MARSH AND GARDNER : RESEARCHES ON THE TERPENES. 285 XXII I.-Reseawhes on the Teypeues, VII. Huloqcn Berivutives of Camphor uizd thew Reactions. By J. E. MARSH and J. A. GARDNER. IN 1882, De la Roykre (Bull. Xoc. Chem., 38,579) described a compound of the formula C,,H,,Br,, which he obtained by the action of phos- phorus trichloride and bromine on camphor. He subsequently showed (BUZZ. Acad. Bely., 9, 565, 10, 759) that two isomeric compounds were obtained, both capable of forming, by loss of HBr, one and the same tribromocamphene, C,,H,,Br,. We had begun to study the action of phosphorus trichloride and bromine on camphor and on various members of the turpene group in ignorance of the work of De la Roykre, and it will be seen that the results which we have obtained in the case of camphor confirm those of the Belgian chemist, differing from his only in points of detail.Action of Phospho~us Tvicl&ride and Bromine on Camphor. The quantities taken were such as to correspond approximately with the equation C,,H,,O + Pc'l, + 3Br, = CloH,,Br4 + POCl, + 2HBr. The phosphorus trichloride and bromine were mixed in chloroform solution, and the camphor added gradually, the whole being kept cool. A t first, a red, crystalline compound is formed, possibly C,,H,,OBr,, which gradually disappears, hydrogen bromide being evolved, and the liquid becoming nearly colourless after a few days. It is then poured on ice, and the chloroform solution, separated from the aqueous liquid, is allowed to evaporate, The crystalline product, which is readily soluble in chloro- form, is dissolved in hot, light petroleum, in which it is only sparingly soluble and, on cooling, the greater part, consisting of the a-compound, crystallises out.It melts at 1 6 8 O , and forms small, fluffy crystals. On evaporating the mother liquor a t the ordinary temperature, the P-modi- fication crystallises out in large transparent plates melting at 1 4 4 O , along with small crystals of the a-compound, from which it can be sepa- rated mechanically. Occasionally, the two forms crystallise together in the form of round, crystalline nodules, apparently homogeneous, and melting about 120" ; they may, however, be sepamted by recrystallisa- tion and mechanical selection of the two kinds as before. The yield is good, although a quantity of oily matter is also formed, which we were unable t o crystallise.a-Trib*omocamp?tene hydrobromide.-This crystallises from light petro- leum in the form of a white powder ; it is much more readily soluble in chloroform, and on evaporation is deposited in large, hard, colourless VOL. LXXI. X286 MARSH AND GABDNER: RESEARCHES ON THE TERPENES. crystals melting at 168". It is dextro-rotatory, [a], = + 90.3" in chloro- form solution. On analysis, it gave the following percentage composition. Carbon. Hydrogen. Bromine. Total. Found. ................. 26.24 3.14 70.46 99-84 Calculated C,oH,,Br, 26.43 3.08 70.48 99.99 fl- 25.ibro?~~ocr~rn;vkere hydi-obromide, -This compound, which is formed in much smaller quantity than the other, crystallises from petroleum in large, transparent crystals melting a t 143-144".It also is dextro- rotatory, but much less so than the a-compound. In cbloroform solu- tion, it gave [ = + 7%". The following are the results of analysis. Carbon. Hydrogen. Bromine. Total. Found.. ................ 26.59 3-24 70.34 100.17 Calculated C,,H,,Br4 26.43 3.08 70.48 99.99 Action of Ylt,osphorzcs Trichloride and Bvonaine o n Borneo 1. Borneo1 was treated in the same way as camphor, the substances being taken in the proportion of 4 mols. of bromine to 1 of borneol and 1 of phosphorus trichloride. Much hydrogen bromide was evolved, and the product, after treatment with water, gave crystals from a mixture of alcohol and chloroform melting a t 166-168". They were recrystal- lised from light petroleum. It gave [a], = + 91" in chloroform solution. Calcnlated Found.C,,H,,Br4. Bromine ........................ 