摘要:

664 LI.-Researches on the Constitution of Azo- and Diazo-derivatives. IV. Diazoamido-compounds (continued). By RAPHAEL MELDOLA F.R.S. and F. W. STREATFEILD F.I.C. IN former papers on the diazoamido-compounds (Chem. Xoc. Trans., 1886 624; 1887 102 and 434) it has been shown that the alkyl-derivatives of /mixed compounds* of the general formula X.N,R'*Y are apparently capable of existing in three isomeric modifications :-* This designation is applied to all diazoamido-compounds in which the two radicles are dissimilar (PhiE. Mag. June 1887 518) and is not restricted as in a recently published paper ( B e y . 1888 lOlS) to those compounds which con-tain one fatty and one aromatic radicle THE CONSTITUTION OF AZO- AND DIAZO-DERIVATIVES. 665 (1.) Produced by the action of diazotised X-NH upon Y-NHR’.(2.) Produced by the action of diazotised Y*NH upon XsNHR’. (3.) Produced by the direct alkylation of the mixed compound As this fact is of the greatest interest in connection with the general subject of chemical isomerism we have lately heen concen-trating our effort,s on this branch of the work with the object of inves-tigating in the first place how far the isomerism in question can be regarded as holding good-whether in fact the isomerism is per-fectly general whatever radicles may be represented by X Y and R’ in the above formula. In order t o decide this point it is evidently essential to prepare as ma,ny distinct triplets as possible so that the individual influence of the different radicles may be studied. It is moreover obvious that at least two of the compounds out of each triplet should be solid so that their identity or non-identity may be established by their melting points as the instability of the diazo-amido-compounds when heated does not permit of the boiling points of the oily compounds being determined.The manifestly complicated practical requirements which have thus to be met must necessarily make the present branch of the investiga-tion a very protracted one because in the first place each pair of amines must be experimented with in order to see whether they can be combined so as to form a diazoamido-compound ; and in the next place the alkyl-derivatives have to be prepared in order to see whether they satisfy the required conditions. As the nitranilines lent themselves most readily to the former investigations the results of which have been made known in the papers previously referred to, we have continued our experiments with these nitroamines.Some practical points in the general method of working may here be alluded to before describing the results in detail. All experience with the mixed diazoamido-compounds having shown us that the alkyl-derivative prepared by the third of the foregoing methods (direct alkylation) was the isomeride of lowest melting point it has been found advantageous in all cases to prepare this com-pound first. If this last alkyl-derivative is solid the two other iso-merides will most probably be also solid. On the other hand if the alkyl-derivative in question is oily the two other modifications may or may not be solid but they must be prepared in order to decide this point before the investigation of the triplet can be considered as complete.The nature of the practical difficulties which have to be overcome will be seen more fully in the course of the description of the experimental results. X*NsH*Y 666 MELDOLA AND STREATFEILD THE CONSTITUTION OF 1. iNethylation of the Compound (m. p . 211") produced by the Action of Diazotised Paranitraniline on Metauitraniline or of Diazotised Netanitraniline on Paranitraniline. The first complete triplet of isomeric alkyl-derivatives described in our former papers was prepared from para- and meta-nitraniline and contained the radicle ethyl. In order to ascertain the effect of sub-stituting another radicle for ethyl the corresponding series of methyl-deriratives has been prepared.The mixed diazoamido-compound was dissolved in alcohol with the addition of the necessary quantity of potassium hydroxide (one molecular proportion) and the solution boiled in a flask with an upright condenser for about 10 hours wxh one molecular proportion of methyl iodide. The methyl-derivative separates out as a yellow crystalline deposit which is collected, washed with alcohol and purified by several cry stallisations from alcohol containing a few drops of caustic soda solution to insure the removal of any unmethylated compound. The purified substance forms small bright yellow needles very similar to the ethyl-deriva-tive in appearance and properties and having the same melting point 148".I. 0.1394 gram gave 0.2672 gram CO and 0.0494 gram H20. The following results were obtained on analysis :-11. 0.1080 , 0.2050 , , 0.0390 ,, 111. 0.1244 IV. 0.1248 , 24.4 , , 12" and 760.3 , 25.4 C.C. moist N at 14" and 754.7 mm. bar. Found. 7 Calculated for (---h-'--NO1.C6H4.N3(CH3).C6H4'~o~. I. 11. 111. IV. C 51-83 52.27 51.76 H . . . . . . 3-65 3.93 4.01 - -N 23.25 - - 23.24 23.20 - -The compound is readily decomposed by strong hydrochloric acid. On standing for about an hour at the ordinary atmospheric tempera-ture under excess of strong acid the substance gradually dissolves, leaving only a very slight residue; the acid solution contains a mixture of meta- and para-nitrobenzene diazochlorides and the two corresponding met hylnitranilines.The decomposition is thus similar to that of the ethyl-derivative and of the mixed diazoamido-compound from which these alkyl-derivatives are obtained. 2. Methylation of Paradinitro-diazoanaidobenzene m. p . 223". The methyl-derivative of the above substance was prepared in the usuul way by dissolving the dinitrodiazo-compound in alcohol wit AZO- AND DIAZO-DERIVATIVES. 667 the theoretical quantity of potassium hydroxide and cohobating with the necessary quantity of methyl iodide. After purification by two or three crystallisations from alcohol made slightly alkaline by caustic soda the methyl-derivative was obtained in the form of small yellow needles melting at 219" :-I. 0-1458 gram gave 0.2778 gram COz and 0.0490 gram H,O. 11. 0.1092 , 0.2086 , , 0.0364 ,, 111.0-1872 , 36.4 C.C. moist N at 13.5" and '776.7mm. bar. Found. Calculated for -7 NO,.C,jH4*NB(CH,)*C,H4*NOp I. 11. 111. C 51.83 51.96 52-10 H . . 3.65 3-73 3.70 -3 23.25 - - 23.45 -The substance is decomposed quite normally by acids. With excess of cold hydrochloric acid it is resolved into paranitrodiazobenzene chloride and methylparanitraniline ; hot hydrochloric acid produces paranitrochlorobenzene and methylparanitraniline. 3. Methylation of Metadinitro-diazoamidobenzene m. p. 194". This methyl-derivative was prepared in the same; manner as the others and after purification by crystallisation from alcohol forms small yellow needles paler in colour and ra-ther more soluble in alcohol than the preceding modifications. The melting point is 127-128" the pure substance giving the following results on analysis :-I.0.1740 gram gave 0.3302 gram CO and 0.0584 gram H,O. 11. 0.1740 , 34 C.C. moist N at 15" and 766.3 mm. bar. Found. Calculated for 7-NO?*C6H4.N3(CH3) *C&4*N02. I. 11. C 51.83 5 1.75 H 3.65 3-73 N 23.25 - 23.07 -The compound is decomposed by cold hydrochloric acid into meta-nitrodiazobenzene chloride and methylmetanitraniline and by hot acid into the same alkylnitraniline and metanitrochloro benzene. 4. The Action of Diazotised Paranitraniline on Methy l~etan~tran~line. Paranitraniline was diazotised in hydrochloric acid solution and added to an aqueous solution of the hydrochloride of methylmeta-nitraniline. If care is taken to have present only a minimum of acid 668 MELDOLA AND STREATFEILD THE CONSTITUTION O F the yellow diazo-compound commences to separate out immediately on mixing the solutions ; if too much acid is used the separation of the diazo-compound takes place very slowly but can be promoted by the addition of sodium acetate.After standing for some hours the mixed solutions become almost solid from the separation of a yellow cryetallinc pulp. The latter is collected washed well with water and purified in the usual way by crystallisation from alcohol with the addition of a few drops of strong caustic soda solution. The pure substance resembles its isomerides in appearance and properties, forming small yellow needles melting at 168" :-I. 0.1084 gram gave 0.2068 gram GO and 0.0372 gram H,O. 11. 0.1232 , 24.6 C.C.moist N at 13.5" and 745.5 mm. bar. Found. Calculated for r-7 NO2.C6H,.N,(CHJ *C&4*NOp I. 11. C . 51.83 52.02 H 3.65 3.81 23.08 N 23.25 --The substance is decomposed by cold hydrochloric acid into para-nitrodiazobenzene chloride and methylmetanitraniline and by hot acid into the corresponding nitrochlorobenzene and the same alkgl-nit raniline. 5. The Action of Diazotised Metanitraniline on M e t ~ y I-para~itraniline. This modification was prepared in a precisely similar manner to the last the purified substance being obtained in the form of minute bright yellow needles melting at 176-1 77" :-I. 0.1146 gram gave 0.2194 gram GO and 0.0396 gram H,O. 11. 0.1134 , 22.4 C.C. N at 17.5" and 760.8 mm. bar. 111. 0.1434 , 28 , 14" , 762-6 ,, Found.Calculated for r--- 7 NO~*C6H~*N~(CH,) .CfiH4*NOp I. 11. 111. C 51.83 52.21 H 3.65 3.83 - -N 23-25 - 22-98 23.04 - -The compound undergoes normal decomposition by hydrochloric acid the hot acid giving metanitrochlorobenzene and methylpara-nitraniline and the cold acid the last compound and metanitrodiazo-benzene chloride. The methyl-derivatives now described furnish a second complet AZO- AND DIAZO-DERIVATIVES. 