View Article Online / Journal Homepage / Table of Contents for this issue

110 GIREVILLE WILLTAW, ON IBOPBERE AND CAOUTCHIX- Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17.

XXIIL- -041 Isoprene and Caoutchin.

By C. GBEVILLEWILLIAMS.

[Abelracted from the “Philomphical Transactions,” 1860, p. 241.1

THE products of the destructive distillation of caoutchouc have been studied by several chemists. The results hitherto obtained View Article Online

GREYILLE WILLIAMS, ON ISOPRENE AND CAOUTCHIN. 111 are, however, quite irreconcilable. Gregory* appears to have been the first to direct attention to them, but it is evident that even at the time of writing his u Handbook of Organic Chemistry” hc was far from considering tlie true nature of the products as having been ascertained ; for, at p. 316,-he says : (‘When exposed to heat, caoutchouc first melts and then distils, yieldin? fi mixture of several oily liquids, all of which, as well as pue caoutchouc itself, are carbo-hydrogens. Somc of tlicve oils boil at 90°, otbers at 680” Fahr.; and at intermediate points I found that one highly rectified oil which bded at 96’, and had tlie cornpositioii of olefiant gas, yieldcd when actcd on by sulpliuric acid, an oil diich boiled at 4-28”, and had the same composition. But most of these oils have tlie composition of oil of turpcntine, C& or C,,H,. One of them, called caoutchin, gives with chlorine an oil, C,,H, + Cl.” Now the above statement contains undoibtedly thevbest account that could be given in so short a space of the results that had been obtained up to the time of publication of the work alluded to; but the account is soniemhat confused, because the various observers all differ immensely in their results. The expressions C,H, and CloH8, as used by Grcgorp, mcrclp indicate arithme- tical relations, and are not to be received as definite forrnulze; for, in the first place, C,H, is, according to our present views, an impossiblc formula?, and in thc ncst, caoutcliin is C20H,6,not C,,H,. No chcmist up to the present time has proved a hydro- carbon of the formula C,,H, (= 4 volumes of vapour) to exist among thc products of the .distillation or caoutchouc. Such a Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. formula, if proved, would bc interesting, because, turpentine being the starting-point for K o pp’s empirical formula for boiling-points, any substance having exactly half the number of atoms of carbon and hydrogen, with the same rapour-volume, mould serve as the point of departure of numerous speculations upon boiling-points generally. The results of 31. BouclisrdatS are curious, and not easy to understand. He obtains olefimts and eupioiic, but no bodies of tlic formula n (C5H4),while IIimly8 obtains bodies of the latter formula and no olefiants. The anliexed table contains the results of tlie tlirec chemists alliiclcd to.

* -4nn. ch. Pharm. svi., 61. + C=6, H=l, O=& z Ann. Ch. Pkarm. ssvii., 30. 5 Ibid. p. 40. View Article Online

112 GREVXLLE WILLIAblS, ON ISOPRENE AND CAOUTCHIR.

Density. Name of observer. I Fluid found. Boiling-point. --- Liquid. I-- Vapour. Gregory ...... CnHn...... 36" C. 0.654 Bouchardat ...... C3HS...... below 0 0'630 at 4" Bouchardat ...... Caoutchhne (C.H,) 14.5 0*650 'Bouchardat...... Eupione (?) ...... 61 0-690 at 16' Himly ...... CiH, (1)...... 30 -40 Himly ...... Caoutchin (C,H,,) 171 0.842 4.461 Bowhardat ...... HBvene (C,H.) .... nbove 315 0-921 at 21"

Reiults so contradictory would almost lead one to suppose that caoutchouc, like castor-oil, yielded different products according to the temperature ; but in that case, the distillate obtained in a care- fully conducted experiment should contain bodies belonging to the olefiants and the terebenes; this, however, is not the case. The weight of evidence appearing to prove olefiants to be the principal products, I had somc hopes that caoutchouc, if carefully distilled, would yield hydrocarbons froin which I might obtain certain at present unknown hydriodic ethers and their derivatives, in accord- ance with reactions first imagined by &I.Berthelot, thup :-

