FOR STOPPING DECAYED TEETH. XXX1I.-On the Chemical Constitution and Nature of Organic Radicals. BY H. KOLBE.,PH,D. F.C.S. Do compound radicals exist in organic combinations exercising therein the functions of simple elements and combining like the latter with oxygen sulphur chlorine &c. in certain atomic pro- portions to form oxides (acids) sulphides and chlorides ? Do these compound radicals consist of unaZterabZe groups of atoms or can the substitution of their hydrogen by chlorine bromine hypo- nitric acid &c. be effected unaccompanied by the distiirbance of chemical equilibrium ? YOL III,-NO. XII. BB 370 DR. KOLBE ON THE CHEMICAL Is it at variance with the principles of the electro-chemical theory that elements of such different chemical properties as hydrogen and chlorine should displace each other ; and that oxygen chlorine and similar bodies of decidedly electro-negative character should rank among the constituents of organic radicals ? The above questions first set on foot by the discovery of chloracetic acid have been answered in various ways and most decidedly and minutely by Berzelius in the correspondence that lasted for some years between himself and Dumas These interesting discussions were unfortunately interrupted by the death of the first-named chemist with- out any understanding or approximation of views being effected so that the opinions of chemists are to this day widely at variance with regard to the above points.Berzelius and with him the greater number of German chemists have as is well known adhered to their opinion of the existence of compound radicals the correctness of which they have considered as proved by the successful isolation and the chemical properties of cyanogen and cacodyl.In addition to this Berzelius obstinately defended the principle against all attacks made by adherents to the theory of substitution that organic radicals were unalterable groups of atoms; that substitution could not be effected in them without a thorough alteration of their chemical constitution least of all the substitution of hydrogen by bodies of such prominently electro-negative properties as chlorine bromine oxygen &c. ; and that the assumption of organic radicals containing oxygen or chlorine was in contradiction to the principles of the electro-chernicat theory.On the other hand Dumas without distinctly stating whether indeed he assumes the existence of organic radicals has maintained the principle that compounds containing an equal number of equiva-lents arranged in a like manner are possessed of the same funda- mental properties and that the function exercised by an element in organic combination is not dependent upon its original properties but rather upon the position which it occupies in the compound. Both chemists agree in the opinion that the assumption of a replace- ment of the positive hydrogen by the negative chlorine in a compound radical is at variance with the electro-chemical theory because according to this theory the nature of a compound cannot be consi- dered as independent of the chemical nature of its constituents.This accordance of opinion with respect to the latter point has been most probably the principal reason why the question as to the immutability of organic radicals has not been made the subject of investigation to the degree that it merits. CONSTITUTION AND NATURE OF ORGANIC RADICALS. 371 It may be considered as beyond all doubt that the theory of radicals in its present state no longer suffices to furnish proper explanations of the innumerable metamorphoses resulting from so-called substi- tution and that by a continued adherence to the unchangeability of compound radicals the secure foundation resulting from the theory of radicals becomes continually weakened.We are indeed at the present moment almost forced by facts to assume that organic radicals are alterable groups of atoms in which chlorine bromine hyponitric acid &c. may enter in the place of hydrogen the molecular grouping of their atoms remaining unchanged and secondary radicals being thus produced which are possessed in part of properties similar to those of the primary ones. Instead of at once discussing the question whether and how far this mode of viewing may be made to accord with the electro-chemical theory I will endeavour in the following pages to apply this hypothesis to a series of organic combinations arid more particularly to those the comportment and metamorphoses of which it has been most difficult to explain by the theory of radicals as it now stands The difficulty of bringing the behaviour of acetic acid with chlorine and its reproduction from chloracetic acid to accord with the assump- tion of an unalterable acetyl-radical of the formula C H, first led Bereelius to construct the hypothesis that acetic acid and chlora- cetic acid might be conjugated oxalic acids the one containing methyl and the other sesquichloride of carbon as adjuncts and that the conversion of acetic into chloracetic acid was solely due to a me-tamorphosis of the adjunct while the oxalic acid itself remained unaltered.Whatever may be the opinion entertained as to the value of this hypo- thesis it cannot be denied that the pre-existence of methyl in acetic acid has received a high degree of probability from facts most of which were only subsequently discovered Independent of the analogy existing between acetic and chloracetic acid on the one hand and between methylo-hyposulphuric acid and sesqui-chlorocarboh yposulphuric acid on the other to which I have already referred in a previous paper,* the conversion of acetate of ammonia into cyanide of methyl and the reproduction of acetic acid from the latter as also the decomposition of acetic acid by the galvanic current into methyl and carbonic acid can scarcely be satisfactorily interpreted otherwise than by the assumption of the existence of methyl as a component part gf acetic acid.* Ann Ch. Pharm. d. LIV 148 ff; Chem. SOC. Mem. 11 360. BB2 372 DR.KOLBE ON THE CHEMICAL It has not as yet been attempted to extend the above hypothesis to the remaining combinations of acetyl which if acetic acid were methyloxalic acid must evidently have a composition corresponding to this view. The result of such an attempt would be the necessity of assuming the existence in aldehyde acetylous acid chloride of acetyl &c. of the partly hypothetical bodies C,O C 0 and C C1 coupled with methyl. We should likewise be forced to assume that in .the terchloride of acetyl (the so-called dichlorinated chloride of ethyl) which according to Regnault may be converted by continuous boiling with alcoholic potassa into acetic acid and chloride of potas- sium the sesquichloride of carbon considered as combined therein with methyl must differ considerably in its properties from the same substance when uncombined as the latter cannot be converted by similar treatment into the corresponding oxygen-compound oxalic acid ;moreover a further extension of the hypothesis to the remaining compounds allied to acetic acid-for example to the other mem-bers of the series (C H,) o, as also to benzoic acid and its homo- logues &c.would show that instead of viewing them as oxides of compound radicals as hitherto we must consider them all as conju- gate oxalic acids and finally that of all the radicals which are partly really isolated there would be few or none remaining. Even caco- dylic acid if the replacement of a portion of its hydrogen by chlorine be accomplished as there is little doubt it will must no longer be considered as an oxygen-compound of cacodyl but perhaps as arse- nious acid conjugated with two equivalents of methyl.If however we adhere less strictly to the immutability of organic radicals another hypothesis very nearly allied to the former of the chemical constitution of acetic acid and the acetyl-compounds in general presents itself possessing the advantages of the former without sharing its defects namely that an acetyl-radical really exists in the acetyl-compounds ;that it must not however 6e consi- dered as a group of four equivalents of carbon and three equivalents of hydrogen the four carbon equivalents of which possess equal func- tions but that it should rather be viewed as a compound of two epui-valeuts of carbon and methyl as the adjunct Acetyl = (C H,)T, in which the C presents the exclusive point of action for the powers of ujlinity of oxygen chlorine &c.According to this view the composition of the acetyl-compounds with which we are as yet acquainted would be expressed by the fol- lowing rational formulz CONSTITUTION AND NATURE OF ORGANIC RADICALS. 373 Subacetylous acid . . . HO . (C H,)-C, 0 (aldehyde). Acetylous acid . . . . ('2 H3)nc2, Acetylic acid HO (C H3)^C,, '2 0, . . . . . (acetic acid). (&her hydro- { Terchloride of acetyl. . . (C H3) ^C2 Cl chlorique-bichlorurd) . Dinoxamide of acetyl . . (C H3)^C2>{ $Ho (acetamide). Oxydichloride of acetyl Oxydisulphide of acetyl .Oxysdphochloride of acetyl (C H3)-C2 (&her chloro- I c1 sulfuri). The assumption of this conjugate acetyl-radical in aldehyde and acetic acid renders it evidently necessary that the interpretation hitherto given to the process of acetification should be slightly modi- fied. The action of the oxygen on the alcohol effects probably at first the splitting of ethyl into methyl and the carbo-hydrogen C H2 the elimination of which from organic compounds belongs to the most common phenomena. At the moment of its liberation it under-goes a further decomposition; as the hydrogen becomes oxidised the two equivalents of carbon immediately reunite with the methyl forming the conjugate acetyl-radical (C2H3)-Cg with which the oxygen of the oxide of ethyl remains combined (C H5)0 .H0+2 O=HO (C HJ-C, 0+2 HO L.