where C1, C2, C3, C4 are the concentration of the four components of the system. If, as Berthelot supposed, the sole product of the action were a 'perdisulphuric acid' of the formula H,S,Og, the equilibrium must correspond with an equation of the second order, and would be represented by an equilibrium curve similar to II in the figure, which was plotted from the formula although this curve is similar in type to the experimental curve, it will be seen that the reversal of curvature is much less pronounced. The effect of lowering the order of the equation is seen in curve I in the figure, which was plotted from the equation in this curve, the reversal of curvature has entirely disappeared and the curve does not bear the slightest resemblance to the experimental curve. It is therefore evident that the experimental curve must be represented by an equation of a higher, rather than of a lower, order than the second. The formula H,SO, (von Baeyer and Villiger, Ber., 1900, 33, 124), which actually gives curve I in the figure, must therefore be rejected. The experimental curve, however, approximates very closely to the curve deduced from an equation of the fourth order, such as and assuming that the chief product of the interaction is a 'persulphuric acid of the series H2O2,nSO, it is evident that it must be the fourth member, that is, its formula must be H2O2,4SO, or H2S4014. The small, but steady, deviations of the experimental curve from that deduced from the simple equation of the fourth order are in the direction of the curves of lower order, and serve to indicate that along with the acid H2SO14 there is produced a much smaller amount of some simpler acid of the series, and it appears only fair to suppose that this is the acid H2SO, which Berthelot assumed was alone formed. The equation for an equilibrium between H2O2, H,S,O,, and H2SO14 would be of the form where k is the equilibrium constant for the 'perdisulphuric' and ką that for the 'pertetrasulphuric' acid. Taking k1 = 0.63 and k10.96, this equation gives results which agree very closely with the observed values for the equilibrium, and there can be little doubt that it represents the experimental data as closely as they can be represented by a simple algebraical equation. There is at present no method known by which the persulphuric acids may be separately estimated, but it will be of interest to record the calculated proportions of H2O, H2SO, and H2SO1, in some of the mixtures. 14 A qualitative test is, however, available for distinguishing the 'perdisulphates' and 'pertetrasulphates,' Caro (Zeit. angew. Chem., 1898, 845) having shown that the 'perdisulphates' produce a brown precipitate in a neutral solution of aniline, but that the neutral salts of the acid formed by the interaction of 'perdisulphates' and sulphuric acid have the property of oxidising aniline to nitrosobenzene. On applying this test, it was found that a mixture of hydrogen peroxide with a large excess of sulphuric acid gave a considerable quantity of nitrosobenzene, but only a trace of the brown precipitate characteristic of the 'perdisulphates'; a mixture of hydrogen peroxide with an equal bulk only of sulphuric acid gave, however, a much more pronounced brown precipitate and only a small amount of nitrosobenzene. There can therefore be little doubt as to the presence of both 'perdisulphuric' and 'pertetrasulphuric acids in the product, the relative proportion of 'perdisulphuric acid' being greatest in the more dilute solutions. It is scarcely necessary to say that the conclusions we have arrived at must be verified by the analytical examination of the product; but as Caro has stated that he is engaged in completing his discovery, we have not thought it right to extend our experiments in this direction. We may mention, however, that solutions containing a very large proportion of the peroxidised acids may be obtained, not only by electrolysing solutions of sulphuric acid, but also by mixing sulphuric acid with hydrogen peroxide, and then concentrating by freezing out the hydrate H,SO,H2O. If our conclusion be accepted, it will follow that the account given by Berthelot, and, in fact, by all who have studied the acid, applies to 'pertetrasulphuric acid' rather than to 'perdisulphuric acid,' which appears under all conditions to be by far the minor product and the less stable, although its salts are more stable than those of 'pertetrasulphuric acid.' The production of perdisulphates' on electrolysing solutions of acid sulphates, whilst solutions of sulphuric acid yield chiefly 'pertetrasulphates,' is of interest as an indication that the electrolysis proceeds on very different lines in the two cases. Attention may also be directed to the fact that, inasmuch as the persulphuric acids are formed by the interaction of sulphuric acid and hydrogen peroxide, it is clear that the affinity of hydrogen peroxide must be greater than that of water for sulphuric anhydride. In conclusion, the authors desire to express their thanks to Dr. Armstrong, at whose suggestion the research was carried out, and to whom they are indebted for much valuable help during the whole course of the work. CHEMICAL DEPARTMENT, CENTRAL TECHNICAL COLLEGE, EXHIBITION ROAD, LONDON, S. W [It is proposed to extend the method here described to the study of the peroxides of other acids. H. E. A.] LXXXV.