153 X.-On the Butyl Compounds derived from the Butylic Alcohol of Fermentation. By ERNESTT. CHAPMANand MILESH. SMITH. (Read March 18th 1869.) C~RCUMSTANCES placed at our disposal a large quantity of fiisel oil,* from which a portion of the amylic alcohol had been removed. The crude oil just as it came from the distillers had been distilled and everything which came over in one dis- tillation below 1.27' collected apart. It was this portion boiling below 127' that we operated upon. It had been standing for a good many years in badly-corked jars and was conse-quently yellow and dirty. It was saturated with water. We operated on seventeen gallons of this material as follows -It was first simply distilled those portions which came over below 115' and those from 115' to 131' being col-lected apart.Everything which came over above 131' was also separately collected. That portion which came over at 115°-1310 was repeatedly distilled the distillation being stopped when the thermometer rose to 131' and that which remained behind being added to that which boiled above 131O. After this process had been repeated about ten times the greater portion of the liquid boiled between 106' and 115O or above 131'. That portion which boiled between 115' and 131' was now fractionally distilled in the ordinary manner. It almost entirely split up into liquids boiling above 13G' and liquids boiling below 115'. That fraction which oliginally boiled below 115' was now repeatedly distilled the distillate being always received in vessels containing fi-eshly-ignited carbonate of potash.This liqnid gradually split up into three fractions the one boiling about 130°,another boiling between 106O and 115' and the third boilingfrorn 79' up to 1013'. This latter fraction was gradually split up into bodies boiling above 106O and a considerable volume boiling below 80'; but large quan- tities remained which no amount of fractional distillation appeared capable of splitting into definite bodiea. As will be * From London Distilleries. VOL XXII. N 154 CHAPMAN AND SMITH ON THE BUTYL COMPOUNDS shown on a future occasion this portion which proved so un-manageable is of special iiiterest. The portion boiling between 106O and 115Owas now frac-tionally distilled.We are afraid to say exactly how many times but the fractionation took up several weeks. Finally we obtained from it about five litres of a body boiling within some &ths of a degree. This substance consists almost entirely of butylic alcohol. It is however contaminated with small quantities of iso-butylic alcohol. From this impurity it is im- possible to separate it by fractional distilliLtion notwithstanding the difference in their boiling points. NoTE.-The fractional distillations were performed in an apparatus like that sketched below. The distilling vessel consisted of a tin can,* such as is employed for holding methylnted spirit or varnish of two gallons capa-city ; it could distil conveniently between six and seven litres ; (e-)-_---* For the information of foreign readers we should perhaps remark that in Eogliah a tin can implies a vessel constructed of thin sheet iron coated with tin.It is known in Germany as *‘btech.” A small mistake in a foreign translation of one of our papers is the cause of the insertion of this note. DERITED FROM BUTYLIC ALCOHOL OF FERNENTATION. 155 its neck was closed by a cork through which passed a wide glass tube bent twice at an obtuse angle. C is the tin can used as a distilling vessel; a a is the tube; c is the connection with the condenser and d the thermometer. Of course great condensation takes place in this long tube and by slow distillation we may be assured of condensirig at least twice as much liquid in the tube as we allow to distil.This condensation very greatly facilitates the separation of the liquids by fractional distillation. The tube employed in most of these operations is about five feet in length about six-eighthe of an inch in internal diameter at the end connected with the distilling vessel and rather narrower-say about five-eighths of an inch-at the other end. It is made of very thin glass. We prefer it to a series of bulbs as recommended by Wurtz because it exposes a greater amount of surface in proportion to the vapour it contains. Moreover it mare readily drains is more easily cleaned and is less fragile. The disad- vantage of its inordinate length is to a great extent got over by bending it in the manner illustrated. Butyl compounds may be prepared from the above mixture of constant boiling point.We 'nave prepared and examined in detail the iodide bromide nitrate nitrite acetate alcohol and mercury compound. Iodide of Butyl is a heavy liquid agreeing accurately with TVurt z's description of it. It cannot be satisfactorily prepared by treating the crude butylic alcohol with iodine and phosphorus ; as under these circumstmces large volumes of butylene are evolved and a very poor yield is obtained. By operating with hydriodic acid this difficulty is entirely obviated. If the crude alcohol be mixed with a large excess of hydriodic acid of specific gravity about 1*85-1*9 and the mixtiire be raised to the boiling point the alcohol is without sensible loss converted into iodide. To prepare the iodide the crude alcohol is boiled for half an hour or forty minutes with four times its volume of hydriodic acid of specific gravity about 1.8 or rather higher.At the end of this period the alcohol is completely converted into iodide which latter will be found to have separated completely from the acid. It ig to be separated off and run into a flask containing carbonate of soda and water. The mixture of crude iodide and carbonate of soda solution is now distilled and the clear colour- N2 156 CHAPMAN AND SMITH ON THE BUTYL COMPOUXDS less iodide separated. It is then dried with chloride of calcium and fractionally didled; but