C,H,+HI = (C$,.,+J +I -__I/ v Olcfian- t pl UE Iodide of alcohol- hjdriodic acid. radical. but I soon found that caoutcb.ouc, when Carefully distilled in iron Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. vessels, yielded no olefiant whatever. This is, moreover in per- fect accordance with experiments made by me some years i.go,* in which I-showed that caoutchouc, like many other oils, decolorized the same quantity of bromine as oil of turpentiue. A repetition of this experiment, made with great precaution to ensure success, will be found further on. The distillation was conducted with care in an iron alembic, the lowe& temperature being used consistent with the distillation of oils. In fact I' even stopped the distillation before the whole of the last oil, h&.Cdue, came over, in order to ensure the decom- position taking place in the simplest manner. The crude distillate

* I' On a Procese for estimating the Equivalents of some Fluid Hydrocarbons by means of Bromine," Chemical Gazette, 1853. View Article Online

CREVILLE W11aLTA3?S, ON ISOPRENE AND CAOUTCHIN. 113

was very fetid, and contained traces of volatile b- derived from the decompositioD of a minute trace of vegetable caseine in the caoutchouc. The bases were rcnloved by agitation of the oil with dilute sulphuric acid; the latter was then got rid of by wash- ing with water, which, in its turn, was subsequently 1-cmoved digestion with sticks of potash. The fluid thus prepared was easily separated by fractional distillation into two portions, Ork boiling between 37' and 44' C., and the other between 170' and 180'. The fractional separation of the hydrocarbons in the crude distillate affords a very instructive example of the extraordinary extent to which boiling-points are lowered by successive rectifia- tions. In the first rectification the product was very small, below 116' C. The greater part distilled between 116" an3 193'. Only one-half came over below 2(?4'. In the second iectification, a large fraction was obtained between 40' and 50°, then scarcely any distilled until between 170" and 180", which latter fraction was very large. At 171°, the thermometer was often perfectly steady while an ounce distilled over, the entire contents of the retort not being much more than one and a half ounce. The lower fraction oscillated between 36" and 44", the variability in its boiling-point being principally due to the presence of oxygen; for, by careful and repeated rectifications over sodium, it may be made to distil almost entirely between 37' and 38'. This fluid lies at the boundary line betweon gases and liquids; it is scarcely possible to wet the hand with it ; and if poured from a moderate height in small drops, it is volatilized before reaching the floor. A very good way of obscrtrng the influence of small pressurea Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. upon its physical condition, consists in passing a few drops into a mercurial eudiometer stanaing vertically. The warmth of the hand applied to the top of the instrument is sufikient to convert it into vapour; but if the position of the eudiometer be then altered so as to stand almost horizontally, the increase of pressure is sufficient to condense the gas into a liquid, and, on the other hand, it is instantly vaporized again on restoring the eudiometer to its original position. This hydrocarbon possesses a property which, even mom than its extreme volatility, renders it troublesome to the analyst. Unlike very volatile fluids in general, it lines with a slight film of carbon the bulbs in which it is contained during combustion. It is true that the amount thus left is very small, but it'is sufficient to render an exact determination of the carbon difficult. It being View Article Online

114 GREVILLE WILLIAMS, ON ISOPRENE AND CAOUTCIIIN. of course necessary to have all the apparatus, oxide of copper, &c., aa cold as possible, it was also not easy to keep the hydrogen as low as could be desired. The substance used in the first three analyses boiled between 37" and 38", and was prepared from caoutchouc. The fourth an4 fifth analyses were made 011 a pro- duct from gutta percha. No. IV. boiled between 42" aud 43", and No. V. between 43" a,nd 44'. Experiment. Mean. --.Calculation. I.-- XI. 111. IV. v. Carbon . 88.3 87;8 87-3 87'8 83-1 88.0 Cl0 60 88.2 Hydragen 12.0 12-0 12.1 12.2 12.3 12.1 IIy --8 11-8 68 100.0 leading to the empirical relation

Although the boiling-point indicated the value ofn to be not greater than 2, it is plain that the liquid might possibly have a higher formula. The atomic weight was therefore determined from the density of the vapour, assuming it to possess the usual condensation to four volumes.

Mean of experiments. Theory, Cl0H8= 4 volumes. 2:44 2.3493 I have given the substance thus examined the name of isoprene. The specific gravity of the liquid at 20" C. was found to be Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. 0.6823. It will be seen that none of the hydrocarbons examined by Gregory, Himly, or M. Bouchardat can be identified with that just described; and albhough'the formulae C,H, and C,,H, have often been vaguely applied to the numerous isomers of tur- pentine, yet none of the hydrocarbons obtairied- by the above chemists have been shown to contain C;,H, for four volumes of uapour. The hydrocarbon discovered by Couerbe* in the fluid obtained by the powerful compression of resin gas, although of the -me composition, differs too much in deusity (0-709)and toiling-point (50") to be considered aa identical.

Ann. Ch. Phya lxix., 184. J. pr. Ch. tviii., 165. Qmelin'e Handbook, x., 411, Carendish Society's Tranalation. View Article Online

QREVILLE WILLIAMS, Oh' IBOPRENE AND CAOUTCHIN. 115

Action of Atmospiieric Oxyyen on Isoprene.

If isoprene be left in a partly filled bottle for Borne months, it gradually loses its fluidity, and at last even becomes quite viscid; at the same time it is found to have acquired powerful bleaching properties. It readily decolorizes sulphate of indigo, and, under certain circumstances, converts sulphide of lead into sulphate. It is, in fact, ozonized. Schonbein was the first to show that oil of turpentine absorbed ozone when kept for a long time ; and I subsequently observed several cases where ozone was capable of existing in contact with organic and inorganic matters without exerting any material action on them so long as the temperature was kept down. In these instances, immediately thc heat rose beyond a certain point, the siibstance in contact with the ozone underwent oxidation. Thus, if a slip of lead test-paper be exposed to the vapours of sulphide of hydrogen until the paper has acquired a deep brown colour, and a drop of an ozonized oil be placed 011 it, no change ensues at ordi- nary temperatures; but if the paper moistened -with the oil be warmed until the latter begins to volatilize, the sulphide iN imme- diately oxidized and blcachcd. But the most remarkable pltenomenon perhaps yet known in connection with the ozonization of essential oils, takes place on submitting ozonized isoprene to distillation. If the temperature be kept as low as is compatible with the volatilization of the unaltered portion of the oil, a colourless limpid fluid distils over, Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. having the composition and boiling-point of pure isoprene. As the operation proceeds, the fluid thickens, causing the temperature to rise somewhat suddenly. The oz'one at this point instantly begins to act with energy, a cloudy vapour rises, accompanied by an intensely sharp odour, and the contents of the retort instantly solidify to a pure white, spongy, elastic mass, having, when suc- cessfully prepared, but slight tendency to adhere to the fingers. When pure, it is opaque; but if allowed to become exposed to the air, especially when warm, it becomes transparent, first on the edges, and subsequently throughout the whole mass. When burnt, it exhales the peculiar odour liitherto considered character- istic of caoutchouc itself. It is not easy to prepare or to preserve this substance of definite composition. If the heat be allowed to rive too rapidly during its formation, it becomes decomposed; and View Article Online

116 GREVILLE WILLIAMS, ON ISOPRENE AND CAOUTCHIN.

if not heated sufficiently, ct portion of the original hydrocarbon obstinately adheres, in spite of its extreme volatility when separate. However prepared, its composition is that of irzoprene plus oxygen. If sufficient care be taken, the oxidation is perfectly definite. A specimen prepared with every precaution, and immediately ana- lysed, afforded, on combustion with oxide of copper and oxygcn gas, numbers agreeing with the formula C,,H,O. As regards the true formula of this substance, thcre is reason to suppose that tlie expression Cl,H80 does not correspond to four volumes of vapoiir ; but as it is not volatilc without decom- position, and as its yropcrties forbid any hope of its yielding com- pounds from which the true equivalent might he deduced, I ain at present compellcd to rcst satisfied with obtaining the simplest arithmetical relation tietwceii the elements. The above-described substance is, it is believed, unique in its characters, whether we rcgRrd its composition as the directly formed oxide of a hydrocarbon, or the mode of its production by the influence of ozone.

