JAPP AND MILLER: KETONES AND DIKETONES. 11 11.-On Additive and Condensation Compouncls of Diketones with Ketones. By FRANCIS R. JAPP, M.A., Ph.D., and N. H. J. MILLER, Ph.D. Iutroduction. IN a former communication by Japp and Streatfeild (this Journal, 1882, Trans., 270), it was shown that phenanthraquinone, acetone, and ammonia react according to the equation- and that, when the compoiind thus obtained is treated with aqueous acids, it takes up water and parts with ammonia, yielding a compound of the formula Cl,H,,O,; this contains the elements of acetone and phenanthraquinone, and may, in fact, be obtained in small quantity by heating these two subst,ances together. It may therefore receive, the name acetonephenanthraquinone. Owing to the ease with which, on heating, it was decomposed into acetone and phenanthraquinone, a12 JAPP AKD MlLLER ON COMPOUNDS OF constitutional formula was assigned to it, in which the union of the two molecules of the generating compounds was represented as taking place by means of the oxygen-atoms.A further study of the reactions of this compound has now shown that it possesses the constitution- C~HT,. C (OH) .CH2. CO.CH3 * I I c6& co Thus when reduced with zinc-dust in acetic acid solution, it yields ;I compound of the formula CI,H,,O :- C~H1403 + H, = C17H120 + 20H2. As the carbon residues of the acetone and phenanthraquinone molecules do not part company during this reaction, it is highly im- probable that they could have heen united merely through the medium of the oxygen-atoms. The chief argument, however, in favour of the constitutional formula above given is to be found in the analogy to several similar compounds to be described in the present communication.Owing to the greater ease with which these compounds can be obtained, their reactions have been more thoroughly studied. The above formula, is also in keeping wit'h the analogy to the additive compound of orthonitrobenzaldehyde with acetone- NO,. CeH*.CH( OH). CHz.CO. CH3, since obtained by Raeyer and Drewsen (Ber. 15, 2858). appears to be- The most probable formula for acetonephenanthraqninoiiimide CsH,. C ( OH). C H,. C 0. CH, C6H4. C (NH) I I That the imido-group attaches itself to the phenanthraquinone resi- due, and not to the acetone residue, is rendered probable by the results which we have obtained by acting with potash solution on a mixture of phenanthraquinone and acetone.In this case no imidogen can replace the oxygen of one carbonyl-group of the phennnthra- quinone : under the influence of the potash, therefore, both carbonyl- groups take part in the reaction, and an additive compound of one molecule of phenanthraquinone with two molecules of acetone is obtained. To this compound we assign the formula- * This formula was first suggested by Dr. Armstrong during the discussion which followed the reading of the above-mentioned paper, but was rejected a t the time by the authors of the paper.DIKETONES WITH KETOSES. 13 C,H*.C( OH). C H2. CO.CH3 C6H,.C(OH>.CH,.CO.C& I I Diacetonephenanthraquinone. Under other conditions, we obtain an additive compound of 2 mols.of phenanthraquinone with 1 mol. of acetone. We also describe con- densation compounds obtained from the above additive compounds by the elimination of the elements of water. I n order to extend the application of these reactions, we have studied the action of potash on mixtures of benzil with acetone and with acetophenone respectively, and have obtained the compounds- C6H5. C(OH).CH,. CO .CH3 CtjH5.C(OEC).CH,.CO.C6Hj C6H5. CO and I C6H5. CO I Acetonebenzil. Acetophenonebeuzil. The compound obtained from acetonebenzil by the abstracbion of water has considerable theoretical interest. Whereas acetophenone- benzil parts with its hydroxjl-group along with a hydrogen-atom from the contiguous methylene-group to yield an unsaturated compound of the formula- C6H5.C : CH.CO.CsHj I C6H5.C0 Y Dehy drace tophenonebenzil.in the case of acetonebenzil the hydroxyLgroup appears to be elimi- nated along with a hydrogen-atom from the methyl-group, so as to form a closed chain componnd of the formula- CsH,. C < : :;> C 0. I C6H,.b0 Dehydracetonebenzil. The evidence in favour of these views will be given in detail in the present paper. 1. PHENANTHRAQUINONE AND ACETONE. a. Action of a Small Quantity of Potash on a Mixture of Phenanthra- quinone and Acetone. 50 grams of very finely powdered phenanthraquinone were intro- duced into it small flask along with 43 grams of pure acetone* ( L e . , in * Many of the reactions deacribed in this paper yield no trace of the desirecl14 JhPP AND MILLER ON COMPOUNDS OF the proportion of 1 mol.of the quinone to 3 mols. of acetone) and 2 C.C. of a strong solution of potash (sp. gr. 1.27) were then added. On shaking the flask, a reaction took place attended with considerable rise of temperature : the phenanthraquinone dissolved, and the liquid became dark-coloured. After standing over night, the whole had solidified to a yellowish-brown mass. This was broken up, then triturated with ether until completely disintegrated, poured upon a, filter, and washed with ether until the liquid ran through only slightly coloured. The treatment of the filtrate will be described later on. The crude substance was thus obtained as a yellowish powder. The first ethereal washings from the above process were almost black, and, as it seemed therefore that the strong potash had exer- cised a prejudicial influence, experiments were made to ascertain whether, by employing a more dilute solution of potash, a purer product could not be obtained.It was found that, under these con- ditions, far less heat was evolved, the solution retained its pale colour, and the quinone, without dissolving, was transformed into an almost white powder. These favourable appearances were, however, entirely deceptive. The white powder, on treatment with solvents, proved to be merely phenanthraquinone superficially coated with the new compound. In order that a complete transformation of the quinone may take place, there must be complete solution, and this, as far as our experiments go, is best effected as above by the use of strong potash, and at the expense of a portion of t’he material.The yellowish powder was dissolved in boiling acetone, in which it is sparingly soluble. From the acetone solution, it was deposited, by spontaneous evaporation,* in well-formed short oblique prisms, which, after recrystsllising two or three times, were quite colourless. The new substance is sparingly soluble in the ordinary organic solvents. Glacial acetic acid and amyl alcohol dissolve it most readily, but by boiling with these solvents it is speedily decomposed. Acetone and benzene were found to be the most suitable solvents. products unless pure acetone is employed, and for this purpose a purification by means of the bisulphite