669 Melting point of methyl-derivative. Dinitrodiazoamido-compound. triplet of isomeric compounds confirming the results obtained with the ethyl-derivatives. The isomerism will be seen from the melting points given in the following comparative table :-Melting point of ethyl-derivative. 2. Action of diazotised metanitraniline on alkyl-paranitraniline -I 1 --176-177" 1'74!-175' 1.Alkylation of the compound produced by the action of diazotised paranitraniline on metanitraniline or vice versa" 5. Alkylation of the compound produced by the action of diazotised metanitrani-line on metanitraniline 148" 127-128" 148" , I I 3. Action of diazotised paranitraniline on alkyl-metanitraniline 168" 18'7" 4. Alkylntion of the compound produced by the action of diazotised paranitraniline on paranitraniline 219' 191-192" 119" The compounds 4 and 5 in the preceding table are normal that is, not mixed compounds but are added in order to complete the list of isomerides as far as our present experiments have extended. In the course of our search for triplets of isomeric alkyl-derivatives having at least two solid members several new diazoamido-com-pounds and their alkyl-derivnt ives have been prepared and descrip-tions of these are now subjoined.6. Dinitrodibromo-diaxoamidobenzene. Paranitro-orthobromaniline [NO Br NH = 1 3 4 m. p. 104O], was dissolved in alcohol with the addition of a little dilute hydro-chloric acid and the necessary quantity of sodium nitrite dissolved in a small quantity of water added to the cold solution. The yellow, flocculent diazoamido-compound commences to separate out at once, and after standing for eight hours the separation is completed by diluting the solution with a large bulk of cold water. The product was collected washed with water and crystallised from hot alcohol, in which the substance dissolves but slightly.Analysis having shown the compound to be impure it was crystallised two or thre 670 MELDOLA AND STREATFEILD THE CONSTITUTION OF times from a mixture of alcohol and toluene and was finally obtained in the form of hair-like orange needles which separated from the solution in rosettes. The melting point is 202". 0.1435 gram gave 0.1207 gram AgBr. Calculated for N02-BrC6H3- N,H.CGHIBr.NO,. Found. Br 35.95 35.79 The substance is strongly acid dissolving in hot aqueous alkalis or in cold alcoholic alkali with a fine crimson colour. The bromonitraniline from which this compound is obtained possesses but very feeble basic properties and is thus almost insoluble in acids. For this reason it is difficult to diazotise or to make it react with other diazo-chlorides without the use of alcohol as a solvent.This bromonitraniline did not appear therefore to be well adapted for the present inquiry and the experiments were discon-tinued. 7. Paradichlorodinzoamidobenzene and its Ethy 1-derivative. Parachloraniline is dissolved in water containing just enough hydrochloric acid to keep the base dissolved in the cold (about 20 grams of the base to 1 litre of water and acid) and the necessary qunntity of sodium nitrite (1 mol. to 2 mols. of the base) dissolved in water is gradually added to the cold solution. The diazoamido-corn-pound soon begins to separate as an ochreous precipitate and the separation is completed by adding a quantity of sodium acetate equiva-lent to the hydrochloric acid used.After being collected and washed, the substance was crystallised from dilute alcohol and finally from a mixture of benzene and petroleum when it was obtained in the form of ochreous needles melting at 130". I. 0.1144 gram gave 0.2263 gram GOz and 0.0353 gram H,O. 11. 0.1382 , 0.1485 , AgC1. 111. 0.2164 , 28.5 C.C. moist N at 12.5" and 749.7 mm. bar. Found. 7 Calculated for r-"-CGH,Cl*N,H*CGH,Cl. I. 11. 111. - - C 54.13 53-94 H 3-42 3-42 N 15.79 - 15-38 Cl - 26-69 - --26.56 -The compound is decomposed normally by cold hydrochloric acid into parachloraniline and parachloro-diazobenzene chloride AZO- AND DIAZO- DERIVA TIVES. 6 i l A silver salt was prepared by adding an ammoniacal solution of silver nitrate to an alcoholic solution of- the substance containing a little ammonia.The salt is at once precipitated as a yellow jelly, which slowly changes to a crystalline condition the crystals appearing under the microscope in the form of fine yellow needles. After being washed with alcohol and dried the salt was analysed the silver and chlorine approximating to the formula C6H4C1*N,Ag.C6H,CI, and the C H and N agreeing somewhat more closely with the calcu-lated numbers :-I. 0.2454 gram gave 0.3552 gram CO2 and 0.0532 gram H,O. 11. 0.0986 , 9.4 C.C. N at 15.5" and 767 mm. bar. Found. 'I Calculated for Qlz~8C1zN3Ag. - C 38.60 39.47 H 2.14 2.41 N 11.26 - 11.24 The difficulty of purifying and analysing these expIosive metallic derivatives of the diazoamido-compounds has been alluded to in a former paper (Trans.1887 445). The ethyl-derivative is prepared in the usual way by boiling a sollition of the substance in (absolute) alcohol with the necessary quantities of potassium hydroxide and ethyl iodide. After cohobat-ing for 1-2 hours the alcohol is distilled off and the contents of the flask diluted with water when the ethyl-derivative separates out as an oil which on standing for two or three days solidifies to an ochreous resinous mass. The latter is dissolved in dilute alcohol, boiled with animal charcoal and filtered when the purified substance separates out on cooling in the form of flat straw-coloured needles having a melting point of 85.5". Two crjstallisations from dilute alcohol were necessary before the substance was quite pure :-13.2 C.C.N at 12.5" and 771 mm. bar. -I. 0.1206 gram gave 0-2.532 gram CO and 0.050 gram H20. 11. 0.1090 7 7 111. 0.1663 , 0.1612 gram AgC1. Found. Calculated for r - - 7 C6H,cl.N,(C,H,) *C~HICI. I. 11. 111. C 5 7-26 - - 37-14 H 4-60 - I 4.42 N . . - 14.55 - 14.28 C1 24.14 - - 23.9 7 The substance is decomposed by cold hydrochloric acid into para-chloro- diazobenzene chloride and ethylparacbloraniline 672 MELDOLA AND STREATFEILD THE CONSTITUTION OF 8. Benzy Lderiva fives of Diazoamido-compounds rtot colztaining Nitro-groups. The chief object in undertaking the next series of experiments was to find if possible a triplet of isomerides fulfillinq the necessary conditions and not containing nitro-groups. The first attempt in accordance with the general method of working already described, was made with the mixed compound obtained from aniline and para-toluidine.The substance was benzylated in alcoholic solution with benzyl chloride and potassium hydroxide in the usual way. On distilling off tbe alcohol and diluting the residue with water the benzyl-derivative separates out as a dark viscid oil which does not solidify on standing for weeks. An attempt was next made to prepare the isomeride by the action of diazotised paratoluidine on benzylaniline.* The latter was dissolved in hot dilute hydrochloric acid and the solution when cold just neutralised with ammonia the contents of the flask being strongly agitated so as to get the benzyl-aniline into the state of an emulsion. The solution of diazotoluene chloride was then added to the benzylaniline emulsion and the mixture allowed to stand for about 12 hours.The ochreous deposit was collected and washed but soon began to agglomerate into an oil, and microscopic examination revealed the fact that the precipitate which at first sight appeared solid was really an emulsion of viscid oily globules with the aqueous liquid. I t appears therefore that the triplet of mixed diazoamido-compounds comprised under the formula C6H,*N3( C7H7)*C6H,.CH3(p) contains at least two oily members and the experiments were not further continued in this direction. The benzyl-derivative of diazoamidobenzene has been prepared by direct benzylation by Friswell and Green (Trans. 1886 749) who state that it is a solid melting at 81".We have prepared the COP-responding derivative of diazoamidoparatoluerie which after purifi-cation by crystallisation from alcohol with the addition of animal charcoal forms ochreous needles melting at 114". I. 0.1164 gram gave 0.340 gram COZ and 0.0704 gram H,O. 11. 0.1430 , 15.8 C.C. moist N at 11" C. and 769.5 mm. bar. Found. Calculated for w - 7 (p)CH3*CGH4.N,(CjH7)*C~H4*CH~( p ) . I. 11. - C 80.00 79.66 El 6.66 6.71 N 13.33 - 13-34 -* We are indebted to Dr. A. Weinberg of the Frankfurter AIiilinfarben-Babrik, Gans and Co. for a quantity of this substance prepared in their factory AZO- AND DIAZO-DERIVATIVES. 673 These benzyl-derivatives do not decompose under the influence of acids in the same definite manner as do the other alkyl-derivatives when treated in the same way.A large amount of tarry matter and intensely-coloured products possibly benzylated amidoazo-compounds, are always formed in this decomposition. 