-v-.-J L-v-Alcohol. Aldehyde. The conversion of aldehyde into acetic acid by the direct assimila- tion of two equivalents of oxygen HO . (C H,)^C, 0+2 O=HO (C H,)^C, 0 L-TA LPTL Aldehyde. Acetic acid. must be considered as of old as a pure process of oxidation. It appears to me that all other relations of acetic acid to different compounds may be explained in an equally satisfactory manner by assuming it to be composed according to the above rational formula. Its formation from cyanide of methyl by boiling with potassa (C H3) C N + KO +3 HO=KO . (C H3)-C, 0 + N H3 L-v-J Cyanide of methyl. Acetate o€ potassa. DR. KOLBE ON THE CHEMICAL the decomposition of acetate of ammonia by distillation with phos- phoric acid into cyanide of methyl and water NH 0 .(C H3)^C2,03 + n PO = (C H,) C2N + 4 HO +n PO L.-v-J --A Acetate of ammonia. Cyanide of methyl. the decomposition of an aqueous solution of acetate of potassa at the oxygen-pole of the galvanic current into methyl and carbonic acid KO (C H3)^C2,0 + O= C2 H3+ KO CO,+ CO L-v-J Ly-’ Acetate of potassa. Methyl. the conversion of acetic acid by heating its lime-salt with hydrate of lime into carbonic acid and marsh-gas (hydride of methyl) CaO . (C2 H,)^C, O,+HO . CaO=H (C HJ+2 (CaO . CO,) -v-3 + Acetate of lime. Marsh-gas. and finally the formation of oxide of cacodyl by the distillation of acetate of potassa with arsenious acid 2 [KO .(C2H3)^C,,0,] +As 03=(C,H3)2As,0+2(K0,C02) + 2C0 L-v--3 Acetate of potassa. Oxide of cacodyl. All these decompositions are most simply and naturally expressed by the above equations. I consider it at present impossible to explain the formation of acetone from acetic acid as we are as yet quite in the dark with regard to its rational composition and indeed have not even established its empirical formula with certainty. The opinion defended with so much warmth by the adherents to the theory of substitution that the so-called dichlorinated oxide of ethyl still possessed the constitution of oxide of ethyl does now scarcely need any refutation. Not only do the close relations borne by this compound to acetic acid into which it is converted by lengthened contact with water almost preclude a doubt of its belonging to the acetyl-series but it is of itself more than probable that the oxide of ethyl would undergo a similar metamorphosis by the action of chlorine as with oxygen.The action of chlorine on oxide of ethyl likewise appears to consist in the first instance of a splitting of the ethyl-group into methyl and C H, which is followed by the immediate decomposition of the latter carbo-hydrogen and the formation of (C H3)-C, together with hydrochloric acid ; this radical remaining in combination with the oxygen of the oxide of ethyl assi- -. CONSTITUTION AND NATURE OF ORGANIC RADICALS. 375 milates in addition two equivalents of free chlorine and thus becomes (C2 H3)-C2,{ g12, i.e. oxydichloride of acetyl according to the following equation C4H50 + 4C1= (C HJC2,{ g12+ 2 HCl U Lv-2 Oxide of ethyl. Oxydichloride of acetyl. The same mode of viewing may be applied with facility ta the decomposition of this acetyl-compound by water or alkalis into acetic and hydrochloric acids -2 Oxydichloride of acetyl. L-and its conversion into oxybisulphide of acetyl (&her bisulfuk) and oxysulphochloride of acetyl (&her chlorosulfur6) by the action of hy drosulphuric acid Oxydichloride of acetyl. Oxybisulphideof acetyl. and L-L-v+ Oxydichloride of acetyl. Oxysulphochlonde of acetyl. To these acetyl-combinations must still be added the dinoxamide of acetyl (C2H,)"C2 { $H the formation of which from acetate of oxide of ethyl and ammonia is explained by the following equation Acetate of oxide of ethyl.Acetamide. Alcohol. It can scarcely admit of adoubt that the product of decomposition of chlorideof ethyl corresponding to the dichlorinated oxide of ethyl namely the so-called dichlorinated chloride of ethyl possesses a con-stitution siinilar to that of the former and is therefore the true DR. KOLBE ON THE CHEMICAL terchloride of acetyl,* (C H,)^C, Cl, although it exchanges its chlorine for oxygen with much greater difiiculty and requires long continued boiling with alcoholic potassa for its complete coiiversion into acetic acid. By the action of chlorine on ether there is formed according to Malaguti in addition to the dichlorinated oxide of ethyl a second body the so-called semi-chlorinated ether which M alaguti did not succeed in separating and purifying respecting the existence of which however his experiments leave no doubt.This substance corresponds with the chloride of aldehyde discovered by Regnault in so far that both may from their empirical composition be considered as ethyl- compounds the former as the oxide of ethyl the latter as the chloride * The substance originally named chloride of acetyl is the gaseous body (C H Cl) formed by the treatment of the oil of olefiant gas with alcoholic solution of potassa. The product obtained by the action of pentachloride of antimony on this gas isomeric with the dichlorinated chloride of ethyl has been named by Berzelius superchloride of acetyl; to the dichlorinated chloride of ethyl itself he has given the name super-chloride of paracetyl.It might appear from this that the superchloride of acetyl of Berzelius obtained from the oil of oletiant gas was the true terchloride of acetyl corresponding to acetic acid while the so-called superchloride of paracetyl obtained from chloride of ethyl was only an isomeric compound not really belonging to the acetyl-series If it be considered however that the latter substance permits the replacement of its chlorine by oxygen thus passing over into acetic acid while the former cannot be directly converted into acetic acid or any other member of the acetyl- series there can be no doubt whatever that the dichlorinated oxide of ethyl (super- chloride of paracetyl) is the true acetyl-compound namely the terchloride of acetyl (C H,)^C, c1,.The gas obtained by the action of alcoholic potassa upon olefiant gas the so-called chloride of acetyl appears to me to belong as little to the acetyl-series as the so-called superchloride of acetyl. The isomerism of its supposed radical with acetyl may however be easily explained if we consider it as hitherto to be a body (C H3) homologous to aflyl (C H5) for which I propose the name of vinyl as suggestive of its origin (comp. Iiadwbrterbuch der Chemie Bd. v. p. 548 ff). The olefiant gas may be considered as the hydrogen-compound of this radical or as hydride of vinyl consisting of + volume of vinyl-vapour and + volume of hydrogen which corresponds perfectly with the observed specific gravity assuming a condensation of 2 vols.carbon-vapour and 3 vols. hydrogen to 1 volume in vinyl 2 vols. carbon-vapour . . . 1.658. 3 , hydrogen . . . . . 0.207. 1vol. vinyl gas . . . . . 1.865. , vinyl gas . . . . . 0933. $. , hydrogen . . . . . 0.034. 1vol. hydride of vinyl . . . 0.967. Assuming with Liebig the oil of olefiant gas to be a combination of hydrochloric acid with the chloride of the radical vinyl and admitting the possibility of a substitution of CONSTTTUTION AND NATURE OF ORUANIC RADICALS. 377 in each of which one equivalent of hydrogen is replaced by one of chlorine (C {:()O semichlorinated ether (C4{:f)C1 chloride of aldehyde. As long however as this view is unsupported by facts I consider it much more probable that these compounds possess a composition corresponding to the hydrochlorates of oxide of vinyl and of chloride of vinyl (comp.note p. 376) being thereforecombinations of hydro-chloric acid with oxide of acetyl (in the semichlorinated ether) and chloride of acetyl (in the chloride of aldehyde) (C H,)-C, 0 .H C1 semichlorinated ether (C H3)-C2 C1. I1 C1 chloride of aldehyde. the hydrogen in vinyl itself by chlorine the composition of the substances derived from olefiant gas may be best expressed by the following rational formulse Vinyl = C H,. Hydride of vinyl . . (C H,)H olefiant gas. ~~~~'~~~~fc~lO:} (C H,) C1 H C1 chloride of elayl (Berzelins). Chloride of vinyl . . (C H3) C1 chloride of acetyl (Berzelius).Hydrochlorate of chlo- ride of ch~orovinyl *} (c {F{>C~ Chloride of chlorovi-}(c4{ 3)Cl H ~1 superchloride of acetyl (Berzelius). nyl ..... chloride of formyl (Berzelius). Hydrochlorate of chlo- H GI ride of &chlorovinyl.