70.68 70.48 The crystals were thus identical with the a-compound obtained from camphor. It should be mentioned that the melting point of the a-compound varies according to the conditions under which the sub- stance is heated. If it is heated very slowly, it melts a t 168", decom- posing a t the same time ; but if it is heated rapidly, the temperature a a y reach 173" before the substance melts. It would appear that 168" is the temperature a t which the compound begins t o decompose, rather than its actual melting point. The action of phosphorus trichloride and bromine on menthone and on fenchone has been investigated, but liquid products only were obtained. On a-brornocamphor of m.p. 76", there is only a slight action, with the production of a compound probably identical with a-tribromocamphene hydrobromide.HALOGEN DERIVATIVES OF CAMPHOR AND THEIR REACTIONS. 287 Action o f Phosphorus TmkJdoride and &*ornine on Tu? pentine. Dextrorotatory turpentine was subjected to the action of bromine and phosphorus trichloride in chloroform solution. Thirty grams of the turpentine, 30 of phosphorus trichloride, and 76 of bromine were em ployed. After treatment with ice, an oil mas obtained which was dis- solved in light petroleum, and alcohol added so as to form two layers. Crystals were obtained which were recrystallised from n mixture of chloroform and alcohol. They form colourless needles melting a t 150". Found ..................19-34 2.52 78.18 100*04 Carbon. Hydrogen. Bromine. Total. Calculated CloHl,Br, 19.54 2-28 78-18 100*00 The action of phosphorus triohloride and bromine was tried on camphene and on turpentine hydrobromide, but liquid products alone were obtained. The action in the case of d-turpentine was very slow, and the yield of crystals small, only 4-5 grams of pure substance being obtained from the 30 grams of d-turpentine taken. All the bromine was used up, but the quantity employed was not sufficient to form a hexa- bromide, as it was not expected that this substance would be formed. T~ibi.onaoccL.~Lne, CloK,,Br3. When a-tribromocamphene hydrobromide from camphor is boiled for several hours with sodium methoxide, it loses hydrogen bromide and forms tribromocamphene.This can be purified by distillation in steam, when it comes over slowly ; if not distilled with steam, it is apt to be coloured yellow. Tribromocamp hene crystallises in long needles from alcohol ; it also crystallises well from ether and from ethylic acetate. It is very soluble in chloroform, and is left on evaporation of this solvent as a porcelain-like mass. It melts a t 75-76', and its specific rotation in chloroform solution is [a] = + 32.5" Carbon. Hydrogen. Bromine. Found ..................... 32.15 3.73 64.29 Calculated CloH,,Br, ... 32.17 3-48 64.34 The /3-tribromocamphene hydrobromide yields the same tribromo- camphene when treated in the same way with sodium methoxide. The specimen obtained from the P-compound melted at 75-76", and its specific rotation in chloroform solution was [a], = + 31.5".The production of the same tribromocamphene from two isomeric tetrabromo-compounds leads us to infer that, in the latter, three of the bromine atoms occupy the same position in the two isomers, and that the isomerism depends on the position of the fourth. The tribromo- camphene is a saturated compound. It neither absorbs bromine nor decolorises permanganate. x 2288 MARSH AND GARDNER: RESEARCHES ON THE TERPENES. Action of Phosphorus Pentacldos-ide on Camphor. Camphor was treated with about 18 times its weight of pentachlo- ride of phosphorus, using Spitzer's method, the whole being kept cool. The camphor slowly liquefies, and the pentachloride gradually goes into solution with slight evolution of hydrogen chloride.After seven or eight days, the mass was treat,ed with ice, care being taken that, during the decomposition of the oxychloride of phosphorus, the temperature should not rise, as, if this happens, a considerable decom- position occurs, and the product of the action is profoundly altered, becoming dark coloured and partly liquid. By operating in the cold, however, a white, solid product is obtained ; this, according to Spitzer, after purifying by crystallisation from ether, is camphene dichloride, a substance melting at 155". By treatment with light petroleum, how- ever, it is possible to separate the original substance into two, one very soluble in petroleum and separating on evaporation in small crystals which aggregate into masses, the other scarcely soluble in petroleum in the cold, but crystallising from the hot solvent in large, hard, transparent crystals.The amount of the latter product is variable, but in two experi- ment's about 