compound is necessary. I n the case of one of these reac- tions, which with acetone from the bisulphite compound gave excellent results, we attempted to employ a sample of acetone which had been repeatedly fractioned with a Le Bel-Henninger apparatus, and showed a constant boiling point ; but hardly any of the compound was obtained.We may remark that various bought specimens of acetone, ostensibly prepared “ from the bisulphite compound,” utterly failed to stand the test of these reaction?, and yielded only resinous products from which nothing definite could be extracted. * In a short preliminary notice of this reaction, forwarded to Berlin in English, and published in the Berichte (16, 282), the words of the English manuscript ‘‘ by spontaneous evaporation ” were transformed by the Berlin translator into ’‘ unter freiwilliger Erwarmung ” (“ with spontaneous rise of temperature ”) .DIKETONES WITH KETONES.15 The melting point cannot safely be employed as a criterion of purity. When heated, the compound turns yellow between 150" and 160", and melts a t 187", a t the same time giving off acetone. The substance which remains behind is impure phenanthraquinone. Analysis gave numbers agreeing with the formula C,,H,oOI :- Substance. CO,. OH,. I ...... 0.1594 0,4321 0.0930 I1 ...... 0.1550 0.4212 0.0908 I11 ...... 0.1713 0.4637 0.09 78 Calculated for Pound. -7 C,OH2004. r-.-A-- rh-- 7 I. 11. 111. C20 .... 240 74.07 73.93 74.11 73.82 H2, .... 20 6.18 6.47 6.50 6-34 0 4 .... 64 19.75 324 100.00 - - - - - Analyses I and I1 were made with one preparation; in I11 a The compound has been formed according to the equation- second preparation was employed. C,aH@, + 2C3HtiO = CmHmOa, Phenanthmquinone.Acetone. Diacetone- phenanthraquinone. and we assign to it the constitution expressed by the formula already given on p. 13. In accordance with the system of nomenclature adopted by us for these compounds, it would receive the name diacetonephenant hraquinone. Action of Acetic Anhydride on Diacetonephenanthrapuinone.-A few grams of the compound were boiled with acetic anhydride. No more anhydride than is sufficient for solution ought to be employed, and the boiling should be discontinued as soon as everything has dissolved, otherwise only a red resin is obtaiiied. On cooling the liquid-an operation which ought to be performed rapidly-white crystals were deposited ; these were washed with ether and recrys- tallised from boiling benzene.The new substance was thus obtained in colourless pointed prisms, melting a t 179-181". A further quantity of less pure substance can be separated from the acetic anhydride mother-liquor by shaking it with water; but if alcohol is employed to destroy the. excess of anhydride, nothing but the red resin is obtained. The colourless compound gave the following results on analysis :-16 JAPP AND MILLER ON COMPOUNDS OF Substance. cop OH,. I ........ 0.1478 0.4232 0.01340 11.. ...... 0-1174 0.3365 0.0628 These numbers lead t o the formula C20H,803 :- Calculated for FDund. CWH1803. r-L 7 C,, ...... 240 78.43 78.13 78-18 HIS ...... 18 5.88 6.31 5-94 0,. ....... 48 15.69 306 100.00 r--,---7 I.11. - - -- -- Different preparations were used in these analyses. The compound is therefore formed from diacetonephenanthra- quinone by the elimination of a molecule of water, and may be named deh y d rodiace ton eph enan t h r a p iizone : - C20Hm04 - OH, = C20H1803. Dehydrodiacetone- phenaiithraqui none. There are various ways in which a molecule of water might be removed from a compound of the constitution of diacetonephenan- thraquinone: our experiments do not enable us to decide between these. We could not succeed in preparing an additive compound with bromine. Taking into consideration the fact that monacetonephenanthra- quinone does not part with water when treated with acetic anhydride, it seems the most probable view that in the foregoing case the dehydration takes place between two hydroxyl-groups.Regarding diacetonephenanthraquinone as a glycol, the compound obtained by dehydration would thus be an analogue of ethylenic oxide :- CGH4.C -CHz.CO.CH, CGH,. C-CHZ. C 0. CH, I I>O Formation of Monacetonephena~ntlwaquinone. - The dark-colonred ethereal washings obtained in the preparation of diacetonephenan- thraqninone were decolorised by shaking with freshly ignited animal charcoal. On spontaneous evaporation, the solution deposited colourless crystals of monacetonephenanthraquinone, recognisable by its melting point (goo), and its characteristic crystalline form.DIKETONES WITH KETONES. 17 b. Action of an Excess of Potash on a Misture of Phenanthrapuinoiae and Acetone. If an excess of the potash solution (sp.gr. 1.27) is employed in this reaction, the yield of diacetonephenanthraquinone is not so good, and the ethereal washings contain a new substmce. The quantity of the latter was small; but sufficient was obtained for a,nalysis. It was crystallised several times from boiling benzene until a constant melting point was obtained. It is deposited from the benzene solution in groups of small colourless needles, which under the microscope appear as long pointed prisms. It is moderately soluble in hot benzene or alcohol; almost insoluble in these liquids in the cold. Ether dissolves it rather readily. It melts a t 195" without evolving gas. Analysis gave figures leading to the formula C1,HI2O2 :- Substance. cop OHZ. I ........ 0.1310 0.3946 0.0591 Calculated for * C17H,202* Found. C, ,........ 204 82-25 82.1 5 HI2 ...... 12 4-85 5.01 0, ........ 32 12.90 248 100.00 - -- - The compound is a condensation-product of phenanthraquinone and acetone, and we therefore propose to name it dehydracetonephenccntkra- pinone :- CiaHsOz + CJ&O = CiJizO, + OH2. Phenanthraquinone. Acetone. Deh ydracetone- phenanthraquinone. It is probably formed from monacetonephenantbraqninone by elimi- nation of a molecule of water. The corresponding transformation in the case of monacetonebenzil is described later on. The difficulty of obtaining this substance in any quantity precluded a study of its reactions. Its constitution, however, is probably analogous to that of the corresponding condensation-product of benzil and acetone ( q . ~ .) . c. Acetonephennnthraguinone. I n order to study the reactions of this compound, a quantity of it was prepared from acetonephenanthraquinonimide by a method mentioned by Japp and Streatfeild (Trans., 1882, 273), but not VOL. XLVTI. C18 JAPP AND MILLER ON COMPOUNDS OF further worked out by them. This method consists in the decompo- sition of the latter compound with an aqueous solution of oxalic acid. The following mode of applying the reaction was found to give satis- factory results: 50 grams of phenanthraquinone were shaken in a flask with 60 grams of acetone and 40 C.C. of strong aqueous ammonia, and the acetonephenanthraquinonimide