by far the greater part of it is found to boil constantly at 121O. Small quantities of liquids of lower boiling point separate out.We prepared altogether about 1,300 grammes of the iodide. The fractional diBtillation was carried on in an apparatus like that previously described but consisting entirely of glass. The same remark applies to all other fractional distillations mentioned throughont t'his paper. Iodide of butyl is a clear colourless mobile liquid; it boils quite constantly at 121'. Its sp. gr. is 1,6301 at Oo 1.6032 at 16O 1.54816 at 50'; from which it follows that 10,000volumes at. 0' become 10,168.at 16O and 10,529 at 50'. The following decompositions of the iodide were observed :-With alcoholic solution of potash or ethylate of soda it is for the most part resolved into butylene and iodide of potassium or sodium a comparatively small portion only being converted into ethyl-bntyl ether.With alcoholic ammonia it is converted into butylaminea little or no butylene being produced. Heated with acetate of potash and glacial acetic acid it is converted into acetate of butyl; a certain amount of butylene is formed at the same time. Heated with bichloride of mercury it yields iodide of mer- cury aid chloride of butyl together with traces of hydrochloric acid and butylene. On heating it with zinc and ether large volumes of gas are evolved and a small quantity of zinc-butyl which however cannot be distilled out of the mixtnre. With dilute sodium-amalgam and acetic ether it yields mercury-butyl not, however without considerable evolution of gas. With cyanide of potassium it yields cyanide of butyl with tolerable facility and without the liberation of much butylene.With sodium it yields much gas consisting in part of buty- lene and in part of a gas which is in all probability hydride of butyl. The radical butyl is produced at the mme time but not apparently in any large quantity. As will be obvious from the foregoing this iodide possesses in a high degree the peculiar property of splitting up into hydriodic acid and olefine. Bromide of BxtyZ cannot be conveniently obtained by the DERIVED FROM BUTYLIC ALCOHOL OF FERMENTATION. 157 action of phosphorus and bromine on the alcohol ; partly because much butylene is eT-olved and partly because bromous substitution takes place to a very great extent. It is best obtained by saturating the alcohol with gaseous hydrobromic acid of which it readily absorbs rather less than ita own weight.This saturated liquid is to be mixed with its own volume of aqueous hydrobromic acid of sp. gr. about 1.6 or rather less ;the mixture is then to be heated in closed vessels up to the boilipg point of water.* The heating must be continued until the oily layer which separates out at the top no longer increases. When this is the case the digestion vessel must be removed from the hot water allowed to cool and opened. The bromide is then separated from the hydrobromic acid by means of the separating funnel. Should it contain phosphorus as it frequently does (hydro-bromic acid frequently having traces of phosphorus adhering to it) it is advisable to wash it with hydrobromic acid containing a little free bromine.So soon as it is slightly tinged with bromine it is run into a flask containing a dilute solution of carbonate of soda and distilled along with that liquid. The distillate is neutral and perfectly colourless ; the oily liquid is separated off from the small quantity of water which accom- panies it and dried over chloride of calcium. It is now frac- tionally distilled. By far the larger portion boils at 92* C. but another poi-tion which has apparently the composition of bromide of butyl boils 16O or 17" lower. Both liquids distil perfectly without decomposition and may be separated by fractional distillation though the process is somewhat tedious. Bromide of butyl is a colourless liquid of sp.gr. 1.2702 at 16'. Q We employed soda-water bottles closing them with corks nhich had been heated in melted spermaceti for half an hour before being used. Corks that have undergone this treatment are readily forced into the necks of the bottles though beforehand it was impossible to force them in at all. The corks were covered with little tin plates used by Messrs. Sandford and Blake in securing their soda-water bottles ;the corks were wired into the bottles the tin plates preventing the cork being cut by the wire. The soda-water bottles hold when full up to the neck between 2'70and 300 c.c. they will therefore take a charge of 200 C.C. conveniently. When charged and wired down they were placed in tepid water which was very gradually raised to boiling point.If the temperature be raised too rapidly the pressure becomes excessive and there is great danger of the bottles bursting. By raising the temperature gradually this danger appears to be entirely averted. At least in charging 15 soda-water bottles as above described not the slightest accident nor even the slightest leakage occurred. 158 CHAPNAN AND SMITH ON THE BUTYL comoms In its reactions it most closely resembles the iodide giVing of€ butylene in almost all decompositions. Its reaction with acetate of potash is dew and requires a tolerably high temperature. At 130° it takes many hours to complete the reaction but a little above this point it goes more rapidly. With sodium amalgam and acetic ether it behaves exactly as the iodide behaves.With ammonia the same remark applies. With cyanide of potassium decomposition is perhaps a trifle slower than with the iodide. With zinc its behaviour is exactly similar to that of the iodide. It does not appear to react with chloride of mercury. At any rate decomposition is very slow. With alcoholic potash it behaves exactly as the iodide does yielding butylene. We made about 1,600 grammes of the pure bromide boiling at 92OCy. A7itrate of ButyZ.