DEBUS ON THE CONSTITUTION ETC. W.-On the Constitution of some Carbon-compounds. By HENRYDEBUS,Pb.D F.R.S. Part I. THEsilver-salt of dibromacetic acid decomposes according to Messrs. Perkin and Duppa,* under the influence of heat and forms argentic bromide and bromogl ycollic acid. The silver-corn- pound of the latter produces under similar conditions argentic bromide and an acid the composition of which is supposed to be represented by the formula C,H,O,. These chemical changes may be represented by the following equations :-C2HBr2Ag02 + H20 = C2H3Br0 + AgBr Bromoglycollic acid. and C2H2BrAg03 + H20 = C2H,04t + AgBr. The conversion of dibromosuccinic acid into tartaric or racemic acid takes place by similar reactions. It appears that Messra.Perkin and Duppa adopted the formula C,H,04 for their new acid without experimental verifica- tion because in many instances organic bodies exchange one atom of chlorine or bromine for the elements HO. I am not aware indeed that the acid or any of its salts have been analysed. It is however well-known that organic bodiesmay also lose their chlorine * Chem. Soc. Qu. J. xii. 5. -f The acid CPH404 may be regarded as standing to the hypothetical glycerin CjHaO& in the same relation as gljceric acid does to common glycerin. VOL. XIX. C DEBUS ON THE CONSTITUTION or bromine according to other modes than the one indicated by the above equations. Ethylenic chloride boiled with an alcoholic solution of potash loses the elements HC1; dibromopyrotartaric acid yields with caustic soda aconic acid sodic bromide and water ; and a simiIar mode of decomposition is exhibited by the bromine- c3mpound of milk-sugar by dichlorhydrin and trichlorhydrin.The silver-salt of bromoglycollic acid might undergo a similar change and furnish the acid C2:,IE203and argentic bromide. The formula C,H203 belongs to glyoxylic acid formed by the action of nitric acid on coamon alcohol. I proposed to myself the following questions :-(1) Which of the two formulz C,H,O and C,H,O belongs to the acid obtained by Messrs. Perkiii and Duppa from dibrom-acetic acid. (2) In case the composition of this acid is represented by C,H,O, is it identical or isomeric with glyoxylic acid ? These questions appeared to me to possess considerable theo- retical interest in consequence of considerations which *ill be described in this paper.I have to thank my friend Mr. Dup pa for the dibromacetic acid employed in the following experiments :-A quantity of dibromacetic acid which boiled at about 225" C. was diluted with water neutralised with calcic carbonate at ordinary temperatures and the lime-salt was precipitated with argentic nitrate. The white crystalline precipitate after previous wabhing with cold water was boiled for some minutes with water. It decomposed easily into argentic bromide and an acid which remained dissolved in the liquid. The argentic bromide was separated by means of a filter and the filtrate neutralised with argentic oxide. The silver-compound thus obtained decomposed at the temperature of boiling water producirig as in the former case argentic bromide and a soluble acid.The clear acid liquid dissolved marble with effervescence and gave after neutralisation and suitable concentration a crop of small prismatic crystals. The latter were placed in hot water in such quantities that a saturated solution was formed and on allowing this aolution to cool the crystals were again obtained in a pure state. These crystals were carefully compared with glyoxylate of lime (prepared from alcohol) and found to agree perfectly with the latter in form solubility and reactions. Their aqueous solution gave a white precipitate with lime-water which precipitate dis- OP SOME CARBON-COMPOUNDS. solved in acetic acid immediately after its formation but lost its solubility in that acid after it had been kept for some time Or after exposure to a temperature of 100"C.The precipitate pro-duced by lime-water in calcic glyoxylate decomposes at 100"into oxalate and calcic glycollate. Zinc acetate and plumbic acetate gave white crystalline precipitates but plumbic nitrate and argentic nitrate failed to produce any change in the aqueous solution. Every reaction was made twice once with the salt from dibromacetic acid and the second time with calcic glyoxylate from alcohol and in each instance both salts behaved exactly in the same manner. The experiments