thus formed was filtered off and washed with ether as described in the above-mentioned paper. Without getting rid of the adhering ether, the crude compound was suspended in water, and the r e d t i n g thick cream poured, with constant stirring, into a solution of 90 grams of crystallised oxalic acid in 800 grams of water (temperature about.25'). Almost every- thing dissolved ; but in a short time the liquid became turbid, and the separation of minute needles of acetonephenanthraquinone commenced. After standing for two days, the compound was separated by filtration, and thoroughly washed with cold water in order to remove any oxalic acid. It was then dried by exposure to the air, and dissolved in ether. By spontaneous evaporation of the ethereal solution, the compound was obtained in large lustrous prismatic crystals. The last ethereal mother-liquors, which are rather dark, may be decolo- rised by shaking with freshly ignited animal charcoal.From 50 grams of phenanthraqixinone, 37 grams of a pure product were obtained. A c t i o n of Nascent H y d r o g e n o n Acetonephenanthraquinone.-A quantity of the above compound was dissolved i n cold glacial acetic acid, and zinc-dust was added in small quantities from time to time, keeping the flask in cold water. After a few days, the whole was poured into water to precipitate the substance and dissolve zinc acetate. The substance was then collected along with the excess of zinc-dust, dried at ordinary temperatures, and extracted from tlie zinc-dust with ether, in which it, is very soluble. The impure substance, remaining after evaporation of the ether, was crystallised from hot alcohol, which removed a quantity of very soluble red gum.It was deposited from the alcoholic solution in long slender needles, which, on recrystallisation from the same solvent, melted constantly at 121". The compound is soluble in almost all proportions in ether and chloroform, readily soluble in boiling alcohol, almost insoluble in cold alcohol. It sublimes without decomposition in feathery crystals. The yield of substance is small. Analysis gave the following results :- Substance. cop OH,. I.. ...... 0.0986 0.3168 0.04aa I1 ........ 0*1130 0.3601 0.0545 These numbers lead to the formula C1,HI20.DIRETONES WITH KETOXES. 1 9 Calculated for Found. 7 C1iH120. (--A- * 1. 11. C1, ...... 204 87-93 87.62 87.68 H,, ...... 12 5.17 5.49 5.40 0 ........ 16 6.90 232 100.00 - - - - These analyses were made with different preparations of sub- stance.I n the formation of this compound, 1 mol. of acetonephenanthra- quinone takes up 1 mol. of hydrogen and parts with 'L mols. of water :- C,7Hi,O, + H2 = c17H120 + 20Hz. When bromine is added to a solution of the substance in chloro- form, the colour of the bromine instantly disappears, and a bromine- derivative-probably additive-separates in slender needles. The quantity obtained was not sufficient for analysis, and the prepara- tion of a larger quantity would have involved the sacrifice of mom phenanthraquinone than we could conveniently spare. Action of Dilute Potash o n Monctcetonephe?zanthrapuino?ze in Alcoholic Solution.-A few drops of dilute aqueous potash were added to a cold alcoholic solution of monacetonephenanthraquinone.The liquid a t once assumed a yellow colour, and minute crystals adhering to the sides of the vessel soon began to form. The separation of crystals was complete in about 24 hours, when the liquid was poured off, and the crystals, which welre yellow, were washed with alcohol, dried, and recrystallised from boiling benzene until they were colourless. Thus obtained, the substance forms minute rhomboidal crystals which, when heated, turn yellow at 150-160", and melt at 190", evolving gas and leaving an orange-coloured residue of phenanthra- quinone. It is very sparingly soluble in all the usual solvents, and is deposit'ed from its benzene solution only after long standing. Analysis gave figures agreeing with the formula C31H2205 :- Substance.co,. OH,. I.. ...... 0-1344 0.3858 0.0617 11. 0.1054 0.3020 0 * 0 45 6 C31H.2205. r-A- 1 F - 7 I. 11. CB1 ...... 372 78.48 78.28 78.14 H2, ...... 22 4-63 4.80 4.80 0 5 . . ...... 80 16.89 I - ....... Calculated for Found. -- - 474 100*00 c 220 JAPP AND MILLER ON COMPOUNDS OF The formation of this compound may be expressed thus :- 2G,Hi*O3 = CSiHzzO, + CsHsO. Monacetonephenanthra- Acetonediphen- Acetone. quinone. anthraquinone. It thus contains the elements of 1 mol. of acetone with 2 mols. of phenanthraquinone, and may receive the name ncetonediphenanthra- quinone. Judging from analogy, it most probably possesses the con- stitution- CsH,. C (OH). CH,. CO .CH,.C( OH). C6H4 CO--CsHr C,H,.CO I I . 1 1 There are thus three distinct, compounds containing the elements of phenanthraquinone and acetone in different proportions :- Phenanithra- Resulting Acetone.quinone. compound. 1 mol. + 1 mol. = C1,H1403 (Acetonephenanthra- qninone) . quinone) . quinone) . - - C20H200a (Diacetonephenanthra- C31H2205 (Ace tonedip henanthra- 2 7 9 + 1 97 1 >, + 2 ,> - - Action of Strong Potash on a Solution of Acetonephenanthraquinone in Acetone.-On adding an excess of strong potash (sp. gr. 1.27) to a cold concentrated solution of acetonephenanthraquinone in acetone, the liquid became dark, and considerable heat was liberated. On cooling, the liquid layer floating on the surface of the potash solidified, and on washing the substance with ether and recrystallising it from acetone, the characteristic crystals of diacetonephenanthraquinone were obtained, melting at 187".The following reaction had there- fore occurred :- C17H1403 + C3H6O = CZOHZOO4. Acetone- Acetone. Dittcetone- phenanthraquinone. phenanthraquinone. Action of Amines on Acetonephenanthraquinone.-It had already been shown (Zoc. cit.) that by the action of ammonia on acetone- phenanthraquinone, one oxygen-atom of this compound could be replaced by imidogen. We, therefore, determined to study the action of amines. The reaction was allowed to take place in the cold in ethereal solution. With ethylamine, nothing b u t a green gummy mass was obtained which turned blue when treated with hydrochloric acid. With diethy lamine, crystals were gradually deposited from the21 DIKETONES WITH KETONES. ethereal solution. These proved to be acetonediphenanthraquinone ; so that the action of diethylamine was identical with that of potash.With aniline, the substance yielded nothing but a green, gummy mass. 2. BENZIL AXD ACETONE. By the action of an aqueous solution Qf potash on a mixture of benzil and acetone, three distiuct products may be obtained according to the conditions of the experiment. a. By acting with a small quantity of potash on benzil dissolved in an excess of acetone, the additive compo,und acetonebenxi,? is formed :- a. Cl4HlO02 + CsH60 = C17H1603. Acetonebeazil. b. By employing an excess of potash under conditions otherwise the same as the foregoing, a condensation-product of 1, mol. of benzil with 1 mol. of acetone is obtained :- c. By acting with a small quantity of potash on acetone mixed with an excess of b e n d , a condensation-product of 2 mols.of benzil with 1 mol. of acetone is formed :- a. Action of a small Quantity of Potash o n a Mizture of Benxil with Zxcess of Acetone. 