-Nitrate of butyl is a colourless liquid of ~p.gr. 1-0384at Oo and 1.020 at 16'; it closely resembles the nitrate of amyl both in its stability and in its odour which resembles that of bugs-it boils at 123'. It has the most extremely disagreeable physiological action when in-haled causing restlessnese and most severe headache.It is prepared by a process strictly analogous to that by which nitrate of amyl was prepared by us. A mixture is made of two volumes of concentrated sulphuric acid and one of nitric acid of about 1.4 sp. gr. and 100 C.C. of this mixture are placed in a beaker surrounded by salt wat,er and ice. About 30 C.C. of the alcohol are now added fkom a small dropping funnel the stem of which is drawn off to a fine point and passes beneath the surface of the mixed acids; it is employed to stir the mixture as well as to add the alcohol. The alcohol must be added dowly and the mixture kept constantply stirred ; the nitrate rises to the surface as a clear colourless oil; it is decanted by means of the separating funnel and run into a retort containing excess of solution of carbonate of soda the process being repeated until a sufficient quantity of the nitrate has been prepared ;then the contents of the retort are distilled.The nitrate passea over quite unaltered along with the vapour of water. It is simply separated from the water and dried with chloride of calcium. If pure butylic alcohol has been wed DERNED FROM BUTYLIC ALCOHOL OF FERMEWTATION. 159 in its preparation it is now chemically pure; if not it must be fractionallj distilled when but 1it.tle difficulty ~vill be found in obtaining it of constant boiling point. Nitrate of butyl is unattacked in the cold by concentrated sulphuric acid; very strong aqueous potash has little or no action upon it ; alcoholic potash to a great extent resinifies it but at the same time a small quantity of ethyl-butyl ether appears to be formed :no trace of lnutylene is liberated and the action is exceedingly sluggish.The vaiioiis reactions which we have described in speaking of the nitrate of aniyl may all be repeated with the nitrate of butyl. It is converted into iodide of butyl with great facility when digested with strong hydlriodic acid binoxide of nitrogen and free iodide being liberated. Nitrite of Butt$.-Nitrite of butyl is obtained by passing nitrous acid into butylic alcohol; it is very desirable that the nitrous acid should be as free as possible from nitric acid. This may to a great extent be effected by never heating the mixture of arsenious acid and nitric acid excepting in the water-bat'h.Under these circumstances the nitrous acid comes over nearly pure and dry. The butylic alcohol must be kept cool by surrounding it with cold water and it is advisable to pass the nitrous acid into it slowly. When it is saturated the passage of nitrous acid must be discontinued and the nitrite thoroughly washed with water with dilute caustic potadi and then again with water. It will now present a brilliant bluish- peen colour of which it cannot be deprived by washing ;it may be dried over chloride of calciiim and fractionally distilled ;the distillate will be yellow. Great volumes of binoxide of nitrogen are evolved during the first portion of the distillation.Nitrite of butyl is a yellow light and very mobile liquid; its boiling point is difficult to determine with accuracy as it appears to undergo a slight decomposition. However its boiling point may be taken at 67O; at least on distilling 120 grammes of it it began to boil at a little over 65'; by the time 2 or 3 grammes had passed over it had arisen to 664O between which temperature and 67&O almost the whole of the remaining liquid distilled a very few grammes of liquid remained in the retort by the time the temperature had reached this point. This liquid is almost insoluble in water and ab- sorbs only the smallest traces of that liquid. It does not dis- solve chloride of calcium Its specific gravity is 089445at 0'; 160 CHAPMAN AND SMITH ON THE BUTYL COMPOUNDS -8771at 16'; and $25638 at 50'; therefore 10,000 volumes at Oo become 10,198 at 16' and 10,833 at 50'.It has the same physiological action that &rite of amyl has though the action is more intense. Its smell is not quite so disagreeable. In almost all its decompositions it closely resembles the before-mentioned nitrite of amyl-thus sodium liberates nitrogen from it ;slightly dilute sulphuric acid transforms it into butyrate of butyl With evolution of binoxide of nitrogen and sulphurous acid. Hydriodic acid converts it into iodide of butyl with liberation of binoxide of nitrogen and iodine. Ethylate of soda converts it in part into ethyl-butyl ether. The yield of nitrite of' butyl is not very good unless very great care be taken in its preparation as the butylic alcohol is easily oxidized into butyric aldehyd and butyiic acid which latter combines with a portion of the butylic alcohol to form butyrate of butyl.Of course the butyric aldehyd is removed during the mashing of the niti-ite and the butpate of butyl during the fiactional distillation ; 100 parts of butylic alcohol may however be made to yield from 105 to 110 of the nitrite. L4cetate of ButyZ is prepared by mixing crude butylic alcohol with glacial acetic acid saturating the mixture with hydro- chloric acid warming in the water-bath and then washing with cold water. The great bulk of the acetate separates at once; a small quantity however remains in the washings which are therefore distilled ; the distillate is treated with carbonate of potash; and the oily layer separated from the aqueous one.The oily layer is again treated with glacial acetic acid and hydrochloric acid and washed as before. The acetate so recovered is about 10 per cent. of the whole preparation. The mixture is now carefully dried fimt by agitation with carbonate of potash and subsequently by long standing over a new portion of the freshly ignited carbonate. It is now carefully fractionally distilled. Notwithstanding the drying the first portions of the distillate are always a little wet ; they are there- fore received in a flask containing more ignited carbonate. Other compounds of lower boiling point than the acetate separate fisom it during the distillation.Acetate of butyl boils at 117.5. Its sp. gr.is 089096at O' -8747 at 16' and 083143at 50'. Therefore 10,000 volumes at 00 become 10,186 at 16' and 10,716 at 50'. The smell of the acetate is fragrant only DERIVED FROM BUTYLIC ALCOHOL OF FERMENTATION. 