} (c {Z&)C~ superchloride offormy1 (Berzelius). ''* Hydrochlorate of chlo-ride of trichlorovinyl. }('4 "3) c12 ide of vinyl Hydrochlorate ,Of .Ox-} (C H3) 0,H C1 oxychloride of elayl (Berzelius). ~:fT:)(C,~~ H H3) Br H~Br bromide ~of elayl (Berzelius). ~ ~ Hydriodate of iodide 1(C H3) I H I of vinyl . . . .J iodide of elayl (Berzelius). Iodide of vinyl . . (C H,)I iodide of acetyl (Berzelius). The conversion of the substance mentioned under the name of hydrochlorate of chlo-ride of trichlorovinyl and discovered by Pierre (Ann.Ch. Phys. [S] XXI. 439) into the so-called protochloride of carbon (C CI, or C Cl,) by boiling with alcoholic potassa is probably accompanied by an alteration in the arrangement of the elements (C C1,) C1 to (C C13)-C, C1 (chloride of trichloracetyl) the latter formula expressing the rational composition of protochloride of carbon as I shall presently endeavour to show. The sulphur-compounds of vinyl require e more careful examination than they have hitherto beeii subjected to. The sulphides of elayl described by Lowig and Weidmann were doubtless impure substances containing an admixture of sulphur. It might be interesting to examine the behaviour of chloride of vinyl with sulphocyanide of potassium as a sulphocyanide of vinyl (C H3) Cy S, coifesponding to oil of mustad (C H5) Cy S, might possibly be produced under these circumstances.DR. KOLBE ON THE CHEMICAL By this assumption the formation of semi-chlorinated ether from oxide of ethyl C H 0 $2 C1=(C2 H,)^C, 0.H C1+ H Cl < 2 L7-7--Oxide of ethyl. Semichlorinated ether. its conversion into oxybichloride of acetyl (bichlorinated oxide of ethyl) by treatment with chlorine Y -J L-yL Semichlorinated ether. Oxybichloride of acetyl. and its decomposition by water into aldehyde and hydrochIoric acid. (C HJ-C, 0 H Cl+ HO=HO . (C,H,)^C, 0 + H C1 L-v-L-v-3 Semichlorinated ether. Aldehyde. may be explained in a simple and perfectly satisfactory manner It may be regarded as self-evident that a hypothesis constructed upon the constitution of acetic acid or any other member of the series of fatty acids must likewise admit of application to the remaining members of that series.If therefore we consider acetic acid as the oxide of the conjugate radical (C H3)-C, it follows naturally that we must assume the existence of conjugate radicals analogous to acetyl in propionic valeric margaric formic acids &c. Indeed in an attentive examination of these compounds I have not met with any fact that was in contradiction with the above hypothesis; and I believe that I do not err in expressing the conviction that the compartment of the fatty acids from formic acid up to melissic acid and their manifold relations to other compounds may be better interpreted by the above hypothesis than by any one of the former views.The decomposition of valeric acid by the galvanic current into butyl (C H,) and carbonic acid and its formation from cyanide of butyl (C H,) C N (valeronitril) are particularly well adapted to support this opinion. The relation of stearic to margaric acid and its bibasic nature may be easily explained without our being obliged to adopt the assertion made by Laurent and Gerhardt that both these acids possess the same composition and saturating capacity and are only two isomeric modifications if we view stearic acid as a double acid composed of two different oxides of margaryl (C 32 €13JnC2 namely H 0 (C32 H,)-C, 0 + H 0 (C32 H3,)’’C2 O,-~S CONSTITUTION AND NATURE OF ORGANTC RADICALS.379 I shall presently endeavour to establish the view that in s similar manner sulphobenzoic and sulphacetic acids are double acids and owe their bibasic property to this circumstance. According to the above hypothesis the composition of the fatty acids would be expressed by the following rational formula Formic acid Acetic acid. Propionic acid. Butyric acid. Valerie acid. Caproic acid. Enanthylic acid. Caprylic acid. Pelargonic acid. Caprinic acid Ricinostearic acid (margaritic acid) Laurostearic acid (pichurimo-stearic acid). Cocinic acid. Myristic acid. Benic acid. Bthalic acid. Margaric acid. Stearic acid. Bassic acid. Behenic acid. Cerotic acid. Melissic acid. A glance at the above table will show that the radicals of the fatty acids contain as adjuncts the radicals of the alcohoIs of which however we are as yet acquainted with onIy a few namely methyl ethyl amyl cetyl and cerotyl.The important discovery made by Dumas that the ammonia-salts of acetic propiouic valeric and other acids may in accordance with the deportment of formiate of ammonia be converted by distillation with phosphoric acid for instance into the cyanogen-compounds of the adjuncts of these acids -for instance cyanide of methyl cyanide of ethyl cyanide of propyl and cyanide of butyl-furnishes us with the prospect of obtaining by this method the whole series of alcohols that arc DR. KOLBE ON THE CHEMiCAL 380 still wanting provided we are successful in converting the cyanides into other combinations of the alcohol-radicals The experiments hitherto made on this subject have however given no results.Among the radicals of the fatty acids formyl H-C stands to a certain extent isolated it being the only one in which the adjunct contains no carbon. As however its compounds bear the greatest resemblance to those of acetyl this would appear to indicate that the nature of the adjuncts of conjugate radicals exercises comparatively little influence over the chemical character of their compounds. The composition of those formyl-compounds of which the consti- tution is established with tolerable certainty may be expressed by the following rational formulze Formylic acid . .. . HO .HnC3 0 Formic acid. Terchloride of forrnyl . . HT, Cl Chloroform. Terbromide of formyl . H-C, Br Bromoform. Teriodide of formyl . H^C2 I Iodoform. Tersulphide of formyl . Sulphoform. Dichloriodide of formyl Chloriodoform. Dibromoiodide of formyl Bromiodoform. Diodocyanide of formyl Formyl-aci-bichlo-ride (Be r z elius). Oxydichloride of formyl (Ether methylique bichlorure'). The analogy of hydrated oxide of methyl to hydrated oxide of ethyl warrants the supposition that the former in its oxidation and conversion into formic acid undergoes a decomposition similar to that of the latter in its transformation into acetic acid; namely that the elimination of C H from the composition of methyl precedes the production of formic acid from methyl-alcohol (C2 H3) 0.HO+4 O=HO -' . HnC, O,+2 HO v Hydrated oxide of methyl. Formylic acid In the same manner the rational formula HnC2 { f or (32 &chlorinated oxide of methyl is based upon the supposition that when two equivalents of chlorine enter oxide of methyl in the place of two of hydrogen it undergoes a metamorphosis correspond- CONSTITUTION AND NATURE OF ORGANIC RADICALS. 381 ing to the conversion of oxide of ethyl into oxydichloride of acetyl namely (C H3) 0+4 Cl=H^C,,{ +2 H C1 + -7-f Oxide of methyl. Oxydichlonde of formyl. Unfortunately the chemical deportment of dichlorinated oxide of mythyl has been too little studied to admit of the deduction of any argument therefrom for or against the above view ; it can only be supposed from its analogy with dichlorinated oxide of ethyl (oxy- dichloride of acetyl) that alcoholic potassa would convert it into formic acid and hydrochloric acid HT, { -i-3 KO =KO.H-C, 0,+ Z KCl + Oxydichloride of formyl. Formiate of potassa. The view that the so-called perchlorinated ether the penta-chlorinated oxide of ethyl no longer possessed the constitution of oxide of ethyl was already defended in opposition to the older theory of substitution by Malaguti in his excellent researches on the chlorinated ethers. Moreover if we consider it established that chlorine like oxygen is capable of converting oxide of ethyl into an acetyl-compound the pentachlorinated ether cannot really be a direct product of decomposition and much less a substitution-product of oxide of ethyl but must rather be considered as a derivative of oxydichloride of acetyl.Indeed its most intimate relation to the acetyl-compounds may be inferred from its chemical comportment and that of its products of decomposition chloracetic acid the so-called chloraldehyde &c. Even the fact that perchlorinated ether chloracetic acid chloraldehyde chloroxethose &c. may be almost directly converted into acetic acid and the latter again into chlora- cetic acid renders it more than probable that all these bodies have a similar chemical constitution. It becomes most clearly perceptible in the study of these com-pounds into what a labyrinth of hypotheses and to what very improbable assumptions we are forced if we forsake on the one hand the theory of radicals and adhere on the other to the immu- tability of organic radicals.It is certainly quite as difficult to agree with the hypotheses of Laurent and Dumas as to adopt the views taken by Berzeliu s* of the constitution of perchlorinated * Lehrbuch 5th Edit. Vol. V. 788-825. '- 382 DR. KOLBE ON THE CHEMICAL ether and its derivatives and of the substitution-products of the ethers in general. I will endeavour to interpret these phenomena of substitution from the point of view above adopted and to base the following considera- tions upon the hypothesis that a substitution of chlorine for hydrogen takes place in the acetyl-radical without dissolving the complex atom three new secondary radicals being produced thereby which contain in the place of one two or three equivalents of hydrogen a propor-tionate number of equivalents of chlorine and which are gifted with properties similar to those of acetyl itself.For the better indication of their relations to the latter I will assign to them the names- chloracetyl dichloracetyl trichloracetyl and express their rational composition by the following formulze Acetyl . (C H,)^C Chloracetyl . . (C { ;f)-'2 Dichloracetyl (C2{ E1a)-'2 Trichloracetyl (C CI,)-C2 Of the various combinations of these three secondary acetyl-radicals those of trichloracetyl have been by far the most accurately studied; I will therefore first submit them to examination.EIydrated oxide of trichloracetyl HO.(C C13)-C2 0 Chloral. Trichloracetylic acid . . . HO.