9. The Action of Diazotised Nitranilines on Pnrachloraniline and of Diazotised Parachloraniline on the Nitranilines. In order to ascertain whether a mixed diazoamido-compound con-taining nitrobenzene and chlorobenzene residues could be obtained, diazotised metanitraniline was allowed to act upon parachloraniline, and diazotised parachloraniline upon metanitraniline in the usual way. A bulky pale yellowish product was obtained in each case, and this although apparently a homogeneous substance was soon found to be a mixture from which metadinitro-diazoamidobenzene (m. p. 194") was easily separable by fractional crystallisation.Attempts made with paranitraniline and parachloranilins led t o similar results whichever order of mixing was adopted. The product was always a mixture containing paradinitro-diazoamidobenzene (m. p. 223") and in some preparations free nitraniline was found. I n one case when diazotised paranitraniline had been made to act on parachloraniline in hydrochloric acid solution (no sodium acetate being used) a very small quantity of a substance having the proper-ties of the mixed diazoamido-compound sought for was isolated from the mixture of products. The substance in question appeared mixed up with the needle-shaped crystals deposited from the alcoholic solution of the original preparation and formed ochreous crystals of ereat brilliancy consisting of small well-formed octohedra which, on account of t'heir much greater density could be easily washed out of the mass of fine needles with which they were entangled.The crystals thus separated had a melting point of 181" which was not altered by another crystallisation from alcohol; at the instant of fusion the compound froths up and decomposes. A chlorine deter-mination gave the following results :-0.1126 gram gave 0.0598 gram AgC1. Calculated for (pjNOa.CGH,.N,~.CGH~Cl( I)). Found. C1 12.83 13.13 The substance possessed the properties of a mixed diazoamido-compound dissolving in alcoholic alkali with an intense red colour, and being reprecipitated by acids in a yellow flocculent condition. VOL. LIII. 2 674 NELDOLA AND STREATFEILD THE CONSTITUTION OF The yield of this compound was too small to make it worth while to prepare it in quantity.The foregoing experiments are of importance in connection with the present investigations as showing that two dissimilar amines may fail to combine to form a mixed diazoamido-compound o r that the formation of such a compound may take place only to a subordinate extent the chief products being normal diazoamido-corn pounds. I n other words there may be a degree of dissimilarity between the two amines which is incompatible with the stability of the mixed diazo-amido-compound. In the present imperfect state of our knowledge of the chemical constitution of the diazoamido-compounds it is of course impossible to follow the precise course of the change which leads to the production of a mixture of normal compounds from a mixed compound but two explanations naturally suggest themselves.Supposing the unstable mixed compound to be momentarily produced, this might be regarded as decomposing into a mixture of free diazo-salts and amines in accordance with known facts :-X*N,H*Y giving X*N2*OH X*NH, Y.Nz.0 H X*NH2. These four products might then be supposed to recombine with the formation of the normal compounds-X*NsH*X and Y*N3H*Y. The alternat,ive explanation is that the mixed compound reacts with itself in accordance with the scheme-ZX*N,H.Y = X*N,H*X + Y*N3H*Y. 10. The Action of Diazotised Metanitraniline on Ethylparachloraniline und of Diazotised Yarachloraniline o n Ethylmetanitraniline.Ethylparachloraniline hydrochloride was dissolved in water and the solution of metanitrodiazobenzene chloride added the mixture being kept well cooled and allowed to stand for about 12 hours. The reddish precipitate was collected and washed and purified by two or three crystallisations from dilute alcohol with the addit,ion of animal charcoal. The pure substance crystallises in long flat lustrous plates of an ochreous colour melting a t 106". I. 0.1647 gram gave 0.3343 gram CO and 0.0651 gram H,O. 11. 0.0668 , 10.85 C.C. moist N at 16.5" and 733.3 mm. bar. 111. 0.2120 , 0.0999 gram AgCl AZO- AND DIAZO-DERIVATIVES. 673 Found. 7 Calculated for r-L-(fn)NO2'C6H4.N,(C,H5).CsH4CI ( p ) . I. 11. 111. c 55.17 55.35 - -H 4-26 4.39 N . . 18.39 - 18.15 -c1 11.65 - - 11.64 - -The substance is decomposed by cold hydrochloric acid into meta-nitrodiazobenzene chloride and ethylparachloraniline.The isomeric compound was prepared in a precisely similar manner by the action of diazotised parachloraniline on ethylmetanitraniline. After purification by crystallisation from alcohol with use of animal charcoal it forms long flat pale-yellow needles melting at 129.5". I. 0.1172 gram gave 0.2364 gram CO and 0.0458 gram H,O. 11. 0-1078 , 0.0521 grain AgC1. Found. r. Calculated for (r))CLC6H,.N,(CaH,).CpH4'NOz(m). I. - C 55.17 55.01 H 4.34 - 4.26 €2 11-65 - 11-94 The compound is resolved by cold hydrochloric acid into para-chlorodiazobenzene chloride and ethylmetanitraniline. The pair of isomerides now described are probably members of a triplet of which the third modification is wanting owing to the cir-cumstance that a mixed diazoamido-compound cannot be prepared from parachloraniline and metanitraniline by the ordinary methods." 11.Quantitative Decomposition of D i a z o a ~ i ~ o - c o m ~ o ~ n d s . In a previous paper (Trans. 1887 438) we made known a method for converting the products of decomposition of the diazo-amido-compounds by hydrochloric acid into stable compounds easily separable and capable of identification. This method gave such satisfactory results for the purposes of qualitative identification, that we have been induced to apply it quantitatively and as will be seen from the numbers snbmit,ted the process has proved equally serviceable from this point of view.In order to estimate the amount of free diazo-chloride formed by the acid the process has been modi-fied as follows:-* It is possible that the desired compound may be obtained by acting with 1 mol. of a solid metanitrodiazobenzene salt on 2 mols. of parachloraniline dissolved in some non-aqueous liquid such as benzene. It is proposed to test this method experimentally. 2 2 676 MELDOLA AND STREATFEILD THE CONSTITUTION OF A weighed quantity of the fjnely powdered substance is placed in a small beaker and after being covered with an excess of strong hydro-chloric acid is allowed to stand with occasional stirring till the whole or nearly the whole of the substance is dissolved. The solution is then diluted with a little water and filtered through a tared filter the residue being washed on the filter dried and weighed and its weight subtracted from that of the substance employed.The filtrate is allowed t o fall into an alkaline solution containing an amount of &naphthol slightly in excess of that required by theory the precipitate is collected washed free from excess of alkaline solution then washed with hydrochloric acid to remove the amine and finally washed with hot water dried and weighed on a tared filter. The weight of azo-naphthol-compound is then compared with that required by theory in accordance with the general equations :-X*N,H*X + HC1 = X.N2-C1 + X*NH2, X*N,*Cl + CloH,.ONa = X*N,*C,,,H,*OH + NaCl. With some compounds it was found advantageous to allow the acid filtrate t o drop into an acid solution containing the ,&naphthol in suspension instead of dissolved in alkali.After being allowed to stand for some hours the liquid is gradually made alkaline and the precipitate collected and treated as before. For the sake of brevity, the words " acid " or " alkaline treatment " will be used to indicate the method employed. I. Compound produced by the action of diazotised metanitr-aniline on ethylparachloraniline. M. p. 106". Alkaline treatment. 0.3580 gram (less 0.0014 gram residue) gave 0.3420 gram azo-naphthol-compound = 95.9 per cent. The azonaphthol after crys-tallisation had a melting point of 194" (metanitrobenzeneazo-/3-naphthol). According t o theory 100 parts of the substance should yield 96.5 parts of azonaphthol-compound.11. Compound produced by the action of diazotised parachlor-aniline on ethylmetanitraniline. M. p. 129.5". Alkaline treatment. 0.5076 gram (less 0.002 gram residue) gave 0.4678 gram azonaph-thol-compound = 925 per cent. According to theory 100 parts of the substance should yield 92.8 of azonaphthol-compound. The substance required 13 C.C. of strong hydrochloric acid and five hours' digestion before the decomposition was sufficiently complete. The product weighed was purified by cry stallisation from alcohol and then formed long needles of a deep orange-red colour; the melting point is 162.5". We now give the numerical results :-Analysis showed the substance to be-Parachlorobenaeneazo-/3-naphthol. 0.2346 gram gave Os1210 gram AgC1 AZO- AND DIAZO-DERIVATIVES.677 Calculated for (2)) ClCGH4.N,.C,oHG.OH~). Found. Cl 12.56 12.75 111. Ethyl-derivative of paradichloro-diazoamidobenzene. M. p. 85.5". Alkaline treatment. 0.5206 = (less 0*0002 gram residue) gave 0.4968 gram azo-naphthol-compound = 95.5 per cent Theory requires 96.1 per cent. The substance required 8 C.C. strong acid and was decomposed in 15 minutes. The azonaphthol-compound was the same as in the last analysis. IV. Compound produced by the methylation of metadinitrodiazo-amidobenzene. M. p. 127 -1 28". Alkaline treatment. 0.2378 gram (less 0.0060 gram residue) gave 0.2246 gram azo-naphthol-compound = 96.8 per cent. Theory requires 97.3 per cent., 5 C.C. acid used ; time required for decomposition 45 minutes. Azo-naphthol - compound identified as metanitrobenzeneazo - /3 -naphthol (m.p. 194"). V. Compound produced by the methylation of paranitrodiazo-amidobenzene. M. p. 219". Alkaline treatment. 0.3758 gram (less 0.007C; gram residue) gave 0.3598 gram azo-naphthol-compound = 97.7 per cent. Theory requires 97.3 per cent. Substance very diflicult to decompose; 15 C.C. strong acid required 24 hours at ordinary temperature ; the decomposition had to be completed by heating at 90" for one hour. Azonaphthol-compound identified as paranitrobenzeneazo-p-naphthoI (m. p. 250'). VI. Compound produced by methylation of the mixed diazoamido-compound from meta- and para-nitraniline. M. p. 148". Alkaline treatment. 0.2816 gram (less 0.0122 gram residue) gave 0.2650 gram azo-naphthol-compound = 98.3 per cent.Theory requires 97.3 per cent. Substaiice decomposes much more easily than the preceding isomeride ; 10 C.C. acid produced decomposition in one hour. Azonaphthol-compound a mixture of meta- and para-nitrobenzeneazo-/3-naphthol. VII. Compound produced by the action of diazotised paranitr-aniline on methylmetanitraniline. M. p. 168". Alkaline treatment. 