DR. KOLBE ON THE ORGANIC RADICALS. VIII.-Or&the Chemical Constit.ution and Nature of Organic Radicals. BY N. KOLBE,Ph.D. F.C.S. (Conclusion.) Of all the chemical compounds with which we are as yet more intimately acquainted no one is more nearly related to the alcohols represented by the general formula C H(n+z)O, than hydrated oxide of phenyl (phenylic acid) (CI2H,) 0.HO. If the general characters of the alcohols consist in. their property of combining with sulphuric acid to form the so-called etherosulphuric acids ; in the capability possessed by the oxides they contain of simulta-neously eliminating their water of hydration and combining on the one hand with acids to form neutral salt-like compounds (the so-called compound ethers) and exchanging on the other hand their oxygen for chlorine bromine sulphur &c.whereby thus combinations similar to chloride or sulphide of ethyl are pro-duced which may be re-converted into the corresponding alcohols ; and finally if the alcohols are characterised by being converted when heated with the hydrates of the alkalies or other oxidising agents into the conjugate acids of their series,-oxide of phenyl may be said to fulfil these conditions inasmuch as it combines with sul- phuric acid to phenylosulphuric acid HO .SO,+ (C12H5) 0.SO, which exchanges its basic atom of water for other bases analogously to sulphovinic acid thus yielding neutral double salts of oxide of phenyl with a metallic sulphate which on being heated produce again hydrated oxide of phenyl.The analogy is carried out farther by the DR. KOLBE ON THE CHEMICAL property of hydrate of phenyl noticed by Laurent and Gerhardt* in their beautiful researches on the phenides of combining in the nascent state with acids for example with benaoic acid to form a kind of ether corresponding to benzoic ether {C12H5) 0.(cl2H,)-C, 0 (benzoate of oxide of phenyl) (C H6) 0.(CI2 H,)^C, 0 benzoate of oxide of ethyl (C12H6) 0.((4 H,)^C, 0 benzoate of oxide of phenyl which by being heated with hydrate of potassa is split again into its constituents benzoic acid and hydrated oxide of phenyl. The latter substance participates likewise in the peculiarity of the alcohols of yielding the haloid-compounds of the radical corresponding to the oxides if we may consider as such a compound the chloride of phenyl (C12H5) C1 obtained by Gerhardt and Laurentf by the action of pentachloride of phosphorus upon hydrated oxide of phenyl an ether-like substance which is re-converted by boiling with solution of potassa into phenyloxide-potassa (phenylate of potassa) With regard to the property of the alcohols of being converted by means of appropriate oxidising agents into definite acids which contain the same amount of carbon but two equivalents less of hydrogen which are replaced by two equivalents of oxygen the acid which corresponds to hydrated oxide of phenyl :HO.(Cl0 H,)^C, O, is as yet unknown; it can however scarcely be doubted that we shall succeed in obtaining it directly from hydrated oxide of phenyl as soon as proper energies have been directed towards filling up this gap.Hydrated oxide of phenyl exhibits on the other hand very re- markable differences from the alcoholsof the fatty acids. The oxides of the latter are mostly gifted with feebly basic properties and com- bine with acids particularly when they meet with them in the nascent state The oxide of phenyl however which certainly forms similar combinations although with much greater diGculty is evidently possessed of a far greater tendency to perform the part of an acid. Like water and many metallic oxides which comport themselves with strong bases like acids and with acids like bases this substance is capable of entering into combination with both ;forming apparently with the acids neutral compound ethers but only when the acid itself is in the nascent state as is for instance the case in the decomposi- tion of the so-called chloride of benzoyl (dinoxy-chloride of benzoyl) by hydrated oxide of phenyl into benzoate of oxide of phenyl.The hydrated oxide of phenyl combines as is well known with much * Gerhardt et Laurent C. R. 1849,429. f Gerhardt et Laurent C R. 1849,435. CONSTITUTION AND NATURE OF ORGANIC RADICALS. 43 greater readiness with the alkalies directly forming pure s&s from which it is separated unaltered by stronger acids in the form of oily drops. Oxide of ethyl does not however appear to be totally devoid of acid properties. The two well-known crystallised salts ethyl- oxide-potassa and ethyloxide-soda are perfectly analogous ta the phen yloxide-potassa.A second far more characteristic difference between the phenyl- compounds and those of the common ether-radicals consists in the proportionately much greater stability of phenyl exhibited in an unequivocal manner by the remarkable metamorphoses whiuh hy- drated oxide of phenyl undergoes when acted upon by chlorine bromine or nitric acid. It can now be scarcely doubted any longer that chloro-bromo-and nitrophenylic acids still possess the original molecular grouping of phenylic acid that they are true substitution-products oxides of secondary phenyl-radicals containing in the place of two three or even more equivalents of hydrogen an equal number of equivalents of chlorine bromine hyponitric acid &c.We are already acquainted with the combinations of the following secondary radicals derived from phenyl of which it is as yet not determined whether they can also exist as such in the free state Phenyl= C, H Radical of Phenylic acid Chlorophenyl , , Chlorophenassic acid . . C12{ Ff Dichlorophenyl . . C12{ 3 } , , Chlorophenessic acid Trichlorophenyl :( C12{. Pentachlorophenyl ? C, Cl ? , , Chlorophenissic acid , , Chlorophenussic acid Tribromophenyl . C,( f:3 } , , Bromophenissic acid Dinitrophenyl . . C12{ ,,“S,} , , Nitropheqessic acid Nitrophenissic acid . . C12{ 3g&,> Trinitrophenyl ’’ ” (nitropicric acid) , , Nitrodichlorophenissic Nitrodichlorophenyl C, { 3*} acid A strong proof of the correctness of the view that tnchloro- phen ylic dinitrophenylic and trinitrophenylic acids are still possessed of the chemical constitution of phenylic acid and endowed with similar properties appears to me to be furnished on the one hand by the fact that the first of these upon treatment with potassium- 4!4! DR.KOLBE ON THE CHEMICAL amalgam exchanges its chlorine gradually for hydrogen and on the other hand by the observation made by Gcrhardt and Laurent that the two nitrophenylic acids form with benzoic acid combinations quite similar to those of phenylic acid compound ethers in which the dinitrophenylic and trinitrophenylic acids occupy the place of the base (C12{z$o,)O. (CI2HJnC2 0 benzoate of oxide of dinitrophenyl (CI2{ $Po> 0 .(C12HJ-C, 0 benzoate of oxide of trinitrophenyl. Like the ether-radicals of the series C H(,+1) which are homo- logous to hydrogen and which we must consider as produced by the combination of C H with H phenyl may also be considered as a repetition of hydrogen namely as hydrogen which has com-bined with the additional carbohydrogen C, H,. If this assumption is already justified by the analogy just demonstrated of the ptenyl-alcohol with the hydrated oxides of the ether-radicals it receives farther confirmation from the fact that corresponding combinations of phenyl to the remaining manifold terms of the methyl- ethyl- amyl-series &c. may be found with scarcely an exception. This analogy is most strikingly proved by the capa- bility of phenyl of combining like methyl ethyl &c.with 2 equiva-lents of carbon to form a conjugate radical the combinations of which exhibit the greatest similarity with those of acetyl propionyl &c. It is true we have not yet succeeded in converting hydrated oxide of phenyl into the acid HO . (CI2HJ-C, 0 in a similar manner to the conversion of hydrated oxide of methyl into acetylic acid or what is equivalent in obtaining directly from hydrated oxide of phenyl the cyanide of phenyl ((4 H5) Cy (benzonitrile) which yields benzoic acid by treatment with the alkalies or acids ;but we must bear in mind that as yet only one method of the many that may lead to this end has been tried namely the distillation of pbenylosulphate of baryta with cyanide of potassium (Hofmann).* A more favour- able result might perhaps be obtained by an appropriate treatment of chloride of phenyl discovered by Gerhardt aiid Laurent with cyanide of potassium.In making the well-founded assumption that benzoic acid bears to hydrated oxide of phenyle the same relation as acetic acid to hydrated oxide of methyl propionic acid to alcohol and caproic acid * Ann. Ch. Pharm. LXXIV 32. CONSTITUTION AND NATURE OF ORGANIC RADICALS. to hydrated oxide of amyl it must be taken for granted that all these acids are analogous in their constitution ;and hence that benzoic acid likewise contains a conjugated radicalal in which phenyl is the ad- junct of C ;benzoyl= (C,,HJnC2. If we adopt this hypothesis which likewise finds support in the arguments set forth in Vol.LXXV p. 233 of Liebig's Annalen the following formuls will be the most simple expressions of the rational compositiou of the known benzoy 1-compounds. Benzoyl = (C1 H,)"C2. Hydrated oxide of benzoyl HO .(C, H5)-C2 0 Oil of bitter almonds. Hydrosulphate of benzoyl. HS . (C, H5)"C2 S Sulphobenzole (C ah our s*). Hydrochlorate of chloride Chlorobenzole (Cahourst). of benzoyl . . . . . -HC1. (CI2 H5)"C2 C1 Dinoxide of benzoyl . . (C12 H5)IC2 O,? Benzoeoxide (Berzelius:). Benzoylic acid . . . . H0,(Cl,H5) C, 0 Benzoic acid. Dinoxichloride of benzoyl (C12 H5)-C, So-called Chlorobenzoyl. { 2 Dinoxibromide of benzoyl (CI2H5)-C2 , Bromobenzoyl. { 2 Dinoxi-iodide of benzoyl . (C12H5)-C2 { , Iodobenzoyl.Dinoxicyanide of benzoyl . (C12H5)-C2 , Cyanobenzoyl. { 8 { 2 Dinoxisulphide of benzoyl (C12H5)"C2 , Sulphobenzoyl. Dinoxamide of benzoyl . (C12H5)-C2r , Beozamide. { 2% SUBSTITUTION-PRODUCTS OF THE BENZOYL-COMPOUNDS. { Ef Chlorobenzoylic acid . . . HO .(C12 )-C, 0 Dichlorobenzoylic acid . . HO. (C12 { Fi )-C2 0 TrichIorobenzoylic acid . . HO. (C12 { )T2 0 Bromobenzoylic acid . . . HO. (C12 { g )"C2 0 . . . HO . (C12 { g4)-Cz, Nitrobenzoylic acid 0 Nitrobenzoic acid. Dioxamide of nitrobenzoyl . (C12 { $)4 >-c2,{ 2H2 Dinitrobenzoylic acid. . . HO . (Clz { &O4)^C2 0 )-(& 0 Benzimic acid. Amidobenzoylic acid . . . Ho .(Cl2 { 2H2 ~ ~ ~acid.