Cadchin. The forinnla C,o11,6 derived by M. Himly from his experiments upon the oil boiling bct\\-een 171" aid 173" is lmfcctly correct. My analyses confirm his results as regards Both thc pr-cciitagc corn- position arid the rapour-density ; the latter is, hy my cxperiments, 4.647 ; by Himly's, 44.61 ; by cnlcnlation ( t. vol.), it is 4.W36. Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. On the Relatior$ between the Jiwt aird second f3ydrocur,.bons produced by distillation of Caozrlchouc. We are now, for the first tinic, in a position to uiidcrstand the true connection betwceii thc two hydrocarbons which haw bcen described. The relation between them is tlie same as that bctwccn amylene and paramylene, and, as in tlic case of those bodies, the second is twice as hea\-y in the state of ~apouras thc first. The boiling-point of the bodies show very strongly thlirl~ited power of empirical formulq even when (like tliosc of Kopp and Gerhardt) fouiided on the accurate observation of it vast, iiumbcr of bodies. Thus, aiiiylcne C20H20,contaiiiiug four atoms of hydr 3geu more than caoutchin, should boil, according to Ger- hard t's law, 15 >: 2=30" lower than that body, whereas tlic Id- in;; poilit is only 11" lower. Again, amplenc, contitinine two View Article Online

GKEVlLLE WILLIAMB, ON I80PRENE AND CAOUTCHIN. 117 iltoms of hydrogen morc tlim isoprene, should boil 15" lower, whereas it boils 2" higher. The reduplication of a formula has not the same relative influ- ence upon the boiling-point of amylene that it has upon that of isoprene ; thus-

l'arilmglcne, CmH?O, boils J 160" Caoutchin, C,,,H,,, boils at 171' Amylcne, CICH boile at 39" Isoprcnc, ClaHR, boils at --37" Differencc . . . -I121 Dithcncc . . 134 Oil of turpentine boils at lGOO, and is taken as the pivot round which the boiling-points of hydrocarbons should arrange them- selves at distances proportiorred to the riurnbcr of atoms of carhon arid hydropi in thcir formulx ; yet caoutchiii, wliich riot only has the saiiic formulib, but gives alniost identical reactions, boils at 171", elcwxi dcgrecs liiglier, aiid very near that of carvene, which is 173".

Cornparatice experiments on the Action of Bromine on Caoutclh and Oil of Turpentine. 3%.Himly had observed that cliloriiic and bromine act upon caoutcliiii, iuid that the products when distilled over a he,afforded axi oil eo~it:tiriitig less hydrogen than caoutchin. He did riot examine tlic nature of the reaction more closely. I rneiitionctl, at the commenccmeiit of' this papcr, that I had several years ago rnadc an cxpcriment on thc mutual action of Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. bromine and caoutcliiii, which led me to believe the latter to bclong, not to the olcfiatits, but to the vast group of hydrocarbons isomeric with oil of turpentine. But that experiment having been madc with the l~rominearid hydrocarbon undiluted, and, more- over, tlic caoutchiri not hving been preparcd by mysclf, I con- sidercd it iicccssary to rcpcat it with several prccautions not theii obscrved. The action of.l)rominc upon the tcrcbenes is peculiar, one atom, or four-voltime cquivalent, always rcqiiiring csnctly four atoms of bromiiic to Iwodiice a colourlcss fluid. If the Iiydrocarbon and the bromine he tlilutcd, the forincr with alcohol, the latter with water, tlic cxperirnent may IIC made with considcrablc precision. &"or the 1)iwpose of comparing the isomers, I prepared some freshly dried and rectified turpentine ant1 caoutchin. They were diluted View Article Online