50 grams of finely powdered b e n d are introduced into a flask along with 30 grams of acetone," and $ C.C. of strong potash (sp. gr. 1.27) is added. The flask is then corked, after which the whole is shaken until the benzil has entirely dissolved, about an hour being required for this operation. The liquid at the same time assumes a reddish colour. If, after standing for two or three days there is no sign of crystallisation, a drop of the liquid should be removed, allowed to solidify by exposure to the air, and the crystalline sub- stance thus obtained added to the contents of the flask.The whole is again allowed to stand as long as the separation of crystals con- * The employment of an acetone purified by means of the bisulphite compound is i n this reaction indispensable-not merely for obtaining a good yield of the com- pound, but in order to obtain any of the compound a t all (see note, p. 13).22 JAPP AND MILLER ON COMPOUNDS OF tinues, a process which is generally complete in about a week. Be- fore pouring off the still liquid portion from the crystals, it is ad- vantageous, especially in warm weather, to allow the flask to remain for some time in a refrigerator. On the other hand, the aid of the refrigerator must not be called in before the reaction is complete, otherwise a separation of unaltered benzil will occur.The crystals, after draining from the mother-liquor, should be washed with a small quantity of ether (which must be free from alcohol, since alcoholic potash has a specific action on the substance), then dis- solved in ether, and the solution allowed to evaporate spontaneously. I n this way the new compound is obtained in large colourless square prisms, frequently a quarter of an inch in thickness, with flattened ends and corners generally cut off. It is deposited from a hot alcoholic solution, on cooling, in small lustrous crystals. It is readily soluble in ether and in hot alcohol; but only moderately in cold alcohol. It melts a t 78". The powdered substance, after drying over sulphuric acid, is electric. A further quantity of the substance can be obtained from the oily mother-liquors, but it was found more advantageous to treat these with an excess of strong potash, and in this way to obtain the condensation-compound dehydracetonebenzil, C1,Hl4O2 (vide i n f i n ) , which, from its sparing solubility, can more readily be purified.Analysis of the substance gave figures leading to the formula CIjHieO3 :- Substance. GO,. OH,. I ........ 0.1190 0.3306 0,0644 11. ....... "0-1 305 0.3634 0.0718 CalcuIated for Found. c ,7 H 160 3' 7- 7 rL-- -3 I. 11. C,, .... 204 76-11 75.76 75.91 Hi, .... 16 5.97 6*@1 6.11 0, ...... 48 17.92 268 100.00 - - - -- The substance is therefore an additive compound of benzil a i d acetone in equal molecular proportion (see Equation a, p. 21). It, would receive the name acetonebenziZ.(For the constitutional formula of this compound see p. 17.) Dilute potash cannot be employed with advantage in the preparation of this compound. The react<ion requires a longer time than with coiicentrated potash, and there is the additional drawback that a larger quantity of uncrystallisable oily substance is formed. A c t i o n of Heat o n AcetonebenziL-A weighed quantity of substanceDIKETONES WITH KETONES. 23 was introduced into a tubulated flask, which was connected with another similar flask, the latter to act as receiver. The flask with the substance was heated in a sulphuric acid bath, and the receiver was cooled with ice. A little below 200°, a few drops of liquid distilled over, and a t 200" the liquefied substance in the flask boiled slowly, whilst a colourless liquid collected in the receiver.The heating was continued until nothing further distilled over. The dark-coloured residue, which solidified on cooling, was weighed; and the distillate was also weighed. The solution, on cooling, deposited characteristic, yellow, needle-shaped crystals of benzil, melt- ing at 94". The distillate had the odour of acetone, and on redistillation it boiled between 56" and 58". Mixed with hydrogen sodium sulphite, it became hot, and, on cooling, the liquid deposited crystals of the acetone double compound. The residue was dissolved in hot alcohol. The following are the quantitative results :- Weight of substance .............. 6.38 grams. ,? residue (benzil) ........ 5.25 ,, Loss.. .......................... 0.03 ,, 9 , distillate (acetone) ......1.10 ?, Supposing the decomposition to have taken place according to the equation- C,,H,ba = C,,H& + C3Hs0, Acetonebenzil. Benzil. Acetone. the weight of b e n d obtained from the above weight of substance ought t o have been 5 grams, and that of the acetone 1.38 grams. For an experirnent of this kind, the above may be regarded as a sufficiently close approximation. The decomposition is therefore analogous to that which acetom- phenanthraquinone undergoes under the influence of heat (Trans., 1882, 274). Oxidation of AcetoneberzziL-The study of the oxidation of this com- pound was undertaken, less with a view of throwing light on its constitution than of comparing its behaviour towards oxidising ngent,s with that of the condensation-product dehydracetonebenzil (4.v.).By oxidation with a mixture of potassium dichromate and dilute sulphuric acid, hhe products obbained were benzoic and acetic acids. The formation of acetic acid is of importance in conneckion with the fact that the condensation-product yields no trace of acetic acid on oxidation. We also attempted, by oxidising with chromic anhydride in acetic24 JAPP AND MILLER ON COMPOUNDS OF acid solution, to obtain some intermediate product, but without success. Action of Ammonia on, Acetonebend.-20 grams of the cornpound were dissolved in ether, and the solution was saturated with dry ammonia. A separation of crystals cominenced during the process, and, on standing, the quantity of crystalline substance increased. The liquid was poured off, the cryst,als washed with ether, and recrystal- lised from boiling alcohol, from which the compound was deposited in groups of small, colourless plates, melting a t 176".In melting, i t turns red and evolves gas. The crystals also assume a faint pink colour by long exposure to the air. With hydrochloric acid and with oxalic acid, they yield a red gum. Analysis gave results agreeing with the formula CI~H~~NO, :- Substance. cop OH,. I.. ...... 0.1316 0,3662 0.07134 11.. ...... 