~ ~ m i a In whatever manner we view oil of bitter almonds as the hydrogen-combination of the group C, H 0,,as the second product * Ann.Ch. Pharm. LXX 41. f Ann. Ch. Pharm. LXX 40. $ Berzelius' Lehrbuch der Chemie 5 Aufl. IV 332. DR. KOLBE ON THE CHEMICAL of oxidation of the radical C, H, or as the aldehyde of benzoic acid its very numerous and complicated phenomena of decomposition and particularly the metamorphoses which it undergoes with ammonia present difficulties to each view which it is impossible at present to remove but which will probably vanish upon a repetition of former researches and a careful investigation of the various statements. At present it appears to me that the manifold relations of oil of bitter almonds’ p.articularly to the remaining benzoyl-combinations may be best explained by the view upon which is founded the forniula HO .(C12H5)^C2 0 ; namely that this substance is the hydrate of the lowest oxide of benzoyl analogous to the aldehyde of acetic acid. Its behaviour with chlorine in its conversion into dinoxichloride of benzoyl and hydrochloric acid HO . (C12 HJ-C, 0 + 2 C1= (C12 H5)^C, { i-H C1. certainly differs from that of aldehyde inasmuch as the latter appears not to yield a dinoxichloride of acetyl but to retain its basic atom of water unaltered passing over finally into chlorale ; this greater stability of the atom of water in aldehyde may however be ascribed to the more powerfully acid properties of the latter and its conse- quently greater affinity for the basic atom of water. It must remain undecided whether the compounds described by Cahours the hydrosulphate of benzoyl and the hydrochlorate of chloride of benzoyl possess a rational composition corresponding to the formulz given above ; their properties are still too little known .to enable us to deduce therefrom reasons favourable to any one view.At any rate the above view furnishes a perfectly satisfactory inter- pretation of the genetic relations of these two compounds to the hydrate of the oxide of benzoyl HO (CI2 H,)-C, 0 + P CI = H C1. (CI2 H5)-C2 C1+ P { a3 L-C,---I Hydrate of oxide of benzoyl. Hydroehlorate of chloride of benzoyl. HC1. (C,2H,)nC2 C1+ 2(KS.HS) = HS. (C,,H,)-C, S + 2KC1+2 HS L,-d L-v--.J Hydrochlorate of chloride Hydrosulphate of benzoyl. of benzoyl. The h ydrosulphobenzoyl produced by the direct action of hydro- sulphuric acid upon oil of bitter almonds appears to be identical with the hydrosulphate of benzoyl.The sulphobenzoic acid discovered by Mit scherlich whichisformed under circumstances similar to those of nitrobenxoylic acid classed above among the substitution-products of benzoylic acid (and which CONSTITUTION AND NATURE OF ORGANIC RADICALS. 47 must be viewed as the oxygen-compound of the secondary radical (Cl2{ 304)^C2) is generally considered as analogous to the latter. The analogy of the mode of formation and the rational composition Qf these two acids would be perfect if the composition of the sulpho- benzoic acid did correspond to the formula NO .(C12{ H* so,)^C2 0, and saturated only 1 equiv.of base. It is however bibasic and contains moreover 1 equiv. more of sulphuric acid. I do not doubt however that sulphuric acid effects a perfectly similar meta- morphosis of benzoylic acid to nitric acid with this difference only that the substitutioll-product corresponding to nitrobenzoylic acid HO . (Cl,{ :da)^C, 0, combines with an equivalent of sulphuric acid forming a double acid in which the two constituents (C12{ FdJnC2 0,and SO, retain their original saturating capacities. The composition of sulphobenzoic acid might therefore be expressed by the rational formula 2 HO . {(C12{ Fd?'Tz,0, and its forma- so3 tion by the following equation HO .(C12 H,)-C, O,+ 2 SO,=2 HO (C12 {f)Ja)^C2 0, I -so --J L-Benzoylic acid.Sulphobenzoic acid. Whether sulphurous acid be really capable of replacing 1equiv-of hydrogen as assumed in the above formula cannot be decided apriori but solely by facts. At any rate this assumption does not appear to be bolder than that of the displacement of hydrogen by hyponitric acid. Why should SO not behave in a similar manner to NO,? In the formation of the so-cdled sulphacetic acid we meet with a perfectly similar process of decomposition. Its rational formula 2 HO (C { H 2 )-C2 0,,explains in a no less satisfactory manner 02 i so, its bibasic properties and its relations to acetic acid HO .(C H,)"C, O,+2 SO+ HO . {(C { f/&)nC, 0 so3 L-7-2 L-2 Acetic acid. Sulphacetic acid. DR. KOLBE ON THE CHEMICAL I do not consider it improbable that many of the organic acids the chemical constitution of which is still unknown-such as malic acid tartaric acid &c.-will prove on closer examination to be conjugate acids as suggested some time ago by Dumas and Piria,* and to which circumstance they will be found indebted for their polybasicity.The above considerations lead to the question whether the ether- radicals are not capable of combining with other elements as they do with the adjunct C, to form similar conjugate radicals. f have no hesitation in answering this question affirmatively and think that before all kakodyl must be viewed as a conjugate radical of this description in which 2 equivs. of methyl form the adjunct of 1equiv. of arsenic kakodyl = 2 (C H3)-As.It has already been shown (in Vol. LXXV p. 218 of Liebig’s Annalen) how easily and naturally the formation of oxide of kakodyl is explained by this hypothesis. It may be applied without difficulty to all kakodyl- compounds which have already been treated on this view in the “Handworterbuch der Chemie,” Vol. IV p. 218 ff I will not omit to mention here that the highly interesting mode of formation of chloride of methyl by heating kakodylate of superchloride of kakodyl agrees with none of the views hitherto adopted of the constitution of the kakodyl-compounds better than with the above assumption that methyl is pre-existing in the radical. The group of conjugate radicals discovered by Frankland in which methyl ethyl &c. occur as adjuncts of the metals zinc tin &c.is nearly related to kakodyl. In methyl-zinc methyl-tin ethyl-zinc as in acetyl and kakodyl the powers of affinity of the conjugate members C, As Zn Sn are not only considerably increased by their combination with the adjunct but their boiling- points are also considerably diminished these phenomena are pro- bably in most intimate connection with each other and may possibly arise from the assumption of a large amount of latent heat. A simple comparision of the boiling-points of the two corresponding acids oxalic acid and acetylic acid HO . (C HJ-C, 0, shows that the boiling temperature of the radical C has decreased about 130° by its combination with 1 equiv. of methyl if a conclusion may be drawn from the boiling-points of the combinations as to those of the corresponding radicals C, and (C H3)-C,.In like manner the arsenic combined with 2 equivs. of methyl (in kakodyl) is found to boil already at 170° and methyl-zinc to be a very volatile liquid spontaneously inflammable in the air.-Although attempts to prepare combinations of methyl-zinc and ethyl-zinc have as yet been unsuc- * Ann. Ch. Pharm. XLIV 70. CONSTITUTION AND NATURE OF ORGANIC RADICALS. 49 cesaful still it appears very probable that these bodies must be viewed as radicals partly from their behaviour and partly froni observations made by Frankland,* on the properties of the com- binations of ethyl-tin (stannethyl) which ace exceedingly similar to those of tin. Conjugate radicals similar to those last mentioned in which phenyl exists as the adjunct of metals have not yet been obtained; the former appears however to exist like the other ether-radicals in conjugated combination with sulphur and to form with the latter a radical corresponding to benzoyl (Ci2 H,)"S2 which may be assumed as pre-existing in hyposulphobenzidic acid.Some years ago I described in Liebig's Annalen Vol. LIV p. 145 four acids very closely allied to each other namely chlorocarbohyposulphuric acid chloroforniylohyposulphuric acid chlorelaylohyposulphuric acid and methylohyposulphuric acid of which the three latter may be produced directly from chlorocarbohyposulphuric acid in a manner similar to the re-production of acetic acid from chloracetic acid.I have noticed in the same memoir the interesting relations existing on the one hand between chlorocarbohyyosulphuric acid and chlora- cetic acid and between methyloxalic acid and acetic acid on the other hand and have made the conjecture that they might perhaps possess an analogous chemical constitution. The less doubtful it appeared to me that ~hlo~oca~bohypos~~lphuric, and methylohypo- sulphuric acid (besides the intermediate acids) contain hyposulphuric acid in conjugate combination with various adjuncts the greater was the support that I believed to have obtained for the view that the same adjuncts were contained in chloracetic and acetic acids only in combination with oxalic acid in the place of hyposulphuric acid. As however these views have been modified since then inasmuch as we no longer consider acetic acidas a conjugate oxalic acid but as the oxygen-compound of the conjugate radical (C H,)-C, it appears reasonable to entertain the opinion that those acids considered hitherto as conjugate hyposulphuric acids may be possessed of a rational coniposition corresponding to that of acetic acid.There is in fact nothing to prevent our assuming in methylohyposulphuric acid the existence of the conjugate radical (C H,)-S, and to ascribe to the adjunct the capability of exchanging its hydrogen like acetyl for an equivalent quantity of chlorine which would lead to H the formation of the secondary radicals (C,( cf)nS2 (C { N c1, and (C C13)-S2 of which chloracetylo- chloroformylo- and chloro- * According to a private communication.VOL. IV.-NO. XIII. E DR KOLRE ON THE CHEMICAL carbo-hyposulphuric acid must be considered as the oxygen com-pounds. As according to Muspratt’s experiments the methylo- hyposulphuric acid obtained by the oxidation of sulphocyanide of methyl by nitric acid is identical with that obtained from chloro- carbohyposulphuric acid the same view may also be extended to the homologous compounds formed by the oxidation of the sulpho- cyanides of ethyl and amyl namely ethylohyposulphuric acid and amylohyposulphuric acid. The conjecture made just now that hyposulphobenzidic acid might also possess a similar constitution to which we may also add hyposulphotoluidic and hyposulphonaph- thalic acids does not appear to me to be met by any difiiculties.Their mode of formation alone differs from that of methylo- ethylo- and aniylo-hyposulphuric acids ; as the carbohydrogens of the series C N,,$-21 corresponding to benzole toluole and naphthalole namely hydride of methyl (methylole marsh-gas} hydride of ethyl (ethylole) and hydride of amyl (amylole) are known not to enter into any com- bination with fuming sulphuric acid. The latter appear indeed to possess in general a much greater stability than the former as is shown particularly by the difference in their behaviour with fuming nitric acid; marsh-gas at least obtained by heating a mixture of acetate of soda and hydrate of lime remains perfectly unaltered on treatment with nitric acid even when passed through a mixture of the latter and concentrated sulphuric acid.The following table of corresponding acids the radicals of which consist on the one hand of C, on the other of S, both combined with the same adjuncts may serve for the farther elucidation of the foregoing Formylic acid. Unknown. €30. H-C, 0 HO . H-S, 0 Acetylic acid. M ethyiodithionic acid. (C H,)^S, 0, HO . (C HJ-C, 0 HO Chloromethylodithionic acid Chloracetyiic acid. { chlorelaylohyposulphuricacid). I-€0 (C,{ ;7-c2 0; HO . (C,{ Ef)-S2 O5 (I Dichloracetylic acid Dichloromethylodithionic acid (unknown). ~chlorofor~ylohyposulphuric acid). Trichloracetylic acid Trichlorom~thylodithionicacid (chloracetic acid). (chlorocarbohyposulphuric acid). HO .