35 per cent. of the weight of the camphor was obtained ; it is not improbable, however, that it may reach 5G per cent., the excess remaining dissolved in the petroleum. We call this compound a-chlo~o- cccmphne hyhochloYide, and the one more soluble in petroleum P-chloro- camphem hydrocldoride. a-CIdorocamphene hyclrocl~lmide. -This compound, as stated above, is left as a white, crystalline powder sparingly soluble in cold, light petro- leum; from the hot solvent, it separates in hard, large, transparent crystals melting at 165" with decomposition, and having the rotatory power [u) = - 2'7.7" in chloroform solution.This substance also crys- tallises well from alcohol in similar hard, distinct, crystals which show no tendency to aggregation. It does not give off hydrogen chloride on keeping ; at looo, it slowly evaporates. Carbon. Hydrogen. Chlorine. Total. Pound ..... ... ......... 58.0 8.1 34.3 100.4 Calculated C,,H,,CI, 58.0 7.7 34.3 100.0 P-Cldorocamphene hydrochZoride.-It is not certain that this sub- stance has at present been obtained free from the a-modification, and we are still engaged in investigating it. I t forms the part more soluble in light petroleum, and is left, on evaporating the solvent, as a white, adherent, crystalline mass. It crystallises well from alcohol, and these crystals also adhere to one another when dry, and stick to glass and paper.It slowly loses hydrogen chloride when kept, so that a bottle containing it, after standing for some time, smells of the gas ; in this, itHALOGEN DERIVATIVES OF CAMPHOR AND THEIR REACTIONS. 289 resembles Spitzer's compound. Its rotatory power is less than that of the a-compound. The value [a]= = - 13%" has been found in chloro- form solution; in one specimen, however, the rotation was as low as - 9". Chlwocc~mphem, C,,H,,Cl. a-Chlorocamphene hydrochloride (I 0 grams), prepared from camphor, was boiled for several hours with zinc dust (7 grams) and glacial acetic acid. On distilling in steam, a solid of low melting point came over which was taken up by ether and distilled. It boiled at about 202" and solidified in the receiver. Its rotation was taken in chloroform solution [a] = - 29.3".The chlorine was determined. It gave, Found. Calculated C,oH,,Cl. Chlorine ..................... 20.7 20.8 P-Chlorocamphene hydrochloride, treated in the same way with zinc and acetic acid, yielded a product which distilled for the most part at about 205", but the temperature rose to 220" before all had come over. The distillate, which was solid, was again treated with zinc and glacial acetic acid, and finally a product obtained which distilled between 199" and 201". It was a solid of low melting point, and gave the specific rota- tion in chloroform solution [ a = - 33.2". A determination of chlorine gave Found. Calculated CloHl,C1. Chlorine ..................... 20.2 20.8 The chlorocamphene thus obtained acts as a saturated compound ; it does not decolorise permanganate of potash at the ordinary tempera- ture, and is acted on by bromine with evolution of hydrogen bromide.It is not improbable that the chlorocamphenes from the a- and @hydrochlorides are identical, but we have not been able finally to establish this point. There seems little doubt that Pfaundler (Anna- Zen, 1860, 115, 36) previously obtained chlorocsmphene, but was wrong in attributing its formation to the proportions of camphor and penta- chloride of phosphorus taken; this must be attributed rather to his having heated the mixture, and thereby decomposed a portion of the dichloride of camphene into chlorocamphene and hydrogen chloride. Spitzer ( A m d e n , 1879, 196, 260) was unable to obtain chlorocam- phene in a pure state by Pfaundler's method, nor did his camphor dichloride yield chlorocamphene when heated with water, or with water and marble, or with aniline at 110" (Sitxungsbe?