0-1468 0.4096 0.0880 111.. ...... G.1228 0.3438 0.0730 IV. 0.1104 gram burnt with cupric oxide in a vacuum gave 5.40 C.C. V. 0.0886 gram gave 4.00 C.C. moist nitrogen at 14", and under moist, nitrogen a t 14', and under 756 mm. pressure. 759 mm. pressure. Calculated for Found.m C',;H17N02. r-- 7Jc-- -7 I. 11. C , 7 . . .. 204 76.40 75-89 76.09 76.35 - - HIT.. .. 1 7 6.37 6.61 6.66 6.60 - - N . . . 14 5.24 0, . 5 . . 32 11.99 u 5.72 5.30 - -- - - -- - - - -- 267 '100.00 Different preparations were employed in these analyses. The following equation expresses the formation of this compound :- C,,H,,O, + NH, = C,,H,,NO, + OH,. Acetonebenzil. Acetonebenzilimide. This formation of acetonebenzilim,ide, as we propose to name the compound, corresponds with that of acetonephenanthraquinoiiimide (Trans., 1882, 274), from acetonephennnthraquinone and ammonia. Action of Hydroxylanzine 07% Acetonebenzi1.-An attempt to prepare a hydroxylamine-derivative by heating the compound in alcoholic solu- tion with hydroxylamine hydrochloride failed.The s o l u t i o n instantly became red on warming, and nothing but red resin was obtained. AsDIKETONES WITH KETONES. 25 this was probably due to the action of the hydrochloric acid liberated in the reaction, the experiment was repeated, employing free hydroxyl- amine. For this purpose a quantity of acetonebenzil was dissolved in alcohol, and to this liquid a concentrated aqueous solution of two molecuIar proportions of hydroxylamine hydrochloride mixed with a slight excess of sodic carbonate, was added. After standing for two days, a consideramble quantity of a white crjstalline substance had separated. An excess of water was added in order to precipitate the organic mbstance and dissolve the inorganic salts. By recrystallisa- tion from boiling alcohol, the new compound was obtained in small colourless crystals, melting at 146".It is also moderately soluble in boiling benzene, but only sparingly soluble in ether. Hydrochloric acid converts it into a red resin. Analysis gave figures agreeing with the formula C17H,,N0, :- Substance. cop OH,. I .... ,. .. 0.1371 0.3612 0.0756 11.. .. .. .. 0.1114 0.2930 0.0616 111. 0.1070 gram burnt witb cupric oxide in a vacuum gave 4.7 C.C. moist nitrogen at 17*7", apd under 763.5 mm. pressure. Calculated for Found. A C17H,, N 03. r- 7 r - F I. 11. 111. C1, .. .. 204 72.08 71.85 71.73 - Hi,.. .. 17 6.01 6.12 6.14 - N .. .. 14 4.95 - 0, .... 48 16.96 -. - - -- 5.10 -- - 283 100.00 Only one molecule of hydroxylamine has therefore taken part) in the condensation, and the formation of the compound is expressed by the equation- C17Hi603 + PJH,(OH) = C ~ ~ H I ~ N O ~ + OH2.All attempts to induce the compound thus obtained to react with a second molecule of hydroxylnmine failed, although acetonebenzil must be assumed to contain two carbonyl-groups. This negative result is: however, in harmony with the investigations, since published, of Ceresole (Bey., 17, 812), who shows that di-carbonyl compounds do not react with two molecules of hydroxylamine, unless the carbonyl- groups are directly united. A c t i o n of Potash on a Xolution of Acetonebend in Acetone.-A small quantity of the compound was dissolved in acetone, and an excess of strong potash (sp. gr. 1.27) added. After standing for a day, the26 JAPP AND MILLER ON COMPOUNDS OF solution, which had become very dark, was poured into water.The substance which separated was dried and then recrystallised from boiling benzene. It was thus obtained in yellow crystals, melting a t 147" , and proved to he dehydracetonebenzil, a condensation-product of benzil with acetone described later on. The reaction had therefore taken place according to the equation- C17H1603 - OH2 = C1'IHIi027 Deli ydracetonebenzil. and no diacetone compound had been formed as in the case of the corresponding phenanthraquinone reaction. Action of Dilute Alcoholic Potash on Acetoizebenzi1.-Prom 4 to 5 grams of the compound were dissolved in sufficient alcohol to keep the substance in solution in the cold, 5 drops of concentrated alcoholic potash were added, and the whole was allowed to stand in a corked flask.The solution assumed a light-red colour, and colourless lustrous crystals were gradually deposited on the sides of the flask. These were washed with alcohol and then recrystnllised from benzene until the constant melting point 194-195" was obtained. Analysis led to the formula C31H2401 :- Substance. cop OH,. 0.1462 0.4522 0.0692 Calculated for C 3 1 ~ 2 . 1 0 4 . rL- 7 Found. CSl ........ 372 80.87 80.62 H,, ........ 24 5.21 5.19 O I . . ........ 64 13.92 460 100.00 - - -- The compound is identical with one described later on (dehydracetone- dibenzil) obtained by the action of potash on a mixture of acetone with excess of benzil (see also Equation c, p. 21). The reaction in which it is formed in the present case differs from the reaction of alcoholic potash with acetonephenanthraquinone (p.20), inasmuch as with the quinone compound there was elimination of acetone only, whereas in the present case both acetone and water are eliminated :- 2Ci7H1,Oa = c31Euo1 + CaHc,O + OH,. Acetonebenzil. Dehydracetone- Acetone. di hcnzil. The action of various other reagents-phosphoric chloride, acetic anhydride, nascent hydrogen-was tried, but without yielding any definite result.DIKETONES WITH KETONES. 27 b. Action of an Excess of strong Potash on a Mizture of B e n d with aw Ezcess of Acetone. 100 grams of pure acetone, 150 grams of finely powdered benzil, and 1 C.C. of potash solution (sp. gr. 1.27) were introduced into a flask, and shaken until all the benzil had dissolved; after which 20-:?0 C.C.of the potash solution were added, and the whole, after thoroughly shaking, was allowed to stand for a day. A t the end of this time, the layer of acetone and benzil floating on the surface of the excess of potash had solidified. The potash was poured off, and the contents of the flask were shaken with hot water, which melted the crude product and removed the remains of the potash. The product, which solidified on cooling, mas graund in a mortar, extracted in a flask with a, small quantity of ether, and then washed on a filter with ether until the filtrate passed through only slightly coloured. I n this way, the dark-colonred impurities were for the most part removed. The yellowish-grey powder thus obtained was crystallised from hot alcohol or benzene until the constant melting point 147" was obtained.The alcoholic solution deposits the cornpound in large, canary-yellowx prisms; from benzene it separates in tufts of needles of the same colour. Animal charcoal had no effect in removing this colour, nor could the above melting point be altered by recrystallisation. The appear- ance of the compound was perfectly homogeneous, and we had no reason to suspect the presence of an