(C Cl,)-C, 0 HO . (C ClJnS2 0 CONSTITUTION AND NATURE OF ORGANIC RADICALS. Proyionic acid. Ethylodithionic acid Ho * (c~ H5)nc2 ‘3 HO (C H,)”S2 0 Caproic acid. Am ylodithionic acid. €30 WlO H,l)^C2 0 HO (ClO Hl,)^S2t 0 Phenylodithionic acid (hyposulphobenzidic acid). €3.0 (C H,)”S, 0 Kreotylodithionic acid* (sulphobenzoic acid). NO * (CL&T-I7)32 0 Naphthylodithionic acid (hypos~phonaphtha~c acid). HO (C, HT)-S, 05. The fact that acetyl and benzoyl combine in several proportions with oxygen justifies the supposition that the radicals of methylo-dithionic acid (C H,)-S, ethylodithionic acid (C H,)-S2 &c. might perhaps also combine with a smaller amount of oxygen. Sulphomethylosulphuric acid (Muspratt “Annalen der Cheniie,” Vol LXV p.261) the sulphethylosulphuric acid of Lowig and Wcidmann and the sulphamylosulphuric acid of Gerathewohl might be looked upon as lower oxides of these radical’s. The mode of formation of these acids which are obtained by the oxidation of disulphide of methyl hydrosulphate of sulphide of ethyl and hydro- sulphate of sulphide of amyl by nitric acid differs so little from that of the above dithionic acids with 5 equivs. of oxygen (which are obtained from the corresponding sulphocyanides) that one can scarcely imagine how different products can be formed under such equal circumstances. Add to this that their properties and those of their salts differ according to the various statements little or scarcely at all from those of the dithionic acids and finally that nearly all the ana- lytical results obtained with the so-called sulphomethylo- sulphethylo- and sulphamylo-sulphates correspond much better with the cornposi- tion of the methylo- &c.dithionates. The above reasons lead me to believe that the view first entertained by Gerhardt that Lowig’s sulphethylosulphuric acid contains 5 equivs. of oxygen (in the anhy- * The conjugate radicals of this second series of acids deserve particular names as well as those of the first. I have called them dithionic acids and have distinguished their adjuncts by placing their names in front. To the radical homologous to phengl C, H, which is the adjunct in toluplic acid I have given the name kreotyl (derived from kreosote) as I am of the opinion that kreosote is the homologous alcohol corre- sponding to hydrated oxide of phenyl ; hydrated oxide of kreotpl having the rational formula (C14H7) 0 .HO. E2 52 DH. KOLWE ON THE CHEMICAL drous state) may be extended to the corresponding methyl- and amyl- compounds although Muspratt (1. c.) has concluded from his corn- parative investigation of sulphethylosulphuric acid and ethylodithionic acid that they arc different substances (comp. ‘‘ Handworterbuch der Chernie,” Supplement p. 73 ff. and 169 ff).-I will not omit to mention that the heavy oily compound the sulphethylosulphurous acid discovered by Lowig and Weidmann into which hydrosul- phate of sulphide of ethyl is first converted by the action of nitric acid and which is transformed by continued treatment with nitric acid into ethylodithionic acid may be viewed according to its composition as a lower oxide of the radical (C HJ-S, possessing the rational formula ((2 H5)-S2 0,.Although methyl ethyl amyl phenyl &c, are of themselves only repetitions of hydrogen (produced by the combination of additional carbohydrogens)* C H, C H, C, HiO,and C, H with H it is remarkable that they seem notwithstanding to possess the capability of combining with another equivalent of hydrogen to form binary compounds hydrides from which they may be made to pass over to other substances as chlorine hyponitric acid amidogen &c. I have already at an earlier period made the conjecture? that marsh- gas did not possess the simple composition assigned to it by Berze- lius but that it might be the hydride of methyl (C H,) H; it appears to me that this view affords the best interpretation of its formation from acetates CaO .(C H3)-C2 0,+ CaO . HO = (C H3) H + 2 (CaO . GO,) c-“ -J hF-3 Acetate of lime. Hydride df methyl. This view has received a new support from the observation lately made by Frankland that iodide of methyl and zinc in the presence * It would he desirable to follow a more definite rule and principle of nomenclature in naming the different carbohydrogens particularly those that occur frequently side by side as C H, C H, and C H,. This might easily be effected by retaining the termi- nation yl to the names of the radicals formed by the combination of honiologous carbo- hydrogens with H by giving to the homologous carbrbohydrogens and the analogues of C H, C H, &c the nanies of the radicals with addition of the terminating syllable me and by denoting the hydrides of the radicals by the terminating syllable ole in the following manner C H Ethylene.C, H Phenylene. C H Naphthalene. C H Ethyl. C, H Phenyl. C H Naphthpl. C H Ethylole. C, H Phenylole (benzole). t I-Iandworterbnch dar Chemir 111 700. C H Naphthalole (naphthalin). CONSTITUTION AND NATURE OF ORGANIC RADICALS. 53 of water or methyl-zinc and water are decomposed into protoxide of zinc and hydride of methyl. Unfortunately the chemical deportment of this hydride of methyl as also that of the hydrides of ethyl and amyl discovered by Frankland,* has been so little studied that no arguments can be derived therefrom either for or against the above view unless I may perhaps mention here an observation made by Varrentrapp and myself namely that equal volumes of dry marah- gas and chlorine yield by exposure to diffused daylight equal volumes of hydrochloric acid gas and a chlorinated inflammable gas of which we have however as yet left undecided whether it is really chloride of methyl or an isomeric combination.A much fitter means of testing this question is presented by the interesting metamorphoses of benzole which stands in the same relation to benzoic acid and phenyl as hydride of methyl does to acetic acid and methyl. Its behaviour with nitric acid and sul- phuric acid clearly shows that one equivalent of hydrogen exists in it in a different form to the remaining five.The question how it is that benzole in exchanging this one equivalent of hydrogen for NO and passing over into nitrobenzole changes its chemical character altogether while by the substitution of a second equivalent of hydrogen a body is formed (dinitrobenzole) bearing the greatest similarity to nitrobenzole may be easily answered if me consider benzole as the hydrogen-compound of the phenyl-radical as hydride of phenyl=(C1 HJ H whereas the assumption expressed by the old formula C, H6 that all six equivalents of hydrogen are of eqiial value renders no account whatever of the above circumstance. It is evident that on the entrance of the first equivalent of hyponitric acid in the place of the hydrogen that is combined with phenyl there is formed nitrite of phenyl= (C12HJ NO, the phenyl itself remaining unaltered.It is only by the action of a hot mixture of sulphuric and nitric acids that the substitution of NO for hydrogen is extended to the radical the nitro-compound of the secondary radical C, { $o namely nitrite of nitrophenyl (C12{NHd) . NO 4 (dinitrobenzole) being produced which differs evidently not more from nitrite of phenyl than nitraniline does from aniline methylodithionic acid from chloromethylodithionic acid or nitrobenzoic acid from benzoic acid. Of all the reactions of dinitro-benzole its-behaviour with sulphide of ammonium shows most clearly that the two equivalents of NO play a perfectly different part ;for * Ann.Ch. Pharm. LXXI 171 ; and LXXIV 41. *-' DR. KOLBE ON THE CHBMICAL on the conversion of dinitrobenzole into nitraniline the equivalent of hyponitric acid which occupies the place of hydrogen in the radical remains unaltered; the conversion of NO into amidogen is confined to the hyponitric acid existing in the compound externally of the radical atom (C12{ &).N0,+6HS=(C12{ $o) .NH2+4HOfBS Dimtrobenzole. The metamorphosis which benzole undergoes by treatment with sulphuric acid namely the formation of sulphobenzide and sulpho- benzide-sulphuric acid is also in perfect accordance with the above mode of viewing if we consider sulphobenzide as phenylodithionic oxide (C12H5).S, O, and sulphobenzide-sulphuric acid as phenylodithionic acid HO .(C, H,)^S2 0 (corresponding to methylodithionic acid &C.) Attempts to prepare the chloride of the phenyl-radical from the hydride by proper treatment with chlorine have not hitherto been successful.The ultimate success of such experiments can however be doubted the less since chloride of phenyl has been obtained in another way by Laurent and Gerhardt by the action of penta-chloride of phosphorus on hydrated oxide of phenyl. Mitscherlich's chlorobenzide is perhaps a substitution-product of this substance namely the chloride of dichlorophenyl (C12 { :$) C1 and chloro- benzin is possibly a combination of the latter with 3 equivalents of hydrochloric acid C, {Fc)Cl .3H C1.The homologues of benzole namely toluole cumole and cymole as also naphthalin are so nearly allied to that substance that I have no hesitation in ascribing to them a similar constitution and to consider them as composed according to the following rational formule toluole = (C14H7) H ;cumole = (C18€Ill) H ;cymole= (C20H13)H ; naphthalin= (C2 €I7) H. The radical C, H is the only one which we have not had as yet as adjunct of C in a compound corresponding to acetic acid and benzoic acid; on the other hand naphthylodithionic acid (hyposulphonaphthalic acid) HO .