161 distantly resembling that of acetate of amyl; it recalls the odour of quince. It is not easily decomposed by treatment with aqueous caustic potash. By sealing up however and so raising the temperature to 140' decomposition takes place very rapidly. With strong aqueous ammonia the ether is slowly decomposed into acetamide and butylic alcohol. With alcoholic potash it is decomposed almost instantaneously. With solution of pbtash in butylic alcohol the decomposition is eqiially rapid. Sodium dissolves in the acetate absolutely without evolution of gas.ButyEic Alcohol.-The pure alcohol is most readily obtained by decomposing the pure acetate with caustic soda. The acetate is poured upon about half its weight of powdered caustic soda. After a few minutes the mixture gets hot and finally boils violently. It should now be cooled by immersing the vessel containing it in cold water. The mixture will now have become a semi-solid mass. Water is now to be added and the whole distilled from the oil-bath. Butylic alcohol ac- companied by acetate of butpl and water distil orer. The distillate is now saturated with carbonate of potash the oily portion decanted and treated as before with caustic soda. In this second operation the conversion into the alcohol is quite complete.The mixture is now treated with water and again distilled from the oil-bath ; the distillate treated with carbonate of potash as before then thoroughly dried by boiling with and allowing to cool over carbonate of potash. It is next treated with a large quantity of caustic lime over which it must either be allowed to stand for gome weeks or else it must be digested along with the lime at a temperature between 65" and 75" for 12 or 14 hours. On now distilling it off the lime in the oil-bath it will be found to be perfectly dry. It boils at 1084O at the normal pressure going quite to dryness below 109". Its specific gravity is $055 at 16O.8. Its smell is quite different from that of the alcohol with a trace of moisture in it.Butylic alcohol cannot be dried by treatment with sodium. It would appear that hydrated oxide of sodium is more or less decomposed by butylic alcohol water and butylate of sodium being the products. At any rate butylic alcohol dried as per* fectly as possible with carbonate of potash was not rendered anhydrous by treatment with 6 per cent. of sodium. This 162 CHAPMAN AND SMITH ON THE BUTYL COMPOUNDS alcohol could not have contained much over 1per cent. of water 60 that there was much more than sufficient sodium to form even anhydrous oxide of sodium with the oxygen of all the water present. When quite dry butylic alcohol will dissolve about half an equivalent of sodium though only with considerable difficulty and with the aid of much agitating and heating.Butylic alcohol readily dissolves chloride of calcium acetate of potash and caustic potash. It does not appear to dissolve notable quantities of the chlorides nitrates or sulpliates of the alkalies. It does not dissolve cyanide of potassium. It is readily mis-cible with glacial acetic acid also with glacial acetic acid to which twice its volume of water has been added It is very readily soluble in hydrochloric acid. It dissolves in about 11 volumes of water. If large volumes of the alcohol be distilled it is possible to obtain a portion of it nearly if not quite an- hydrous. The best plan of drying really large volumes of the alcohol appears to be to distil it repeatedly always treating the first portion of the distillate with carbonate of potash.The alcohol appears to form a definite hydrate which however cannot exist at the boiling point of water. Treatedwith iodide of phosphorus or iodine and phosphorus the alcohol is con-verted partially into butylenc and partially into iodide of butyl. We have never been able to obtain a conversion of' more than 60 per cent. of the alcohol into the iodide. With bromine and phosphorus a perfectly similar decomposition takes place. If the alcohol be cooled to -15" or -16" C. and then poured slowly into concentrated sulphui-ic acid kept at the same low temperature the alcohol appears to be totally or almost totally converted into butyl-sulphuric acid. If the mixture be effected at the ordinary temperature polymerised butylenc appears to be the piincipal product.If the sulphuiic acid be diluted with a third of its weight of water and the mixture gradually effected at the common temperature but$-sulphuric acid is the principal product. If butyl-sulphuric acid be diluted and gently warmed and excess of crystaked sulphate of soda added a colourless liquid of most extraordinary odour rises to the sur- face. This liquid is soluble in water in all proportions. It appears to be a hydrate of butylic alcohol. At any rate on distillation it splits up into water and butylic alcohol. Mercury-ButyZ.-Mercirry-~utyl is eaaily prepared by the DERIVED FROM BUTYLIC ALCOHOL OF FERMENTATION. 