made to determine the solubility of the sub-stance in water gave the following results :-(a) 7.371 grms.soiutim of the salt made from dibromacetic acid gave after evaporation at 100" C 0.053grm. solid residue. (b) 4.97 grms. solution of the lime-salt made from alcohol gave after evaporation at 100" C. 0.0355 grm. residue. The solutions had been prepared at the same time and under precisely the same conditions at a temperature of 18" C. 100 parts of water dissolve therefore of substance (u) 0.719 grm. of substance (b) 0.714 grm. According to these results I believe I am justified in assuming that the acid prepared from dibromacetic acid is identical with glyoxylic acid prepared by the oxidation of common alcohol. The formula of the latter is C211203,and therefore the decompo- sition of the silver-salt of bromo-glycollic acid may be represented thus :-C2H,AgBr03 = C,H,O + AgBr.Qlyoxylicacid. If we consider the compounds which are derived from ethylic bydride by the gradual addition of oxygen to the latter our attention is in the first instance arrested by the fact that each mode of reaction takes place twice. If ethylic hydride is represented by the formula and if I% denotes one atom of hydrogen and one of oxygen both c2 DEBUS ON THE CONSTITUTION of the same chemical value which attaches to them in the ordinary water-type where they are written separately by the side of the radical which stands for a second atom of hydrogen and so com-pletes the molecule then the oxygen-compounds in question may be represented by the following series :-Alcohol Aldehyde = C2{ggH Acetic acid = C2{ z:H Glycollic acid = C2{ Glyoxylic acid I C {gg Oxalic acid = C2{gg The changes of the group C are therefore the following : G (I).Addition of one atom of oxygen; (2). Loss of two atoms of hydrogen; (3). Combination with one atom of oxygen the pro- ducts of the reactions being alcohol aldehyde and acetic acid. The repetition of these processes produces glycollic acid glpoxylic acid and oxalic acid. Instead of allowing these changes to take place in succession we may consider them to take place simiiltaneously; we may add at once two atoms of oxygen to ethylic hydride; and instead of subtracting twice each time two atoms of hydrogen we may at once remove four atoms of this element and finally add two atoms of oxygen to the residue.In this manner we arrive at the following compounds :-HHH Ethylic hydride = C2{ HHH Glycol Glyoxal Oxalic acid = C2{0gc OF SOME CARBON-COMPOUNDS. and herewith the list of the oxygen derivatives of ethylk hydride is complete. It is worthy of notice that the successive steps from ethylic hydride to oxalic acid as represented in the foregoing tables show that the six atoms of hydrogen in the ethplic hydride arrange themselves in two groups each group containing three atoms of the element and undergoing the same transformations. The introduction of oxygen and the removal of hydrogen fo1lows a certain rule which must be dependent on the constitution of ethylic hydride and the nature of oxygen The most simple view of the constitution of ethylic hydride in accordance with the above facts follows from the following considerations.Modern chemistry assumes that chemical reactions take place between molecules ; the determination of the molecular weight of bodies is therefore one of the most important problems of the science. The molecular weight of marsh gas is represented by the formula CH,. Instead as is commonly done of contemplating compound bodies as originating by union of their atoms we adopt the opposite method and consider them to be derived from a series of molecules. A molecule may be looked upon as a system of atoms governed by certain forces which system must be in a state of equilibrium if no external influences produce a disturbing effect.If such a molecule is approached by a second molecule one of the following effects may take place. Either the two molecules simply unite and form one new molecule or as a consequence of action and reaction two or morenew molecules result Examples of the first kind are the formation of chloride of ammonium from hydro-chloric acid and ammonia the double salt of bichloride of platinum aud chloride of potassium the combinations of water waith many salts and other similar cases. The second mode of action may be exemplified by the formation of acetamide and hydrochloric acid from