(C2 H,)"S, O, in which naphthyl is the adjunct of S, corresponds to methylodithionic acid and phenylodithionic acid (hyposulphobenzoic acid.) The numerous derivatives of naphthalin obtained therefrom by treatment with chlorine bromine nitric acid &c.merit in a high degree the attciition of chemists. There is perhaps no class of CONSTITUTION AND NATURE OF ORGANIC RADICALS. bodies so well adapted to demonstrate in a convincing manner the insufficiency of the assumption of the immutability of organic radicals. All attempts to interpret the modes of formation of these deriva- tives of naphthalin their metamorphoses and relations to each other according to this view may be considered as having totally failed. I shall therefore confine myself to giving an arrange-ment of those formulae which I consider as the most probable expression of their rational composition; they are grounded upon the hypothesis that naphthalin contains the radical C, H (naphthyl) and that the latter admits of the substitutions of its hydrogen by chlorine hyponitric acid &c within a certain limit without under- going a farther change in the molecular grouping of its atoms.Naphthyl = C, H,. Hydride of naphthyl . . . . (C H,) H Naphtalin. Chloride of naphthyl . . . (C H?)C1 Chlonaphthase. Souschlorure de napht alaline Hydrochlorate of chloride of naphthyl (C20 H,) C1. HCl chloride of naphtalin (Berzelius.) Bromide of naphthyl . . . (C H,) Br Bronaphthase. Hydrobromate of bromide of naphthyl (C% H,) Br .HBr Unknown. Nitrite of naphthyl . . . (C H,) NO Nitronaphthalase. Naphthylamine . . * (C H,)NH Naphthalidine. Naphtliylodit~onic oxide . . . (C H,) SO Sulphonaphthalin (Berz.) Naphthylodithionic acid .HO (c Hyposulphonaphthalic acid 20 7) 2 5 (Berz.) SUBSTITUTION-PRODUCTS. Chloride of chloronaphthyl . . . (C { Ef ) C1 Chlonaphthese. Chlorure de naphta- Dihydrochlorate of chloride of chloronaph-(c { €Ifc1. HC1 line &loride of naph-thy1 . . . . . . . thaline (€3 erzelius.) Chloride of dichloronaphthyl . . (C20{ Fi) Cl Chlonaphthise. Dihydrochlorate of chloride of dichloro-Chlorure de chlo- naphthyl . . . . . . ('20 { Ff2) ' HC1 u't pht ase. Chloride of trichloronaphthyl . . . (C20{ tt)C1 Chlonaphtose. Dihydrochlorate of chloride of trichloro-cl. HC1 Chlorure de chlo- naphthyl . . . . . . { Fc) C1 naphtese. Chloride of quintichloronaphthyl . . (C Chlonaphthalase. Chloride of hypachloronaphthyl . . (C C1,) C1 Chlonaphthalise.{ !:) Br Bromide of broponaphthyl . . . (C Bronaphthese. { i:; Br Bromide of dibromonaphthyl . . (C Bronaphthise. DR. KOLBE ON THE CHE3lICAL Bromide of tribromonaphthyl . . . (C,* !;> { Br Hydrobromate of bromide of tribromo-(C { g:3)B2.HBr Bronaphthose. Sousbromure de bra-naphthyl . . . . . . naphtise. H Br HBr Bromure de bro- Dihydrobromate of bromide of tribromo-. (c20 { Br,) naphthyl . . . . . naphtese. Dihydrobromate of bromide of quadribro-. . . . . (C { !:>Br .2 WBr Br:g:st: monaphthyl bra-Souschlorure de Hydrochlorate of bromide of chloronaphthyl (C20{ Ef ) Br * HCl bronaphtase. Chloride of chlorobromonaphthpl . . (CZ0{ zf) C1 Chlorebronaphtise. Br Chloride of chlorodibromonaphthyl .. C1 Chlorebronaphtose. Dihydrobromate of chloride of chlorodibro-cl I.IBr Broniure de chlo- naphthyl . . . . . . naphtese. Chloride of dichlorobromonaphthyl . . C1 Chloribronaphthose. Bromure de chlo- Dihydrobromate of chloride of tribromo- H cl. HBr robronaphtese. monaphthyl . . . . . . (‘20{ Br) Dihydrochlorate of bromide of dichlorobro-Br HCl Chlorure de bro- monaphthyl . . . . . . { 3)Br naphtese. Bromide of trichlorobromonap~~thyl. . (Czo Bromechlonaphtnse. Dihydrochlorate of bromide of trichlorobro-Br. HC1 Chlorure de bro- monaphthyl . . . . . . mechlonaphtise. Nitrite of nitronaphthyl . . . (C ) . NO Nitronaphthalese. { a4 . . . (C { s4 Nitrite of dinitronaphthyl ) . NO Nitronapbthalise. { Ei ) C1 chlorwk. Chloride of dichlorodinitronaphthyl .. (C Binitronaphtaline bi-2 NO { F! chlor6. Chloronaphthylodithionic acid . . HO . (C )“S2 0 Acide sulfonaphtalique acid . HO . (C,{ zi )-S2 0 Acide sulfonaphtalique ~ichloronapt~thylodithionic bichtork. Trichloronaphthylodithionicacid . HO . (c2,)Ff3 )“S, 0 Acide sulfonaphtalique { trichlor& H Quadrichloronaphthylodithionic acid . €10. (c { ,-,i’ )-s, 0 hide sulfonaphtalique qt,adrichlork CONSTITUTION AND NATURE OF ORUANIC RADICALS. 5'1 Acide sulfonaphtalique Bromonaphthylodithionic acid . . HO . (C,{ i; )^S, 0 brom& !:2{ Acide sulfonaphtalique Xtronaphthylodithionic acid . . HO . (C { Z8,)IS2 0 nitr6. Acide sulfonaphtalique bibrom& Dibromonaphtbylodithionic acid . HO .(C )-S2 0 Acide thionaphtalique Thionaphthalic acid* .. 2 HO. [('%I :)&{ 5303 )-'27 '5 (Laure nt.) Napthtinhyposulphuric acid (Berzelius.) The two highly remarkable compounds containing chlorine and oxygen but no nitrogen which Laurent has obtained by the action of nitric acid on the dihydrochlorate of chloride of dichloro-naphthyl (C,o{ 2)Cl . 2HC1 and on chloride of quintichloro-naphthyl (C2 C1 possessing the empirical formula?C, H C1,0, { 3) 5 and C, C1 0, besides the acids obtained from the latter by treat- ment with potassa HO . C, H C1 0 (chloronaphthalic acid) and HO . C, Cl 0,,appear to possess a rational composition corresponding to the salicyl- and anisyl-compounds (see further on) and to belong H to the conjugate radicals (CIS { cf } O,)-'C and (Cis C1 O,)-C,.Since Strec ker? has called attention to the similarity existing between this chloronaphthalic acid and alizarin both with respect to their composition (alizarin = HO . C, H50,)and properties I scarcely consider it too bold an assumption to suppose that the H primary radical of the above hypothetical radical (C18 { cf } O,)-C, namely (CI H O,)-C, exists in alizarin. I propose for it the name alizyl. The following rational forrnulx?appear to me to afford the best interpretation of the mutual relations of the above compounds Alizyl = (C18 H5 OJ-C Dinoxychloride of alizyl . . (CIS H 02)%,,{ 3 Unknown. Alizylic acid . . HO . (CIS H 0,)^C2 0 Alizarin. Dinoxychloride of chloralizyl . (CIS{ F; } 0,)T2 Oxide (La { 0 tose de chloroxknaph-cj t.) * Conjugate acid of thionaphthylodithionic acid and sulphuric acid.f-Handworterbuch der Chemie von Liebig Poggendorff and Wohler IV 598. Art. Madder. DR. KOLBE ON THE CHEMICAL Chloralizylic acid . HO . (C18{ Ef} O,)-C, 0 Chloronaphthalic acid. Dinoxychloride of pentachloralizyl. (C, CI O,)^C, { Ei o%(&de chloroxQnaphta-(L au rent.) Pentachloralizylic acid. HO . (CIS C1 O,)-C, 0 Acide chloroxdnaphtalt5-siqae. (Laurent.) Hofmann in his latest excellent researches on the volatile bases,* has already expressed the opinion that there are probably two corresponding series of organic bases amidogen- and ammonia-bases and that the existence of amidogen-bases in which we assume with Liebig amidogen in combination with a compound radical does not necessarily preclude the existence of ammonia-bases in which we imagine ammonia as pre-existing in combination with a compound body.While thiosinamine thialdine the ureas and many vegetal bases the chemical constitution of which is as yet perfectly un-known may perhaps belong to the latter series Hofmann is cer-tainly right in viewing the volatile bases described by himself and Wurt z as the amidogen-combinations of compound radicals as ana- logues of ammonia of which the one (radical-) equivalent of hydrogen is replaced by a compound radical Ammonia . H .NH Methylamine . (C H5) *NH Phenylamine . . (C12 H,) .NH (Aniline). &C. &c. Another argument might be added to the many proofs already established by Hofrnann in favour of the view that ethylamine aniline &c.are really amidogen-combinations of the radicals C,H5 and C1 H, and have not according to Berzelius’ mode of viewing to be considered as conjugate ammonias (with the adjuncts C H and C, HJ. The important discovery that bromide of ethyl and am- monia are transformed into hydrobromic acid and ethylamine (C HJ Br+H .NH,=(C H5) .NH, I3Br +c7-.-v-3 Bromide of ethyl. Ammonia. Hydrobromate of ethylamine. and that nitrous acid reconverts etbylamine into an oxygen-com-pound of ethyl namely into nitrite of oxide of ethyl appears to me to place the question of the chemical constitiition of ethylamine parallel with the question whether we should consider bromide of ethyl as the bromide of the ethyl-radical or as hydrohromic acid * Ann.Ch. Pharm. LXXIV 125. CONSTITUTION AND NATURE OF ORGANIC RADICALS. combined with C H, viz. as C H,. H Br. It would according to the above facts be perfectly inconsistent to view ethylamine as ammonia conjugated with C H, and as expressed by the rational formula (C H,) .NH, or (C H,) (H .NH,) ;bromide of ethyl on the other hand being assumed as the ethyl-radical combined with bromine. An objection that has been repeatedly raised against the view that organic nitrogenised bases are amidogen-compounds is that amidogen has not the property of forming bases in support of which have been quoted amide of potassium oxamide oxamic acid &c. of which it cannot be doubted that they contain amidogen none of which are however endowed with basic properties.This kind of argument might certainly be classed with the no less paradoxical assertion that hydrochloric acid contains no chlorine because the latter forms neutral or indifferent bodies with potassium carbon &c. The diffe- rence between chloride of potassium chlorobenzoyl (C12H,-C, Ct { c, and hydrochloric acid is not less than that between amide of potassium benzamide ((4 H,)-C,, '2 { NH2 and ammonia. It cannot be doubted that the chemical nature of the amidogen- compounds exactly as with the chlorine-compounds is essentially modified as well by the nature of the body with which aniidogen combines as by the part which it has to play in such combinationsi While in amide of potassium it appears to play the part of a salt- former it is found in oxamide C2 { benzamide acetamide &c.to occupy the place of oxygen. From all these compounds to which may be added the amidogen-acids it may be expelled with facility as ammonia and oxygen may enter into its place. In amidobenxoylic acid HO .C, { EhJnC2,0 (benzimic acid) the amidogen appears to occupy the place of hydrogen as the chlorine does in chloro- H benzoylic acid NO.