80 TILDEN ON THE RYDROCARBONS OBTAINED FROM XI1.- On the Hydrocarbons obtained from the Pinus Sylvestris, with Remarks on, the Constitution of the Terpenes. By WILLIAM A. TILDEN, D.Sc. (Lond.). I. Russian Turpentine Oil. SOME months ago, by the kindness of Dr. Armstrong, I received a quantity of Russian turpentine oil, with the information that it would distil chiefly between 168" and 180°, that is a t a temperature considerably above the boiling point of the terpenes of ordinary Ameri- can and French turpentine. The sample contained a small quantity of acetic acid and empyreu- matic products, and thcre can be no doubt, therefore, that this oil is obtained, not from the natural exudation from the living tree, but as a bye-product during the distillation of the wood for the purpose of preparing tar, as practised extensively in the north of Europe. Ac- cording to H a n bury and Fluckiger (Pharmacographia) the trees used for this purpose are chiefly the Pir~us syhestris and Pinus Lede- bourii.The Russian turpentine oil as it came into my possession had the specific gravity 0.8682, and 100 mm. of the liquid, t'ested by Wild's instrument, using the soda flame, were found to rotate the polarised ray about 17" to the right. It has a yellowish colour, and a peculiar pleasant odour, quite distinct from that of common turpentine oil. After sLaking with solution of caustic sr;da to remove acetic acid and creosote, the oil was submitted to fractional distillation, the frac- tions uItirnately obtained of nearly constant boiling point, being finally distilled once or twice from sodium.There are four principal ingre- dients in this turpentine- 1. The most volatile constituent is a terpene boiling at the same temperature as australene (from American oil), but its action on the polarised ray is much greater t*han that of australene. 100 mrn. give a rotation of + 23.3". There can be no doubt, however, that it is identical from a chemical point of view with austlralene, since itTHE PINUS SYLVESTRIS, ETC. 81 behaves in the same manner as that hydrocarbon when acted upon by nitrosyl chloride gas. This terpene constitutes 10 to 15 per cent. of the turpentine oil. 2. A second terpene present in Russian turpentine oil is an entirely different substance. After a most tedious series of fractionations and examination of the products, I have come to the conclusion that the pure hydrocarbon boils under ordinary atmospheric pressure at 171", or thereabouts.A considerable quantity of it, boiling at 171" t o 171*5", was collected. It possesses a characteristic odour. 100 mm. rotate the polarised ray 17" to the right. It gave by combustion 88.09 per cent. of carbon, and 12.14 of hydrogen, and by Hofmann's process a vapour density =65*8 (H = 1). The formiila ClOHl6 requires 88.23 per cent. of carbon, 11.76 per cent. of hydrogen, and vapour density 68. This terpene occurs in the turpentine oil to the extent of about two- thirds of its volume. The product I have just described is, I believe, as pure as it is possible to get it. Nevertheless, I think it contains traces of cymene, a hydrocarbon of which this turpentine contains a considerable quan- tity, and from which it would be very difficult to separate it com- pletely, in consequence of their boiling points lying so close together.3. Cymene was isolated by the usual process from the two fractions boiling at 1'72-173" and 173-175". These were mixed together. A portion of the mixed liquid was oxidised by chromic acid, when it gave about 3 per cent: of a mixture of toluic and teTephthalic acids. The rest was divided into two portions. 50 C.C. were cooled by ice and salt, and then mixed very gradually with more than an equal bulk of oil of vitriol. After standing for twenty-four hours the mixture was diluted, and from the separated oil a quantity of cymene was obtained equal t o about 79 per cent.by weight of the liquid operated upon. The second part of this liquid, consisting as it did mainly of the new terpene, was submitted t o the action of bromine in order to re- move if possible two atoms of hydrogen and convert it into cymene. In order to moderate the action of the bromine upon the hydrocarbon, the latter was diluted with about twice its bulk of chloroform. 30 C.C. of terpene and 60 C.C. of chloroform were mixed, and the liquid cooled t o - 10" before the bromine (rather less thac 28 grams) was gradually added. The resulting nearly colourless liquid was distilled, the chlo- roform being collected separately, and the distillation carried on till the thermometer went up to 240", and only a few drops of black liquid remained. Much hydrobromic acid was evolved throughout the opera- tion.The liquid was returned to the flask and redistilled five or six Its specific gravity was 0.86529 at 15".82 TILDEN ON THE HYDROCARBONS OBTAINED FROM times till the fuming ceased. It was then distilled several times from sodium. About 14+ grams were collected, corresponding to 56; per cent. by weight of the mixed hydrocarbons taken. The cymene was in both these experiments