*. Akad.TVGn, 1880, 596). We had made several experiments on the dichloride ob- tained by Spitzer's method, with a view to the production of chloro- camphene, before we found that this dichloride was a mixture ; our ex- periments show, however, that although the dichloride readily loses some290 MARSH AND GARDNER: RESEARCHES ON THE TERPENES. of its hydrogen chloride, it is difficult to remove the whole so as t o form monochlorocamphene. Not unfrequently, the product obtained gave an amount of chlorine just midway between the monochloride and the dichloride, as Spitzer himself found.For instance, the dichloride (34 grams), mixed with aniline (30 grams), was boiled for half an hour; after the aniline had been removed, the solid product contained 27.5 per cent. of chlorine, distilled between 205" and 230", and had a rota- tory power in chloroform of [a],= - 27.0". Some of the dichloride, after being boiled with aniline, was separated and heated in a sealed tube with quinoline at 250" ; when the product was distilled after removal of the quinoline, the greater portion passed over between 198" and 200°, and became solid. It gave 20.5 per cent. chlorine (calculated for C,,HI,Cl, 20-S), and its specific rota- tion in a mixture of alcohol and chloroform was [.ID = - 5". It was found that the dichloride, when distilled by itself, lost hydro- gen chloride, and that its rotatory power increased with each successive distillation.Thus, a specimen of the dichloride having the specific rotation [ a ] D = - go, when distilled, gave off hydrogen chloride, the boil- ing point rising to about 225" ; this distillate had the specific rotation [ a ] , = - 19" in chloroform. After a second distillation, the specific rotatory power rose to [ a ] , , -- - 30" in chloroform, and after a third dis- tillation to -35". The distillate was now crystallised from alcohol, and the crystals gave 23 per cent. of chlorine. Another quantity of dichloride distilled in a current of hydrogen chloride did not show this increase of rotatory power. After three distillations in a current of the gas, the distillate, after crystallisation from alcohol, had the specific rotation [ a ] D = - 7.7" in chloroform, and contained 31.1 per cent.of chlorine. Hydmxycccnzphe.e, 01' Camphenol. Chlorocamphene (p. 289) is very stable under the action of the usual reagents employed to replace chlorine or to remove hydrogen chloride. Thus, we found that quinoline at 250" did not remove hydrogen chloride. Moreover, when chlorocamphene is heated with potassium acetate and glacial acetic acid a t 220°, it is recovered apparently quite unchanged. Chlorocamphene, however, dissolves in cold, sulphuric acid, giving off torrents of hydrogen chloride and forming an orange-red liquid, a sub- stance being produced from the chlorocamphene by replacement of the C1 atom by OH. There is a certain quantity of tarry matter produced at the same time, but the yield is very good if the sulphuric acid has been previously diluted with about 5 per cent.of water, and if care be taken to keep the mass cool when the sulphuric acid solution is after- wards diluted with water.HBLOGEN DERIVATIVES OF CAMPHOR AND THEIR REACTIOXS. 291 The process adopted is as follows : Chlorocamphene (34 grams) was added to strong sulphuric acid (340 grams) previously diluted with 5 per cent. of water ; the action was brisk, much hydrogen chloride being evolved, with frothing. The red liquid was then poured into excess of water, and distilled with steam, about 20 grams of oil being obtained in the distillate. This mas extracted with ether, distilled, the distillate dissolved in sulphuric acid containing 10 per cent.of water, and the solution shaken with light petroleum, ot remove any chloro-compound or other impurity not dissolved in the sulphuric acid. The acid solution was again diluted with water, distilled with steam, and the oil, which is lighter than water, was extracted with ether and distilled. range of lo", the boiling point being at about 230". A number of analyses were made of different specimens of hydroxy- camphene, which, however, were found still to contain traces of chlorine, nearly 1 per cent. being found in one specimen, This chlorine does not appear to be removed by the action of sulphuric acid, and is probably due to the presence of some higher chlorinated substance in the chlorocamphene used, as chlorocamphene carefully puri6ed by frequent fractionation was found to yield hydroxycamphene quite free from chlorine. I.gives the analyses of 5 specimens of hydroxycamphene, all of which contained traces of chlorine. 11. Analysis of hydroxy- camphene free from chlorine. It all came over within Calculated Found. Ci0H160* Carbon ......... 78.3 78.5 78.0 77.8 77.5 78.9 Hydrogen ...... 