(C C13)-C2 0 Chloracetic acid. Oxide of trichloracetyl . . (C C13)1C2 0 Chloroxethose. Chloride of trichloracetyl . . (C Cl,) C, C1 Chlorethose(ch1or.ofcarbon). Oxydichloride of trichloracetyl . (C CI3)-C2 Chloride of chloroxethose { g, Oxydibromide of trichloracetyl . (C ClJ-C, Bromide of chloroxethose. { ir2 Dinoxychloride of trichloracetyl . (C C13)-C2 Chloraldehyde. { Ef Dinoxamide of trichloracetyl . (C Cl,)-C, Chloracetamide. { 2H Dinoxyphosphamide of trichloracetyl (C C13)"C, { O2 Chloracetyphide. PH Cblorodibromideof trichloracetyl (C Cl,)-C, { g Bromide of chlorethose. We may easily perceive that those of the trichloracetyl-compounds above-mentioned which are also represented in the acetyl-series differ in various points in their properties and chemical comportment from the corresponding members of the latter series.The acetyl has evidently undergone no unimportant modifications by the exchange of its hydrogen for chlorine. In these instances as in the substi- tution-products of aniline the chlorine to however great an extent it CONSTITUTION AND NATURE OF ORGANIC RADICALS. 383 appears to have lost its specific properties imparts some portion of its chemical character to the combination in which it has entered the place of the hydrogen. This influence is rendered very distinctly evident in chloraniline and trichloraniline by the weakening of the basic properties of the aniline; in the compounds of trichloracetyl it is not so obvious although it is always rendered to a certain extent perceptible by a modification of the original affinities of acetyl.It should therefore not appear surprising that chloracetic acid when boiled for instance with potassa deviates somewhat in its com-portment from its analogue acetic acid or that oxydichloride of trichloracetyl (perchlorinated ether) the analogue of oxydichloride of acetyl (dichlorinated oxide of ethyl) should be decomposed with much greater difliculty by similar treatment with potassa than the latter. It is worthy of note that the conversion of the chlorides (oxychlorides) of acetyl and trichloracetyl into the corresponding acids is effected with greater ease in proportion to the number of atoms of oxygen that the compound contains.While the terchloride of acetyl which coiitains no oxygen is but very slowly converted into acetic acid by boiling with alcoholic potassa the oxydichloride of acetyl is converted into that acid with the greatest ease by the mere action of water; and doubtless the dinoxychloride of acetyl which is as yet unknown but the production of which may be expected with tolerable certainty to result from the treatment of concentrated acetic acid with pentachloride of phosphorus will be a liquid much more easily decomposed by water. On the other hand dinoxychloride of tri-chloracetyl (chloraldeh yde) is equally distinguished from the oxydi- chloride of trichloracetyl by the facility with which it is converted into chloracetic acid.-M alagut i* has left it undecided whether the small quantity of formic acid which he detected in the fluid after continuous boiling of perchlorinated ether with alcoholic potassa owed its production to the action of the potassa on the alcohol or to the conversion of perchlorinated ether into cloracetic acid (C C13)-C3 { +3KO =KO (C C13)-C2 0,+2 KCl.v Perchlorinated ether. Chloracetate of potassa. The simultaneous deposition of chloride of potassium gives at any rate a greater probability to the latter view. The accompanying equations will show that by the assumption of -2 * Ann. Ch. Phys. [S] XVI 19. 384 DB. KOLBE ON THE CHEMICAL a trichloracctyl the most simple and natural explanations are furnished 1.Of the decomposition of chloracetic acid into terchloride of forrnyl and carbonic acid by boiling with potassa HO (C ClJ-C, 0 + 2 KO = HT, C1 + 2 (KO CO,) L-v--J 7 Chloracetic acid. Chloride of formyl. 2. Of the formation of chloracetamide by the action of ammonia on chloracetate of oxide of ethyl -v-d L-v-.-J Lv-d Chloracetic ether. Chloracetamide. Alcohol. 3. Of the conversion of chloracetate of ammonia by phosphoric acid into the substance termed by its discoverers Dumas Malaguti and Leblanc chloracetonitrile and which is probably a cyanogen- compound (C Cl,) . C N NH 0 (C CI3)-C2 O,+ nPO,= (C CI,)^C N -t4 HO nPO L-v-& L-v--' Chloracetate of ammonia. Chloracetonitrile. 4. Of the reproduction of' chloracetic acid and ammonia from chloracetonitrile by boiling with potassa (C C13)"C2 N + KO + 3 KO = KO .(C CI,)"C, 0 + NH,. Lv-d L-.-vp2 Chloracetonitrile. Chloracetate of potassa. 5. Of the transformation of perchlorinated ether at a temperature of 300°C. into sesquichloride of carbon and dinoxychloride of tri-chloracetyl (chloraldehyde) L-WU -yy-' L7-Perchlorinated ether. Sesquichloride Chloraldehyde. of carbon. 6. Of the metamorphosis which the latter body undergoes when heated with concentrated sulphuric acid 0 0 (C C13)"C2 { c1+ HO . SO,= (C,C 1,)^C2 { ci -t HC1+ SO -L-Y---J Perchlorinated ether. Chloraldehyde. 7. Of the decomposition of chloraldehyde by water into chloracetic acid and hydrochloric acid CONSTITUTION AND NATURE OF ORGANIC RADICALS.385 (c c13)_c, { Ef + 2 HO = HO. (C Cl)-C, 0 +-HCl. L-v-J L---v-a Chloraldehyde. Chloracetic acid. 8. By alcohol into chloracetic ether and hydrochloric acid (C Cl,) C, { 4+ (C HJO. HO=C H 0. (C CI,)-C, 0,+ HCI. LFV-d \-Chloraldehyde. Alcohol. L,-v--) Chloracetic ether. 9. By ammonia into chloracetamide (C C13)nC2,{ Ef + 2NH = (C Cl,) C, { $H + NH,Cl 2 c-v-2 L-v-J ChIoraldehy de. Chloracetamide. 10. By terphosphide of hydrogen into chloracetyphide (C C13)-C2 { gf + PH3= (C2 C13)-C2 { pH+ HCl. L-v-3 L-v-”J Chloraldehy de. Chloracetyphide. and finally the formation of chloraldehyde by a simple change in the arrangement of the elemeots of the so-called perchloracetic ether effected by passing the vapour of the latter through a dull red- hot tube L v------d L-v-J Perchloracetic ether.Cloraldehyde. Chloraldehyde exhibits in its behaviour a remarkable similarity to the so-called chloride of benzoyl (CI4 H 0 Cl). The reason of this analogy between two such heterogeneous compounds may be easily arrived at if the benzoyl-radical (free from oxygen} be viewed as a conjugate radical like acetyl and trichloracetyl having the rational formula (C12 HJC,. All three radicals possess in common the term C, forming the real point of attack for the powers of affinity of the negative elements while the adjuncts alone are different. Here then the subordinate part played by the adjuncts compared with that of the body with which they are conjugated and their comparatively slight influence over the chemical nature of the compounds of conju- gate radicals is again most strikingly exhibited.The above similarity is however not merely confined to chlorobenzoyl and chloraldehyde it exists equally between benzoic acid and chloracetic acid benzamide and chloracetamide benionitrile and chloracetonitrile and will doubt-VOL. IJI.-NO* XII. cc DR KOLBE ON THE CHEMICAL less extend to the sulphide of benzoyl and cyanide of benzoyl and the terms corresponding in the trichloracetyl series with which we are yet acquainted. The analogy of these compounds is best rendered evident by the following comparison of their rational formuh '-G2 H5)^C ('2 c13)-c2 Benzoyl. Trichloracetyl.-U v + Benzoic acid. Chloracetic acid. -Chlorobenzoyl. Chloraldehyd. -v Benzamide. C hloracetamide. PI2 H5)'c2 N (C Cl3)* c2 N - Chloracetronitrile. Benzonitrile. The difference between the behaviour of oxydichloride of trichlor-acetyl (perchlorinated ether) with sulphide of potassium and that of the corresponding member of the acetyl series (the oxydichloride of acetyl) is worthy of remark. Whilst the latter when treated with hydrosulphuric acid exchanges 1 or 2 equivalents of chlorine for sulphur the former undergoes a simple reduction by treatment with an alcoholic solution of sulphide of potassium and is con-verted into the oily compound called by Malaguti chloroxethose (C2 CI3)-C2 0 (oxide of trichloracetyl) sulphur and chloride of potassium being separated.0 (C2 C13)-C2 { c1 + 2 KS= (C C13)-C2 0+ 2 KC1+S v2-Perchlorinated ether. Chloroxethose. The mode of formation of chloroxethose together with its property when brought into contact with chlorine (or bromine) of combining directly with two equivalents of these elements and thus being reconverted into perchlorinated ether (or the corresponding bromine- 0 compound (C C13)-C2 { Br LONSTITUTIQM AND NATURE OF ORGANIC RADICALS. 387 -v Chloroxe t ho se. Perchlorinated ether. as also its comportment with chlorine in the presence of water whereby it is partially converted into chloracetic acid (C C13)"C2 0+2 Cli-3 HO=HO. (C ClJ-C, 0,+2 H C1 T L-v-3 Chloroxethose. Chloracetic acid. are in perfect accordance with the view expressed by the above formula namely that chloroxethose is the oxide of trichloracetyl.The above considerations lead to the conjecture by no means unfounded that the chloride of carbon C C1 (Malaguti's chloride of chlorethose) which in its comportment exhibits so remarkable an analogy with oxide of trichloracetyl is the chlorine-compound of the same radical corresponding to this oxide namely chloride of trichloracetyl (C CI3)-C2 Cl. We may perhaps even go farther and regard the sesquichloride of carbon obtained from this chloride of trichloracetyl by direct assimilation of chlorine as the higher chlorine- eompound of the same radical namely as terchloride of trichloracetyl (C C13)-C2 Cl, or at least assume the existence of two isomeric coin-pounds C C1 (sesquichloride of carbon trichloromethyl) and (C ClJnC2 Cl (terchloride of trichloracetyl).The principal objection against the assumption that the chloride of carbon produced by the combination of chloride of trichloracetyl with chlorine is the terchloride of the latter radical lies in the circum- stance that we have not yet succeeded in replacing by oxygen the three equivalents of chlorine considered as combined with trichloracetyl. Alcoholic solution of potassa is certainly not quite without action upon it as after