chloride of acetyl and ammonia. If we imagine one atom of hydrogen to be removed from the molecule of marsh-gas it appears as a natural consequence that the equilibrium between the atoms of the molecule must be destroyed and that a resultant of given magnitude and direction must originate.The same process may be conceived with a second molecule of marsh-gas and a similar resultant must be obtained. If now the tworesidues of marsh-gas are placed together so that these two equal re- sultants counteract each other a stable molecule must be obtaiued. This is commonly expressed by saying that u1 atom DEBUS ON THE CONSTITUTION of methyl has replaced an atom of hydrogen in marsh-gas and that this atom of methyl plays the part of an atom of hydrogen. Ethylic hydride would according to the above considerations be composed of the two residues CH and CH of marsh-gas; it would be identical with methyl. Schorlemmer has prepared ethglic chloride from methyl and thus the above view appears to be confirmed.CH,.CH + C1 = CH,.CH,CI + HC1. Ethylic chloride. The compound molecule ethylic hydride (or methyl) which con- sists of the two residues CH and CH can evidently not remain in equilibrium if from one or the other residue an atom of hydrogen is withdrawn just as some forces which are around a given point in equilibrium cannot remain SO if we snppose one of the forces to be removed. If however of these forces two are equal and act in opposite directions both may be removed from the point without; disturbing the equilibrium of the remaining forces. In a similar manner a stable molecule may be derived from ethylic hydride if each of the residues of which it is composed loses one atom of hydrogen.Thus we obtain CH,.CH, or ethylene. The two chemical units which have thus been removed from ethylic hydride in order to obtain ethylene may be added again and therefore we say that ethylene is diatomic. It is not necessary that these two units shoulcl be hydrogen they may be chlorine or bromine forming chloride or bromide of ethylene. CH,.CH + C1 = CH,Cl.CH,Cl. Chloride of ethylene. CH2.CH2 + Br = CH,Br.CH,Br. Bromide of ethylene. Each of the two marsh-gas residues present in ethylene may lose an atom of hydrogen and thus the stable molecule of acetylene results which because four units may be added to it is tetratomic. OF SOXE CARBON-COMPOUNDS. CH,.CH CH.CH Ethylene.Acetylene. CH.CH + Cl = CHCl,.CHCI Acetylene. Chloride of acetylene. The view on which the foregoing considerations are based leads to the conclusion that CH and CH in ethylic hydride CH and CH in ethylene and CH and CH in acetylene are the proximate constituents of the molecules of these bodies. It is apparent that when oxygen is introduced into the molecule of ethylic hydride the hydrogen ought to comport itself as if it were arranged in two groups each consisting of three atoms and each change produced in the molecule ought to take place twice. Three couples ot' bodies are thus produced the members of each couple possessing similar chemical properties. Alcohol CH,.CH,H ; CH,B.COB Glycollic acid Aldehyde CH,.COH ; COH.COH Glyoxylic acid Acetic acid CH,.CO&; C0H.COH Oxalic acid Perkin and Kekul6 have called attention to the alcoholic pro-perties of glycollic acid.I have shown that glyoxylic acid possesses the characteristic properties of an aldehyde and the relation of oxalic to acetic acid is self-evident from the above table. The properties of alcohol aldehyde and acetic acid are dependent on the change which the group Ca3 has undergone; it shows alco- holic properties when it has become CH,$€,-the properties of aldehyde in the state of' COH,-and of acid when it has been transformed into COH; and becanse the gronp CH occurs twice in ethylic hydride therefore two bodies with alcoliolic properties two aldehydes and two acids may be derived from it. If the changes which the two groups CH, CH may undergo in succession in ethylic hytlride take place simultaneously in both residues the following bodies result :-Ethglic hydride ..............CH,.CH Glycol.. .................... CH,.HCH,A Glyoxal .................... COH.COII Oxalic acid.. ................ COH.COB In the former cases the derivatives of the two groups CH, CH 24? DEBUS ON THE) CONSTITUTION were in each substance not the same; in glycollic acid CH,A.COf€; for example one of the residues COH contributes the acid and the other CH,H the alcoholic properties; whereas in the cases mentioned in the last table the derivatives of the two residues CH, CH are in exactly the same condition and therefore we have so to say a double alcohol a double aldehyde and a double acid viz.glycol glyoxal and oxalic acid. From considerations of a similar nature to those hitherto employed it is easy to foresee the probable existence of tbe body CH,lX.COH a substance which would possess the composition of acetic acid but the properties of alcohol and aldehyde and would stand to glycollic acid in the same relation as common alde- hyde to acetic acid. From the point of view here adopted it is also easy to recog- nise why cyanide of methyl and the nitrile of acetic acid must be identical. CH,CH Ethylic hydride. CH,COH Acetic acid CH,.CN Nitrile of acetic acid-cyanide of methyl. With regard to the bodies which may be looked upon as deriva- tives of ethylic hydride and chlorine it is at once seen that there must be at least two isomeric series of chlorinated bodies.Hy-drogen may be replaced in CH, CH atom after atom first in one group CH, and then in the other. Or the substitutions may proceed in both groups at once. The following table represents the two modes of substitution :-Ethylic chloride CH,.CH,Cl 1 CH,.CH,Cl Ethylic chloride CH,.CHCl 2 CH,Cl.CH,Cl Chloride of ethylene CH,.CCl 3 CH,Cl.CHCl CH,Cl.CCl 4 CHC1,.CHCI2 CHCl,.CCl 5 CHCl,.CCl CCld.CC1 6 CCl,.CCl Theory asserts that if all the hydrogen atoms in CH . CH are of the same chemical value the formulz C,H,Cl C,HCI, and C,CI represent only one compound while the formulze C2H,C1, C2H,C1, and C,H2C1, on the other hand represent each two compounds C,HCI, prepared by the action of chlorine on chloride of ethyl seems indeed to be identical with the body OF SOME CARBON-CaMPOUNDS.C2HCl from chloride of ethylene. The chlorinated chloride of ethylene boils at 154",and has a specific gravity of 1.662;the isomeric body from chloride of ethyl was not obtained quite pure and in this impure state was found to boil at 146' and to have the specific gravity 1.644. And inasmuch as they appear to com- port themselves in the same manner with alcoholic potash-solu- tion I think these bodies may be considered to be identical. Hubner* asserts the existence of three bodies of the formula C2H,C1,. Two are the well-known derivatives of chloride of ethyl and chloride of et,hylene and the third was obtained by him from pentachloride of phosphorus and chloride of acetyl.The exist- ence of this third chloride of the formula C,H,Cl is however very doubtful because Hiibner obtained by accident only one drop of it which appeared to boil at 60° and just served for one chlorine determination. Such evidence does not appear to me to be sufficient for the admission of new bodies into the scientific system. The constitution which has been adopted for ethylic hydride and its derivatives in these pages is also confirmed by the syuthesis of acetic acid from cyanide of methyl or from potassium-methyl and carbonic acid. The same theory explains also why the products of the decomposition of chloride of ethylene by caustic potash are identical with those obtained by the action of caustic potash on chloride of ethglidene.The chloride of ethylene may be repre- sented by CH,Cl.CH,Cl the chloride of ethylidene by CH,.CHCI,. The results of their decomposition by caustic potash are water chloride of potassium and chloride of vinyl C,H,Cl. The latter therefore originates by the subtraction of HCl from the organic chlorides. But whether we take HC1 from CH,Cl.CH,Cl or from CH,.CHC12 the result must in both cases be the same viz. CH,.CHCl. We assume that if an organic body loses chemical units such units cannot be taken out of the same carbon group which forms one of the proximate constituents of the body but are supplied by different carbon groups of which the body happens to be composed. If for example ethylic hydride by the removal of two atoms of hydrogen is to be converted into ethylene or the latter into acety- lene each of the residues CH in CH,.CH must lose 1 at. of hydrogen. Thus CH,. CH would become CH,. CH, or CH. CH * Ann. Ch.Pharm. CXE. 330. DEBUS ON THE CONSTITUTION respectively and not CHCH, or C,C!H,. This view is supported by the fact that hitherto the attempts to prepare methylene CH have been unsuccessful and that bodies of the molecular