(C12{ C14)-C, 0, and it cannot be detected in this compound by the ordinary means any more than the chlorine can in the last-named acid. It is well known that on boiling these substances with solution of potassa neither ammonia is given off on the one hand nor chloride of potassium formed on the other.Can it appear strange after these statements that amidogen should combine and form bases with hydrogen and its analogues methyl phenyl &c.; that it exercises the functions of a base-former in organic DIt. KOLBE ON THE CHEMICAL bases ? The analogy of the different functions which amidogen and chlorine exercise would be perfect if the chlorides of methyl and ethyl in fact all the chlorine-compounds of the homologues and analogues of hydrogen were equally similar in their properties to hydrochloric acid as methylamine ethylamine phenylamine &c. are to ammonia. Methylamine and ammonia which according to Wur t z corre-spond so closely in their chemical behavioixr and even in their physical properties that their compounds can only be distinguished from each other by analysis may be compared in this respect to acetic and formic acids which contain the same radicals as adjuncts and do likewise not differ much more in their properties than those two bases Benzoic acid .HO. (C,H ) C, 0 (C12H ) .NH Aniline. Toluylic acid . HO. (C14H7 )-C, 0 (C14 H ) .NH Toluidine. &C. &C. On a closer examination and comparison of the above compounds it will be seen that what has been said (in Liebig’s Annalen pp. 225 and 233) of the coniparatively subordinate influence exer- cised by the adjuncts of the above acids over the chemical character of the latter may be almost equally applied to the amines of the same carbohydrogens For with whatever carbohydrogen (radical) the arnidogen may be combined,-provided that it is an analogue of hydrogen,-the resulting amines possess the peculiar chemical cha- racter of their prototype ammonia to such a degree that they would be always regarded as imitations of ammonia even if their rational composition were unknown.In the same manner as the member C, occurring in the above acids in combination with different adjuncts gives the principal stamp to the character of the resulting conjugate radicals formyl acetyl &c. and their combinations the correapond- ing amines appear to owe their basic nature almost exclusively to amidogen. Hofmann’s important discovery that several equivalellts of hydrogen in the radical of aniline niay even be replaced by chlorine iodine hyponitric acid &c.and that their products of substitution chloraniline (el$ ) .NH, dichloraniline (C12{ Et).NH, nitra- (‘ONSTITUTION AND NATURE OF ORGANIC RADICALS. niline (Ci2 { E.0) .NH, &c. are still possessed of basic properties proves most indisputably how little in general the basicity of the arnines is dependent upon the nature and composition of their radicals. These facts appear at it cursory glimpse to be favourable to the hypothesis constructed by the adherents to the theory of types that the part which an element plays in organic Composition is not dependent upon its original properties but solely upon the position which it occupies in the combination. Upon closer investigation of the question it will however be found that Hofmann’s investiga-tion of the chlorinated bases contains indeed the refutation of those extreme views.That chemist has shown that the basicity of aniline decreases in an inverse proportion to the number of the equivalents of chlorine and bromine introduced in the place of hydrogen. Brom-aniline is still a tolerably powerful base though weaker than aniline itself dibromaniline likewise still possesses basic properties its salts exhibit however but very slight stability ; tribromaniline finally is no longer basic but is like trichloraniline a perfectly indifferent body. Bromitie has therefore evidently imparted a portion of its original (negative) character to the compound into which it has entered in the place of hydrogen; the fundamental properties of aniline among which its basic nature must unquestionably be reckoned as of the highest importance are more or less modified in consequence of the above process of substitution; they are therefore not exclusively conditional upon the equality of the position of the elements.H ofmann’s latest and highly important discovery that all three equivalents of hydrogen in ammonia may be successively replaced by the so-called ether-radicals without the original basic character of the ammonia being lost is not only of thegreatest interest to the radical theory inasmuch as it appears to nie to place beyond all doubt the existence of compound radicals but promises also to throw quite unexpectedly new light upon the chemical constitution of many organic bases hitherto not properly understood.The re- markable isomerism of toluidine (C14H,) .NH, and methylaniline (C,2 H5). N {CH3} would if we were unacquainted with the mode of formation of the latter be to us equally enigmatical to the hitherto inexplicable isomerism of aniline and pico1ine.-As we have well-founded reasons for viewing aniline as composed according to the rational formula (CI2HJ .NH, we may imagine picoline to be a methylated base analogous to methylamine perhaps DR. KOLBE ON THE CHEMICAL If this assumption be well founded picoline would according to the experience we possess up to the present time only admit of the entrance of one atom of an ether-radical in place of its hydrogen ;this might be easily decided by a simple experiment.Since the one (radical-) equivalent of hydrogen in ammonia is pos- sessed of fiinctions different from those of the two others in amidogen the question arises which of the three equivalents of hydrogen is first replaced by ethyl in the conversion of ammonia into ethylamine (by treatment of the former with bromide of ethyl) ;or in other words -. CH whether the rational formula (C H,) .NH, or H .N { 5} must be assigned to ethylamine. The facility with which arnidogen may be transferred from the radical-hydrogen to other corn binations for instance in the formation of oxamide bcnzaniide &c. evidently imparts the greatest degree of probability to the first formula which is also generally adopted. The formation of aniline from nitrobenzole can scarcely be made to accord with any other formula for aniline than (Cl €i5).NH,.If therefore we should succeed at some future period in producing this base in the same manner as ethyl- amine-perhaps by the action of ammonia on the hitherto unknown bromide of phenyl (C12H5) Br,-which camot be regarded as unlikely or in producing from hydride of ethyl a nitrite of ethyl (C H5) .NO, and from this ethylamine,-the correctness of the view that the first substitution of the one equivalent of hydrogen in ammonia occurs with the radical-hydrogen might be considered as pretty well proved. The extent to which methyl ethyl and the so-called ether-radicals in general are capable of playing the part of hydrogen in organic eoimbinations is most clearly shown by the remarkable combination of iodide of ethyl with triethylamine discovered by Hofmann the hydriodate of ammonia in which all the equivalents of hydrogen are replaced by ethyl (C H5).N 2(C H5) (C H5) I as also by the oxygen-base obtained therefrom (C H5).N 2(C H5),(C H,) 0.HO which represents in the ethyl-series the member that is wanting in the ammonia-series hydrated oxide of ammonium.There can I think exist no more simple or decided proof of the existence of compound radicals than these methylated or ethylated bases. It is a circumstance worthy of notice that met'nylaniline although it contains the elements of C H more than aniline has a boiling- CONSTITUTION AND NATURE OF ORGANIC RADICALS.63 point only 10' C. higher than the latter (192"),while the isomeric toluidine boils at 200O. Again ethylaniline only boils 22O higher than aniline notwithstanding the difference existing between them of 2 (C HJ. Diamylaniline which contains 10 (C H,) more than aniline and should therefore have a boiling-point about 200" higher than the latter boils at a temperature only 100' higher than aniline ; lastly the boiling-point of amylethylaniline (262") which contains 7 (C H,) more than aniline exceeds that of the latter only by 80° instead of 140°,which is therefore likewise only half the ordinary increase. It hence appears as though the entrance of (C H,) into organic combinations effects at times only half the increase in the boiling- point that one is accustomed to expect and as though the form in which the homologising carboh ydrogen is added to the compound had no unimportant influence over this point.The above observa- tions are at least in favour of the view that where this carbo- hydrogen in the form of methyl ethyl &c. replaces the hydrogen exterior of the radical the boiling-ppint of' the combination rises only one half the amount that it does in cases where the carbohydrogen enters into the place of the radical itself. At any rate it appears to me to show in a most decided manner that the regularity of the increase in boiling-points caused by the entrance of certain elements into organic combinations is subject to many encroachments. The conversion of organic amines by the removal of 1 equivalent of hydrogen into compound amides which bear the same relation to the former as the simple amides to ammonia can be explained in a simple manner only by the assumption that this elimination of hydrogen takes place in the radical combined with the amidogen the amidogen itself remaining unaltered ; that therefore there is formed from qniline (C13H5) .NH, anilide of the rational composition (C12HJZNH ;from naphthalidine (C20H,) .NH, a naphthalidide= (C,* HJZNH ;from nitraniline (C12 { H4 ) .NH, a nitranilide = NO4 H (c, { Nb,)zNH,. The question now at once arises how it can be explained that bodies of so similar a composition as aniline and anilide possess such different properties ; that amidogen in combina- tion with the one carbohydrogen C, H should form a base while it forms with the other C, H, a body beariug a much greater resemblance to chlorine and oxygen ? This question may be easily answered if the fact be borne in mind that besides hydrogen itself DR.KOLBE ON THE CHEMICAL only those compound radicals that are homologous or analogous to it are capable of producing bases with amidogen or in general only radicals that can play the part of hydrogen as ethyl phenyl naphthyl &c. among which must also necessarily be included the secondary radicals derived from thcni lor instance chlorophen yl C, { tf} nitrophenyl C, { :b,} &c. The so-called addi- tional carbohydrogens C H, C H, C, €I4,C, H, &c. by the combination of which with 1 equivalent of hydrogen we imagine the above radicals produced and which evidently fulfil functions quite different from the latter must not be confounded or placed in the same rank with them.Incapable themselves of playing the part of hydrogen or indeed of a radical and therefore incapable of forming bases in combination with amidogen but apparently only destined to produce series of analogous combinations these carboh ydrogens only give rise by combination with amidogen to a series of different amides which still possess to a full extent the general characters of the simple amides. Anilide (C, H,)SNH is a true amide the properties of which are not altered to a greater extent by the addition of C, H, than those of formic acid WO . H-C, O, in its conversioii into benzoic acid = HO .