continuous boiling chloride of potassium is depo- sited the liquid becoming brown ;neither chloracetic acid terchloride of formyl nor formic acid can however be found among the products of decomposition.It is possible that in this experiment the terchlo- ride of trichloracetyl gives off two equivalents of chlorine to the potassium without any replacement ensuing the oxygen of the latter attacking the alcohol and oxidizing it to aldehyde; this would also account for the brown colour assumed by the alcoholic potassa- solution. It must however be borne in mind that terchloride of acetyl too is only attacked with difficulty by boiling alcoholic potassa md that the action of the latter is likewise obstinately withstood by the oxydichloride of trichloracetyl. On the other hand the conjecture that the so-called chloride and sesquichloride of carbon are chlorine- cc 2 388 DR. HOLBE ON THE CHEMICAL compounds of trichloracetyl is supported by a vast number of facts.The formation of the chloride of trichloracetyl from the so-called chloride of carbon by passing thevapours of the latter through a red- hot tube 4 c c1 = (C C1Jnc2 c1+4 c1 c-y-J Perchloride of Chloride of trichloro-carbon. acetyl. may be coupled with many similar examples in which compounds poor in carbon yield by the action of a high temperature less volatile substances richer in carbon as for instance the formation of naphthalin benzol &c. The solid chloride obtained together with the chloride of trichloracetyl doubtless owes its production to the further action of the large quantity of liberated chlorine upon the liquid chloride and must therefore be considered as a secondary product of decomposition of the dichloride of carbon.The remarkable similarity exhibited between the so-called sesqui- chloride of carbon and the oxydichloride of trichloracetyl (perchlo- rinated ether) in their outward appearance as also in their behaviour with hydrosulphate of potassium a point to which attention was first called by M alagut i finds a surprisingly simple explanation in the assumption of the rational formulE (C C13)'-C2 Cl, for the first named body. A simple comparison of the rational formulae of these two substances 0 (C cl,) -C2 { c12oxydichloride of trichloracetyl (C Cl,) 712 Cli terchloride of trichloracetyl (sesquichloride of carb.) and of the corresponding members of the trichloracetyl-series that are poorer in chlorine namely of oxide of trichloracetyl and chloride of trichloracetyl (C C13)-C2 0 oxide of trichloracetyl (chloroxethose M alaguti) (C Cl,) C, Cl chloride of trichloracetyl (chlorethose M alaguti) is only needed to account for the concordant behaviour of the two former compounds with hydrosulphate of sulphide of potassium and of the two latter with chlorine and bromine and particularly to explain the remarkable relations of the two latter to chloracetic acid.With regard to the tran;sformation of chloride of trichloracetyl into chloracetic acid by exposure to the action of chlorine in the presence of water under the influence of solar light it is partly due to the decomposition of water into hydrochloric acid and oxygen and partly CONSTITUTION AND NATURE OF ORGANIC RADICALS.389 to the power possessed by chloride of trichloracetyl of combining with oxygen directly as it does with chlorine at least when the former is in the nascent state. Dinoxychloride of trichloracetyl (chloraldehyde) is next formed and this being readily decomposed by water is con- verted into chloracetylic acid. (C C13)”C2 C1+ 2 HO + 2 C1= (C C13)T2,{ Ef -f-2 H C1 -y-d Lp-d Chloride of tnchloracetyl. Dinoxychloride of trichloracetyl. (c,c13)-c2,{ 8+ 2 NO= HO . (c ci,)-c, 0 + H ci L-7-V Dinoxychloride of Trichloracetyl. trichloracety1. The conversion of the greater part of the chloride of trichloracetyl into terchloride of trichloracetyl in the above process is doubtless owing to the slowness with which the oxygen is liberated from water by the chlorine.If a similar condensation of the elements be assumed in trichlora- cetyl as in acetyl one volume of trichloracetyl containing therefore two volumes of carbon-vapour and three volumes of chlorine ; 2 vol. carbon-vapour . . . 1.658 3 , chlorine . . . . 7.846 1 vol. trichloracetyl . . . 9.004 and if the chloride of trichloracetyl and terchloride of trichloracetyl be regarded as containing according to the usual modes of conden-sation half a volume of trichloracetyl combined respectively with half a volume or one and a half volume of chlorine the calculation of the specific gravities of these bodies in the gaseous form will furnish numbers corresponding very closely with the observed vapour densities of the so-called protochloride of carbon = 5.82 and sesqui- chloride of carbon = 8.157 4 vol.trichloracetyl . . . . . . 4.502 8 , chlorine . . . . . . . . . 1.224 1 vol. chloride of trichloracetyl . . . 5.726 4 vol. trichloracetyl . . . . . . 4502 -$ , chlorine . . . * . . . . . 3.672 1 vol. terchloride of trichloracetyl . 8.174 DH. KOLBE ON THE CHEMICAL It is worth mentioning here that the decomposition of oxydichloride of trichloracetyl into two equivalents of sesquichloride of carbon and one equivalent of dinoxychloride of trichloracetyl at a temperature of 300° C. may be even more simply expressed by the following equa- tion than by the mode of viewing adopted at page 384 Oxydichloride of trichloracetyl. Terchlonde of trichloracetyl.trichloracetyl. According to the view expressed in this equation the metamor- phosis of oxydichloride of trichloracetyl into terchloride and dinoxy- chloride would be due to a simple transposition of the negative elements combined with the radical. The remaining modes of formation of the chloride of trichlor- acetyl for instance by the action of chlorine on chloride of ethyl and hydrochlorate of cloride of vinyl (oil of olefiant gas) appear to me to accord likewise with the above assumption; in the first case the terchloride of trichloracetyl is not formed as a direct product of decomposition of chloride of ethyl but as a substitution-product of the terchloride of acetyl into which the chloride of ethyl is first con- verted; in the second case the formation of the terchloride is evi- dently preceded by that of the chloride of trichloracetyl which may again be viewed as a product of decomposition of the hydrochlorate of chloride of trichlorovinyl (C C1,)Cl.HCI lately discovered by Pierre from which it is probably produced in consequence of the separation of the apparently but loosely combined hydrochloric acid by the transposition of the elements from (C,Cl,) Cl into (C,C13)-C2 C1. (Compare the note page 376.) In speaking some time back of the chemical constitution of alde- hyde and chloral,* I advanced the assumption that they might be conjugate combinations of formylous acid respectively with methyl and sesquichloride of carbon C H,. C HO aldehyde. C Cl,. C HO chloral.According to this view by which the close relations of these two compounds is very well expressed and a satisfactory interpretation given of the decomposition of chloral by alkalies into terchloride of * Ann. Chem. Phaim. LIV 184. CONSTITUTION AND NATURE OF ORGANIC RADICALS 391 formyl and formic acid the oxidation in the conversion of aldehyde into acetic and of chloral into chloracetic acid would extend to the common constituent formylous acid which would be transformed into conjugate oxalic acid with formation of water. If however the readiness with which the elimination of hydrogen-equivalents pro- ceeds at least in aldehyde be taken into consideration it must appear strange that the action of chlorine should confine itself solely to the hydrogen of the adjunct and that the formylous acid so readily acted upon by oxygen should remain unaltered.This difficulty is imme- diately overruled by assuming with Liebig the pre-existence of a basic atom of water in chloral as well as in aldehyde and by viewing the former as hydrated oxide of acetyl and the latter as hydrated oxide of trichloracetyl HO. (C H,)^C, 0 aldehyde. HO. (C C13)-C2 0 chloral. The formation of chloracetic acid in the oxidation of chloral by means of fuming nitric acid would then be as in the conversion of aldehyde into acetic acid the result of the direct assumption of two equivalents of oxygen HO .(C C13)-C2 O+Z O=HO. (C C13)"C2 0 c-v-3 L-v-3 Chloral. Chloracetic acid. This hypothesis likewise furnishes a proper explanation of the decomposition of chloral by solution of potassa into formic acid and terchloride of formyl as is shown by the following equation HO.( C CI,)^C,+ KO. HO =KO. H-C, O,-i-HT, Cl Chloral.T Formiate of Terchloride of potassa. formyl. With regard to the interesting metamorphosis which hydrate of chloral undergoes by treatment with sulphuric acid other expressions may be easily given of the composition of chloralide accounting equally well for its formation and chemical comportment as the formula 2C HCl .3 C 0 constructed by Stadeler,* which it is difficult to bring into accordance with the above view of the constitution of chloral It might with equal justice be considered as a combination of two equivalents of oxide of trichloracetyl with one equivalent of hydrate of formic acid =2 [(C C1 .C, O)] +HO .H-C, 0,,or as a double * Ann Ch.Yharm. LXI 104. DR. KOLBE ON THE CHEMICAL compound of chloral and formiate of the oxide of trichloracetyl =NO. (C C13)'T2 0 + (C HJ-C, 0. H-C, 0,. It is left for future researches to decide which of these formulz can claim the advantage or which view of the rational composition of chloralide is the correct one. It has already been stated that we are as yet only acquainted with a few compounds of the intermediate secondary acetyl-radicals pre-ceding trichloracetyl namely chloracetyl (C {%)^C, and dichlo- racetyl (C { CH1,)-Cd. One of the most interesting is the acid of chloracetyl corresponding to that of trichloracetyl namely chlora- cetylic acid HC).