weights CH and CH, are unknown. It also appears worthy of notice that the glycol of the methylic and the glycerin of the ethylic series are still unknown. The experiments of Butlerow and of Simpson,* undertaken with the desire to obtain these bodies are known to chemists. If the glycol of the methyl series existed its formula would be "E:} O, or CH,BA.As soon as we attempt to replace the iodine in CH,T, by && water is eliminated and dimethylenic oxide C,H40 is pro-duced. It would therefore appear as if the group H could not exist twice in combination with 1at. of carbon. If this is the case the failure of the experiments to obtain the glycerin of the ethyl-series is explained. Ethylic hydride contains only two atoms of carbon and therefore the group H can only be introduced twice into the place of hydrogen. Propylic hydride contains 3 atoms of carbon and accordingly in glycerin which may be considered as a derivative of propylic hydride we have 3G ; and as a further confirmation we find the first tetratomic alcohol to contain four atoms of carbon.The following table contains the formuls of the alcohols derived from some of the hydrocarbon-compounds of the series CnH2ni-2 :-Methylic seGes. Ethylic series. Propylic series. Butylic series CH,H C,H,H C,H,H C,H,H C,H4HA C,H,Afi C4H,HB C,H,Hga C,H f-If-Ia We perceive by this table that in no compound does the number of the groups $I exceed those of the carbon-atoms. I believe this rule holds good in all native compounds. Not long ago Cariust described as propyl-pliycite a body which had been prepared from epichlorhydrin and to which he gives the formula c3H4}04. This body would form an excep-*4 * Jahreabericht 1857 461. Ann. Ch. Pharm.cvii. 110 ;cxi. 242 ; cxiv. 204 j cxv 322. .t. Ann. Ch. Pharm. cxxxiv 71 OF SOME CARBON-COMPOUNDS.tion to what appears in the above and many other examples aa a rule. The propyl-phycite is however described M an amorphous Viscous mass and the same properties belong to its derivatives. Even if it were possible to obtain such bodies in a suficiently pure state for analysis it might still be doubted whether the for- mula C3H4] 0 was the correct representation of the composition H* of the propyl-phycite because it may have been in the state in which it was analysed a hydrate. Pseudopropylic alcohol forms hydrates with a definite boiling point. Carius has however prepared by the action of sodium-alcoho-late on dichlorbromhydrin two bodies which he considers to be the ethers of the propyl-phycite H 0 and c3H4 10,.From (C2H5)3 c3*4] (C2H5)4 these bodies we cannot deduce the formula of the corresponding alcohol. Williamson and Kay* obtained from chloroform and sodium- alcoholate a substance which may be regarded as the ether of cH”’}03, methyl- glycerin (C2H5.X and Mr. Rassettt has prepared from chloropicrin oythocarbonate of ethyl ci’)04. The (C2H5)4 corresponding hydrogen-compounds however cannot be obtained from these ethers. Another circumstance deserves attention. Carius could not replace more than two hydrogen-atoms in propyl-phycite by acid radicals. The treatment of dichlorbromhydrin C3H,.fi.C12Br with sodic acetate and acetic acid failed to produce the ether ‘SH4 }04 but gave instead a number of substances which (C2H30)4 must be regarded as the result of a very complicated reaction.From the above it will be seen that our knowledge of the propyl- phycite is still very incomplete and that its formula has not been established on a satisfactory basis. If we consider it as a hydrate of the formula c3H3} 03,H20 the rule which obtains H3 with regard to other artificial and natural alcohols would also apply to the propyl-phycite. The rule which asserts that the atomicity with regard to the Ann. Ch. Pharm. xcii. 346 .i. Chem SOC.J. (2) ii. 198. DEBUS ON THE CONSTITUTION group aof a carbon-compound cannot be greater than the number of carbon-atoms present holds good also for the acids. The methyl-series contains only monoatomic acids for the body COna is unknown. The ethyl-series contains diatomic but no undoubted triatomic acid ; whereas in the propyl-series glyceric acid is certainly triatomic &c.