0.2658 gram (less 0.0088 gram residue) gave 0.2478 gram azo-naphthol-compound = 96.4 per cent. Theory requires 97.3 per cent. 11 C.C. acid required and three and a half hours at ordinary tempera-ture. Azonaphthol-compound identified as paranitrobenzeneazo-@-naphthol (m. p. 250'). VIII. Compound produced by the action of diazot!ised metanitr-aniline on methylparanitraniline. M. p. 176-177". Acid treatment.0.2567 gram (less 0.0156 gram residue) gave 0.2302 gram azo-naphthol-compound = 95.4 per cent. Theory requires 97-3 per cent 6 78 THE CONSTITUTION OF AZO- AND DIAZO-DERIVATIVES, 12 C.C. acid allowed t o act for six hours then filtered into acid solu-tion containing p-naphthol and allowed to act for 12 hours after which the mixture was made alkaline and treated as usual. This compound is the only one which has given unsatisfactory results by the alkaline method the percentage of azonaphthol-corn -pound always coming out too high. It is probable that the precipi-tate retains methylparanitraniline which is imperfectly removed by acid washing owing to its feeble basic properties. The result obtained by the acid method is nearly 2 per cent. too low although every precaution was taken to insure accuracy.With this exception, however the numbers obtained are sufficiently exact to enable us to say that the decomposition of the diazoamido-compounds by cold hydrochloric acid takes place quantitatively in accordance with the general equation-XsN3H.X + HC1 = X*N2*C1 f XNH,. The value of this method would of course be greatly enhanced if it were possible to ascertain exactly in the case of nzized diazoamido-compounds how much of each radicle is set free by the decomposition. Thus the azonaphthol-compound in such a case consists of the mixture-X*N,* C ,,H,* OH + Y *N,* C 1oHS.O H. It is only necessary therefore to find some method of estimating the amount of X or Y in the azonaphthol-mixture in order to get at the desired result. Experiments in this direction with the mixed compound (p)ClC,H,-N,H*C,H,.CH,(p) are now in progress. The triplet of methyl-derivatives of dinitrodiazoamidobenzene described in the present paper furnish another instance towards the generality of the isomerism which is now undergoing investigation. Our chief object in extending these researches will be to secure if possible triplets of mixed isomerides free from nitro-groups and experiments having this end in view have already been commenced. It gives us great pleasure in conclusion to express our thanks to Mr. E. H. R. Salmon who has given us material assistance by making most of the analyses and by preparing many of the materials used in the course of the investigation. Finsbury Technical C'dlege, May loth 1888

ISSN:0368-1645    

DOI:10.1039/CT888530664b

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

6 i 9 LIL-The Optical and Chemical Properties of Caoutchouc. By J. H. GLADSTONE Ph.D. F.R.S. and WALTER HIRBERT F.I.C. IN a paper on Essential Oils published in the Trans. 1886 609 one of us gave some observations on the refraction and dispersion of caoutchouc as well as on isoprene and caoutchene derived from it by distillation. As these seemed to show that the caoutchouc is a hydro-carbon in which a t least two pairs of carbon-atoms are doubly linked, it was deemed interesting to pursue the inquiry further. We have met with great difficulties part,ly from the doubtful purity of the substance partly from the readiness with which it undergoes transmutations. We think however that some of the results arrived at may not be without value. The earlier workers on the subject were able to get the natural juice of the tree but the analyses including those of Faraday and Greville Williams indicate the presence of some impurities.At the present day the juice is no longer imported. The substance we used was the best Para rubber kindly placed at our disposal by Mr. Willoughby Smith as well as another specimen of it and of Penang rubber obtained from Mr. Hancock. Para Rubber. This iwbber was pale in colour in pretty uniform lamine and had a sp. gr. of about 0.92. On analysis it contained as might have been expected from its origin some inorganic constituents. In Willoughby Smith’s specimen this was found to be 0.46 per cent. and in Hancock’s 0.38 per cent. Another constituent was water and to an extent that) was not expected by us.Willoughby Smith’s specimen gave 10-6 and 11 per cent. Hancock’s 7.1. As this water came off a t loo” it probably indicates nothing but insufficient drying. It is well known that part of the caoutchouc is soluble in benzene and other solvents and that a part is insoluble. Faraday also speaks of albumin as one of the constituents of the juice. We found the best solvent to be chloroform. This very slowly dissolves the whole of the hydrocarbon leaving a sort of network of the nitrogenous body. In one experiment this residue was found to be about 4 per cent. of the whole weight. We are inclined to believe that the less soluble modification of the hydrocarbon which has been frequently observed, is produced during the drying of the juice for we find that if the more soluble part be heated it is more or less changed and less susceptible of subsequent solution.The change increases as th 680 GLADSTONE AND RIBBERT THE OPTICAL AND temperature is raised and also with the length of time during which the heat continues. This will account for the very varied results obtained by different experimenters as to the proportions of soluble and insoluble caoutchouc. We made many attempts to separate these two modifications and to remove the oxidised product which commonly occurs with them. The most promising method seemed to be to dissolve the rubber in cold chloroform and precipitate partially with a little alcohol. The precipit,ate obtained was generally dried in a vacuum in order to avoid oxidation. We failed however to produce any good separation and the results of analyses gave usually about 3 per cent.of oxygen nor was this oxygen removed by the action of sodium on the solution. One small specimen however gave the following numbers which closely approach those deduced from CIOH16 :-0.1164 gram gave 0.3733 C02 and 0.1258 H20. Found. Calculated. Carbon 87.46 88.24 Hydrogen . 12.00 11-76 99-46 100~00 - -Optical Analysis. Accepting it as a fact that the carbon and hydrogen in the principal constituent of caoutchouc are in the same proportions as in the essential oils there remains the question as to the arrangement of the elements. It is quite easy to conceive that the CIoH16 is so built up that only one pair of carbon-atoms is double linked as in the terpenes, or two pairs as in the citrenes or there may be three pairs.Nor does it follow that CloH16 is its molecular formula C5H8 exists as one of its distillation products and C15H24 is not uncommon among essential oils. Now a knowledge of the refraction and dispersion of the substance would probably determine the first question and might throw light on the second. The theoretical refraction and dispersion equivalents of C~OH,~ in the three conditions above mentioned would be as follows assuming that each pair of doubly-linked carbon-atoms would produce the same increment as in the aromatic series. pPA - P F ( H - PA. C10H16. d o d. I pair of doubly-linked carbons 73.0 4.0 2 7 97 ,) . . 75.2 4.8 3 7 9 I> , . . 77.4 5. CHEMICAL PROPERTIES OF CAOUTCHOUC. 681 Refraction equivalent.In order to determine this point a number of specimens prepared in various ways were dissolved in benzene and the optical determina-tions made as usual. The following are the whole of the results obtained with the exception of three which were known to be untrustworthy. Dispersion equivalent. TABLE I. ' 5-13 ' 4.90 5 -02 I 5-31 4 *87 5 -01 5 *60 5 *01 5 -19 5 *67 5.34 5 -44 5 -40 6-13 5 -69 5 -63 5.18 No. of prepara-tion. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Strength of solution. 16 44 16 -66 13 -2'7 19 -71 11 -44 30-40 5 -83 14 -141 12-83 18 *24 17 -58 8 *24 15 -38 10.32 12 -18 9 -42 13 *92 76.02 76.90 74.20 73 -42 '73 -23 74 *58 75 -11 76 -94 75.69 76 -87 77 *59 75 -43 75 a69 74 -46 76 *02 76 *52 76 *01 Nos.1 and 2 were made from the original rubber as dissolved in chloroform and dried at 100". The others from 3 to 13 were made from the substance that was precipitated by alcohol from the chloro-form solution. This precipitate was dried in different ways No. 3 at 100" in hydrogen gas ; No. 4 in a Sprengel vacuum at 40" ; No. 5 over sulphuric acid for seven days ; Nos. 6,7,.8 were parts of the same preparation of which the carbon and hydrogen constituted 97.2 per cent. No. 10 arid 11 art alluded to below. Nos. 14 and 15 were the first and second fractions of a precipitate obtained by adding alcohol to an ethereal solution of the original Para rubber. Nos.16 and 17 were respectively the first and second precipitates similarly obtained from an ethereal solution of Penang rubber. It will be observed that there is no well-marked difference between these first and second precipitates. Now although the refraction and dispersion equivalents in the above table vary more than is easily accounted for they all agree in showing that the C,oH,6 must have more than one pair of carbon-atoms doubly linked. All the dispersion equivalents and 11 of the No. 8 was No. 7 concentrated by evaporation. No. 13 was dried in hydrogen at 100" 682 GLADSTONE AND HIBBERT THE OPTICAL AND refraction equivalents also exceed what theory requires for two