(C2{ :i)^C, 0,,which Leblanc* prepared by the action of dry chlorine-gas on concentrated acetic acid in diffused daylight . cf -C2 O,+HCl NO. (C H,)-C, 03+2Cl=HO. (c( ("> L-v--3 L--.-\.---.-J Acetic acid. Chloracetylic acid. We must also include among these bodies the two substitution- products of terchloride of acetyl described by Regnaul t,f and resulting together with chloride of acetyl from the decomposition of the so-called chloride of aldehyde (first substitution-product of chloride of ethyl) namely the terchloride of chloracetyl(C ,{c"12> ^C 3 Cl, (trichlorinated chloride of ethyl) and terchloride of dichloracetyl (Cz{tl)-C2 C1 (tetrachlorinated chloride of ethyl.) The substi- tution-products corresponding to oxydichloride of acetyl namely the oxydichlorides of chloracetyl and dichloracetyl have not yet been obtained.The vapour-densities of terchloride of acetyl (= 4*530),of ter- chloride of chloracetyl (= 5*799) of terchloride of dichloracetyl (= 6*983) and terchloride of trichloracetyl (= 8*157),as deter-mined by Regnault are in perfect accordance with the above assumption if we consider 8 vol. of each of the radicals of these bodies combined with Q vol. of chlorine to form 1 volume accord- ing to the usual mode of condensation. Acetyl (C H3)-C,. Tercbloride of acetyl (C H3)^C, C&. 2 vols. carbon vapour. . 1.658 Q vol. acetyl gas . . . . 0.932 3 , hydrogen . . 0.207 Q , chlorine . . . . 3.604 1 vol. acetyl . . . . 1.865 1 vol.terchloride of acetyl. 4.536 * Ann. Chim. Phgs. [3] LXXI,353. (found . 4.530) + Ibid X. 212. CONSTXTUTION AND NATURE OF ORGANLC RADICALS 393 Chloracetyl (C { Ef}-Ce. Terchloride of chloracetyl (C L c1)-C,0,C13 2 vols. carbon-vapour. . 1.658 Q vol. chloracetyl-gas . . 2.123 2 , hydrogen . . . 0.138 +j-, chlorine . . . . 3.672 1 vol. chlorine . . . 2,449 1 vol. chloracetyl . . . 4.247 1 voI. terchloride of chlor- acetyl . . . . . . 5.795 (found . 5.799) Dichloracetyl(C { tf)-C2 Chloride of dichloracetyl (C { F1)-C2 C 2 vols. carbon-vapour . 1.658 vol. dichloracetyl-gas . . 3.314 1 , hydrogen . . . 0.069 3 , chlorine . . . 3.672 2 , chlorine . . . 4.898 1 vol. dichloracetyl . . 6-628 1vol. chlor. of dichloracetyl 6986 (found .6.983) Trichloracetyl (C C13)-CT Chloride of trichloracetyl (C Cl,)-C, Clg 2 vols. carbon-vapour 1.658 4 vol. trichloracetyl-gas . 4.502 3 , chlorine . . 7.346 + , chlorine . . . + 3.672 1 vol. trichloracetyl . . 9*004 1 vol. chlor. of trichloracetyl 8.174 (found . 8.157) The metamorphoses which chloride of ethyl undergoes consecu-tively by the action of chlorine may be exhibited by the following forrnulze of the resulting products of decomposition Protochloride of ethyl . . . . (C H5) C1 . . &her hydrochlorique bydrochlor. Hydrochlorate of protochloride of acetyl (C H3)-C, C1 HCI { monochlorurk. { ether hydrochlor. Terchloride of acetyl . . . . (C HJ-C, C13 bichlorur6. Terchloride of chloracetyl . . trichlorurk. Terchloride of dichloracetyl .. . (C { t()-C2 C1 { &her hydrochlor. quadrichlorurk. Terchloride of trichloracetyl . perchlorurk. It still remains for us to discuss the chemical constitution of those bodies which are produced by substitution from the combinations of the other radicals homologous with acetyl. With regard to chlorobutyric and chlorovaleric acid as also to nitropropionic acid the assumption of the conjugate radicals (C H5)-C2 (C H,)-C, and (C H,)-C in propionic butyric and valeric acids and the analogy of these substances with acetic acid scarcely admit any other explanation of the above processes of substitution than that the 394 DR. KOLBE ON THE CHEMICAL hydrogen in the adjuncts of these radicals is substituted by chlorine hyponitric acid &c.new secondary radicals being thus produced which still possess the constitution of the primary radicals. Hence the cornposition of nitropropionic chlorobutyric and chlorovaleric acids is expressed by the following rational formulz. HO .(c4( Eb4)-C2 0 Nitropropionic acid. HO ,(C iH ) cc -C2 0 Dichlorobutyric acid. HO . (C cf4 -C2 0 Quadrichlorobutyric acid. {H ) NO.(C cl' -C2 0 Dichlorovaleric acid. {" ) HO C {li5)-C2 0 Quadrichlorovaleric acid. ( c, It is much more difficult to explain satisfactorily the metamorphoses which the formyl-compounds undergo by the action of chlorine. Although I do not dispute the existence of a secondary formyl- radical of the composition Cl-C, but am on the contrary convinced of its existence in chlorinated formiate of oxide of ethyl and of oxide of methyl yet the assumption of its existence in the substitution-products of the simple formyl-compounds for instance in chlorinated chloride of formyl oxydichloride of formyl &c.appears to me exceedingly bold and improbable. Even the behaviour of hydrated formic acid with chlorine by which it is converted not into cbloro- formic (HO .Cl-C, 03),but into hydrochloric and carbonic acids unmistakeably indicates the action of aenities different from those exerted in compounds the radicals of which contain a carbo-hydrogen as adjunct. Terchloride of formyl undergoes an analogous decomposition being split up as is well known into two equivalents of perchloride of carbon by the action of chlorine H-C, C1 +2Cl= 2 C C1 + H C1 + w Terchloride of formyl.Perchloride of carbon. It would at any rate be difficult to deduce from the chemical comportment of perchloride of carbon any argument in favour of that substance being terchloride of chloroformyl (C1-C % C1,) or chloride of trichloromethyl (C Cl, Cl) or of its possessing any other rational formula.-The chlorinated oxydichloride of formyl (the final product of the action of chlorine on oxide of methyl having the empirical formula C C1 0),has been unfortunately too little studied to afford by its comportment any conclusion as to its constitution. It may perhaps CONSTITUTION AND NATURE OF ORGANIC RADICALS. 395 be a simple combination of perchloride of carbon with chloro-carbonic acid having the rational formula C Cl, Cn gl* This pro- duct appears at any rate no longer to possess the constitution of oxide of methyl nor of oxybichloride of formyl; at least the observation made by Regnault that the condensation of the so-called oxide of perchloro-methyl in the gaseous form is only half as great as that of the compound from which it is directly derived appears to indicate that the conversion of oxydichloride of formyl (dichlorinated oxide of methyl) into oxide of perchloromethyl is accompanied by a change in the relative position of the atoms.It is at present impossible to decide whether the final product of the action of chlorine on sulphide of methyl is analogous in compo- sition to the foregoing substance or whether its composition may be S expressed by the rational formula C C1,.C { cl Malapti,* in his admirable researches on the chlorinated ethers was the first to advance and support with powerful arguments the view that the various compound chlorinated ethers examined by him (in which all the hydrogen was substihted by chlorine) contain no perchlorinated ether in the form in which it is known in the isolated state (oxydichloride of trichloracetyl) but that they more probably have the same constitution as the normal ethers of which they are the derivatives. The composition of perchloracetic ether would then be expressed by the rational formula (C Cl,) 0 (C C1JnC2 0,. If it were assumed on the other hand that the oxide of ethyl in acetic ether in its conversion into perchloracetic ether underwent the same metamorphosis as it does in the free state namely that it was converted thereby into oxydichloride of trichloracetyl the perchloracetic ether would then have to be viewed as trichlo-racetate of oxydichloride of trichloracetyl = 0 { Cl,.Irrespective of the argumenis maintained by Malaguti against the existence of perchlorinated ether as it is known in the free state in perchloracetic ether it appears to me in itself but slightly pro- bable that in a metamorphosis so completely altering the molecular arrangement of atoms as for instance in the conversion of carbonate of oxide of ethyl (C H5) 0 CO, into carbonate of oxydichloride of * Ann Chein Phys. [3] XVI 4. 396 DE. KOLBE ON THE CHEMICAL trichloracetyl (C C13)-C2.