&c. The view adopted in the present paper assumes that all the hydrocarbons of the general formula CnH2n+2consist of as many residues of marsh-gas molecules as there are carbon-atoms in the compound or in other words contain these residues as their proximate constituents. The alcohols aldehydes and acids of these carbon and hydrogen compounds are produced by the substitution of oxygen or 8,or of oxygen and $€ for hydrogen. If we compare the formuh of formic acid acetic acid and oxalic acid Formic acid ............ H.COfi Acetic acid .............. CH,.COfi Oxalic acid .............. CO$€.COB it appears that the basicity of an acid is equal to the number of times the group COB occurs in it.I think this rule applies to all organic acids derived from the hydrocarbons C,H,+,. Thus propionic acid CH,.CH,COH is monobasic ; malonic acid COQ. CH,COB bibasic ;mesoxalic acid COI'I.CO.COfi bibasic ;tartaric acid COH.CHH.CHH.COQ bihasic and tetratomic. The group COH is derived from CH, and therefore a tribasic acid must be derived from a hydrocarbon-compound in which the group CH occurs three times. Propylic hydride is CH,.CH,.CH, and there- fore no tribasic acid can occur in the propylic series. Citric acid is a tribasic acid and its corresponding hydrocarbon is represented by the formula CsHl,. This hydrocarbon could be derived from six molecules of marsh-gas.The first five molecules would produce CH,.CH,.CH,.CH,.CH,. In order to obtain 3CH3 in C6H14,the sixth molecule CH must attach itself not to one of the two CH, but to one of the residues represented by CH,. Thus because the residue and the marsh-gas molecule must each lose one atom of hydrogen we obtain CH,.CH,.CH,.CH CH3 (CH,--and by substitution OF SOME OARBON-COMPOUNDS. 29 COBCH,.CH,.CA { ggg citric acid. Black oxide of manganese oxidises the three groups COB and the residue CH,.CH,.CH takes up one atom of H and forms acetone. Tribasic acids which may be considered as derived from one of the hydrocarbons CnH,,, must contain the residue CH or C amongst their proximate constituents. The rational formula of propylic hydride is CH,.CH,.CH,.or CH,.CH c-v-J If in each of the residues 1at. of hydrogen is replaced CH CH,Cy .CH,Cy by cyanogen we obtain-; and replacing the ni-CHCy trogen of the cyanogen by On we have COB,CH,CH,COft CH,CO$ = carballylic acid .* The corresponding hydrocarbon of carballylic acid would be CH,.CH,.CH.CH,.CH,.CH, or accord-ing to one of the more common expressions The atomicity of an acid or an alcohol depends on the number of times the group ft is contained in it. Tartaric acid contains 6 atoms of oxygen 6 atoms of hydrogen and 4 of carbon which elements according to the chemical properties of tartaric acid must be arranged in the following manner C0fi.CHfi.CHfr. COH; and in an analogous manner succinic acid = CO&.CH,.CH,.COH and malic acid = COH.CH$f.CH,.COfi. Succinic acid may be prepared from butyric acid the latter from butylic alcohol and butylic alcohol from butylic hydride. Thus we have the following series :-CH,.CH,.CH,.CH Butylic hydride CH,.CH,,CH,.CH,a Butylic alcohol CH,.CH,.CH,,COH Butyric acid CH,&CH,,CH,.COB CO&CH,,CH,.COH Succinic acid COH.CH.H.CH,.CO~~ Malic acid CORCHXI.CH.~~~.COQ Tartaric acid * Simpson. Proceedingsof the Roy. SOC.xii. 236. MATTHIESSEN ON THE EXPANSION OF WATER. The manner in which the proximate constituents of butylic hydride become changed in consequence of the gradual introduc- tion of oxygen into its molecule is clearly perceptible in the above table. The question very naturally arises can the second atom of hydrogen in the residues ,..CHQ.CHn .. .be replaced by H and the bodies COH.Cfifi.Cfifi.COf€ = C,H60 and COft.C8€& CHQ.COfi = C,H,O be obtained? According to our present experience it appears that this cannot be done. Tartaric acid is hardly attacked by chlorine and nitric acid removes one of the groups CHH altogether and produces tartronic acid COH.CH$€.COH= C3H40,. Also tartronic acid is not succeeded by an acid wherein two atoms of oxygen have joined the group CH,. The composition of such an acid would be = C0fi.CHH.COfi = C,H40B. The ammonia-salt of mesoxalic acid is = C,(NH,),O,. The relations of this acid to various members of the uric acid series also tend to establish the formula C3H,0S. Therefore we pass from tartronic acid to mesoxalic acid by the addition of 1 at.0. to the group CHH and the sub-traction of H,O. COIJ[.CHIJ[.COIJ[ Tartronic acid coH.co.co~ Mesoxalic acid The latter may be converted into tartronic acid by means of sodium-amalgam. The properties of the group CH, as exhibited in the above examples occur again in many other cases.