pans of carbon-atoms so combined. They fall short of what would be required by three pairs at any rate in the case of refraction but it must be remembered that the presence of a little oxygen in the substances examined will pull down the refraction and dispersion considerably.3 per cent. of oxygen f o r instance would make a reduction of at least 1.4 in the refraction column and 0.13 in the dispersion column and if the addition of oxygen caused some of the carbon to be saturated, the reduction would be still greater. It is worthy of remark that the lowest figures are in No. 5 where there was along exposure to air during drying. The best observations are Nos. 10 and 11. For these the chloro-form solution was very carefully mixed with alcohol which caused the caoutchouc as it separated to rise to the surface like cream. After separating this cream was heated in hydrogen at 140° and after-wards dried in a Sprengel vacuum a t 125".I n this way we succeeded in preparing a substance of which the hydrocarbon formed 99.46 per cent. The solution was made in the case of No. 10 by dissolving a known weight of this caoutchouc in a known weight of benzene. I n the case of No. 11 the strength of the solution was determined by subsequent evaporation. The observations show a heightening of both optical properties in fact, the mean refraction equivalent 77-23 and dispersion equivalent 5-riO almost coincide with the calculated figures 77.4 and 5.6 given above. We have little doubt therefore that the main constituent of caoutchouc is a compound which for C,,H, has three pair of carbon-atoms doubly linked. If this be the case the molecular formula cannot be C5Hs like isoprene or CI5H21 like cedrene as these would give respectively one and a half and four and a half pairs of carbon-atoms united by double linking.It cannot contain the hexagonal ring but must be expressed graphically by a chain formula. This may account for the wide difference of properties between caoutchouc and the various essential oils. The analysis has already been given. Halogen Compounds. An endeavour was made to test the previous conclusions by chemical methods. The chloroform solution appeared to offer a good oppor-tunity of studying the action of halogens on caoutchouc and the results were considering the small amount of oxygen always present, sufficiently definite. We did not succeed in obtaining a simple additive compound with chlorine.On passing the gas into a chloroform solution of caoutchouc containing about 1 per cent. of the hydrocarbon-the experiment bein CHEMICAL PROPERTIES OF CAOUTCHOUC. 683 made in weak diffused light-fumes of hydrocliloric acid were at once evolved and this continued for a very long time. The general results of many experiments indicate that substitution and addition were taking place simultaneously. On evaporating the solution a residue was always obtained in slightly yellow scales. Different specimens did not give the same results on analysis. The best preparation which we succeeded in making gave the following figures :-0.0575 gram gave by Carius’ met,hod 0.1515 AgCl. 0.1067 , 0.1108 GO and 0.0305 H,O. Carbon.28.32 Hydrogen 3.1 7 Chlorine 65.18 96.67 The presence of oxygen renders the true interpretation of these figures somewhat uncertain but they agree fairly with the ratio CloHI4C1, which will require for 28.32 per cent. of carbon 3.3 per cent. hydrogen and 66.95 per cent. chlorine. Such a result is in accordance with the view that the hydrocarbon contains six un-saturated atoms of carbon and is not easily explained on any other supposition. Action of Bromine.-Bromine acts energetically on caoutchouc dissolved in chloroform fumes of hydrobromic acid being evolved. But a weak solution of bromine in chloroform while acting at once, does not give rise in the first instance t o a very rapid separation of hydrobromic acid though upon standing the gas slowly appears.This suggested the making of a synthetic and volumetric determination of the proportions in which caoutchouc and bromine combine. Standard solutions of each were therefore prepared and the latter was added from a burette to a measured quantity of the former. As the close of the experiment was approached the bromine disappeared a little less quickly. The flask containing the liquid was then shaken, and a drop taken out by means of a glass rod was brought into contact with a drop of potassium iodide starch solution. When the blue iodide of starch appeared. the presence of free bromine was shown, and this was taken to indicate the completion of direct combination bet ween the bromine and caoutchouc. The following are the figures obtained :-The solution of bromine contained 0.03609 gram per C.C.The solution of caout.chouc contained 0.02458 gram per C.C 684 GLADSTOKE AND HIBBERT THl3 OPTICAL AND TABLE 11. Bromine used calcu-lated for 136 parts Caoutchouc taken. Bromine used. of caoutchouc. 0.1229 gram 0.2670 gram 295.5 0.1229 , 0.2634 , 291.5 0.4917 , 1.0570 , 292.4 Mean 293.1 This number (293) was judged to be somewhat too low since the very faintest indication of free bromine had been accepted as an indi-cation of completed action. A better and quicker method of deter-mining the value was founded on t-he subsequent observation that iodine did not act on the solution of caoutchouc in chloroform. It occurred to us that by first adding an excess of bromine and then a strong solution of potassium iodide we should get a quantity of iodine set free equivalent to the excess of bromine and the free iodine could be estimated at leisure by thiosulphate.On trial of this method very fairly accordant results were ob-tained :-The solution of bromine contained 0,03684 gram per C.C. The solution of caoutchouc contained 0-0281 gram per C.C. TABLE 111. Caoutchouc taken. -gram. 0 -141 0 -141 0 *141 0 -1692 0.1692 0.141 0 -141 0 *141 0 -282 0.141 Bromine added. --gram. 0.3684 0-3684 0 '3684 0 *4420 0 -4052 0 -3352 0 -3205 0 -3168 0 * 6447 0 '3278 Excess of bromine. .- I 0.05157 0.05268 0.05157 0.0663 0 * 0405 0'0221 0.0051 0 -00368 0.0173 0.0081 Bromine required. -0 *3168 0.3157 0 -3168 0 -3757 0 *3647 0.3131 0.3151 0 -3131 0 %274 0 -3197 Bromine required for 136 parts caoutchouc.305 -5 304 ' 5 305-6 302 -6 293 '1 301 *9 303 -9 301.9 302.5 308 * 3 The average result is that 136 grams of caoutchouc combine with 303 grams of bromine." Supposing that 1 mol. of C,,H1 combined * Greville Williams has employed bromine to test the saturating power of caoutchene one of the hydrocarbons obtained by the destructive distillation of caoutchouc. The properties of caoutchene allowed a simpler procedure to be followed. His experiments show that 136 grams of caoutchene combine with 315 grams of bromine CHEXICAL PROPERTIES OF CBOUTCHOUC. 685 directly with 4 atoms of bromine to form the tetrabromide CloH16Br4, it would require 320 grams of bromine.When we consider that the caoutchouc experimented on contained about 3 per cent. of oxygen, the figures obtained appear as satisfactory as could be expected. The result does not agree of course with the constitution assigned above to caoutchouc and can be reconciled only on the supposition either that the action of the halogen has produced an isomeric change, or that the tetrabromide formed is not yet a saturated compound. Attempts were made to isolate this compound by evaporating down the solution or by precipit'ating with a large excess of ether. The precipitate was a white solid body but it always proved to be unstable on drying with loss of hydrobromic acid. Prolonged Action of Brorrzine.-A fair excess of a solution of bromine in chloroform was added to a solution of caoutchouc in chloroform and allowed to act at ordinary temperatures for three days.Fumes of hydrobromic acid were evolved all the time. The solution was then poured into a considerable quantity of ether which caused a white precipitate. This was separated by filtration and washed with ether. When dried at 60° it was quite white and easily powdered. 0.1312 gram gave 0.1091 CO and 0.0348 H,O. 0.1226 , 0.2095 AgBr 0.1054 , 0.1819 ,, 0.1270 , 0.2182 ,, Estimated by Carius' method. 7 Calculated for -L-1 Found. I. 11. 111. IV. C10H1.5Br5* Carbon . . . . 22.67 - - - 22-43 - 2.80 Bromine . . . 72.71 73.43 73.10 74.76 Hydrogen . . 2-94 - -Remembering that; some oxygen was present the experimental results agree sufficiently well with those calculated from the formula CloHl,Br5.I t seems not improbable that this body is formed from a compound, CloH16Br6 by the elimination of HBr. Iodine.-We have already stated that iodine has little or no action on caoutchouc dissolved in chloroform. It is perhaps worthy of note that Adriani (Chem. News 2 278) says that the action of iodine on the globules in the natural juice is more marked than that of bromine. Action of Heat. The ordinary statement is that caoutchouc when heated softens and melts at about 200" ; that if cooled again it never becomes hard as before 68 6 GLADSTONE AND HIBBERT THE OYTICAL AND We found that the fusing point depended partly upon the exposure of the subst'ance to air oxidation affecting the result.0.5 gram of caoutchouc was heated a t 200" for two hours in a sealed tube filled with hydrogen. There was only a slight superficial fusion and scarcely any decomposition. A solution of caout'chouc in toluene was also heated in a sealed tube a t 200" for two hours and the solvent then distilled off at 112" in a Sprengel vacuum. It left a hard residue and there was no odour of decomposition. Even when heated at 210" in a Sprengel vacuum, caoutchouc was found to be only superficially melted though after that treat,ment i t dissolved very slowly in benzene. I n order to ascertain whether the optical properties of caoutchouc were changed by heating we dissolved some of Hancock's Para rubber in toluene determined its refraction and dispersion then exposed the solution in a sealed tube to a temperature of 200" for three hours and after cooling examined it afresh.The solution was found far more limpid than it had previously been. The other properties were as follow. The solution contained 12.85 per cent. of caoutchouc :-Temperature. Sp. gr. PA. P H . Before heating 19.4" 0-8706 1,4896 1-5295 After , . . 