{El CO, the two members-carbonic 2' acid and oxydichloride of trichloracetyl should remain combined. The objection may be raised here that it is still enigmatical why the action of chlorine on oxide of ethyl combined with acids produces no change in the molecular arrangement of the atoms while free oxide of ethyl under similar circumstances immediately passes into an acetyl-compound; this may be answered by re-ferring to the many cases in which a compound owes its existence olely to its combination with another body or even to the mere presence of such a substance. Carbamic acid cyanic acid and many others are well known to exist only in combination with bases; they split up on the moment of their liberation into simpler groups of atoms.We know likewise that the addition of a few drops of solution of potassa to the neutral aqueous solution of a large quantity of sulphovinate of potassa is sufficient to prevent perfectly the easy decomposition of that substance by boiling. If moreover the view be adopted without hesitation that in hyposulphobenzolic and napthalic acids the hyposulphuric acid assumed as existing therein acquires a stability perfectly foreign to it when in the free state by the assimilation of the adjunct it is certainly not less justifiable to assume that the elements of oxide of ethyl in compound ethers are held together with a greater force by their combination with acids than is the case with free oxide of ethyl so that the original grouping of atoms remains the same even if chlorine be replaced for all the hydrogen-equivalents of a compound ether.The property possessed by so many binary compounds of gaining increased stability by their combination with a third body may be compared to the action of a powerful magnet on two weaker ones which alone have not power sufficient to support each other but to which this power is imparted on the approach of a more powerful magnet. It must not however be considered strange that we have not as yet succeeded and perhaps never shall be successful in converting the chlorinated compound ethers into the normal ethers by a process similar to that by which chloracetic acid is reconverted into acetic acid and chloraniline into aniline. The attachment of their con-stituents the chlorinated acid and the chlorinated oxide of ethyl appears weakened to such a degree by the entrance of chlorine in the place of hydrogen that a comparatively slight impulse such as is imparted by the action of heat alkalis alcohol &c.is required to effect an essential change in the molecular arrangement of their atoms. All the endeavours to reproduce the original compounds CONSTTTUTION AND NATURE OF ORGANIC RADICALS. 397 from the chlorinated ethers have met with no result and it is probable that the slight stability of the latter compounds will always present an insurmountable obstacle to the success of such experiments. The highly interesting and numerous metamorphoses of chlorinated compound ethers for the knowledge of which we are principally indebted to Malaguti and Cahours furnish undoubtedly the most appropriate means of deciding the question of the rational composition of the latter substances.We will therefore examine the explanation furnished of these phenomena by the above mode of viewing. The metamorphoses of the various chlorinated ethers containing chlorinated oxide of ethyl by potassa ammonia alcohol methyl- alcohol &c. exhibit a remarkable analogy in the circumstance that the resulting products of decomposition are the same as would be furnished by a combination or a mixture of the particular (chlori- nated) acid with dinoxychloride of trichloracetyl (chloraldehyde). It may hence be assumed with tolerable certainty that the contact of these substances with the above agents always effects in the first instance their separatisn into the two groups of atoms composing them namely into chlorinated oxide of ethyl and the chlorinateh acid.It appears however that a chlorinated oxide of ethyl cannot exist as such but exhibits a tendency to pass over into the metameric compound oxydichloride of trichloracetyl (C Cl)-C 2 {O while the c1 acids separated from chlorinated oxide of ethyl (with the exception of carbonic acid) cannot exist in the anhydrous state. It is therefore probable that at the moment of separation an exchange is effected of one equivalent of oxygen for one of chlorine in the chlori- nated oxide of ethyl while it undergoes the above metamorphosis; whereby there is formed on the one hand an oxychloride correspond- ing to the hypothetically anhydrous acid or a direct product of decomposition produced by a transposition of the atoms and on the other hand dinoxychloride of trichloracetyl.This metamorphosis may be best understood by means of the following general equation in which Ac represents the acid-radical Indeed the term (C cl,) C { (312 (the so-called chloraldehyde) is found to be a constant product of decomposition by heat of all com- 398 DR. KOLBE ON THE CHEiMICAL pound ethers containing chlorinated oxide of ethyl accompanied by a compound corresponding in composition to the formula Ac '2 {Cl Or at least by a direct product of decomposition of the latter. The chlori- nated ethers of oxide of methyl undergo a perfectly analogous meta- morphosis with the only difference that in this case chlorocarbonic acid is formed instead of (C ClJ-C, {gf (C Cl,) 0.Ac 0,=2 C{ 0 Cl+Ac{ gf The above rationale holds good likewise with respect to the con-version of perchloracetic ether into two equivalents of the dinoxy- chloride of trichloracetyl (chloraldehyde) having the same composition as the former by the passage of its vapours through a red-hot tube L 7-) =(C2 C13)^C {02 Perchloracetic ether. 2' c1 Chloraldeh yde. the decomposition of chlorinated formiate of oxide of ethyl into chloraldehyde and chloracarbonic acid L-v-' Perchloroformic ether. Lyp---J Chloraldehyde. that of perchloroxalic ether into chloraldehyde chlorocarbonic acid and carbonic oxide (the two latter being doubtless products of decomposi- tion of C {3(still unknown) L-77-2 Perchloroxalic ether.v Chloi aldehyde. as also with the decomposition of the perchlorinated acetate of oxide of methyl isomeric with perchloroforrnic ether into chlorocarbonic acid and chloraldehyde CONSTITUTION AND NATURE OF ORGANIC RADICALS. 399 2 v L v Chlorinated acetate of oxide of methyl. C hloraldehyde. of chlorinated formiate of oxide of methyl into four equivalents of the isomeric compound chlorocarbonic acid L-v-3 u Chlorinated formiate of Chlorocarbonic oxide of methyl. acid. and of the chlorinated oxalate of oxide of methyl into chlorocarbonic acid and carbonic oxide Chlorinated oxalate of oxide of methyl.The metamorphosis of perchlorocarbonic ether by exposure to heat into chloraldehyde sesquichloride of carbon and carbonic acid = (cs c1,) 0 .C02 = (c c13)ncp{ L-v-d Chlorocarbonic ether. h-v-J Cblorafdehyde. exhibits a slight deviation from the above which is evidently based partly upon the property of carbonic acid of existing in the anhy- drous state and partly upon its volatility. The latter is probably the principal cause why the carbonic acid at the moment that it should exchange an equivalent of oxygen for an equivalent of chlorine with the intermediate product oxydichloride of trichloracetyl (C,Cl3)T2,ig12(which is undergoing transposition) evades this metamorphosis SO that the latter substance alone undergoes a further change; and thus we obtain instead of chloraldehyde and chlo- rocarbonic acid carbonic acid and the products of decomposition of oxydichloride of tricbloracetyl namely chloraldehyde and sesqui- chloride of carbon.400 DR. KOLBE ON THE CHEMICAL We find in accordance with the above assumptions that perchlora- cetic ether (C GI,) 0 . (C C1,JnC, 0 = 2 (C2 Cl,)"C, {4) behaves with potassa ammonia and alcohol exactly like dinoxychloride of trichloracetyl (chloraldehyde) ;potassa forming with it trichlorace-tate of potassa and chloride of potassium ;ammonia dinoxamide of trichloracetyl (chloracetamide) and chloride of ammonium ;alcohoI trichloracetate of oxide of ethyl and hydrochloric acid.-In like manner chlorinated formiate of oxide of ethyl and the metameric acetate of oxide of methyl behave with these reagents like a mixture of chlorocarbonic acid and chloraldehyde (compare their metamorphoses at a high temperature p.399). Solution of potassa converts them nto trichloracetate and carbonate of potassa and chloride of potas-sium ; ammonia into chloracetamide and chloride of ammonium (and doubtless into carbamide or its products of decomposition) ; alcohol into trichloracetate and oxychlorocarbonate of oxide of ethyl and hydrochloric acid ; methyl-alcohol into the corresponding methyl- ethers.-Chlorinated carbonate of oxide of ethyl which is converted by potassa into chloride of potassium carbonate and formiate of potassa (the latter being evidently a product of decomposition of the previously formed trichloracetate of potassa) ;by ammonia into chlo- ride of ammonium chloracetamide (and probably carbamide and car- bamate of ammonia) and which furnishes with alcohol trichloracetate and carbonate of oxide of ethyl besides hydrochloric acid resembles in these metamorphoses an instable compound of chloraldehyde and 0 0 chlorocarbonic acid (C Cl,) 0 .GO = (C ClJ-C, { cF + C { ci or a compound of oxydichloride of trichloracetyl with carbonic acid (C Cl,) 0 CO = (C ClJT, {gl .CO, if we may assume that 2 oxydichloride of trichloracetyl at the moment of its formation can pass over with the elements of water into trichloracetylic and hydro- chloric acids analogously to the dinoxychloride of trichloracety1.-The chlorinated oxalate of oxide of ethyl which yields with potassa trichloracetate and oxalate of potassa and chloride of potassium ; with ammonia chloracetamide chloride of amrnoniurn and pro-bably oxamide ; with alcohol trichloracetate and oxalate of oxide of ethyl together with hydrochloric acid and chloride of ethyl comports itself like a compound of chloraldehyde with a hypothetical oxalo-dinoxychloride corresponding to oxa-{2, mide C, which may be presumed to be decomposed by ammonia into chloride of ammonium and oxamide.In the decompo- CONSTITUTION AND NATURE OF ORGANIC! RADICSLS. 