impurity, especially as analysis gave figures agreeing well with those required for a condensation- product of 1 mol. benzil with 1 mol. of acetone, formed with elimina- tion of 1 mol. of water. I n an experiment, however, to be described later, in which the substance was oxidised with chromic anhydride in acetic acid solution, there was obtained, along with a new acid, a colourless neutral substance, which was deposited from benzene in forms indistinguishable from those of the above yellow compound, but melting a t 149".On analysis this colourless compound gave figures agreeing with the formula deduced f o r the yellow compound. The oxidation had therefore removed from the supposed yellow com- pound a coloured impurity, and had at the same time slightly raised the melting point. In order to dispel any doubt as to the identity of the white and yellow compounds, a mixture of the two was dissolved in benzene. Crystals of a p a l e r yellow colour were deposited, not a mixture of white and yellow crystals. The following are the analytical results, which lead to the formula C17H1102 :- See, however, following paragraph.28 JAPP AND MILLER ON COMPOUNDS OF Substance.coz. OHz. I ...... 0.1176 0.3502 0-0614 I1 ...... 0.1234 0.3679 0.0638 I11 ....... 0.1330 0.3978 0.0686 IV ...... 0.1348 0.4034 0.0700 Calculated for Found. C17H1402. r- 7 Cl,.. .... 204 81.60 81.21 81-31 81.57 81.61 Hid.. .... 14 5.60 5.80 5.74 5.74 5-77 0, ...... 32 12.80 250 100*00 I. 11. 111. IV. rL- 7 - - - - -- -- Analyses I and I1 were made with the yellow substance ; I11 and The formation of the compound is expressed by the equation- I V with the colourless substance. In all probability acetonebenzil is formed as an intermediate product. The conversion of acetonebenzil into this compound has been already described (p. 25). As the new compound is derived from acetonebenzil by the removal of a molecule of water, we propose to name it dehydracetwdenzil.It is worthy of note that this con- version cannot be effected by means of acetic anhydride. As regards the constitution of dehydracetonebenzil, apparently the most natural supposition would be to regard it as an analogue of Claisen and Ponder's benzalacetone, C~H,F,.CH CH.CO.CH3, (Annulen,, 223, 138), obtained by the condensation of benzaldehyde wit'h acetone under the influence of dilute caustic soda. According to this view, it would be a benzoyl-derivative of benzalacetone and would possess the formula- The behaviour of this compound with bromine and with oxidising agents is, however, quite incompatible with this view. Its beha- vinur with oxidising agents can best be accounted for by the supposi- tion that it is a closed-chain compound of the forrnula-DIKETONES WITH KETONES. 29 C6H,.C0 Deh ydracetonebenzil.(See also p. 13.j Action of Bromine o n Dehydracetonebenzil.*-lO grams of the compound were dissolved in su5cient chloroform to keep the whole in solution in the cold, and a solution of bromine in chloroform was gradually added. The colour of the bromine did not disappear. After standing for some time, fumes of hydrobromic acid were given off, and a crystalline substance was deposited. Both the chloroform and the bromine had been carefully dried before using. The crystalline substance was washed with chloroform and recrystallised from hot glacial acetic acid, from which it was depo- sited on cooling in slender colourless needles, melting a t 172” with blackening and decomposition.It is only sparingly soluble in alcohol. Two bromine determinations (method of Carius) gave figures pointing to a monobromo-substitution compound :- Substance. AgBr. I ........ 0.2872 0.1627 I1 ........ 0.2538 0.1458 Found. Calculated for f------ 7 C1,H,3Br02. I. 11. Br in 100 parts .......... 24.21 24.10 24.U In order t o make perfectly sure that this monobrominated com- pound had not been obtained from a dibromide by decomposition during recrystallisa,tion, a fresh quantity of the substance was pre- pared, washed thoroughly with cold chloroform, dried a t ordinary temperatures, and then analysed. This preparation gave 26.2 per cent. of bromine, showing it to be merely an impure monobromo- compound. A compound of formula (I) ought, judging from the analogy of benzalacetone and the other compounds prepared by Claisen, to yield a dibromide.That dehydracetonebenzil has not this formula is rendered still more probable by the fact that dehydracetophenone- benxil ( q . ~ . ) , in which the substitution of phenyl for methyl appears to have prevented the formation of a closed chain, and which has a constitution corresponding with formula (l), readily forms an additive compound with bromine. * For all reactions here dcscribed, the yellow compound was employed.30 JAPP AND MILLER ON COMPOUXDS OF Oxidation of D~hlldmcetonebenzil.--20 grams of the yellow com- pound were dissolved in glacial acetic acid, and an equal weight of chromic acid-also dissolved in acetic acid-was gradually added. The mixture, which became slightly warm, was finally boiled with a reflux condenser in order to finish the reaction. It was then poured into water, which occasioned a separation of organic substance.The whole was extracted with ether, and the ethereal solution was shaken with a solution of sodium carbonate in order to remove acids. On evaporating the ether, some unattacked dehydracetonebenzil was obtained, but in a colourless condition. The sodium carbonate solution was acidified with hydrochloric acid and extracted with ether. On distilling off the ether, an acetic acid solution of a new orgarh acid remained behind, and, by allowing the acetic acid to evaporate in a desiccator over lime, the new acid was obtained in almost colourless crystals, whilst any benzoic acid that had been formed remained in the mother-liquor.The acid was recrystallised from boiling benzene until it showed a constant melting point. Thus purified, it forms tufts of colourless needles, melting at 152'. It is readily soluble in boiling benzene, but separates almost entirely on cooling. Boiling water dissoIves it sparingly. Analysis gave the following results :- Substance. ( 2 0 2 . OH,. I.. ...... 0.1331 0.3682 0.0672 11.. ...... 0,1276 0.3530 0.0642 These numbers lead to the formula C16&03 :- Calculated for Found. C16H1403* rd- 7 TL-- I. II. C,6 .... 192 7 5 - 2 75.44 75.44 H 1 4 .... 14 3-51 5.61 5-59 O3 ..... 48 18.90 254 100.00 - - - -- A siZver s d t was prepared by precipitating a solution of the ammon- The dry salt It gave the following figures on combustion :- ium salt with silver nitrate.is electric. It forms a white powder. Substance. cop OH2 4 5 0.1418 0.2752 0.0466 0,0424DIKETONES WITH KETONES. 31 Calculated for C16H1303Ag* w- Found. C I S . . a . 192 53.18 52-92 Hi3 ...... 13 3-60 3.6 4 Ag ....... 108 29.92 29.90 O3 ........ 