-Dimethyldiacetylacetone, Tetramethylpyrone, and Orcinol Derivatives from Diacetylacetone. By J. N. COLLIE, F.R.S., and B. D. STEELE, B.Sc. (Melbourne), In a paper published last year (Trans., 1899, 75, 710), one of us showed how dimethylpyrone was capable of acting as a basic substance, forming salts with various acids, and the idea was put forward that oxygen under certain conditions could replace phosphorus, nitrogen, sulphur, VOL. LXXVII. 3 U or iodine in basic organic compounds owing to its quadrivalence, these compounds being supposed to be derived from the theoretical bases: NH,OH. PH,OH. SH2OH. ондон. IH2OH. The present work was undertaken in order to see whether the property of forming these salts was peculiar to dimethylpyrone alone, or was shared by other pyrone compounds. Diacetylacetone, when allowed to react with sodium ethoxide, at once yields a disodium derivative. This, when treated with methyl iodide, should yield dimethyldiacetylacetone, which should give tetramethylpyrone when boiled with acids or when heated: The change undergone by dimethyldiacetylacetone when boiled with acids is due to molecular rearrangement and subsequent elimination This tetramethylpyrone should have properties closely resembling those of dimethylpyrone, and we expected that, owing to the greater number of methyl groups, it might form salts with even greater ease than the latter. We also hoped that the yield would be nearly quantitative. In both cases we were disappointed; the yield is by no means good, several other compounds being produced at the same time, and the basic properties of tetramethylpyrone do not seem to be so well. marked as those of dimethylpyrone. The platinichloride separates at once when a moderately strong solution of platinic chloride is added to a solution of the base in hydrochloric acid, but with other acids crystallisable salts are not easily formed. Although the original object of the investigation has not been attained to the extent that we had hoped, yet the side issues that have opened out are of considerable interest, and the other compounds produced by the reaction show how easily, by means of another kind of condensation, benzene and naphthalene derivatives can be formed. The substances separated from the product after the action of methyl iodide on disodium diacetylacetone are the following: (1) Dimethyldiacetyl acetone, CH140g, m. p. 86-87°. (2) Tetramethylpyrone, C,H12O2, m. p. 92°. 14 3' (3) Hydrate of tetramethylpyrone, C,H140 ̧, m. p. 63–64o. 14 14 (7) Dimethylacetodihydroxynaphthalene, C1H1403, m. p. 183–184°. The reaction is one, therefore, of some complexity, but is an excellent example of how simple condensation of molecules containing the complex CH, CO or its enolic form ⚫CH:C(OH). may be brought about in an alkaline solution. EXPERIMENTAL. When diacetylacetone is dissolved in absolute alcohol and a molecular equivalent of sodium ethoxide added, an insoluble salt is not obtained, but with more sodium ethoxide, a white, crystalline sodium compound separates. Fifty grams of diacetylacetone were dissolved in absolute alcohol, and sodium ethoxide made from 20 grams of sodium (calculated amount for 2Na = 16.2 grams), dissolved also in absolute alcohol, was added. The sodium salt that crystallised out was washed and dried; when pure, it has a blue fluorescence. On analysis : Found, Na = 22.6 and 22.8. C-H1004Na, requires Na = 22.5 per cent. This sodium salt was suspended in alcohol and boiled with 105 grams of methyl iodide (an excess of 5 grams) until the contents of the flask were neutral. The spirit was removed by distillation, and the residue treated with water and extracted with chloroform. After the chloroform had been distilled off in a water-bath, the residue was fractionated under 15-20 mm. pressure. Several fractions were obtained; the bulk of the distillate (25-30 grams), however, was collected between 125° and 150°. This portion, on standing, became semi-solid. It was transferred to a porous plate, and the resulting crystals recrystallised from water. When pure, it melted at 86-87°, and boiled with much decomposition at 230-240° under ordinary atmospheric pressure. On analysis, the substance proved to be dimethyldiacetylacetone : Found C 631; H=8.2 per cent. = The results of several other experiments were as follows: 10 grams of diacetylacetone gave 6 grams boiling at 120-150° under reduced pressure; 30 grams gave 17 grams, and another 30 grams yielded 19 grams. During the reaction between methyl iodide and disodium diacetylacetone, it is absolutely necessary that the alcohol and other substances should be perfectly dry; if this is not the case, considerable quantities of a yellow resin, mainly dimethylacetodihydroxynaphthalene (Trans., 1893, 63, 334), are formed. |