401 sitiou of chlorinated oxalate of oxide of ethyl by alcohol Malaguti obscrvcd besides the formation of oxalic ether and chloracetic ether the production of the body C Cl 0 (Cbloroxethide M) correspond-ing in composition to the formula (C Cl,) 0.2 C 0 (?). In the same manner the decomposition of chlorinated oxalate of methyl by potassa into chloride of potassium oxalate and carbonate of potassa,-by alcohol into hydrochloric acid oxalate and chlorocar- bonate of oxide of ethyl,-and by ammonia into chloride of ammonia and carbamide (and doubtless oxamide) corresponds perfectly with the comportment of a compound of chlorocarbonic acid with the above-mentioned hypothetical oxalo-dinoxychloricle. It might be expected from the analogy of chlorinated formiatc of oxide of ethyl with the other compound chlorinated ethers that it would comport itself with the above reagents like chlorocarbonic acid into which it is converted by the action of a high temperature 0 (C Cl,) 0.ClnC2 0 = 4C { cl* The formation of oxychlocarbonate' of oxide of ethyl and hydrochloric acid by treatment of the chlorinated ether with alcohol is likewise in accordance with this supposition; instead however of yielding chloride of ammonium and carbamide by treatment with ammonia it is stated by Cahours to yield dinox- amide of trichloracetyl (chloracetamide) . Although the percentage-composition obtained by Cahours for the latter compound agrees exactly with that of chloracetamide I cannot refrain from doubting the above statement and presuming that it must be based upon some error. Additional weight is given to this supposition by the fact that in endeavouring to construct the formula of chloracetarnide from that of the Chlorinated formiate of oxide of methyl there remain two atoms of oxygen of the use of which no account can be rendered as will be seen from the following equation (C Cl,) 0.Cl-C, 0,+2 NH = (C CI3)-C2,{ $-Ha+N H,Cl+20 3 L-v-/ I-v-Chlorinated formiate of Chloracetamide.oxide of methyl. It is very desirable that Cahours should repeat this experiment and remove the doubts on the subject. 1 shall refrain from making any observations on chlorosuccinic ether and its complicated decompositions described by 31alagu ti as it is my belief that the metamorphoses which free succinic acid under- goes by the action of chlorine must be first examined and determined VOL.111.-NO XII. Dn DR. KOLBE ON THE CHEMfCAL before we can furnish any account of the rational composition of chlorosuccinic ether or of its products of decomposition. The perfect similarity exhibited by chlorinated acetate of oxide of methyl and chlorinated formate of oxide of ethyl as far as our present knowledge of them extends has induced most chemists to consider them as identical. While the most different views exist respecting the manner in which the elements are grouped in these two compounds their supposed identity has been employed as an argument against the assertion that compound clorinated ethers still possess the constitution of the normal ethers. I myself am far from concluding that because the chlorinated ethers which have hitherto been more carefully examined exhibit an analogy with those from which they have been produced,-the molecular arrangement of the atoms arid of the proximate constituents must likewise remain unaltered in the action of chlorine on all other compound ethers; on the contrary I consider it more than probable that ethers of complex composition such as chlorinated amylic-ether valerianic or even niargaric ethers cease to possess the constitution of the original ethers long before the whole of the hydrogen is replaced by chlorine.Perhaps even the chlorinated succinic ether belongs to this class as its complex composition and metamorphoses lead to the assumption that the exchange of its hydrogen for chlorine is followed by a far more intricate decomposition.The identity of chlorinated formiate of oxide of ethyl and chlorinated acetate of oxide of methyl even if it were established could therefore not serve absolutely to determine our views concerning the nature of other chlorinated ethers. But I am not of opinion that the two latter compounds have the same chemical constitution and are really identical because they are alike in their comportment. If we view these two others as composed according to the rational formuh (C CI,) 0 .Cl-C ,-O3 chlorinated formiate of oxide of ethyl (C C13) 0 (c C1,) C ,O3 chlorinated acetate of oxide of methyl and consider at the same time that their proximate constituents- (C Cl,) OandClnC2,0 on the one hand (C2CI3) 0,and(C2 Cl,)"C,O on the other-are held together with but feeble affinity-if we bear in mind moreover that as no one of these adjuncts exists in the uncom- bined state at the moment that their chemical equilibrium is disturbed by the influence of any agent their further decomposition must be preceded by a transposition of their atoms after the following manner -\-,,-A L-v-J Chlorinated formiate of oxide Chloraldehyde.Chlorocarbonic of ethyl. acid. CONSTITUTION AND NATURE OF ORGANIC RADICALS. 403 L-v--d I-v-d -Chlorinated acetate of oxide of methyl. Chloraldehyde. Chlorocarbonic acid. (these metamorphoses being produced in reality by the action of a high temperature) ;-if all these circumstances be duly considered it will be easily understood why these two bodies exhibit a similar comportment or at any rate yield the same products of decomposition with such agents as are capable of effecting this disturbance in the chemical equilibrium of their constituents Their correspondence besides in specific gravity boiling-point &c.must least of all excite surprise as even the two normal ethers from which they are produced exhibit a perfect similarity in these respects. Acetate of oxide of methyl and formiate of oxide of ethyl are well known to differ only in their behaviour when boiled with alkalis; they evidently owe this mark of distinction to the greater stability of their consti- tuents without which they would probably be likewise considered as identical. I feel convinced that by a careful investigation and com- parison of these two chlorinated ethers for instance with regard to their refractive power or to their behaviour with such bodies as do not effect their decomposition their relative solubilities in ether benzol and so on some slight varieties will be discovered between them which will lead to a proof of their dissimilarity by experimental means.If the compound ethers which have been the subject of our con- sideration in which all the hydrogen is replaced by chlorine still possess the constitution of the original substances it follows neces- sarily that the molecular arrangement of the atoms in the intermediate products must likewise be the same. The dichlorinated acetic ether which is decomposed by water and potassa into acetic and hydro- chloric acids corresponds therefore in composition to the formula (C { 3)0.(C2 H,)^C, 0,. Its conversion into these products is probably preceded as with perchloracetic ether by a separa-H tion of the two members and as (C fCf) 0 cannot exist in the 2 free state by a transposition of these atoms to (C H3)-C2 { c12 (oxydichloride of acetyl). The trichlorinated acetate of oxide of ethyl produced by a continuation of the process of substitution is either H acetate of oxide of trichlorethyl (C,( cf) 0. (C H3)-C2 0, or 2 DD2 404 DR KOLBE ON THE ORGANIC RADICALS. monochloracetate of oxide of dichlorethyl The deliquescent potassa-salt-containing chlorine obtained therefrom according to Leblanc by treatment with solution of potassa is probably monochloracetate of potassa mixed with acetate of potassa ; and the oily body simultaneously deposited is possibly a secondary product of decomposition produced from monochloracetic acid in a manner similar to the formation of trichloride of formyl from trichloracetic acid.By the distillation of trichloracetic acid with alcohol containing sulphuric acid as also by the treatment of dinoxychloride of trichlor- acetyl with alcohol an ether isomeric with the above compound is produced namely trichloracetate of oxide of ethyl differing greatly in its behaviour from the former and particularly in the circumstance that it yields alcohol again by treatment with boiling potassa. A comparison of their rational formulae will suffice to account for their isomerism (C {*3) 0.(C ,{,H,2)^C, 0 trichlorinated acetic ether.a CI (C H5) 0. (C ClJ-C, 0 trichloracetic ether Among the intermediate substitution-products of compound ethers in which only a portion of the hydrogen is replaced by chlorine and which have generally speaking been less carefully studied the dichlo- rinated acetate of oxide of methyl and the dichlorinated formiate of oxide of ethyl are particularly distinguished not merely by being iso- meric like the final products above-mentioned but also by correspond- ing in their cheniical comportment without however being identical. If they be coiisidered as composed according to the rational formulz (C,{ $2 0 .(C H,)-C, 0 dichlorinated acetate of oxide of methyl 0 .H-C, 0 dichlorinated formiate of oxide of ethyl (c4{2) the decomposition of both by potassa into acetic formic and hydrochlo- ric acids admits of a ready explanation. In the first compound the decomposition is probably preceded by the transposition of H (C,{ c,J 0 into H-C, { 0 (oxydichloride of formyl) and in the c], second case by the metamorphosis of (C c1, ''3{ { ) 0into (C H,)^C, 0 c1, (oxydichloride of acetyl) . The formic acid is therefore prodcced in the first case from the basic member and in the second case from the acid member of the dichlorinated ethers j on the other hand the MR. BRODlE ON THE ALCOHOL RADICALS. acetic acid produced in this reaction from the dichlorinated formiate of oxide of ethyl must be considered as a product of decomposition of the member (C4{ H cl”,’ 0 while it is pre-existing as such in the dichlorinated acetate of oxide of methyl.(To be concluded ila the next Jozsmal.)