48 13-30 361 100.00 - - -- The barium salt was obtained by boiling the acid with barium car- bonate, and allowing the solution t o evaporate over sulphuric acid. Like all the soluble salts of this acid which we examined, its difference of solubility in hot and cold solutions is very slight. It was obtained in rosettes of flat prisms of the formula (c16H1303),Ba,20H,. It parts with its water of crystallisation at 100". 0.3064 gram of air-dried salt lost at 100" 0.0160 gram, and the resulting 0.2904 gram anhydrous salt gave 0.1042 gram barium sul ph ate.Calculated for (C16H1303),Ba,20H2. Found. OH, in 100 parts ....... 5.30 5.22 Ba in 100 parts ........ 21.30 21.09 Calculated for (C16H1303)2Ba* Found. The acid is formed from dehydracetonebenzil according to the equatian- Ci,Hi,O, + 3 0 = Ci6H1403 + COz. Supposing dehydracetonebenzil to possess the constitution repre- sented by formula (l), the oxidation of such acompound to an acid of the formula ClsH140, would be very difficult t o account for. It would be necessary to assume that the methyl-group at the end of the chain is oxidised away, the carbonyl-group converted into carboxyl, and that then the two unsaturated carbon-atoms, under conditions which generally lead to the separation of such atoms, take up, in presence of a powerful oxidising agent, hydrogen and become saturated.On the other hand, by adopting formula (2) for dehydracetone- bend, the formation of the acid can be explained. In the first place by a separation of the carbonyl from one of the methylene-groups in the closed chain, and by conversion of carboxyl-groups, an acid of the formula ............... COO H C6H5*7 < c H,. c o OH C6&.CO these separated groups into3-2 JAPP AND MILLER ON COMPOUNDS O F would be obtained. It is well known that ketonic acids vary greatly in stability according to the class to which they belong. Those ketonic acids are stable in which carboxyl and carbonyl are directly united, as in pyruvic acid, or in which carboxyl and carbonyl are attached to different carbon- atoms, as in lawulic (/%acetylpropionic) acid ; those are unstable, in which carboxyl and carbonyl are attached to the same carbon-atom, as in the case of acetoacetic acid: such acids readily part with carbonic anhydride, yielding a ketone.An acid of the above formula would unite in itself the properties of two of the above classes; it would be unstable as regards the carboxyl-group attached to the same carbon-atom as the benzoyl-group ; and it would be stable as regards the ohher carboxyl-group. Under the conditions of the oxidation experiment, this acid would part with carbonic anhydride from the first of these carboxyl-groups (as indicated in the formula), yielding the monobasic acid :- C6Hs.CH.CH2.COOH This acid is a bibasic ketonic acid.I C6H5. C O which would thus be P-benzoylhy drocinnamic acid. Although we regard the above as the most probable of the various constitutional formula that might be suggested for the acid, we must call attention to the fact that the only reactions which we have tried in confirmation of this coiistitution have yielded negative results. Thus, by the action of nascent hydrogen we hoped to obtain a lactone ; but after subjecting the acid for some days to the action of sodium-amalgam, nearly the whole of the original substance was recovered unchanged, and only a trace of an indifferent oil was formed, which, however, did not appear to be a lactone, as it did not dissolve in caustic alkalis on heating." We further hoped, by the action of hydroxylamine, to prove the ketonic character of the acid; but no action took place with hydroxylamine hydrochloride in aqueous alcoholic solution a t 100".The various other constitutions that might be suggested for this acid-thus that it is an acid of the glycidic type, or that it is an unsaturated acid containing an alcoholic hydroxyl-group-are still less in keeping with its reactions and mode of formation, The chief obstacle to a thorough study of this acid is the di&culty of obtaining it in any considerable quantity, the yield being very small. As a dehydracetonebenzil of formula (1) ought to yield acetic acid on oxidation, it seemed of importance to show that this acid was not * Of course this insolubility in caustic alkalis does not absolutely prove that the oil was not a lactone.DIKETONES WITH KETONES.33 formed, especially as acetonebenzil readily yields acetic acid. A quantity of dehydracetonebenzil was therefore boiled with a mixture of dilute sulphuric acid and potassium dichromate, until all action had ceased. The liquid was then distilled until about a third had passed over; the distillate was filtered from benzoic acid, neutralised with sodium carbonate, and evaporated to dryness. Not the slightest trace of acetic acid could be obtained from this residue. Another portion of dehydracetonebenzil was oxidised with a 5 per cent. permanganate solution in the cold. The filtrate from the man- ganese dioxide, when acidified and extracted with ether, yielded as chief product benzoic acid, along with a small quantity of benzoylforniic acid, identified by means of the characteristic thiophene reaction.We satisfied ourselves that p-benzoylhydrocinnamic acid does not give this reaction. The action of the following reagents upon dehydracetonebenzil WRS also tried, but without definite result :-acetic anhydride, alcoholic ammonia, potash (fusion) , hydriodic acid and amorphous phosphorus, zinc-dust a t higher temperatures, zinc-dust with acetic acid. c. Action of a Xmall Quantity of Potash on a 2Cli'zture of Acetone with Excess of Bend. 50 grams of finely powdered benzil were introduced into a flask with 20 grams of pure acetone and 4 C.C. of potash solution (sp. gr. 1.27). The flask was shaken until all the benzil had dissolved, this process requiring about an hour. After standing for a day, the con- tents of the flask, which were almost solid, were shaken with ether.A small quantity of acetonebenzil went into solution and a sparingly soluble white crystalline powder remained, which, after washing with ether, was recrystallised from benzene until the constant melting point 194-195" was obtained. Besides acetonebenzil, the ethereal washings contained an uncrystsl- lisable gum, which, however, by shaking with more potash, could be converted into the sparingly soluble crystalline compound. Analysis gave results agreeing with the formula C31H,a0, :- Substance. CO,. OH,. I.. ...... 0.1194 0.3520 0.0550 11. ....... 0.14'20 0.4182 0.0666 111.. ...... 0.1656 0.4874 0.0790 VOL. XLVII. D34 JAPP AND MILLER ON COMPOUNDS OF Calculated for Found.C3lH2.104 * I----- 7 r--L-- I. IT. 111. C,, .. .. 372 80.8; 80.40 80.32 80.27 H,, .. .. 24 5.21 5.30 5-21 5-30 0 4 .... 