163 general method given by Frankland and Duppa for the pre-paration of the mercury-compounds of the alcohol radicala.Five granimes of sodium were dissolved in over 2,000 grammes of mercury and agitated with an equivalent quantity of iodide of butyl to which about one-tenth of its weight of acetic ether had been added. The process was carried on in a stout glass- stoppered bottle the stopper of which was removed and its place supplied by a bored cork carrying a long wide upright. glass tube; this served as a condenser. When the mixture in the bottle was vigorously shaken the mercury which at first remained fluid soon divided itself into an infinite number of minute globules which with the liquid and iodide of sodium formed a grey pasty mass in which after a while not a single globule of mercury could be discovered.On further agitating the mixture as the reaction slackened the mercury coalesced again into a liquid mass. Great heatwas given out during the reaction. When the bottle was nearly cold it was found that the mercury could be poured out in a perfectly clean state with- out the smallest difficulty the whole of tht mercury-butyl iodide of butyl &c. remaining in the form of a paste along tVit,h the iodide of sodium formed and a small quantity of mercury. The mercury was at once re-amalgamated poured into a clean bottle and again treated with iodide of buts1 and acetic ether a8 before. During the progress of the second operation the first bottle was washed out the washings being collected in a retort ; this bottle was nciw dried and was ready to receive the charge of mercury from the other bottle by the time the operation going on in it was completed.The process was continued in this way until a sufficient quantity of crude mercury-butyl had been obtained. The contents of the receiving retort consisted of two layera an aqueous layer and a layer of dense liquid at the bottom. Most of the water in the retort was removed and the contents of the retort were distilled from the oil-bath. Mercury-butyl passes over readily with the vapour of water. The distiIlate consisted of two very well-defined layers the lower being mercury-butyl contaminated with iodide of butyl and acetic ether. To fi*eeit from these two latter impurities it was placed in small retorts and a current of steam driven through it ; this was continued until no iodine could be detected in the distillate.The liquid in the retort was now separated from the water accompanying it and dried with chloride of calcium. 164 CHAPMAN AND SMITH ON THE BUTYL COMPOUNDS Mercury-butyl thus prepared is a colourless tramparent liquid of sp. gr. 1*7469 at O" and 1.7192 at 16". It cannot be distilled by itself but it will stand a temperature of 130" without undergoing much decomposition. It combines with iodine and bromine with characteristic violence. Boiled with hydrochloric acid it liberates a gas apparently hydride of butyl. It is attacked with comparative facility by zinc forming zinc-butyl; no great disengagement of gas takes place in the transference. It has the peculiar disagreeable yet perfume-like smell which appears to be characteristic of the whole of this class of bodies but we did not experience the slightest ill effects from working with it.Its smell is one the dislike to which is soon converted into absolute disgust. We are not acquainted with any sub- stance excepting these mercury-radicals the smell of which we cannot more or less accustom ourselves to; but though at first the smell does not strike one as being peculiarly disagree- able after a few days' work we seemed instinctively to shrink from it ; we have prepared all the known mercury-compounds of the alcohol radicals and have always found that they pro- duce this extreme disgust. We may here observe that we do not think that there is any great danger in working with these compounds now that their poisonous nature is 60 well esta- blished.The butyl spoken of in the above paper is isopropyl- rn e thy 1;the alcohol therefore would be in K o1 b e's nomen- clature isopropyl carbinol and is represented by the for- mula-(OH The alcohol yields on oxidation iso-butyric acid. We defer a detailed account of' its oxidation products to a future com- munication. The following is a tabular statement of the boiling-points and specific gravities of those butyl-compounds examined both by W iir t z and ourselves. With regard to the specific gravity of the iodide we have here inaerted a specific gravity observed at19" so that the humbers may be strictly comparable. It will be Been that whilst DERIVED FROM BUTYLIC ALCOHOL OR' FERMXNTATION.165 the iodide and alcohol agree closely as to boiling-point alI the other compounds differ notably. -4s we have made many pre- parations of all these compounds and have operated on a very large scale we think the numbers we give may be implicitly depended upon :-Boiling point Co. Specific gravity. Name of substance. c.& 5. Wurtz. c. & 5. I Iodide of butyl.. ...... Bromide of ditto.. .... 121 89 121 92 1.604 at 19 C. 1.274 at 16" C. 1 .59C3 at 19"C. I -2702 at 16OC. Nitrate of ditto ..,,,. Acetate of ditto ...... Butylic alcohol. ....... 130 114 109 123 117.5 ioa-5 heavier than water ,8845 at 16"C. *8032at 18*5" C 1 *030at 16"C. -8747 at 18" C. -804at 18PC. ~ APPENDIX Containing the evidence on which the above account of the butyl-compounds rests.A. The following is the proof on which we base our assertion that we are dealing with butyl-compounds First. A combustion of the alcohol. Burnt with chromate of lead. 03507 grammes of the alcohol yielded 4348 grammes of carbonic acid and 440 grammes of water from which the follow- iiig percentages of carbon and hydrogen are calculated :-Theory. C ................ 64-92 64.8'7 H, .............. 13.94 18-51 O................ 21.62 Second. A combustion of the iodide obtained fiom the alcohol was made. Burnt with chromate of lead and copper-turnings 0844 grammes of the iodide yielded -807grammes of carbonic acid and 03775 of water. These iodine determinations were made in three different samples of the iodide prepared at dif-ferent times.I. 1.257 grammes of the iodide yielded 1.6016 of iodide of silver. 