18.5 0.8708 1-4895 1.5294 These figures are practicaIly identical and it may be concluded that if the heating has produced any change it has not affected the pro-portion of carbon-atoms doubly linked. It is evident also that its lnolecular volume in solution has not been affected. The optical properties of the caoutchouc as deduced from the above figures are refraction equivalent 76.46 ; dispersion equivalent 5.40, which confirm the observations made from the solutions in benzene.When caoutchouc is heated considerably above 200" it is converted into a mixture of liquid oils which are believed t o be isomeric or polymeric with the original substance. From the distillate have been separated and more or less fully described isoprene caoutchene and heveene. Isoprene C5H, has been already shown to contain two pairs of doubly-linked carbon-atoms (Trans. 1886 619). Caoutchme CI0Hl6 was shown at the same time to have two pairs of doubly-linked carbon-atoms. The ratio of unsaturated carbon to molecular weight is here only half that of isoprene. We prepared two specimens of caoutchene by repeated fractional distillation ; the first from a quantity of oil made by ourselves from Para caoutchouc the other from oil obtained from Messrs.Hopkin and Williams. The examination of these gave the figures shown i CHEMICAL PROPERTIES Ol? CA OUTCHOUC. 687 3'. 1 '4807 1 *4791 1 '4793 1.51'72 1.5265 1 * 5249 1.5371 Tables IV and V Nos. 1 and 2. The second specimen when distilled from sodium several times gave the figures of No. 3 showing that it was practically unchanged notwithstanding the appearance of a little red flocculent matter. Eeveene.-From the same two mixtures we also separated a liquid boiling at about 315". It was dark in colour and became darker o n standing. The results of observation are given in Nos. 4 and 5 and are very accordant. The observations made upon it are given in No. 7. The sp. gr. is increased, but the specific optical properties remain practically the same as those of heveene.As the refraction equivalent in these cases was a little lower than we should have expected on the supposition that these bodies of higher boiling point were similar in composition to caoutchouc we thought they might contain some quantity of an oxygen product. On treating our heveene with sodium we found that some hydrogen was liberated and it was therefore redistilled from sodium several times. A nearly colourless oil was obtained having an intense blue fluores-cence. This caused the spectrum to be nearly cut off between G and H. The results of observation are given in No. 6 showing that it has been only slightly increased in refraction and dispersion. Several attempts to obtain the vapour-density failed apparently through some molecular change taking place in the vapour itself when highly heated.The product boiling above 320" was still darker in colour. H. --1.4950 1 *4938 1 * 4939 - -1 - 5421 -TABLE IV. -NO. -1 2 3 4 5 6 '7 7 Substance. -Caontchene . . . . . . . . . . . . . . . . . . Heveene . . . . . . . . . . . . . . . . . . . . . . . . Higher product . . Boiling point. 174-1'7'6" 1'73-17'7 173-178 300-320 310- 316 312-318 above 320 -Temp 18.0" 19 *o 22.0 17 -8 16.5 22 *6 16 -3 -0 *a377 0 Ti361 0 -8350 0 *9146 0 *9291 0 * 9245 0 *9487 Refractive index. A. 1 -4657 1 *4644 1 -4638 1.5009 1 *5090 1 *5074 1 -5190 After distilling off the highest product there remained a quantity of a substance not volatile at low redness and which on cooling became quite hard.As the whole had been originally distilled we may suppose that this product had been formed by the subsequent action of heat ; indeed there were other indications that polymerisa 688 PROPERTIES OF CAOUTCHOUC. T.ABLE V. NO. Refract ion equivalent. I 1 Caoutchene . . . . . . . . . . . . Heveene ,’ Higher product 75 *GO 75 *54 75 -54 74.48 74-51 74 *64 74.40 -Dispersion equivalent. F - A . 1 H-A. I-- -2 -443 2.39 2 ’53 2 ‘43 2 -56 2 ‘58 2 *60 tion took place during redistillation. It is probable also that there are bodies produced which are intermediate in boiling point between those examined. The general results of the destructive action of heat upon caout-chouc may therefore be considered as involving no change in the proportion of carbon to hydrogen but changes of structure in the hydrocarbon which are best represented by the following optical con-stitutional formuke. In this table C” is used for carbon having the refraction equivalent 6.1 and dispersion equivalent 0.66. Optical constitutional Substance. formule. Caoutchouc. mC’’&qH1~ Caoutchene . c’’pC~H16 Isoprene C”aCHs Heveene nCfr2C3H,. Since the paper was read at the Society’s meeting we have eiideavoured to determine the molecular weights of caoutchouc and heveene by Raoult’s method. For heveene we obtained depressions of 1-85’ C. and 1.52“ C. with solutions containing respectively 9.37 and 7-68 grams to 100 of benzene. These figures give 248 and 247.5 for the molecular weight of beveene which strongly indicates the formula C20H32. As to caoutchouc the depression obtained with a solution contain-ing 9 parts to 100 of benzene was so very small that its molecular weight must be at least 50 times that of hereene if the method holds good

ISSN:0368-1645    

DOI:10.1039/CT8885300679

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

LIII.-On an Apparatus f o r Maintaining n Constant Pressure when BistiZling under Reduced Pressure. By W H. PERKIN Ph.D. F.R.S. THE great value of conducting distillations under reduced pressure in the separation of volatile products from non-volatile ones in fractioning substances decomposable at high temperatures &c. is now so well recognised that it is unnecessary to comment upon it, and as it is a method of working so much resorted to at the present time anything that can render it more simple and manageable is of service. Those who have conducted distillation under reduced pressure know that the constant watching which is usually required to keep the pressure uniform very much interrupts the attention which i t is necessary to give to the distillation itself and as the work I am engaged in frequently requires me t o resort to distillations of this kind a variety of experiments have been made t o obtain some simple and effective automatic arrangement for this purpose and, thinking the results might be useful to others I have ventured t o lay them before the Society.The first idea which naturally suggested itself was to regulate the pressure by means of a column of merciiry so arranged that air would be drawn through it as soon as the desired reduction of pressure was attained. Arrangements on this principle have been used by Lothar Meyer and others (AnnuZen 165 303 ; 198 218), but my experiments in this direction especially for large reductions of pressure were not satisfactory and some more mechanical method was sought for.The use of a valve opening inwards and regulated by a lever and weight was next tried. This however,a-as also found to work unsatisfactorily because when the regulating point was reached a state of equilibrium existed and directly the air commenced entering vibration set in the valve chattering and not coming to rest again anti1 a considerable change of pressure had taken place. This showed that what was wanted was a valve that would act not only with considerable force and decision both as to opening and closing when the desired reduction of pressure was obtained but also with very small variations of pressure. This was attained eventually by means of a valve weighted to make it close quickly and capable of being opened by an electr~ma~gnet the action of the latter being inflnenced by a column of mercury making electrical contact with a platinum mire as soon as it reached the desired height.An apparatus made on this principle has been in use in my laboratory for several VOL. LIII. 3 690 PERKIX APPARATUS FOR RTAINTAINING years and worked satisfactorily. The following is an account of it given in detail as there are several points which require attention to make it work well. The description of the apparatus will be under-stood more easily by refereiice t o the diagrams which are given. At first, valves of metal and also of caoutchouc were experimented with b u t as they did not work very well glass was eventually used. The shape of the valve is that of part of a sphere a (Fig. l) half an inch in diameter ; it fits into a glass seating b t o correspond both surfaces The first thing to be considered is the regulating valve.F I G . 1. being very carefully ground and polished together ; there is a hole passing through the seating for the inlet of air. The valve has a hole drilled through it into which a wire with hooks on both ends is cemented. One hook is to carry tt weight c the other to suspend the valve to an electromagnet. The valve and seating is fixed into the movable top of a brass cylinder d in which the weight attached t o the valve hangs. A tube e is inserted into the side of the cylinder to connect it with the part of the apparatus which is to be exhausted by the water-pump 01' other means. The upper hook on the valve is attached by means of a copper wire,f to a ball and socket joint provided with a long screw passing through the armature of th r-4 I d' Fe.3 A CONSTANT PRESSURE. 