64 13.92 - - - -- -- 460 100.00 The compound is identical with that obtained by the action of dilute alcoholic potash on acetonebenzil (p. 26). I t s forniation in the present case occurs by the abstraction of 1 mol. of water from 2 mols. of benzil and 1 mol. of acetone. It woixld therefore receive the name dehydra~etonedibenzi1:- 2CiaH1002 + CJLO = CsiH2aOa + OH,. Dehydracetonedibenzil is almost insoluble in cold benzene and alcohol; even boiling alcohol dissolves it but sparingly. The best solvent is boiling benzene, from which it is deposited, after some time, in well-formed colourless crystals. From the alcoholic solution, it separates with 1 mol.of alcohol of crystallisation, which is retained at loo", but given off at 120". When the crystals containing alcohol of crystallisation are heated in a capillary tube they melt at 158- 160". yesult :- Alcohol of cry stallisation was 0.2410 gram air-dried substance, tion, lost a t 120" 0.0222 gram. determined with the following containing alcohol of crys tall isa- Calculated for ~31H'3404,C2H60. Found. C,H,O in 100 parts . . . . . . . . 9.09 9.21 3. BENZ~L AXD ACETOPHENONE. a. A c t i o n of Potash iu the Cold on a Mixture of B e n z i l a n d Acetopherzone. Equal molecular proportions of acetophenone and finely powdered benzil were shaken in a flask with an excess of strong potash (sp. gr. 1.27) and allowed to stand. At first the potash solution remained in suspension, but after a few days it separated, whilst the organic substance formed a solid cake on the surface. This cake was ground with water, thoroughly washed, and, after drying, shaken with ether.The greater part dissolved, leaving a yellow powder, which proved to be dehydracetophenonebenzil, a compound to be described later. On spontaneous evaporation, the ethereal solution deposited large colourless oblique prisms, and the mother-DIKETONES WITH KETONES. 33 liquor from these crystals yielded a further quantity of the same substance, contaminated however with unchanged benzil. By crystal- lisation from alcohol, the colourless substance was obtained pure in flat oblique prisms, melting at 102'. It is readily soluble in ether and in boiling alcohol, sparingly soluble in alcohol in the cold. When heated above its melting point, it gives off acetophenone, Rhich may be recognised by its odour.Analysis gave numbers agreeing with the formula C,,H,,O, :- Substance. co:. OH:. I . . ...... 0.1426 0.4 170 0.0724 11.. . . . . . . 0.1498 6.43378 0.07G6 The formation of this compound is expressed by the equation- Ci,Hio02 + CJ&O = CyZHir303, and it would receive the name aceto~~1~e~zoizelieri;iZ. Its constitotional, formula would be- C6H5.C (0 H).CH2.Co.C,H, I CsH,. c 0 b. Action of Potash, aided by Heat, 011 a Mixtiire of Betwil a h c J Acetophenone. The ingredients were mixed as in the preceding expei*iment; but heat was applied until the whole of the benzil had fused, after which the flask was allowed to stand for some hours at a tein- psratnre sufficiently high to prevent solidification.On allowing the flask to cool, the layer of organic substance floating on the surface of +he potash solidified. The solid cake was treated as in the former experiment. This time, the ethereal extract contained only a small quantity of a reddish oil and a trace of maltmered benzil. No aceto- phenonebenzil was formed on this occasion. The portion undissolved by the ether was recrystallised from hot alcohol, until it exhibited the constant melting point 129". It crystszllises in tufts of flat pointfed yellowish needles, which are very sparingly soluble in ethey and in cold alcohol, but dissolve readily in boiling alcohol. u 236 JAPP AND MILLER: DIRETONES AND KETONES. The analytical figures agreed with the formula C,,H,,O, :- Subatance. con.OH,. I.. . . . . . . 0.1582 0.4900 0.0748 11.. 0.1996 0.4010 0.0628 . . . . . . Calculated for Found. C25)3[1602. 7<-7 rd- -7 I. 11. C2.2 . . . . 264 84.61 84.47 84-38 Hlti . . . . 16 5.13 5-25 5.38 02.. .... 32 10.26 ' - - -- - 312 100.00 This compound is formed by the condensation of a molecule of benzil with a molecule of acetophenone :- It may therefore receive the name dehydracetophenonsbenxil. Acticn of Bromine on DehydracetophenonebenzX--The compound was dissolved in cold chloroform, avoiding an excess of the solvent, and one molecular proportion of bromine was added. On standing, the colonr due to the bromine gradually became much fainter, without however entirely disappearing, and a bromine-derivative was deposited in large crystals. There was no evolution of hydrobromic acid. The crystals were of a reddish colour, and, when exposed to the air, gave off a faint odour of bromine, even after standing for some days, at the same time becoming opaque. As it was found impossible to re- crys tallise this substance without decomposition, the freshly prepared crystals were washed with chloroform, exposed for a short time to the air, powdered, the powder dried for two hours over sulphuric acid, and in this condition analysed. A bromine estimation (Carius) gave figures which were somewhat too high for a tetmbromide-an entirely unexpected result. 0.1864 gram of substance gave 0,2278 gram of silver bromide. Calculated f o r CT4Hl6O2Br4. Pound. Br in 100 parts.. ...... 50.63 52.00 Heated in a capillary tube the substance becomes dark at about rO , turns pale again at aboiit 80", and melts between 110" and 115". The bromine was in a state of very unstable combination. A por- tion of the substance which had been allowed t o remain for some weeks in a desiccator over lime, had parted with nearly the whole f - 0SORABJI ON SOME NEW PARAFFINS. 37 of its bromine, and was found, after recrystallisation from alcohol, to have been reconverted into dehydracetophenonebenzil. We are unable satisfactorily to explain the formation of a tetra- bromide. A dehydracetoph enonebenzil of the formula- C,H,.C CH.CO.C,&, C,H,.CO ought to yield a dibromide, and it is conceivable that this dibromide might form a molecular compound with a second molecule of bromine, similar to the molecular compounds of acetic acid with bromine a i d hydrobromic acid. In any case, the action is nnomaIous and deserves further study. For the present, however, we regard this reaction as sufficient evidence of the unsaturated character of dehydracetophenone- benzil, and the foregoing is the only probable constitutional formula which would represent it as an unsaturated compound. The fusing points of dehydracetonebenzil and dehydracetophenone- benzil also render it probable that these compounds do not belong to one and the same category. Whereas acetophenonebenzil, a com- pound which may be regarded as derived from acetonebenzil by the substitution of phenyl for methyl, fuses higher than acetonebenzil, clehydracetophenonebenzil fuses 20" lower than dehydracetonebenzil. The high melting point of the latter compound is probably due to the fact that it is, as assumed in this paper, a closed-chain compouiid. Norrnnl School of Xcience, So 11 th Kensing ton.