11. -6904 of iodide yielded -8839 of iodide of silverd 111. *8830 of iodide yielded 1.1253 of iodide of silver I66 CHAPMAN AND SMITH ON THE BUTYL COMPOUNDS From the above the following percentages are calculated :-I. 11. 111. c c ............ 26.08 - L -H ............ 4.97 - I ............ -6S*86 69-19 68-87 Theory. Found. C4........ 48 2 6.09 2 6.08 H9.. ...... 9 4.89 4.97 I ........ 127 69-02 68-91 (mean) 184 100. 100.02 These iodine determinations were made by decomposing the iodide with alcoholic soda free fieom chlorine (obtained by dis- solving sodium in alcohol).The alkaline iodide was rendered acid by dilute nitric acid and precipitated with nitrate of silver. The iodide of Eutyl was also titrated in the manner described by Professor Watnklyn the digestion being carried on in sealed tubes. I. 2,471 gramrnes of iodide digested with alcoholic soda neutralized as much alkali as 13.5 C.C. standard sulphuric acid would have done. 11. 5.384 grammes of iodide neutralized soda equivalent to 29.3 C.C. of standard acid. These titrations correspond to the following percentages of iodine :-I. 69.38 11. 69.11 Theory 69.02. When we recollect that the atomic weights of propyl butyl and amyl are as follows C,H =43 C,H9 = 57 and C,Hll =71 we see that a determination of the atomic weight at once points out with which radical we are dealing.Now the atomic weights as deduced from the three iodine determinations are respec- tively :-I. 57043; 11. 56.55; 111. 57.4. The atomic weights deduced &om the titrations are- I. 56.05; 11. 56.76. The above appears to us to be ample and more than ample proof that we were dealing with butyl compounds ;incidentally it establishes also the purity of the alcohol and the iodide. B. Proof of Pudy of the Iodide of ButyL-This depends lst DERlVED FROM BUTYLIC ALCOHOL OF FERMENTATION. 167 on the combustion given above; 2nd on the three iodine determinations given above ; 3rd on the two titrations given above; 4th on the yield of iodide from the alcohol. 80 grammes of the alcohol were digested with a very large excess (500 cc.) of hydriodic acid of apecific gravity 1.8 for forty minutes.The mixture was then distilled the distillate rendered alkaline with carbonate of soda and again distilled. The heavy oily layer of the distillate was pipetted off and while still wet weighed. Weight 198.5 grammes. To ascertain how much water hung about it 200 grammes of the pure dry iodide were distilled from carbonate of soda solution and the distil- late weighed &c. as before; it now weighed between 201.5 and 201.6. We may therefore assume that the increase of weight due to moisture was about 1-5; this gives us 197 grammes as the yield of iodide from 80 grammes of alcohol therefore-100 parts of alcohol yield 246.25 parts of iodide. 100 parts of butylic alcohol yield 248.65 parts (theoretically.) 100 parts of propylic alcohol yield 383.33 parts (theoretically.) 100 parts of amylic alcohol yield 225 parts of iodide (theo- retically.) We have therefore obtained 99-03 per cent.of the yield of iodide theoretically obtainable from butylic alcohol. C. Proof of purity of Bromide of ButyL-This rests entirely on the yield under the circumstances a perfectly sufficient datum. Two experiments were made on 80 grammes of the alcohol. I. yielded 147.4 grammes of bromide; 11. yielded 146.7 grammes ; or calculating from these yields- By I. 100 parts of alcohol yield 184-25 of bromide. 9 9, ? 11.9 ,) 183-38 , Theoretically , , 185.14 of bromide of butyl. We have therefore obtained reBpectively- I. 99-52 per cent.; II.99.05 per cent. of the theoretical yield. The conversion of the alcohol into the bromide was effected in soda-water bottles. The weighing of the bromide was in this instance effected dry; the contents of the soda-water bottles were distilled the distillate re-distilled from carbonate of soda solution the oily layer pipetted 0% and dried over chloride of calcium; it was now decanted from the chloride of calcium and weighed. Water was now added to the chloride of calcium and the mixture distilled whereby a few grammes 168 CHAPXAN AND SMITH ON THE BUTTL COMPOUNDS of bromide were recovered ; these were decanted and weighed wet and their weight added to the weight of the dry bromided As in neither case did the weight of this wet portion amount to 4 grammes and as bromide of butyl does not contain more than 97 per cent,.of water the error introduced by weighing this bromide wet could never amount to 0-03 grammes a quantity totally without influence on the result and indeed only influ- encing a figure in the third or fourth decimal place. D. Proof of purity of Nitrate of ButyL-This depends also on the yield. From the nature of the method of preparation a very sharp result could not be expected. Two determinations were made :-I. SO grammes of the alcohol yielded 126.5 grammes of the nitrate. 11. 100 grammes of the alcohol yielded 158.5 grammes of the nitrate. Calculating fiom I. 100 parts of the alcohol yield 158.13 of the nitrate; and by II. 100 parts of the alcohol yield 158.5 of the nitrate.Theoretically 100 parts should yield 160-8. From I. therefore we have obtained 98.35 per cent. of the theo- retical quantity and from IT. 98.57 per cent. E. Proof of purity of Acetate of ButyZ.-In this case we could not depend upon the yield because it is necessary to wash the acetate many times and because it is difficult to ensure the total convergion of the alcohol into the acetate. Still by dis-tilling the washings and treating the mixture of acetate and alcohol obtained from them with glacial acetic and gaseous hydrochloric acids and again separating the acetate so obtained we succeeded in obtaining between 96 and 97 per cent. of the theoretical yield. The proof of the purity of the acetate how-ever rests on titrations of which two were made.I. 3-9246 grammes of acetate neutralized as much caustic potash as 42.2 C.C.of standard acid could have done. 11. 