(C12 H,)-C, 0, by the assumption of the same additional carbohydrogen in the adjunct of its radical. The following very simple expressions are obtained for the rational composition of the corresponding amides and anilides aniidic and anilidic acids by representing anilide (C12HJTH by And in the same manner as amidogen is represented by Ad. AMIDES ANILIDES. Carbamide . . . . Oxamide . . . . . C2{ % Oxanilide . . . . Benzamide . (C12 H )-C, :%{ Benzanilide . (C,,H )-C { kd Suberamide . (c €3 ):C2{ ?& Suberanilide (C H )zC2{ 2d &C. &C. CONSTITUTION AND NATURE OF ORCQANIC RADICALS. AMIDIC ACIDS. 0 Sulphamic acid . . HO. S { Aa; SO 0 Carbamic acid . . HO. c {Ad; CO Oxamic acid .. HO .C,{ 0 C 0 &C. ANILIC ACIDS. Carbanilic acid . . HO. c {zd; co oxanilic acid . . HO .c2{kd; C 0 &C. It is a remarkable and certainly not purely accidental circum- stance that a portion only of the acids that yield amides form also apidic acids and imides. With the exception of benzimide of which it is more than doubtful whether it may be reckoned among the imides there exist neither amidic acids nor imides of the fatty acids and those nearly related to them such as benzoic acid toluylic acid &c. On examining the series of known amidic and anilidic acids imides and aniles it will be observed that these compounds are formed exclusively from those acids which are capable of existing in the anhydrous state and are at the same time particularly inclined to form acid salts such as sulphuric acid carbonic acid oxalic acid besides the whole series of acids so nearly related to the latter succinic acid suberic acid camphoric acid phtalic acid &c.It must remain undecided whether the peculiarity of these acids of being capable of existing even without their basic atom of water is necessarily in some connection with their capability of forming amidic acids and imides ;we may perhaps expect some further elucidation of this subject as soon as we are better acquainted with the nature of these combinations particularly of the imides. With regard to the chemical constitution of the imides which we do iiot yet understand the interesting observations made by Dumas Malaguti and Leblanc that the amides are converted into cyanogen-compounds by the abstraction of water-acetamide VOL.I%'.-NO. XIII. F 1)H. KOLBE ON THE CHEMICAL { 2H2 (C H3)-C2 into thc cyanide*) of its adjunct cyanide of methyl (C H,)"C N; benzamide (CI2 HJ-C, { O2 into cyanide N*, of phenyl (C12 H,)-C N (benzonitri1e)-appears to afford a new point of view from which to investigate this question. I consider it as not improbable that in the conversion of amidic acids into imides by the elimination of 2 equivalents of water the process by which it is effected is the same as in the formation of cyanide of methyl from acetamide or in other words that the imides consist of I equivalent of the hydrated acid combined with a cyanogen-com-pound.If "we express the rational composition of an amidic acid of which there also exists an amide by the general formula { 2H HO .PIC, ; PSC, 0 (in which P represents the adjunct of 2 the acid) the formula HO. PZC N; PIC, 0 would be the rational expression of the composition of the corresponding imide and the conversion would proceed according to the following eqna- tion Nb2; HO.PZC,{ 0 PLC, 0 = HO .P3 N ;PZC 0,. + 2 HO. 3 c v 3 L v Amidic acid. Imide. This hypothesis by which the acid properties of the imides are likewise accounted for requires that we should assume in the acids yielding imides namely snccinic acid suberic acid phtalic acid camphoric acid &c. 2 equivalents of carbon combined with an adjunct as in the fatty acids.These adjuncts differ however mainly from those in the fatty acids in not consisting of ether-radicals but of the homologising carbohydrogens themselves C H, C H, &c In the same manner therefore as acetic acid butyric acid &c. may be viewed as repetitions of formic acid the former may be considered as repetitions of oxalic acid Oxalic acid HO. C, 0 HO. HT2,03 Formic acid. Succinic acid HO.(C H2)'3, 0 HO.(C H3)"C2 0 Acetic acid. Rdipinic acid HO.(C H4)ZC2,0 HO.(C,H,)-C,,O Propionic acid. Suberic acid. HO.(C H(?):C, 0 H0.(C6H7)T2,03Bntyric acid. &C. &C. * There might possibly exist two isomeric combinations of the empirical composition C W N one of which might be the nitro2e~i-combi1iation of the conjugate radical (C H3)-C, namely (C €13)-C, 3' (cyanogen the carbon of which has the adjuuct methyl) j the other the real cyanide of methyl (C H,)"C N =(C2 H,),Cy.In like manner benzonitrile might perhaps be viewed as (C,? H5)-C2 N differing from the still unknown true cyanide of phenyl {G,2 Hi), Cy. CONSTITUTION AND NrATURF OF ORGANIC RADICALS. fi7 Many chemists who consider the acids analogous to oxalic acid asalso the combinations belonging to this series camphoric acid phtalic acid pyrotartaric acid &c. as bibasic acids double their atomic weight. If‘ this view which is supported by a number of facts should be correct the kind of composition with which we have already become acquainted in speaking of sulphobenzoic and sulph- acetic acids may perhaps furnish us with pkey to the proper under- standing of the chemical constitution of those conjugate oxalic acids.These may be viewed in a similar manner to the above-mentioned as compound acids of oxalic acid and another conjugate oxalic acid in the following manner Oxalic acid ....... HO . C 0 Succinic acid .......2 HO .{(C H c4 04 )zC 0 Pyrotartaric acid ......2HO .{(C2 zr”)CC20, c6 o6 Adipic acid .......2HO .{(C2 os >cC203 c8 H3 23 Pimelic acid .......2 HO .{tC10 HI0 >CC2o3 2 3 Phtalic acid Suberie acid 2 HO .{((312 .......2 HO .{@I2c ....... 2 Hi2)zOd o3H )CC,O 3 Camphoric acid ......2 €10 .{ 2 ,H14)32 O3 3 Hyposulphosuccinic acid . .3 HO . According to the above view the composition of their amidic acids and imides would be expressed by the following rational forniulze Succinamic acid ..HO .(C H )3,{ gh,};C 0 Phtalamic acid ...HO .(CI2H4 )SC,{ $H2 .C2 0 Camphoramic acid ..HO .(C16H14)1C,( $H2} ;C2 0 &€.Succinimide ..HO .(C4 H )zC2 N ;C 0 Phtalirnide ...HO .(C12H >3, N; C 0 Camphoriniide . .HQ .(C, TI,,)zCz N ;C 0 &c. P2 DR. KOLBE ON THE CHEMICAL I do not deny that difficulties are met with in following up this hypo- thesis still farther. Among others it would be necessary to assume in the aniles for instance in succinanile the existence of the very hypo- thetical compound NC(C, H4) combined with (C4 H4)r3C2 and forming a peculiar cyanogen in which the carbon as well as the nitrogen are each joined with an additional carbohydrogen ; thus succin-anile would be = MO .(C H4)zC2NZ(C, H4) C 0,; phtalanile= NO. (C, H4)1C2 N\7(C1,H4) C 0, &c. Such compositions as these do certainly not appear adapted to increase our confidence in the probability of this hypothesis on the constitution of the imides. Nevertheless I consider it not impossible that these difficulties may be removed by future researches. A fresh investigation of this subject which is yet so little exhausted might perhaps lead to no uninteresting results. Among the various hypotheses that may be advanced upon the chemical constitution of anisylic acid and particularly for the expla- nation of the wonderful and manifold correspondence exhibited by anisylic acid and anisole on the one and benzoylic acid and benzole on the other as regards their chemical behaviour none appears to me more simple or probable than the view expressed in the formuh HO (C14 €3 O,)-C, 0 (anisylic acid) and (C14 H 0,) H (anisole); namely that their combinations possess a chemical composition similar to that ascribed to benzoylic acid and benzole and that they are only distinguished from the latter compounds in containing instead of C, H, the complex atom (C14H 0,).Objections may perhaps be raised as to the capability of the group (C14H7 0,)of playing the part of a radical and exercising functions similar to those of the adjunct C, H of benzoyl or the carbohydrogen C1 H in toluole as assumed by this hypothesis. The following comparative table of correspond- ing anisyl- and benzoyl-compounds and the corresponding bodies derived from both will be best adapted to weaken any objections made from a theoretical point of view against the plausibility of the assumption that oxygen can be the constituent of an organic radical Anisyl.Benzoyl. Pl* H7 02)-C 1 (Cl% H,)^C% I Anisglous acid. i Oil of bitter almonds. WO. (C14 H 02)-C, 0 1 I10 . (C12 H5)-c2 0 Anisylic acid. Benzoylic acid. 110 . (CI4 If; O2)-c?,0 HI0 (C, H,)-C, 0 '3 CONSTITUTlON AND NATURE OF ORGANIC RADICALS. Chloranisylic acid. Chlorobenzoylic acid. Nitranisylic acid. Nitrobenzoic acid. Ho ' ('14 {:8,)02)^'2 Dinoxybromideof anisyl. Dinoxybromide of benzoyl ('14 H7 '2)-'2 {gi DinoxychIorideof anisyl. Dinoxychloride of benzoyl.Dinoxamide of anisyl. Dinoxamide of benzoyl. ('14 H? 03)-c2t {&Tb20 Anisole. Benzole. (c14 H Oi?> Sulphanisolide. Sulphobenzide. ('l* H 02) so2 (C12 J35) 802 Sulphanisole-sulphuric acid. Sulphobenzole-sulphuric acid. HO .