69 1 magnet a s seen in Fig. 2 ; this screw is provided with a milled head, g so that the valve may be adjusted in relation t o the armature. There is also a nut above this and working on this screw which when tightened up to the armature locks the screw so that the adjustment may not alter by vibration. The electromagnet and other parts of the apparatus are fixed on a rigid mahogany stand as seen in the diagram and need no explanation. The armature is kept in its place at one end by a steel spring la the other resting in a brass support provided with an adjusting screw i so that the distance of the armature from the electromagnet may be amanged when necessary.This apparatus is kept in a box so that it may not suffer from dust and laboratory fumes a hole beipg pierced in the side of the box t o receive the connecting tube whilst binding screws are fastened on its top to connect with the necessary wires. The arrangement by which the desired reduction of pressure can be regulated consists of a graduated glass tube about 7 or 8 mm. in diameter and about 800 nim. long. A piece of tube about 70 mm. long and 12 in diameter is fused on to the tup and from its side an ordinary piece of quill tubing passes being afterwards bent twice at right angles ( j j j Fig. 3). This is connected with a T-piece k one arm of which passes t o the receiver consisting of a strong bottle of about 4 or 5 litres capacity the other arm being connected with the regulating valve by the tube 1.A stout copper wire m about 600 mm. long passes down the graduated tube ; this is provided with a platinum point screwed into it at the lower end and a binding screw on the upper end. Near the lower end a piece of copper is screwed on with three projectirtg points so as t o keep the copper wire in a central position in the tube j . This wire fits air-tight into the enlarged upper part of the tube by means of a vulcanised cork. The lower end of the graduated tube is drawn off sideways as in Fig. 4 leaving an aperture of about 0.5 mm. diameter (the object of this will be described further on) and is placed in a vessel of mercury m"" Fig. 3. A thin copper wire m' passes from the mercury in this vessel t o one of the binding screws of the regulating valve.Another in" passes from the binding screw on the thick copper wire to the battery and a third m"' from the second binding screw of the battery to the second one on the regulating valve. The working of the apparatus will now be easily seen. The copper wire 132 is drawn up till its point corresponds pretty nearly to the height of the column of mercury necessary t o be used for the desired pres-sure. On setting the water-pump in action the air being gradually removed from the glass bottle the mercury will rise in the graduated tubej but the moment it touches the platinum point the circuit is completed and the electromagnet lifts the glass valve so that some 3 A 692 PERKIN APPARATUS FOR MAINTAINIKG air enters ; this causes the mercury column to fall but the moment it does so the circuit is broken and the valve closes.It then rises FIG. 4. again ; the circuit is again completed and the valve opens and so on. I n this way when all the adjustments referred to further on are attended to a nearly perfectly regular pressure may be obtained not varying more than a fraction of zi millimetre. When first trying this apparatus the graduated glass tube j j j, was not cont'racted at the bottom but it was found that owing t o the sparking which takes place when contact is made between the wire and the mercury thus heating the rarefied air in its vicinity and the action taking place a t nearly equal intervals oscillations were set up in the mercury column and the apparatus therefore worked un-steadily but by contracting the end of the tube sufficiently as in Fig.4 the oscillations are almost prevented and the fluctuations of the columns which take place are only sufficient to regulate the pressure that is provided the regulating valve be properly adjusted, because it is evident that if it opens too wide it will let in so niuch air that the mercury will fall some distance before it closes again and if it does not open suGciently there will not be enough air admitted to counteract the excess of pumping power of the water-pump ; but with a little attention this matter can be arranged by turning the milled head g Fig. 2 on the regulating valve so that the quantity of air admitted only slightly influences the column of mercury.When read-ing the height of the column of mercury in the graduated tube j j j , the depression of the mercury in the reservoir 172"" must not be forgotten to be taken into account. As the column of mercury in the graduated tube cannot obviously be kept quite quiet a second tube was originally placed a t its sid A CONSTANT PRESSURE. 693 standing with its open end in the mercury reservoir the other end being connected with the exhausted receiver and this mas used as a barometer to measure the pressure and when the apparatus was working well no appreciable fluctuations mere noticed in this tube ; when measuring the pressure in this way however it was always neces-sary to compare it with an ordinary barometer giving the atmospheric pressure and subtract the one from the other to get the pressure the apparatus was working at.It was therefore thought that it would be more simple to connect it with a barometer direct instead of using this second tube and thus get the correct pressure at once. The most convenient way of doing this is by means of a syphon barometer (n Fig. 3) with the return tube made more than long enough to pass the zero point of the graduation. To the side of this, and near its end a tube 0 is fused to connect it with the glass bottle used for the exhausted receiver the open end of the tube p being closed with a cork. To make the apparatus more easy to move about, the syphon barometer is provided with a glass stop-cock q in the lower part of the return tube so that by tilting i t fills with mercury ; the stop-cock being then closed it can be carried about without danger.* When first trying this barometer it was found to woi+ very unsteadily.Owing to the vacuum in it it was found to be excessively sensitive t o minute changes of pressure especially when occurring at regular intervals oscillations setting in but it was eventually found that by turning the stop-cock p partially off taking care not to prevent the action of the barometer these oscillations practically were stopped; or the same thing can be effected by drawing out the tube which connects the barometer with the exhausted receiver to a fine point; when well adjusted only slight fluctuations will be seen on the meniscus of the mercury. This apparatus works successfully at a pressure of 60 mm.this being about the limit of the water-pump with which it has been used, but no doubt it would do equally well for lower pressures. For most work a pressure of 200-210 mm. is most convenient as the boiling is then generally pretty steady and there is no need to make arrsnge-ments for the admission of minute quantities of air its when using very low pressures. This pressure gives a reduction of over 40" in the boiling point. The apparatus employed in conducting the distillations is practically the same as that described by my late assistant Dr. L. T. Thorne (Trans. 1883 301). It has been in use in my laboratory ever since it was first devised and is found to be extremely convenient especially for fractional distillations.It is attached to the tube s. Q This would probably be a convenient adjunct to ordinary syphon barometers 694 APPARATUS FOR MAIXTAINING A COSSTANT PRESSURE. The battery power used for the regulating valve apparatus should only be sufficient to cause the apparatus to work with decision other-wise if strong sparking takes place between the mercury and the platinum point the surface of the mercury becomes dirty and the apparatus does not work well ; but if the power be kept down this takes place but slowly if the tube is of the diameter given above. A bichromate battery as shown a t r is convenient the power being regulated by the distance the zinc is allowed i o go into the chromic mixture. The tube j j j in which this sparking occurs should be attached tc the stand in such a way that it can be easily removed for cleaning when necessary.Experiments were made with floats covered with platinum foil so that there should be no sparking on the mercury. It was found that the best floats to work in a tube were small flat discs of metal covered with platinum upright ones attach themselves t o the side of the tube. These work pretty well but not SO sharply as the mercury itself a certain amount of pressure being sometimes necessary to make good contact and this necessitates a greater rise in the mercury than is necessary with mercury only the fluctuation, however is but small ; nevertheless I prefer to use the mercury alone, and to clean the tube when necessary. The use of a good sized bottle for the exhausted receiver (about 4+ litres) is advantageous as it assists in keeping the pressure equal and prevents any considerable alteration in it when changing receivers as in fractional distillation the small amount of air which enters having but little influence.If using this apparatus in connection with any powerful arrange-ment for pumping out the air it would be necessary to restrict the connection with it by means of a stop-cock capable of pretty fine adjustment or by a tube of fine bore because it is evident that if the pumping power is in excess of the capability of the valve to let in air it can no longer act as a regulator. The use of the weight on the valve is not necessary when large reductions of pressure are being maintained as the atmospheric pressure is sufficient to close it but for small reductions it is believed to be of service. Since devising this apparatus my attention has been drawn to a very ingenious pressure regulator devised by F. D. Brown (Proc. Phys. SOC. Lond. 1879 3 68) which has some points in common with mine but it has the disadvantage of requiring mot'ive power to work it as well as a battery