4-9846 grammes neutralized as much potash a8 53.7 C.C. standard acid; of this acid 1 C.C. equals 0*03118 grammes of potasaium. From these numbers we calculate that according to-I. It yields 51.44 per cent. of acetic acid; and according to II. 51.53. Theoretically it should yield 51.72. F. Proof of purity of Mercury Butyl.-This rests on a deter- mination of the mercury contained in the compound. DERIVED FROM BUTYLIC ALCOHOL OF FERMENTATION. 169 0.370 of the compound yielded 0.235 of metallic mercury therefore 63-51 per cent. Theory for (C H9),Hg requires 63.69 ,) G. Proof of purity of Butylic AZcohoL-This rests first on the combustion quoted above; secondly on the yields of iodide bromide and nitrate ;lastly on the purity of the numerous butyl compounds obtained from it as shown by the iodine determina- tions and other analyses quoted above.H. The annexed table contains the equivalent of but71 deduced from the various analyses mentioned above. It is appended chiefly because in this manner and in this manner only can the exact comparative value of the different estima- tions be seen since the intervals between the percentage composition of different compounds vary so greatly that what is a sufficient approximation to the trut8h in one case has com- paratively little meaning in another ; thus in comparing one combustion with another for example a combustion of butylic alcohol with one of iodide of butyl an error of 0.25 per cent.in the carbon in the case of the alcohol would only correspond to an error of 0.1 per cent. of carbon in the iodide. If there- fore we were to argue from two combustions one of the alcohol with an error of -5 per cent. and one of the iodide with an error of -25 per cent. we should arrive at a more accurate result by using that combustion with the largest appa.rent error. TABLEshowing the Atomic Weight of Butyl as deduced from the following 13 determinations. Atomic weights deduced from. Nature of Determination. 3ets of Expe-solatedExpe-riments and riments. neam of sets. (1.) By combustion of alcohol C + H found = 79.86 per cent.100 -(C + H) found 22 14. 2214 79.86 : 16 = (Eq 0) atomic weight.. .... .. 66 *27 (2.) Bycombuebion of iodide C + H found = 31-05 per cenf. 100 -31.05 per cent. = 68-95 :. 68.95 31.6 : 127 = (Eq I) ....,.....,,.................q .. 67 .I9 VOL. XXH 0 170 CHAPMAN AM3 SMITH ON THE BUTYL UOMPOUNDS Atomic weights deduced from. Nature of Determination Sets of Expe-solated Expe- riments and riments. neana of seta By estimation of iodine in iodide as Ag I-(3.) 1. ......................................... 57 *43 (4.) If. ........................................ 56.55 (5.) 111.. ....................................... 57 -40 Mean.. ...................................... .. 67 -13 By estimation of iodine in iodide by titration- (6.) I......................................... 66-05 (7.) 11. ........................................ 56 *76 Mean. ....................................... .. 56 -41 (8.) By yield of iodide of butyl from alcohol ........ .. 58 -22~ By yield of bromide-(9.) I. . ...................................... 67 *77 (10.) 11. ....................................... 58 *50 Mean.. ...................................... .. ti8 -17 By titration of aeehte-(21.) I. ........................................ 6'1 -64 (12.) IT. ....................................... 67*42 Mean.. ...................................... .. 67 *63 (13.) By estimation of Hg in mercury butyl. ........ .. 67-46 Mean of all determinations ...................... 67.30 DISCUSSION. Dr. Odling objected to the definitions commonly given especially in text-books of primary secondary and tertiary alcohols these definitions being based upon the manner in which the carbon atoms are supposed to be grouped whereas it would be much better that they should be founded upon reactions. The definition of a body should in fact be that it is one which behaves in a particular way from which its constitution is after-wards inferred. Thus with regard to the particular case under consideration an alcohol may be classed as primary when it * We should perhaps here point out that the method of exhibiting the resuIts o€ analysis in atomic weights whilst it shows the relative value of anaiyses better than any other plan exaggerates the apparent error very much.Thuu for example if we exhibit the percentage composition of iodide of butyl 88 detetmined by the yield we obtain 68-57 per cent. of iodine a result which is-&G bf the iodine in Prror' but the atomicz weight is 58-22 a result fiin error. DERIVED FROM BUTYLfC ALCOHOL OF FERMENTATION. 171 yields by oxidation an acid containing the same number of carbon atoms. Hydrocarbons may be formulated first according to the residues which they contain viz. :-Methyl.. ............ CH,’. Methylene .......... CH,”. Formyl ............ CH“‘. Carbon ............ C””. Secondly according to the number of each of these constituent residues; and thirdly according to the manner in which they are tied together.The hydrocarbon C4H1,, from which the butylic alcohol under consideration is derived is admitted to contain the residue CH three times its formula being CH(CH,), and the piimary alcohol derived from it by substitution of OH for €3. in one of the groups CH will be-CH CH, r3 CH,OH. Such formulae have this advantage over those now generally in use that they do not involve any particular assumptions as to the superiority in importance of one hydrocarbon residue over another and its consequent selection for the post of honour as the typical basis of the formula. Professor W ankl y n said that the classification of alcohols according to the linking of the carbon atoms appeared to him to be indirectly a classification according to reactions but that he preferred to classify according to the linking and to have criteria by which to recognize that linking.The butyl alcohol