10.3 10.2 10.5 10.4 10.8 10.5 Calculated Found. C10H160* Carbon .................... 78.7 78.9 Hydrogen 10.7 10.5 11. { ................ Hydroxycamphene, or camphenol, is a colourless liquid which becomes slightly yellow on standing. It has a fragrant, camphorous smell, and burning, aromatic taste. It boils at about 230°, and the density CL is d 18.5"/18*5" = 0.9347." It is insoluble in water, but dissolves ap- parently unaltered in strong sulphuric acid, in strong nitric acid, and in a saturated solution of hydrogen chloride in water.Bromine acts on it with evolution of hydrogen bromide. Sodium attacks it with evolution of gas and formation of a solid compound. Acetyl chloride does not act readily on the substance when cold, but on boiling it with benzoyl chloride, hydrogen chloride is evolved in abundance ; phosphorus pentachloride also acts violently with evolution of hydro- gen chloride. When acted on by potassium dichromate and dilute * The specimen used contained traces of chlorine.292 MARSH AND GARDNER: RESEARCHES Oh’ THE TERPENES. sulphuric acid, part of the substance is destroyed, but a large part is recovered unchanged. There is no evidence of the formation of a ketone in this reaction.We are still engaged in the further study of this compound ; we may, however, put forward the opinion that, taking into account the various properties of the substame, there seems no donbt that it is an alcohol, most probably a tertiary alcohol and a saturated compound. Nomenclccture. I n giving names to the compounds described in this paper, we have considered them as derivatives of camphene. This seems t o be the simplest mode of regarding them, and the most suitable for purposes of nomenclature. Moreover, camphene itself is derived from cam- phor in a way similar to that by which these compounds are them- selves derived, and for the sake of simplicity, as well as from the point of view of general analogy, this method of nomenclature appears to be the most suitable.Conclwion. The views which we have put forward from time to time as to the constitution of camphor and of camphene appear to us to be supported by the facts ,described in this paper. We bave insisted on the saturated character of camphene as well as on that of camphor. J u s t as we have camphene derived from camphor through bornyl chloride, so we have a saturated tribromo-compound and a saturated monochloro- compound derived from camphor in a very similar way. We have, in a previous paper, endeavoured to explain the relationship of camphene to camphor, and the saturated character of each of them, by the assumption of the existence of two ring formations in camphor and of not less than three ring formations in camphene, the mutual con- version of compounds of one class into those of the other being effected by the making or breaking of a ring formation.The nature of the compounds can hardly be reconciled with formuls for camphor, such, for example, as those proposed by Bredt and by Tiemann, formulae which, as we have already pointed out, do not seem to us to be sustained even by the evidence brought forward by their supporters. Besides the saturated character of camphene, we have, in particular, the production of one and the same tribromocamphene from two different tetrabromides, and probably the production of the same monochlorocamphene from two different dichlorides, and, further, the production of an isomeride of camphor from chloro- camphene. Tn the latter instance, we should have expected, from Bredt’s formula, to have obtained, not an isomer of camphor, but cam- phor itself.HALOGEN DERIVATIVES OF CAMPHOR AND THEIR REACTIONS.293 We append to this paper notes on the crystallographic characters of some of the cry&& kindly contributed by Prof. Miers and Mr. H. L. Bowman. CHEMICAL LABORATORY, OXFORD. UNIVERSITY MUSEUM, C R Y S T A L LO a R A P H I C DET E R M I N A T I o NS. BY PROF. MIERS AND MR. H. L. BOWMAN. a- Fribvomocamphene Hydrobromide (Crystccllised fiom Ether). System anorthic- Axial angles a = Z 12'22'. p= 114'58'. y = 73'18'. Axial ratios, cb : b : c = 1.1135 : 1 : 1.0189. Observed forms, P{100}, e(010), g{OOl), h ( i f l } , n{llO}, a{Oii}, 1(430}, m(410) ? FIG. 1. a-Tribromocamphene hydrobromide crystallised from ether. Angles.Pe 1OO:OlO ge 001 :010 Pg 100:001 ~h ioo : iii eh oio : Ti1 Pn 100 : lL0 Pa 100 : 011 PI 100 :430 ga 001 : o i l Pm 100 : 410 1 Observed. 