recognised by smell, optical inactivity, boiling point, and oxidation products. In order to represent the amount of cymene produced by the splitting np of the dibromide, the quantity which was obtained in the first ex- periment, and represents pre-existent cymene, must be subtracted from the quantity obtained in the second: that is, 56+--i* = 49 per cent., OP nearly half its weight.I have not succeeded in obtaining from the liquid, which I believe to be the nearly pure terpene, any solid hydrochloride by the action of hydrochloric acid gas upon the hydrocarbon alone, or upon its solution in ether. Nitrosyl chloride gas passed into the hydrocarbon, whether pure or diluted with chloroform or with alcohol, produces no crystalline nitro- sochloride as in the case of all other terpenes hitherto examined. After passing the gas into a mixture of the terpene with chloroform and allowing the solution to evaporate spontaneously, a few scaly crystals were deposited. These, when collected and dried by pressure, were found to be very easily soluble in rectified spirit, and the solution when set aside deposited nothing but oily drops.4. The presence of cymene in the Russian oil has been already proved. 5 . The oil contains small quantities of viscid hydrocarbons boiling at high temperatures. Within the past few months an examination of Swedish turpentine oil has been published by A. A t t e r b e r g (Dezk Chem. Ges. Ber., X, The one boils a t 156*5-157*5", and has the specific rotatory power + 36.3". The other is said to boil a t 173" to 175", with a specific rotatory power = + 1 9 * 5 O , and a specific gravity ~8612 at 16". The latter is called by A t t e r b e r g '' sglves- trene." I believe it to be identical with the terpene I have described, partly on account of their close agreement in boiling point and rota- tory powers, partly because tthey are almost certainly derived from the same source.I have not, however, been able to obtain a crystallised dihydrochloride in the manner described by A t t e r b e r g. 1202). He finds in this oil two terpenes. 11. Oleurn foliorum Pimi sylvestris. Under this name an oil is prepared for use in medicine by dis- I believe it to be tilling the leaves of the Scotch fir with water. identical with the " Fir-wool oil " imported from Germany.1WE PIKUS SYLVESTRIS, ETC. 83 The specific gravity of the sample I examined was 0.8756, and 100 mm. of the liquid turn the plane of polarisation 5i0 to the right. When distilled it began to boil at 80", owing, as I afterwards found, to the presence of a small quantity of alcohol. The temperature went up rapidly to near 165", below which point about one-fourth of the whole passed over.Between 165" and 175" a distillate was obtained equal to about two-thirds of the original liquid. By careful fractionation this oil was found to contain, beside small quantities of other bodies, two terpenes corresponding in boiling point with those obtained from Russian turpentine oil. The first smells like common turpentine, boils at 156" to 159", and a column of the liquid 100 mm. long rotates the plane of polarisation -k 18' 48'. It is, therefore, like australene dextrorotatory, but more strongly so than the usual variety. It possesses, however, the same chemical constitution, for it behaves with nitrosyl chloride in precisely the same manner. The second terpene, which constitutes about two-thirds the bulk of the oil, boils at 171" or it little above.It is laworotatory. The rotation per 100 mm. is about - 4", but this number can be considered only as an approximation, since the liquid with which the determination was made was undoubtedly contaminated with traces of cymene, as well as of another compound present in the oil, and it was found impossible to purify it completely. This hydrocarbon possesses the sameodour and behaves in the same manner towards reagents as the liquid boiling at the same tempera- ture obtained from Russian turpentine. It also has the same specific gravity at 15", viz., *86529. No solid hydrochloride could be prepared by saturating with hydro- chloric acid gas and exposing the resulting liquid to spontaneous evaporation, or to a very low temperature in a freezing mixture.In addition to these two terpenes, the essential oil of the Pinm syZvestris contains a small quantity of cymene and of pleasant-smelling liquids of higher boiling point. 111. Remarlcs o n the Constitution of the Terpenes. In discussing this question, several facts must be taken into con- sideration : 1. From experiments carried out in conjunction with Mr. Shen- stone (J. Chem. Soc., 1877, i, 554), I find that the number of isomeric terpenes is certainly very small, and according to my belief, there are amongst the natural terpenes only three isomerides essentially dif - ferent in chemical constitution. The known varieties of these differ84 TILDEN ON THE HYDROCARBONS OBTAINED FROM only i x i odour, in action on polarised light and other optical properties,* arid these differences can be fully accounted for without assuming any but mechanical differences of constitution.Thus I find that the terpenes from the following sources yield the same nitroso-derivatives. The terpenes thcmselves boil a t 156" to 159", and have nearly the same density, viz., about 0.860. Dextro-rotatorg. .... Russian turpentine oil ...... Natural order. Conifera?. American turpentine oil.. Oil of leaf of Pinus sylvestyis.. Lmvorotatory. French turpentine oil .......... N.O. Coniferae. Oil of sage .................. N.O. Labiatae. Oil of juniper ................ N.O. Coniferae. To these I feel pretty confident may be added the non-rotating hydrocarbon, terebene, which boils at 155-156", and has the specific gravity 0.860 at 20" (Riban).Like terebenthene, it yields a crystal- line monohydrochloride. The sem i-hy drochlor ide which this substance was supposed to form, has been shown by Riban to be a solution of the monohydrochloride in cymene. To the presence of cymene may also probably be ascribed my failure to obtain a crystalline nitroso- compound from the terebene upon which I operated. The second group of terpenes comprises those which boil a t or near t o 174", and which have a density somewhat below -85 at 20". These all yield the same nitroso-substitution compounds having the same melting point and crystalline form. They are obtained from the fol- lowing volatile oils :- Dextrorotatory . l, Natural Order Oil of sweet orange peel Oil of lemon .............. Oil of bergamot ............ 1 Aurantiaceae.Oil of caraway ............ N. 0. Umbelliferae. The third class comprises the two terpenes of higher boiling point described in this paper as derived from Russian turpentine ........ }N. 0. Conifer@. Oil of leaf of P. syZuest& . z . . * The slight differences observed in density and boiling point in the several members of the same group of terpenes are probably i? the main owing to the fact that in very few cases have the substances operated upon been in a pure state.THE PINUS SYLVESTRIS, ETC. 85 With the former of these, as already stated, Atterberg’s “syl- vestrene ” will, I believe, be identified. They all boil at 171-175”, and have zt specific gravity higher (-865 at 15”) than that of any of the others, whilst they yield no crystalline monohydrochloride or nitroso- chloride.2. All known terpenes yield, by the action of bromine, one and the same cymene, and that a-cymene, the constitution of which is admitted to be that of methyl-propyl- benzene, probably containing iso-propyl. 3. Terpenes are almost entirely broken up by oxidation into carbonic and acetic acids, and yield neither toluic nor terephthalic acid when pure. In those cases in which minute quantities of these acids have been obtained, their production is to be attributed to the presence of traces of cymene in the terpene operated upon. This has been already pointed out by Dr. C. R. A. W r i g h t , and I agree with him upon this point. The production of toluic and terephthalic acids by oxidation of terpin (Hempel, Liebig’s AnnaZen, clxxx, 71) seems to me to throw no light upon the question.I shall refer to this again presently. Terebic and terpenylic acids, from what is known of them, are not benzene derivatives. And, lastIy, it must not be forgotten that con- siderable quantities of oxalic acid are always produced when turpen- tine is oxidised by nitric acid. 4. Cymene cannot be made to combine with hydrogen, so as to pro- duce a terpene. 5. Terebene has been obtained from diamylene by removal of four atoms of hydrogen. Von Richter’s formula for diamylene is repre- sented as follows :- (C,H,) H H CHS H H, I I I I I I I H-C-C-C-C-C=C I I I I H H H H + u It was the consideration of this last fact which chiefly led me to observe that 0 p p e n h e i m’s formula for turpentine, which is identical with the formula for terebene proposed by v.Richter, is not the only nor even the most probable formula which presents itself for these bodies. That formula represenh turpentine as containing the benzene ring of six carbon atoms, as being in fact cymene with two atoms of hydrogen added. Hz l3 H, C-C ,c<c - c>c-cH3 CSH, H-H ThiR formula agrees with my hypothesis that there exist but three VOL. XXXIJI. H86 TILDEN ON THE HYDROCARBONS OBTAINED FROM isomeric terpenes, assuming that the propyl group is always the same. It also agrees with the fact that cpiene can be obtained from tur- pentine by the removal of two units of hydrogen and two of bromine from the dibromide. But it is very difficult t o reconcile with the fact that the same cymene is obtained from all the known terpenes, the isomerism of which must be explained, according to this hypothesis, by the assumption that the hydrogen symbols change their position in the formula. It is also difficult t o explain by this formula why, as I maintain, toluic and terephthalic acids are not obtained by the action of oxidising agents upon the terpenes.The formula which I propose for discussion is derived very simply from the formula for diamylene; for on removing the two pairs of hydrogen symbols, bracketed together in the formula