(C12 H7 0,)-S2 O5 HO a ((42 H,)^S2 0 Nitranisole. Nitrobenzole. (Cl*H7 02) NO4 w12 H5) NO4 Dinitranisole. Dinitrobenzole. ('14 {:6)O2' Anisidine. Aniline. ('14 H7 02) NH2 Nitranisidine. Nitraniline. &C. &C. Independent of the proof which the above comparison appears to me to afford that oxygenated organic radicals are capable of playing the part of ordinary radicals consisting merely of carbon and hydro-gen I think that after the assumption of secondary organic radicals no doubt can exist on this subject.If the existence of hypothetical DR. KOLBE ON THE CHEMICAL radicals is assumed in the nitrophenylic acids containing four eight and even twelve atoms of oxygen we can have no hesitation in ascribing to the complex-atom (C14H 02)the properties of a radical. It is certainly difficult to conceive in what form the two equivalents of oxygen are contained in the adjunct of anisyl and the radical Qf anisole anisidine &c. One might imagine them to play a similar part to the two equivalents of chlorine for instance in dichlo- H robenzoylic acid HO . (C12{ c$)-C2 0, or in dichloraniline fC12 { 8) NH, assuming therefore the grouped atom (Ci4 H70,) to be a secondary radical derived from the primary one (C14 H9)* &~kia substitution of 4 volumes of hydrogen by 2 volumes of oxygen is however in itself rather improbable and unsupported hitherto by any data Maq- chemists consider anisole as a similar and homologous corn-to po~~nc~hydrated oxide of phenyl (phenylic acid) from which indeed it differs only in containing C H more.This view is supported particularly by the fact that anisole is formed from anisylic acid under the same circumstances as hydrated oxide of phenyl is produced from salicylic acid which is homologous to anisylic acid. This fact can however only be considered as a slight proof as all other circumstances are favourable to the assumption that anisole possesses a constitution similar to benzole. Neither anisole nor its derivatives dibromanisole or trinitranisole possess the acid properties of phcnylic acid dibromophenylic acid and trinitrophenylic acid.On the other hand however chrysanisic acid lately discovered by Cahou1*s,* which is formed together with trinitranisole by the action of fuming nitric acid upon anisylic acid and is isomeric with trini- trailisole possesses exactly the properties which one would expect from a homologous combination corresponding to trinitrophenylic acid. r> lhere can be no hesitation in the decision of the question whether chrysanisxc acid or trinitranisole is most nearly related to trinitrophe- nyk acid since the first of the two possesses all and the second ilone of those properties so characteristic of trinitrophenylic acid. are moreover acquainted with a substance isomeric to anisole namely kreosote which appears at the same time to possess the greatest similarity to hydrated oxide of phenyl to which it is at any rate much more closely related in its properties than to anisole SO that it may be assumed with some degree of probability that krcosote and llot anisole is the compound honiologons to hydrated oxide of AIIA.Ch.1’11):. [3j xsl 11 CONSTITUTION AND NATURE OF ORGANIC RADICALS. phenyl sought for. A farther support for this view is furnished by the boiling-points of these bodies Hydrated oxide of phenyl boils at 187' the homologous compound containing C H more should therefore boil as usual 19O higher therefore at 206'. The boiling- point of kreosote approaches the latter number very closely it being 203' while anisole boils at as low a temperature as 152'.A similar regularity in the difference of boiling-points is likewise exhibited by anisole and its homologue phenetole (C ahonrs loc. cit.) obtained by the distillation of salicylate of oxide of ethyl with caustic baryta which has according to Cahours a boiling-point 20' higher than anisole (172O) and differs in its composition by an increase of C H, and but slightly in its cheniical behaviour from the latter substance. Hydrated oxide of phenyl and kreosote are evidently the alcohols of the homologous radicals C12 H and C, H, the former being HO . (C12H,) 0 the latter HO (C14 H,) 0 hence trinitrophe-nylic acid is composed according to the rational formula and chrysanisic acid according to the formula The following are the rational formuls of the isomeric combina- tions in question together with those of the members as yet un-known L--v-J Hydrated oxide of phenyl.Trinitrophenylic acid. L-v-3 Unhnown. L-'v-2 Chrysanisic acid. L-v-3 hnisole. Trinitranisole DR. KOLBE ON THE CHEMICAL L-v-J Unknown. L-v-3 Trinitrophenetole. It can scarcely be considered doubtful that salicylic and salicylous acids cmrespond in their rational composition to anisylic and anisy- lous acids Salicylic acid . . HO . (GI H O,)-C, O3 Salicylous acid . . HO . (C12H O,)-C, 0 Anisylic acid . . . HO . (GI*H7 O,)-C, 0 Anisylous acid . HO . (C14 H O,)-C, 0 and that the hydrogen-combination of the radical (C12 H 0,) isomeric with hydrated oxide of phenyl will likewise be discovered.The difference in the behaviour of salicylic and anisylic acids when heated with baryta is certainly highly deserving of attention; it is how- ever not more strange than many other reactions in which homo- logous combinations so frequently differ from each other. Since H ofmann’s researches on methylaniline diethylamine &c. have taught us that the ether-radicals can replace hydrogen in organic combinations I consider the existence of a compound similar and analogous to anisole having the rational formula (C1213 0,) . (C H3) as not improbable. It is possible that the anisole obtained from salicylate of oxide of methyl may possess such a composition (C H 0).(C12H O,)^C, 0 + 2 BaO = (CI H 0,) . (C 13,) + 2 (BaO CO,) and that as it differs probably but little from the anisole prepared from anisylic acid (C14 H 0,) H these slight differences have been overlooked. If formobenzoic acid may be considered as a double-compound of oil of bitter almonds and formic acid its isomerism with anisylic acid and gaultheric acid may be easily explained by their different rational formuls HO * (C14 H7 O,)^C, 03 Anis ylic acid. ((32 H3) 0 (C12 H O,)^C, 0 Gaultheric acid. HO . H-C, 0,; HO . (C, H,)-C, 0 Formobenzoic acid. rT I he following rational formukc afford likcmisc an explanation of CONSTITUTION AND NATURE OF ORGANIC RADICALS. 73 the isomerism of the four combinations composed according to the empirical formula C, H NO Salicylamide .. . . . (C12 H502)-C2 {2H2 Nitrotoluole . . . . . (Cl4 H7) . NO Benzamic acid . . NO . (C12 { NHHa)^0 Anthranilic acid (car- } HO . C (And; 0 CO%.* banilic acid) . . . By these formuh the chemical character of each individual com- pound is so definitely expressed that a cursory examination thereof is sufficient to enable us to understand for instance why salicyla- mide yields upon being heated with potassa ammonia and not aniline ; why anthranilic acid yields under similar circumstances neither salicylic acid nor ammonia but carbonic acid and aniline; and why benzamidic acid without being identical with anthranilic acid yields the same products HO C{ (Clz i)zNH2}; C0,+2KO=(C,H,).NH,+2(K0.C02) L---Anthranilic acid.HO . (Cl,{ 3H)-C2 0 + 2 KO = (CI2 H5).NH + 2 (KO. CO,) Y Benzamidic acid. I have in the foregoing endeavoured to show by a series of examples how easily the facts which appeared hitherto incompatible with the radical-theory and have been principally employed by the opponents of this theory as arguments against the existence of compound radicals may be made to accord with the above theory as soon as the idea of the immutability of these radicals is set aside and exchanged for the opinion that they are complex atoms in which certain atoms may be substituted by others. I have also endeavoured to establish the view that new secondary radicals are thus produced many of which still possess properties similar to those of the primary ones; that however not all radicals are equally adapted for con- version into secondary radicals (e.y.ethyl being less so than acetyl) ; that indeed one and t’ne same radical according to the manner in which it is combined possesses this property in different degrees (ethyl for instance possesses it to a higher degree in the compound * hiid = (Ck2 IJ4)ZNI12 (anilide). DR KOLBE ON THE CHEMICAL ethers than it does in the simple ones) Lastly I have deemed it necessary to distinguish between two classes of organic radicals the true repetitions (analogues) of hydrogen (the ether-radicals) and the conjugate radicals in which the former exist as adjuncts for instance of C, As Sb Sn &c. Besides these there appears to exist a third class formed by the combination of homologising car- bohydrogens C H, and similar compounds with simple substances for instance with C in the radical of succinic acid with Pt in elayl-platinum &c.and which may therefore be viewed as homo- lopes of these elements. The question still remains to be answered whether and how the assumption of alterable chlorinated radicals can be made to accord with the electro-chemical theory. According to this theory the nature of a chemical combination is dependent on that of its con-stituents arid hence the substitution of the negative chlorine for the positive hydrogen would be impossible. It appears to me necessary before all to examine carefully whether the hydrogen replaceable by chlorine really is the electro-positive constituent of the radicals and whether the chlorine which exercises the functions of hydrogen retains the negative properties generally ascribed to it.In simple inorganic combinations it is in most cases easy to determine which element is the positive and which the negative constituent but in organic compounds it is much more difficult to decide this question.-Berzelius adhered to the rule that the application of what is and will still become known with respect to the mode of combination of the elements in inorganic nature serves as a guide for the proper judgment of their combinations in organic nature by which we may hope to arrive at correct and con- cordant conceptions of the mode of composition of those bodies which are produced under the influence of vital processes as also of those which are produced by the metanioryhoses of these bodies by chemical means.” This principle led to the hypothesis that organic combinations must contain compound radicals exercising therein functions similar to those of simple radicals in inorganic nature and that in such a complex atom playing the part of a simple radical electro-positive properties are predominant as in the latter by which the combination with the negative elements is effected; no farther point of view is however furnished us by this hypothesis from which to arrive at a decision of the question which electro- chemical properties are possessed by the elements composing those groups of atoms as there is no analogy in inorganic chemistry with regard to this point.CONSTITUTION AND NATURE OF ORGANIC RADICALS. We must therefore content ourselves at present with the con- viction that there were in the elements constituting organic radicals electrical oppositions at the time of their combination as required by the electro-chemical theory and must leave it to the future to discover by other means which of those constituents are the positive and which the negative ones. Even in the single and most simple case in which two elements carbon and nitrogen combine directly to form an organic radical it is quite out of our power to determine with any degree of probability to which of the elements cyanogen owes its predominant negative properties and whether the nitrogen or carbon constitutes the positive (with respect to the negative) constituent.The solution of these questions becomes much more difficult with reference to the radicals con-sisting of carbon and hydrogen which are produced under far more complicated circumstances and cannot be