·CO2H, is obtained by boiling the ester with alcoholic potash for 1 hour. After the alkaline solution has been freed from alcohol as completely as possible by distillation on the water-bath, water is added to the residue, and the solution acidified with sulphuric acid. The white precipitate, thus formed, is almost insoluble in water, but dissolves readily in ether or alcohol, and crystallises from the latter in colourless needles which melt at 163° with evolution of gas. On analysis :

0.1574 gave 0.4060 CO2 and 0·0685 H2O. C=70·34; H=4·83. C16H12OS requires C-70-31; H= 4.68 per cent.

15

The solution of the acid in ammonia gives with copper sulphate a light green precipitate which does not blacken at 100°. On adding silver nitrate to the same solution, a white silver salt is formed, which decomposes when heated on the water-bath, but can be dried in a vacuum over sulphuric acid. On analysis:

0.3993 left on ignition 0.1193 Ag. Ag=29.87.

C15H11O,SAg requires Ag = 29.75 per cent.

Thiophenylstyrene, CH, C(S.CH):CH2.

B-Thiophenylcinnamic acid, like the corresponding phenoxy-acid, on heating, loses carbon dioxide, and thiophenylstyrene is formed. It is a yellowish oil with a peculiar aromatic odour, boils at 174-175° under 14 mm. pressure, and has the density d 28°/28° 1·1024. On analysis:

0.1590 gave 0.4603 CO, and 0.0820 H2O.

C-78-95; H=5·73.

C14H12S requires C=79.24; H=5.66 per cent. Thiophenylstyrene, when heated with mineral acids, is decomposed, yielding only thiophenol and acetophenone.

Ethyl Thiophenylfumarate, CO2C2H ̧•C(S•C ̧H2):CH•CO2C2H ̧· The union of thiophenol with ethyl acetylenedicarboxylate takes place with development of heat, when the unsaturated ester (1 mol.) is carefully added to sodium thiophenolate (1 mol.) suspended in toluene. The oil which remains behind on treating the product of the reaction as in the former cases, is subjected to fractional distillation in a vacuum. The largest portion boils under 12 mm. pressure at 201-202°; then the thermometer rises, and between 220° and 260° a very viscous substance passes over, which partially solidifies. The fraction of lower boiling point has a yellowish colour and its density d 21°/21° is 1.1618. On analysis: