5. On the Tartrarsenites. By G. G. HENDERSON, D.Sc., M.A., and A. R. Ewing, Ph.D. Arsenious oxide dissolves readily in a boiling solution of sodium hydrogen tartrate, and on concentration and cooling a compound of the formula CHO.ASONa. 24 Aq. crystallises out in aggregates of needles or prisms. The salt is quite stable when dry, and may even be heated for several hours at 185° without undergoing further change than loss of water of crystallisation. It has a sweetish, not unpleasant taste, and is easily soluble in water, but it appears to be decomposed slightly by a large excess of water. It crystallises from dilute alcohol in colourless plates. The corresponding ammonium salt, CH,O,ASONH. Aq.. which is prepared in a similar way, crystallises in small glistening needles which are easily soluble in water. The crystals effloresce slowly, and appear to undergo partial decomposition on standing for some time. The potassium salt is not so easily prepared, owing to its instability in aqueous solution. It is obtained by adding arsenious oxide to a boiling concentrated solution of potassium hydrogen tartrate so long as it dissolves, filtering, cooling the filtrate, and then adding two volumes of alcohol to it. The white, finely crystalline precipitate which is formed is washed with alcohol and dried on a porous plate. Analysis of this compound gave results agreeing fairly well with the formula CHO ASOK. Aq. When the salt is treated with water, even in the cold, it decomposes into arsenious oxide and potassium hydrogen tartrate, but it may be recrystallised from dilute alcohol, from which it separates in long needles. Slight decomposition occurs in this case also. When a dilute solution of barium chloride is mixed with a dilute solution of the sodium salt, delicate glistening needles of the barium salt are gradually formed. It has the formula (CHO, ASO), Ba. Aq., and is only slightly soluble in hot water. It is decomposed to some extent by boiling with much water. The corresponding strontium and calcium salts are obtained in a similar way, but the mixed solutions are boiled for a short time. They crystallise in small shining cubes, apparently isomorphous, and are more soluble than the barium salt. Other compounds of this series have not yet been prepared. All of the compounds described above are decomposed by excess of mineral acids, with liberation of arsenious oxide. If, however, the salt is kept in excess, a substance is obtained in solution which has the properties of an acid, but which has not been isolated owing to its instability. When the barium salt, suspended in water, is decomposed by sulphuric acid (taking care to keep the salt in excess), barium sulphate is precipitated and a clear solution is obtained, which remains unaltered even after standing for several weeks. The solution has a strong acid reaction, but contains no free sulphuric acid, and with hydrogen sulphide it gives a copious precipitate of arsenious acid. If heated, or allowed to evaporate spontaneously over sulphuric acid, or mixed with alcohol, it is decomposed into arsenious oxide, which precipitates, and tartaric acid, which remains in solution. It undergoes the same change at once if a drop or two of a mineral acid is added. It decomposes the carbonates of the alkalies and of the alkaline earths, carbon dioxide being evolved, and the salts described above being formed. A quantitative examination of the solution showed that it contained exactly the quantity of arsenic required on the assumption that a substance of the formula ASC,H,O, was present. It may be concluded, therefore, that there is a definite compound of this composition in the solution, and that it is stable at ordinary temperatures if the solution be not too concentrated. Taking its properties into account, this substance CHO i.e., as a derivative of the may be regarded as a tartrarsenious acid As Кон ОН hypothetical orthoarsenious acid As(OH),, and the compounds described above may be considered to be the salts of tartrarsenious acid, or the tartrarsenites. Arsenic acid and various acid oxides likewise appear to form definite compounds when treated with the acid alkaline salts not only of tartaric acid but also of other organic oxyacids, but the investigation of this subject has not proceeded far enough to justify the publication of results. 6. On the Constitution of the Acid Amides. Formation of the Acid Amides.-These compounds are usually obtained by three methods, to which the following equations are assigned: 1. R or H. NH2+ R or H. CO2H = R or H. NH. CO. R or H+ H2O = R or H. NH.CO.R + HCI 3. R or H. NH2+ R or H. CO2R = R or H. NH.CO.R or H+R.OH In 2 and 3 the mode of formation would point to the following constitution for the acid amides, which until recently has been generally accepted: R or H. NH R or H. C:0 The first equation might be construed so as to yield a body of the constitution R or H.N I R or H.C.OH Derivatives of this class have actually been obtained by Pinner and others, and termed imido-ethers, and are therefore isomeric with derivatives of the substances having the first formula: H.N R.C.OR H.N.R R.C:0 Evidence upon which the choice of formula of the acid amide itself rests is of a very unsatisfactory kind. The action of conc. HCl, conc. NaOH, PC, P2O,, or Br would satisfy either formula. One reaction appears definitely in favour of the carbonyl formula, whereas there are two which point equally distinctly to the hydroxyl formula. Caustic soda and sodium ethylate unite with a few acid amides to form Na derivatives, in which the Na may be replaced by an alkyl group by the action of alkyl iodide, and yields a compound in which the new group is undoubtedly attached to the N atom. H.N.R2 R'. C: O. On the other hand, many acid amides unite with silver oxide to form silver compounds in which one atom of H is replaced by Ag, and these bodies treated with alkyl iodide yield the isomeric imido-ethers. Beckmann's reaction, which consists in a molecular change produced by PCI, and other reagents on ketoximes forming isomeric acid amides, points at least to the intermediate formation of a compound of the hydroxyl formula. Since working on the aromatic amides it has frequently struck me as curious that of the series formanilide, acetanilide, &c., including benzanilide and oxanilide, the first formanilide-should possess chemical and physical properties so totally different from the others of the series. Formanilide crystallises in long prisms from alcohol, whereas the others, including benzanilide, form glistening plates all so similar in appearance that it would be impossible to identify them by their exterior alone. Formanilide, of which the formula is usually written CHI,NH. COH, yields with NaOH and Ag2O, Na and Ag compounds, which none other of the series do. This evidently points to a fundamental difference in constitution. Recent Views on the Constitution of the Acid Amides.-Tafel and Enoch showed that the Ag compound of benzamide differs from the Na compound by the fact 1894. SS that by the action of alkyl iodides the former yielded the imido-ether and the latter the isomeric alkyl amide thus: 1. C.H,.C. OAg+IC2H1 = CH ̧. C. OC,H,+AgI and Comstock shortly afterwards showed that similar reactions occurred in the case of formanilide. Formanilide was therefore regarded as an example of tautomerism; a view which is now generally held in regard to the constitution of the acid amides. Leaving for the present the consideration of the Na compounds of the amides, I will take the larger class of acid amides in which an atom of hydrogen is replaceable by an atom of Ag. The following form Ag compounds, and the solids crystallise in needles or prisms: The following do not form Ag compounds, and crystallise in plates : In order to account for the different properties of the two groups, we must fall back upon Hantzsch's researches on the stereo-isomerism of the nitrogen compounds. Hantzsch's Theory.-Hantzsch has shown that bodies of the following general formulæ exist in two modifications: The two modifications vary in stability according to the relative attraction of the groups. The following is the order of attraction of various radicals for the OH group, according to Hantzsch: CC,H.CH,; CO,H; CH.; CHX; CH.S; CnH+1; CH, The first attract and the last repel. The group CH, also repels NHR'. certain of the metallic compounds of the first three groups, and by the action of a metallic oxide the one configuration is converted into the second more stable one. Application of the Theory to the Acid Amides.-Suppose we reverse the configurations 1 and 2, we get a formula which is capable of forming two configurations: Svn Anti HO.C.H HO.C.R R'N R2.C.OH These formulæ will fully account for the constitution of the acid amides. Before proceeding to apply the theory, I will assume that the acid amides, which form Ag compound and crystallise in prisms, belong to the 'anti,' and those which do not, belong to the 'syn' configuration, and crystallise in plates. We can now predict which of the acid amides belong to the former class and which to the latter group. 1. All the acid amides derived from ammonia or fatty amines must be 'anti' compounds of the general formula H or CnH2n+1N H or CnH2n+1 or C,H,. C. OH 2. Bodies derived from aromatic amines must have the 'syn' configuration: C,H,N HO.C.H or CnH2n+1 or CH, &c. Formanilide forms the only exception. This substance should belong to Group 2, but evidently belongs to Group 1. The other modification, which should be easily obtainable, will crystallise in plates, and will not form an Ag compound without reversal into the anti' form. In anthranil, oxindol, and hydrocarbostyril, we have inner acid amides corresponding to the first three anilides. In these compounds it is obvious that the hydroxyl group must occupy the anti-configuration, and should unite with silver oxide and crystallise in prisms. This is the case, although in anthranil the Ag compound rapidly undergoes reduction : The same principle may be applied to the closed chains in the fatty compounds, which should give stable Ag compounds: It also accounts for cases of isomerism, such as the two succinamides, which will probably have the configurations Formation of the Sodium Compounds. In the course of numerous attempts to prepare pure Na acetanilide, one experiment gave an unexpected result, and it appears to me not only to bear out the above view of the constitution of these bodies, but to throw some light on the formation of the sodium compounds. If acetanilide is dissolved in dry ether, and one molecule of solid sodium methylate is added, the liquid becomes turbid, and a compound of the formula CHAN. C(OH)CH,.CH,ONa separates out in needles. The two molecules are very loosely attached, and the acetanilide may be easily dissolved out from the other substance. The formation of this compound points to the following: 1. Acetanilide is an unsaturated compound. 2. The formation of the sodium compounds in the case of anti compounds is preceded by the formation of addition compounds, with subsequent loss of one inolecule of water, thus: CH ̧N + NaOH=CH,N. Na = C ̧H ̧N. Na+ H2O The experimental part of the work has been carried out in conjunction with Mr. Archdeacon, and will shortly be communicated to the Chemical Society. 7. Report of the Committee on the Bibliography of Spectroscopy. See Reports, p. 161. WEDNESDAY, AUGUST 15. The following Reports and Papers were read:- 1. Report of the Committee on the Action of Light on Dyed Colours. See Reports, p. 238. 2. Report of the Committee on Isomeric Naphthalene Derivatives. See Reports, p. 268. 3. A Discussion took place on Dr. J. B. COHEN's Paper on the Constitution of Acid Amides.-See p. 625. 4. On Certain Phenomena occurring during the Evaporation of Salt Solutions. By Dr. W. MEYERHOFFER, Vienna. When a solution is partially evaporated it is not necessarily the case that the solid dissolved salts are also precipitated. From the closer examination of what |