written above, we come to the following expression :- Terpefie a. C,H, H H CH, H H H-C==C-CC---C-CC----C-H I I I I I I It is now clear that this admits of two other modifications, as repre- sented below, viz. :- Terpene p.H C3H7 H H CH3 H H-C=C-Cv-C - C-C-H I I I I I I Teiperze 'y. H H C3H7 H H CH, I I I I I I H-C---C-C=C-C---C-H The isomerism arises, therefore, from the varying posit'ions of the propyl and methyl groups upon this chain of carbon. So long as these two radicles maintain their distance from one another in the formula, the remora1 of one hydrogen symbol from each end of the chain and the linking of the terminal carbons together, result in a formula which is always the same, and which agrees with the received formula for ordinary cymene. The following diagrams represent the most probable expressions for the dibromides of the respective ter- penes, and their conversion into cymene by loss of 2HBr. From a-Terpene. CaH, H H CH, H H I J I I I I Dibromide C-C-C-C-C-C I H A.I H Br BrTHE Cymene Dibromide Cymene D i bromide Cymene PINUS SYLVESTRIS, ETC. C3H7 H H CH3 H H I I I I I I c..c-c=c-c=c From (3-Terpene. H C3H7 H H CH3 H I I I I I I c - c - c ~ c - c ~ c /\ I B Br Br I H H C3H7 H H CH3 H I I I I I I c ~ c - c ~ , c - - - c ~ , c I I From q-Terpene. H H C3H7 H H CH3 c=c-c=c-c-c I I I I I I I H I /\ Br Br H 87 These three formula for cymene are evidently identical. Hempel has found that terpin hydrate yields, when oxidised, toluic and terephthalic acids, and hence concludes that the production of these acids by oxidation of ordinary turpentine cmnot be wholly attributed to the existence of cymene or other impurities in the turpentine. Now when terpin hydrate is heated with diluted nitric acid, the first effect i6 the removal of the elements of water and production of terpinol, which, if it really has the formula usually assigned to it, namely, C,H,,O, or (&HS3(OH), is a condensation product.Since it con- tains but one oxygen-atom, and that supposed to be in the form of hydroxyl, it must evidently be generated by the union of carbon to carbon in some way at present unknown. On the other hand, there is the not improbable suggestion put forward by Gerhardt, that ter- pinol may be a hydrocarbon, an isoterpene. This supposition derives support from the fact, that the terebene semi-hydrochloride, C,H3,C1, to which terpinol was supposed to correspond, does not exist. Whatever be the real nature of terpinol, my object in referring to it is only t o point out that the products of its oxidation must not be assumed without further evidence to be the same as those of a ter- pene.88 TILDEN OX THE HYDROCARBONS, ETC.I have only further to say, with regard to this question, that the formulz proposed in this paper for the terpenes are brought forward with the utmost diffidence. But although I do not insist upon them, it seems to me that, in view of the dead-lock a t which we have arrived in regard to the question of the constitution of these compounds, the time has come when we may fairly test the capacities of some hypo- thesis different from that which has hitherto chiefly found favour. From the terpenes to camphor is a comparatively short step; and although I have no experiments of my own to offer, I cannot forbear a few words upon the subject.The following facts seem to have been established :- 1. Camphor, identical with ordinary camphor in all but rotatory power, has been obtained by Riban by oxidising the lsvogyrate cam- phene from Brench turpentine oil. Camphor is also said to have been formed by oxidising turpentine by permanganate (B e r t h elo t). 2. Camphor, by losing the elements of water, yields ordinary cymene identical with that obtained from the terpenes. These two circumstances seem to connect camphor with turpentine, and at first sight it would appear that the grouping of the carbon symbols must be the same in the formulz of the two compounds. But although I have represented the terpenes as consisting of an open chain of carbon, I feel disposed to consider camphor as a benzene or cymene derivative, and chiefly for the following reasons :- 3. Camphor yields substitution-derivatives with comparative ease. Thus the following compounds have been obtained :- Bromo-camphor, ClOHI5BrO. Chloro-camphor, CloH15CI0. Nitro- camphor, CloH,, ( NOz) 4 0 . 4. Sodium camphor, treated with carbonic anhydride, yields the sodium salt of camphocarbonic acid, a compound which is easily re- solved again into GO, and camphor. This reaction is parallel to the production of salicylic acid from phenol. CloH1,O + C02 = CioH,5O.COOH CsH,O + C02 = C,H,O.COOH 5. By the action of nitric acid, camphor yields, amongst other pro- ducts, a large quantity (more than half its weight) of camphoric acid, which contains as many C and H atoms as camphor. 6. By loss of H,, camphor is converted into carvacrol. These last characteristics, it may be observed, are unlike those of ketones in general, a class of compounds to which camphor, on account of its relations to borneol, is very commonly referred.