98"42&' 72"38' 69'19g 69'55' 61'37' 53'334' 64'1 5' 34'55' 53"4' l7"40' Limits. 98'28'-98'57' 72'25'-73"5' 69'9' -69'27' 69'33'-70"18' 61 "1 8'-61'5 6' 53'3 3'-53"34' 63'35'--64"57' - - - No. of edges. 7 5 11 8 3 2 3 1 1 1 Calculated. - ._ - - - 52'15' 64'8' 34'45' 51'46' 15'42'294 MARSH AND GARDNER: RESEARCHES ON THE TERPENES, 56'299' 35'46' 65'22' 70"12$' 55"4' Traces were seen of a face on the edge between (100) and (001) making an angle of about 55;" with (100) ; also of a face on the edge between (010) and (110) making an angle of about 33" with (010). The crystals were extremely imperfect, and variable in habit ; some appeared to be twinned, the plane of union being parallel to IL (iil), and the crystal being traversed by a lamina parallel to that face.Birefringence strong ; an optic axis emerges obliquely through P (loo), the plane containing the optic axis and the normal to P being nearly parallel to the edge Ph. 56'29'-56'30' 35'43'-35"54' 6 5'20'-65 "24' /3-Tribs.omocccrnphenne Hydsdwornide. System orthorhombic- Observed forms, cc(lOO), bfOlO), m{110), 7-{01l},f{102), x{120}, x may possibly be hemihedrally developed, appearing sometimes at cc : b : G = 0.7203 : 1 : 0.6621. only one end of the brachydiagonal a. FIG. 2. B-Tribromocamphene hydrobromide, 111. p. 144". br 010:011 am 100 : 110 af 100:102 wzf 110 : 102 ax 100:120 0 bserved. Limits. No. of edges. Calculated.I - - 65'25' 70'16' 55'14' Optic axial plane (001) ; acute bisectrix perpendicular to the face Optic axial angle, as measured through a natural crystal immersed The crystals are tabular owing to the large development of the (010). Birefringence strong, negative. in cedar oil, was found to be 53"45' for sodium light. faces (100).HALOGEN DERIVATIVES OF CAMPHOR AND THEIR REACTIONS. 295 ~ ~ 3 2 110 : i i o 88 011 :011 ST 011 : l o 1 ms 110:011 T~ibs.oll~ocuin2~lene (csystallised fq-orn dcohol). By s t em or t h orhom bic- CL : b : C = 1.0410 : 1 : 0.6164. Observed forms, m2(110}, s{Oll}, ~f101). a7"42' 87"7' -88"4' 6 63"ia' 63'15'-63'22' 3 1 42'57' - 67'45' 67"42'-67"48' 2 F I G . 3. Tribromocamphene crystallised from alcohol. Angles. Observed. l l Limits. No.of edges. Calculated. - - 42'549' 67"46' Optic axial plane (001). An optic axis emerges through each face rn, and at an inclination of 6" to the normal of nz, on the side towards the crystallographic axis b. Birefringence very strong ; the bisectrix which coincides with the crystallographic axis b is negative. The crystals from which the above results were obtained are acicular, the substance having been crystallised from alcohol. When crystallised from ether, the crystals were of quite different habit, being tabular, owing to increase of two opposite faces m. Some other faces occurred, possibly of the forms (120) and (2211, but these were too rough to give trustworthy measurements.296 MAltSH AKD GARDNER: RESEARCHES ON THE TERPENES. I;C-C'hlo,~ocanaphene Hychochlo?*itle. System or t ho rhombi c -- Observed forms, nfl@@), b{010), mfllO}, still), ~ ( O l l ) , x(lOl), cc : b : c = 0.9169 : 1 : 0.5906. q{ 223;. FIG. 4. a-Chlorocamphene hydrochloride. ~ Angles. 7nb 110:010 llLS 110 : 111 712a 110 :lo0 222s 110 : iii bs 010:l~l ss 111:ill ss 111:111 9mn 110 : 110 snr 11O:Oll 7nq 110:223 br 01O:Oll Observed. 47"29' 48'51' 42'334' 63'34' 51'40' [ 57"53'] 85'2' 86048' Limits. 47"22'-47"36' 48'42'-49'5' 42"15'--42"55' - - - - a4°50~--850ii~ - :75:31 [ 59"45']-[59"47'] 59'184' I 59'15'--59'22' No. of edges. 25 9 11 1 1 1 K1 [I1 23 2 2alculated. - - 42'31' 86'44' 63'36' 52'48' 58'2' 69'54' 59'26' 8502' 59'46.4' Measurements inclosed in square brackets were only made by means of the Optic axial plane (010). A negative bisectrix perpendicular to (OOl), Birefringence strong. The faces (100) are usually narrow. The face (111) is much smaller than (171). The face (223)q may possibly occur hemihedrally, since it mas, in more than one instance, developed on one corner of the crystal done. These crystals appear to be identical in form, though not in habit, with those described by F. v. Spitzer (Berichte Akad. Wien, 1880, 71, p. 596). maxinium illumination of the faces. the optic axes visible through (001) making a wide angle. MINERALOGICAL DEPARTMENT, UNIVEESITY MUSEUM, OXFORD.