which forms the subject of Mr. Chapman’s paper is a primary alcohol that is an alcohol in which the carbon which is united with hydroxyl is united directly with only one atom of carbon. The criterion by which we recognize such an alcohol is that by oxidation it gives an acid containing the same number of atoms of carbon as the alcohol itself. The alcohol that Mr. Chapman haa examined fulfils this character. By oxidation it gives isobutyric acid which has the same number of atoms of carbon as the alcohol. In order to get this isobutyric acid we have to oxidize carefully but still we can get an acid containing the 02 172 CHAPMAN AND SMITH ON THE BUTYL COMPOUNDS same number of atoms of carbon as the alcohol.This alcohol would be called a pseudo-primary or an iso-primary alcohol. The proof that this alcohol is not the normal butylic alcohol c{r ,is that the four-carbon acid formed from it by oxidation is not common butyric acid C 0 ,but iso-{F2cH3 butyric acid C 0 ,identical with that which is produced r3), by the action of alkalies on cyanide of isopropyl. The salts of these two acids exhibit considerable diversities of character. The calcium salt of isobutyric acid is much more soluble than ordinary butyrate of calcium and isobutyiic ether boils at 112" whereas butyric ether boik at 119'. Mr. Chapman observed that a further difference between these two iaomeric acids is afforded by the fact that isobutyric acid breaks up on oxidation whereas normal butyric acid does not.Dr. Odling said that it wa8 not the ordinary mode of classi- fshg the alcohols that he objected to but the definitions. He thought as a matter of logic that the definition should have reference as nearly as possible to the criteria or characteristic properties of bodies 80 that in saying that a Body belongs to a particular class we should be understood as implying that it behaves in a particular way. Dr. Crum Brown said that he perfectly agreed with the last remarks that had fallen from Dr. Odling though he could not quite agree with him respecting the classification of the alco- hols. Some alcohols when oxidized yield an aldehyde and mme aldehydes (those namely to which the term is generally applied) yield by oxidation acids containing the same number of carbon atoms.NOW the alcohols which yield aldehydes converted by oxidation into the corresponding acids are those which we call primary alcohole. Then again there are alcohols convertible by oxidation into aldehydes (ketones) which when subjected to the action of oxidizing agents split up into two acids; theRe are the secondary alcohols. And lastly there are alcohols which do not yield aldehydes at all but split up at once by oxidation into acida containing smaller num- DERIVED FROM BUTYLIC ALCOHOL OF FERNENTATION. 173 bers of carbon atoms. So far as we can at present Bee &om observed facts primary aecoudary or tertiary alcohol is a carbinol in which 1 2 or 3 atoms of hydrogen are replaced by alcohol radicals.By classifjTing in this way we get rid of the danger attending the too free use of the atomicity theory which must be regaded as at best only a temporary expedient destined Booner or later to be replaced by a theory more in accordance with the true principles of physics. Theories in fact are but mere scaffolding; let them be used as such and knocked away when the proper time comes. Dr. Guthrie inquired what was the precise origin of the fusel oil upon which Mr. Chapman had been working. It would be interesting to know whether it was obtained fiom potatoe spirit or grain spirit. Mr. Chapman expressed his regret at not being able to give an exact answer to the question and indeed he thought it would be difficult to obtain a precise answer as many distillers are in the habit of using mixed grtius.The fusel oil upon which he had worked was obtained fiom Bowerbank‘s and fiom Whatney’s and unfortunately the two had been mixed before he got them. He had however examined many other specimens of fusel oil and they all appeared to contain butylic alcohol. Returning to the question of formulae and classification Mr. Chapman said ‘‘1 do not see any objection to the mode of defining the several claases of alcohols proposed by Dr. Odling and Dr. Crum Brown which indeed is quite in accordance with observed facts. There is however one car- dinal point about alcohols or the compounds derived fiom them.Some alcohols split up directly into olefine and water and some do not. Now can we not base some kind of classification upon this fact or can we not make it a portion of some classi- fication? I of courae have never seen tvha,t would be called normal primary butylic alcohol. But if it should turn out that on treating normal iodide of butyl with alcoholic potash no butylene whatsoever is evolved and that on treating this isopropyl car- binol with alcoholic potash butylene is evolved as we know it is I think we should there have a very fair ground for beginning to make a very wide division of the alcohols and that we should perhaps be justified in calling one set of them hydrates of olefines and another alcohols giving them two generic names.There 174 STOKES ON A CERTAIN REACTION OF QUININE. is however another question not immediately connected with this. Before we can argue very directly as Dr. Odling and Dr. Crum Brown appear to wish from the reaction to the nomenclature we must have a very much more careful study of the reactions than has hitherto been made. At the prerJent moment we do not know whether on treating iodide of amyl with alcoholic potash any amylene is evolved or not or whether ontreating common iodide of ethylwith alcoholic potash anyethy-lene is evolved. I have however ascertained that on treating normal bromide of proyyl which I have obtained from another portion of this fusel-oil with alcoholic potash no olefine or only an infinitesimal quantity is evolved.”