118 GREVILLE WILLIAJIS, ON ISOI'ItENl3 ANI) CAOUTCUIN. with sufficiciit dcolic~lfor cach of the resulting liquids ,to contain canctly tcii pcr cent. of hydrocarbon. The bromine cinploycd was in tlic state of a weak solution in water, tweuty cubic centi- metres containing 0.2527 of 1)romine. A small flask contaiuiiig t\wnty cubic ccntiiiictrcs with tlic stoppcr in its place, was eni- ploycd cacli tiiiie to mcasurc the hrotiiiuc-\vnter, wliicli was trans- ferred to a large stoppered flask, cvcry prccaution being takctn to prcveiit loss. The diluted liydrocarboii was thcn gradually dropped into the bromiiic-water from a Scliiistcr's alkalimetcr corithinirig a weiglicd quautity. Consi&rablc agitation mas applied aftcr each addition. When thc bromine-wstcr was ncarly decolorizcd, thc liquid was added very cautiously, and an intcrd of about half minute allowed to elapse betwcen each addition. Wheu the fluid had bccoine perfcctly colourless, tlic alkalimctcr was rcwiglicd. The following numbers were obtained in eight cxpcriniciits niade in tlrc inaiiiier described :-

Mean of turpentine experimcnts. Mcan of cnoutchiri crperiniciits. 0.10i4 0-1091 The identity of behaviour of caoutcliin and oil of turpcntiiie bccomes very evident from the ahovc experiments. Tlie mean of the caoutclrin results is ncwrtliclcss somewhat liiglicr tliaii that of tlie turpcritinc series; I ascribe this to cnoutcliiii being, to a very slight dcgree, morc sluggisli iii its actioii than oil of turpentine. Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. I was exceedingly desirous. of making espcrimeuts with weighed quantities of isopreiic aid hromiiie, in order to coinparc it with caoutchin. With this iiiteiit I niadc it solution of it in alcohol, the strength being tlie same its tlie turpciitiiie and caoutcliiti solu- tions. But it was iinpos6ible to obtniii the wished-for results, owing to tlie maiincr in wliicli tlie isoprene volatilized from the solution while pouring it into the bromiiie-water. Isoprene combines explosively with bromine, even in presencc of seventy or eighty tinies its yolunie of water. When the vessel containing the bromine at the bottom and tlic isoprcnc on.the surface is shaken, union talies place with a loud noise and con.. siderable evolution of heat. A fine mobile bromine-compound is the result. Whcri cohobated with cxcess of hydratc of potash, the View Article Online

GREVI1,LE WILLIAMS, ON ISOPRENE AND CAOUTCHIN. 119 greater portion is entirely decornposcd, yielding a black mass, the distillate having the same odour as that evolved under similar circumstances by the higher olefiants, but accompanied by a sub- stance so irritating to the eyes as to render it very distressing to manipulate. On placing about one ounce in a flask, and adding a fragment of hydrate of potash to neutralize it, an explosion took place, the fluid being thrown in my face, and causing the most intense pain in the eyes. For several days no person could enter the laboratory without incurring great pain, accompanied by profuse lachrymation. The substance liaving in this unfortunate manner been lost, I am compelled to defer a more minute investigation of it.

Conversion oj Caoutchin into Cymole and Puracymole. In Gerhardt’s classification of organic bodies, oil of turpentine is made to belong to the cymenic series; and it is even stated that Deville, by passing turpentine and carboriic acid through a tube heated to dull redness, obtained an oil which appeared to contain C,,H,, ; carbonic oxide, water, and empyreumatic pro- ducts being formed simultaneously. Disregarding the emp.vreu - matic bodies, the followiDg equation may be considered to explain the nature of the reaction :- C2,H16 + C20, = C2,H,, + 2110 + 2CO w L7-J + +’ c-Y--J 4 vola oil of 4 vols. carbonic 4 vole. 4 vols. 4 vo18. carbonic turpentine: acid. cymole. watrr. oxide.

Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. Some experiments made by me a considerable time ago with oil of turpentine, led me to conceive the idea of converting it into cymole, or an isomer, by the alternate action of bromine aud a highly electro-positive metal, and the identity in certain cases of the reactions of oil of turpentine and caoutchin rendered it pro- bable that a similar product would be yielded with the latter. Experiment has completely confirmed this supposition. If to a quantity of bromine and water in a stoppered flask ctlout- chin be gradually added, and the flask be shaken after each addition, a sharp hissing sound is heard, and combination takes place with great energy and considerable rise of temperature; at the same time a heavy oil, red at first but finally colourless, sinks to the bottom. The product when cold is as viscid as treacle. Its characters do not offer any great guarantee for purity; and the production of hydrobromic acid at the same time indicates View Article Online

120 GREVILLE WILLIAMS, ON ISOPRENE AND CAOUTCHIN. tht while one portion of the halogen combines with the caout- chin, another seizes the most easily removable of its atoms of hydrogen, and yields the acid which is found dissolved in the water. I examined the reaction which takes place with turpentine uuder these circum3tances some years ago, but at that time I was not fully alive to the true explanation, because I had not treated the resulting oil with alcoholic potash or sodium. .The analyses I then made indicated the formula of the oil from turpentine to be C,H,,Br. 2Br. 2H0. One cquivalent of liydrobrornic acid was always found in the watery fluid. Assuming this to be correct, as I have no doubt it is, we have

C2,H,6+4Br+2EIO=C,,H,,Br. 2Br. 2HO+HBr.

Every four equivalents of bromine yield, therefore, one eqniva- lent of hydrobromic acid ; and in an experiment mhcre 33.2 parts of bromine were employed, 5.2 of hylrobromic acid were fouiid ia the aqueous fluid. If thc above cquation be correct, 8.1 should have been obtained. The oil produced as ahovc froin caoutchin yields, when hcntcd, torrents of hydrobromic acid, and mi oil distils over containing less bromine. and hydrogen. On coliob7ting it with sodium in excess, a very fragrant niid perfcct ly colonrlcss hydrocarlxm mas produced, distilling betwccn 171" and 2000. It was received in four portions : a, betwccm 171' and 177" ; b, 177" to 181"; c, 181" to 186"; and d, 186" to 200". A yellowish oil remained behind Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. at this point ; it requircd a very inuch higlicr temperature for clis- tillstion, and on cooliiig deposited a few very minute crystals. Caoutchin arid turpentine behave in a precisely similar manner with the above reagents; and it is worthy of observation that the resulting hydrocarbon has the same peculiar odour, from which- ever 'source it is prepared.* Adpsis yielded numbers agreeing completely with the formula C20H14,which 'is that of cpmole. Moreover, the density of the liquid, 0.8664 at ZOO, was almost the same as that given by Gerhardt, namely, 0.8610 at 14'.

* Since writing thc above, I have compared the cymole from caoutchin, ancttllat from oil of torpentine, with a specimen of the same hydrocarbon, prepared with great care from oil of cumin by my friend Mr. Church ; the odour in cach case was absolutely identical. View Article Online

OREVILLE WILLIAMS ON ISOPRENE AND CAOCTCHIN. 121 I could have wished to give equations expressive of the entire decompositions which take place, and it would be easy to do ao were I to diurcgard, or consider as mercly accidental, the sub- stances formed at the same time. When the brominized oil is distilled, before or after treatment with alcoholic potash, nearly fifty per cent. of loss is incurred. The retort contains a black carbonaceous mass, occupying two-thirds of its capacity. A similar phenomenon occurs in the treatment with sodium; and at the same time, as already observed, a certain portion of caoutchin or turpentine, as the case may be, is regenerated. The yellow oil of high boiling-point previously alluded to exists in very small quantity among the products formed by the action of sodium on brominized caoutchin; probably not more than one- tenth of the distillate boils above 200' C., whereas the yellow oil boils only a little below 3Oo0 C. The quantity in my possession was far too small to allow of a determination of this point. Iri composition, homcver, it agreed with that of cymole. The crystalline substance was in exceedingly minute quantity. It is far from impossible that it may be the radical thymyl formed from its hydruret cymole by the action of the bropine. As it would have the composition C40H26=2C20H,Yths probability is that it mould be it crystalline solid. The cxcccdingly high boiling-point of the yellow oil is sucn strong evidence of its being a polymeric condition of cymole, that I liavc provisionally named it purucymoie.

Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. Coizrcrsion of the Cyniole from Turpentine and Caoutchin id0 Irrsolinic Acid. Altliorigli the cspcrimcnts detailed left little doiibt of tlic itlcritity of the cymole from turpentine and caoutchin with that from cumin oil, still the number of isomeric hodies coritiuiially Lcirig discovered is so great, that it was highly dcsirahlc to nialie a crucial esperimeiit. Thc exceedingly cha- ractcr istic propcrtics of isolinic acid rendercd it peculiarly suit- able in tlic present instance. Dr. IIofmann has shown it to be a product of thc ositlntion of cumiuic acid and cymole under the influcricc of chromic acid. 'Plie quantity of cymole in my posses- sion was so small (allout two draclims), that there mould have bccn little hope of success had the reaction been less definite. On coliobntilig it with 8 parts of sulphuric acid, 8 pts. of bi- View Article Online

122 QICEVILLE WILLIAYB ON ISOPRENE AND CAOUTCIIIN. chromate of potash, and 12 pts. of water, for a few minutes, white flocks Iiccarnc apparmit ; thcy contiriually iucrcascd until their bulk cquallcd that of the cymole employed. The crude acid was dissolved in ammonia, and, after filtration, precipitated by hydro- chloric acid. It was washed, first with warm water, then with boiling alcohol, and fiually with ether. Thus purified it was quite colourless. It was then dissolved in ammonia, and the solution evaporated until crystals of the acid salt began to'form. The solution, rcndercd faintly alkaline by the cautious addition of ammonia, was prccipitated by nitrate of silver, carc being taken to keep it slightly alkalinc, to prcvcnt the separation of any insolinic acid. The salt, after drying at 100" until it ceased to lose weight, was ignited, during which operation it evolvcd the peculiar aro- matic odour so cliaractcristic of all the imolinaies when heated strongly. In this mariner I prepared thrce specimens, one from turpcntine, another from caoutchouc, and a third from gutta percha. All three gave, by analysis, results corresponding SUE- ciently with the formula C,,(~~,AS,:0,.

Action of Sulphuric Acid on Caoulchin. 'On slowly adding caoutchin to sulphuric acid (sp. gr. 1.845) in great cxccss, the latter becomes warm, and the greater part of thc hydrocarbon dissolves, traces of sulphurous acid bcing at the same time cvolvcd. On the addition of matcr, the greater part of the caoutchin rises in an altcrcd state to the surfacc; it is in the Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. form of a thick acllicsirc fluid of the consistencc of molasses. On saturating thc acid solution ivitli chalk, a very small quantity of thc liinc-salt of a corijugatc sulpliuric acid was obtained. It is csccssivcly soluble in matcr, arid dcposits small granular crystals from a conceritratcd soluti-on. Dried at 200" C. it gavc, by ana- lysis, results agrecing witli thc formula C2,€I1,CaS2O6.

012 the Conposttion of Cuoutchouc. The only annlyscs of caoutchouc wliicli I have scen are those of Faraday aiid tlic lstc Ill*. Ure. That of the latter clicmiet is obviously iiicorrcct, as hc obtaiiicd IfO pcr ccrit. of carhori aiid only 9.1 of liytlrogcn, a rcsult iricornpatiblc witli tlic phenomena which take placc uilcr thc iufluencc of hat. View Article Online