formed like cyanogen directly from their elements. In this difficulty the phenomena of substitution afford appropriate means with which at any rate to impart some probability to the conjecture that the hydrogen for instance in acetyl is possessed of electro-chemical capacities differing from that contained in water or hydrochloric acid. We are also justified in doubting that the chlorine entering acetyl or aniline in the place of hydrogen carries with it the powerful electro-negative properties of which it is pos-sessed in its inorganic combinations.It can indeed be no longer doubted that in the simple substances the direction of the powers of affinity peculiar to them in their so-called normal state is altered under certain circumstances and frequently by apparently unim-portant causes and that therefore they assume at times an electro- chemical character differing from that they usually possess. The red amorphous phosphorus produced by heating phosphorus to 240' C. is certainly still the same substance but with respect to its chemical properties it is quite a different body. This element most remarkable next to the alkali-metals for its affinity for oxygen possesses singularly enough after exposure to the requisite heat not a trace of this affinity.The red phosphorus differs so perfectly from the normal phosphorus in its physical properties and in the altered direction of its powers of affinity that if its relations to colourless phosphorus were unknown it would certainly be con-sidered as a new element. These facts appear to me to warrant the assumption that hydrogen may also exist in organic combinations in a state in which it is possessed of properties differin9 from those of the ordinary electro- DR. KOLBE ON THE CHEMICAL positive element. Does not hydrogen exhibit a great difference with regard to its aflinities in the free and the nascent state? In the same manner I consider it not impossible that the chlorine entering into the place of hydrogen in compound radicals assumes by this act of substitution a less negative character than that peculiar to it in its normal state.It would not according to this be the positive hydrogen and the negative chlorine bromine &c. that re-place each other but bodies differing perhaps no more in their electro-chemical properties than oxygen and sulphur. This mode of viewing evidently still exhibits great deficiencies it shows however at any rate that it is not impossible to make the enigmatical phenomena of substitution accord with the radical-theory. “A good theory,” as Hofmann* very aptly remarks (;is more than a temporary expression of the state of science collecting under a general view the facts acquired up to the moment of its birth.It must not like ephemeral hypothesis vanish before the light of suc-ceeding discoveries but expanding with the growth of science it must still correctly represent the known facts though of necessity modified into a more general expression.” Who could doubt that the radical of theory is capable of such an expansion ? If the conception hitherto adopted of the immu- tability of organic radicals be only done away with the explanation of the phenomena of substitution will present no farther difficulties. It would be ridiculous to allow a single fact difficult of expla-nation to induce us to throw aside at once a theory which has served us for so long a period as a trustworthy guide in the difficult field of organic chemistry and has preserved us most securely from the errors of a code of laws like that which has been laid down by Laurent and Gerhardt-unless indeed we had some better theory to substitute for it.The radical-theory has already outlived the theory of metalepsy and the theory of types of which at the present time scarcely mention is made in the researches of French chemists; the nucleus-theory and the ingeneous invention of whole numbers of atoms will likewise disappear as rapidly from the field; for chemistry is indeed something better than a mere arithmetical problem into which Laurent and Gerhardt erideavour to convert it. By founding the hypothesis that replacements of hydrogen by chlorine hyponitric acid &c. occur in compound radicals the question whether oxygen can constitute the component of a radical is likewise answered.It is easy to render to oneself an account of the form in which oxygen exists in the radicals of nitrobenzoic and * Phil. Tians. 1851 I 96. CONSTITUTION AND NATURE OF ORGANIC RADICALS. nitrophenylic acids ; in many other cases as in anisyl salicyl we are as yet unable to do so and must therefore content ourselves for the present with the establishment of the mere fact. Perhaps these are also secondary radicals which we have not as yet been able to trace back to their primary ones. Since Frankland succeeded in isolating from their combinations the ether-radicals ethyl and amyl to which are added the radicals methyl and valyl obtained by the electrolysis of acetic and valeric acids various opinions have been put forward and supported with regard to the nature of these carbohydrogens and of these the view that they are not the true radicals but isomeric modifications of them appears to have become the one most generally entertained.According to Laurent and Gerhardt they possess double the atomic weight ascribed to them by Frankland and myself because the formulz C H, C H, &c. are at variance with their law of whole numbers of atoms. I acknowledge that it is not in my power to follow these chemists in their train of argument. The circum- stance that none of these radicals re-combines directly with oxygen chlorine sulphur &c. to form simple ethers is however more deserv- ing of attention. The importance attached to this circumstance is the greater because other radicals particularly kakodyl enter with such facility into combination with the above-named metalloids ; upon this circumstance endeavours have been made to establish the view that these indifferent carbohydrogens are not the true radicals contained in the ethers.It is a tolerably general but certainly unfounded opinion that the organic radicals in the free state must be gifted with powerful affinities. Many have expected particularly of the ether-radicals that their properties would not be much unlike those of potassium. This prejudice has evidently originated with the great diflicitlties hitherto met with in the isolation of the ether-radicals from their combinations. This is however at present effected in the most simple manner ;and since contrary to expectations methyl ethyl amyl &c.differ much in their combining powers from potassium as well as from kakodyl one is too much inclined to regard their indifferent properties as the most certain proof that they are not the ether-radicals Those who argue thus forget that hydrogen and platinum also belong to the radicals and that these two substatices exhibit just as little tendency to combine directly with oxygen chlorine &c. without the mediatiq infi uenee of another agent. Nothing justifies us in classing the ether-radicals with potassium ; on the other hand it appears to me quite natural to compare them with DR. KOLBE ON THE CHEMICAL hydrogen as a repetition of which they must be regarded.In fact they differ no farther in their properties from the latter than is con- ditional upon their mere complex composition. Hydrogen and methyl both possess under ordinary circumstances no affinity for oxygen iodine sulphur &c. ; when mixed with chlorine in the dark they both remain unaltered combination only ensues under the influence of light. It is however no more strange that chloride of methyl is not formed under these circumstances which the analogy with hydrogen would lead us to expect than that on igniting a mixture of methyl-gas and oxygen no oxide of methyl but water &C. is formed. It is easily conceivable that the stronger affinities of oxygen and chlorine for the constituents of methyl must predo- minate as is the case likewise when kakodyl is brought together with free oxygen or chlorine.The deficiency exhibited by the ether-radicals in strong powers of affinity does not therefore appear to me any reason whatever why they should be excluded from the series of radicals; in that case hydrogen must also not be allowed to stand as a radical. And finally the circumstance that kakodyl zincethyl stibethyl and as may be predicted with tolerable certainty even acetyl the lower oxide of which aldehyde possesses so powerful an affinity for oxygen are much more nearly related to the alkali-metals as regards their powers of combination cannot be regarded as a rule for the ether- radicals as these possess a constitution totally different from the above conjugate radicals.The principal support for the view extensively entertained that methyl ethyl butyl amyl and capryl in the isolated state possess double the atomic weight assigned to them that therefore their com- position should be expressed by the empirical formulae C H (methyl) c,HI0 (ethyl) Cl Hl (butyl) c, H,2 (amyl) c, H, (capry]) is found in the circumstance to which attention was first called by Hofmann that the boiling-points of butyl am$ and capryl differ from each other successively by 40° on account of the difference in their com- position of 2 C H, while if they were composed according to the formulae C H, Cl0 Hll C, H13 their boiling-points would only differ successively 20". This observation appears to me worthy of the greatest attention; I cannot however consider it as decisive.Hofmann himself by the discovery of the methylated- and ethyl- ated aniline-bases has brought to light facts which rob the assump- tion that the boiling temperatures of organic combinations increase 20" for every additional C M of its general validity. As the boiling-points of those aniline-bases increase almost regularly by only CONSTTTUTION AND NATURE OF ORGANIC RADICALS. loo,instead of 20° for every equivalent of C H added to the aniline the difference of C H in the ether-radicals may likewise just as easily effect a difference of 40' in the boiling-point. The question whether these carbohydrogens are the real ether- radicals or merely isomeric combinations appears to me at any rate to be of but little importance with respect to the radical-theory.They may be shown by future researches to be combinations of a different nature without endangering the radical-theory in the slightest degree. In the present state of science however the preference must undeniably be given for its greater simplicity to the assumption of the identity of those carbohydrogens with those of the radicals hypothetically assumed in the ethers