GBEVILLE WILLIAMS ON ISOPRENE AND CAOUTCRTN. 123 Faraday's analysis, made in the year 1826, probably exhibits almost exactly the truc composition of pure caoutchouc. It gave carbon 87.2 and hytlrogen 12.8 per cent. 'l'lics(: rititii1)crs woul(1 require to be slightly modified, in coiiscquc'uc(: of thc atomic weight of carbon, 11s now received, having a somewhat difl'erent value to that which was admitted at the time of theanalysis. I analysed two small specimens (probably less pure than that exa- mined by Faraday), which had formcd in phials of the juicc, with the anncxcd results. Thcp wcrc dried at 100". The first speci-. men mas dark brown, thc second pale straw-colourcd and almost transparent. The first gave 86.1 p. c. C, 12.3 11, 0.9 ash, and 0.7 nitrogen, oxygen, and loss; the second 87.2 C, and 12.8 €1. I am anxious to call attention to the fact that the atomic con- stitution of caoutchouc appears to bcar some simple rclation to the hydrocarbons resulting from its decomposition by heat. The composition of caoutchouc coincides with that of isoprene and caoutchin, as found by analysis, to a degree which is remarkablc, wlicn we consider that caoutchouc, in addition to bcing non-crystal- line, is scarcely capable of purification by chemical means. Atltli- tional evidence may be found in thc smallness of thc residue which is left on submitting caoutchouc to heat. The folloming.rcsulh of the analyses will show the amount of confidence which may he placed in this idea :-

Isoprene. Caoutchin. Caou t c h o u c.

Mcan. Mean. Faraday. @. (f. W.. Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. Carbon . . 88.0 813.1 87.2 86-9 87.3 Hydrogen , 12.1 12.0 12.8 12.4 12.1 The following table contains a summary of the physical pro- pcrties of isoprene and caoutchin. Table of the Physical Properties of Isoprene and Caoutcliin.

Vapour-density. Name. Formula. Boilingpoint. Specific gravity. Experiment. Calculation. -- --- 1 Isoprene I c,,,n3 37" C. 0-6823at 20" C. 2-349 Caoutchin 1 cl0irlG 171 0.8420 2444.65 1 4699

Ash deducted. View Article Online

124 GREVILLE WILLTAMS ON ISOPRENE AND CAOUTCIIIN.

It will be seen from what has been Eaid upon the composition of the above hodirs, that thcy may eq~dlybe obtained from caoutchouc or gutta pcrcha. IVhcn the latter substance is sub- mitted to careful distillation, tlic phenomena are csseirtially the same as with the former. The distillate contains a small quantity of water, which, instead of being alkalinc, as with caoutchouc, is powerfully acid. The acid is volatile, and appears to belong to one of the lower mcmbers of thc series Cn€In04. On neutralizing the acid liquid with potash or soda, the odour of a volatile base becomes very perceptible. The oily distillate, as with caoutchouc, consists principally of the thick uninviting fluid called by Boucliardat, h6v&ne. I have not as yet minutely examincd it, but believe it to bear a poly- meric relation to caoutchin. The amount of isoprcnc in the crude distillate is very small, probably not more than five pcr ccnt., and, as will readily be sup- posed with so volatile a fluid, much is lost in the subsequent purification. It is this circumstance which has prevented me from examining it so closely as could be desired. The caoutchip constitutes about twenty per cent. of the distil- late; it was seusibly less pure than that procured from caout- chouc, and more rectifications over sodium were necessary before it was obtained in a sufficiently pure state for analysis. This partly arises from the tenacious manner in which the volatile acids above alluded to adhere to it, and render treatment with alkalies essential previous to rectification. It is also accompanied by what we are too much accustomed to vaguely term empyreumatic pro- Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17. ducts. These are of readily oxidizable nature, and may be got rid of by rectification two or three times over hydrate of pot ash . The following is a brief summary of the results of the above investigation :- 1. ‘l’be isolation of isoprene, CIOHB,the cxistence of which among the produc+ of the destructive distiilation of caoutchouc had not been proved. 2. The production of an apparently definite oxide of the hydro- carbon CIo€Is, by the agency of ozone spontaneously generated. 3. The determination of the polymeric relation between isoprene and caoutchin. 4. The conversion of turpentine and caoutchin into the hydriiret of thymyl or cymole. View Article Online

FIELD ON THE SULPEJIDES OF COPPER AND IRON. 125 5. The production of paracymole. 6. The determination of the products of tile destructive distil- lation of gutta percha. Published on 01 January 1862. Downloaded by University of Chicago 24/10/2014 23:36:17.