ISSN:0368-1645    

DOI:10.1039/CT8885300689

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

LIV.-CChlorofu,maric and Chloromaleic Acids and the Magnetic Rotatory Power of some of their Derivatives. By W. H. PERKIN Ph.D. F.R.S. IN a paper published in the Journal of this Society by the late B. F. Duppa and myself in 1861 (J. Ohem. Xoc. 13 9) an account is given of some experiments upon the "Action of Pentachloride of Phosphorus on Tartaric Acid." The results of these mainly consisted in the formation of a chloride which with water yielded an acid and to the latter we gave the provisional name of chloromaleic acid for reasons given in the paper. Later on it was shown that this acid yielded succinic acid when treated with nascent hydrogen (J. Cherri. Xoc. 1863 16 198). Being desirous of examining the magnetic rotatory power of the derivatives of this chlorinated acid and of getting further chemical information in coniiection with it fresh experiments were made and the following results obtained.Tartaric am? Racemic Acids with Pewtachloride of Phosphorus. About 225 grams of phosphorus pentachloride and 38 grams of finely-powdered tartaric acid were mixed and heated on the water-bath in a large Wurtz flask until the mixture became liquid. On examining this with the polariscope i t was found t o have a per-manent rotation of 7' 6" for a length of 102 mm. It was distilled until the temperature of the vapour rose t o 130" and then transferred to a smaller retort and again distilled until the temperature of the vapour was 160" so as to remove as much oxychloride of phosphorus as possible. Finally the residue was fractioned under reduced pressure (210 mm.) when nearly all distilled over at 142" a little at last coming orer somewhat higher leaving a small carbonised residue in the retort.In a second experiment 280 grams of phosphorus pentachloride were used to 38 grams of tartaric acid ; this was a considerable excess of the former and after the reaction had been completed on the water-bath and the fluid allowed to cool part of it crystallised out. The liquid had a smaller permanent rotation than that from the previous experiment amounting t o only 1' for a column of 102 mm. It was treated as in the previous experiment and gave a product boiling at about 144' at 210 mm. The yield obtained was about half that required by theory. Several similar preparations were made with racemic acid dried at 100".The reaction appears to go more neatly with this acid and the I t had a very small permanent rotation 696 PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. product obtained is of better quality. It boiled at 141-143" at 210 mm. a little guid of higher boiling point coming over at last, leaving a small quantity of carbonised residue. The yield was nearly the same as the weight of racemic acid used. This substance as will be seen further on is chlorofumaric chloride and not chloromaleic chloride. It is a very slightly yellowish liquid boiling at 142-144" corr. under 210 mm. but at 184.5-187-5" corr. under 760 mm. with slight decomposition the distillate being yellow, the colour becoming darker towards the end of the operation and the thermometer rising t o 195" hydrogen chloride also being given off and slight carbonisation taking place.When cooled with ice and hydro-chloric acid it does not solidify. Ih fumes slightly in the air and when added to water sinks as a heavy oil which gradually dissolves on agitation leaving any neutral chlorinated oil it may contain. The chloride prepared from tartaric acid usually contains a small quanticy of a neutral chlorinated oil attacking the eyes very much ; that pre-pared from racemic acid was free or nearly so from this product. This oil boils at about l80" it was not obtained in sufficient quantity for examination. If some quantity of the chloride be agitated with about its own bulk of water the mixture gets hot and boils. Chlorofumaric chloride reacts with strong ammonia with great violence and the resulting solution deposits a crystalline substance on standing.If t o an ethereal solution of the chloride an ethereal solution of aniline be added energetic action takes place a solid substance separating out ; this when crystallised from alcohol in which it is not very soluble was obtained as pale yellow needles melting at The following density determinations were made of three different 186-188". preparations of the chlorof umaric chloride :-I. From tartaric acid. B. p. 141 -6-143 -6". 210 mm. I T . From tartaric acid. B. p. 143-145". 210 mm. 15" d- 4" 1.5956, 4 d- 1.5826, 15" 20" 20" 25" 25 d7 1.5764, d- 1.5717, 15" 15 25" 25 d- 1.5807, d, 1.5690. III. From racemic acid.B. p. 14a-144°. 210 mm. & 1.5890 4" 15" 15" d- 1.5731, 25" 25 d- 1.5623 PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. 697 t. 1 sp. rotation. The following numbers were obtained f o r the magnetic rotation of this substance Specimens I1 and 111 being used. Mol. rotation. Specimen II. From Tartaric Acid. 21 '0° 21 -5 22 -0 Average 21.5 1 '5025 9 -951 1 '5019 9 *950 1 *5015 9 *951 1 *5017 9 -951 -, Sp. rotation. 1 '5250 1 *5249 1 -5269 1 -5256 Specimen III. F~om Racemic A c i d . Mol. rotation. ---__ 10 '134 10 *133 10 -145 10 * 13'7 --______ t . 20 -2" 20.2 20 *2 Average 20.2 The average of these two sets of determinations is 10.044. Chlorof zcnzaric; Acid. This acid previously called chloromaleic acid is obtained from the chlorofumaric chloride by the action of water.The aqueous solution is separated from any oily products evaporated t o a small bulk and allowed to crystallise then drained and pressed ; if crystallised twice more it is pure. The crystallisation is much facilitated by the addition of some strong hydrochloric acid chlorofumaric acid being much less soluble in water in presence of this acid ; this is also true of its alcoholic solution. Chlorofumaric acid forms very minute prismatic crystals and when dry looks like a white amorphous powder. It is easily soluble both in water and in alcohol. It gave the following numbers on analysis :-0,2023 gram of substance gave 0.2376 gram of CO and 000.398 Theory for CqH3C104. Pound. Carbon 31.@9 31.58 Hydrogen .. . . . . . . 2.18 OH gram. 1 . 9 698 PERKIN CHLOROFUMARIG AXD CHLOROMALEIC ACIDS. It melts at 189-190" or at 191m5-192*5c corr. Rauder who has also examined this acid (J. pr. Chem. [Z] 31 28) gives 191". If kept heated so that it gently boils it at first sublimes to a small extent then gires off gas regularly consisting of hydrogen chloride and about equal volumes of carbonic anhydride and carbonic oxide, most of the product disappearing. Heated with bromine in a sealed tube at lo@" it does not combine with it. Hydrogen Yotassizm Ch1orofumarate.-This salt was prepared from the chlorofumaric acid obtained from tartaric acid and also from that from racemic acid to see if they were identical no difference could be detected.Determinations of their solubility in water gave the fol-lowing numbers. Temp. = 15". Salt from tartaric acid ; 100 parts of sol. gave 3.852 of salt. , racemic , 9 7 7 3.834 7 9 For the following measurements of the crystals I am indebted to Herr. V. Ussing. Potassium Clzloro %marate C.rystallised from Water. Asymmetrical. a b c = 0.5223 1 0.7910. a = 91" 22'; /!I = 105" 5 5 ' ; y 53" 47'. Forms (010) mPm; (101) mPm; (001) OP; (110) mP ; (011) Thin in the direction of the brachypinacoid, In individual crystals the two vertical pinacoids are equally ,p'm ; (102) g,P,m. elongated in the direction of the axis c. developed. The hemiprism was only rarely observed. Observed. Calculated. (100) (010) 122" 26' -(100) (001) 71 11 -(010) (001) 100 16 -(001) (ioe) 77 28 -(001) (011) 48 7 c (ioi) (011) 55 10 54" 57' Cleavage perfect in the direction (loo) less perfect in the direc-tion (ioa).The direction of extinction on (100) is inclined 15" to the left from the vertical axis on (010) it is also inclined about 32" to the left. The optical axis cuts (100) i n a line which forms an angle of 75" with the crystallographic axis c to the right above ; below to the left PERKIN CHLOROFUMARIC AND CHLORO,\lhLEIC ACIDS. 699 The trace on (010) forms an angle of about 89" with the axis a, to the right above ; below to the left. I n Schneider's apparatus both axes are perceptible in a cleavage plane parallel to (100). The angle for the sodium line in glass is 79". The bisectrix is almost vertical to the macropinacoid.Strong negative double refraction. For particulars of the preparation and properties of this salt see J . Chem. Xoc. 13 10. Ch lorofurnarate of Ammonium.-Chlorofumaric acid combines with energy with ammonia and if an excess of the latter be left with it it does not cause any decomposition by removing chlorine as it does in the case of the ether. On slowly evaporating the aqueous solution of this salt it separates out in beautiful transparent well-defined crystals. They are moderately soluble in water and neutral to test-paper dried at 100" they do not lose weight a determination of the chlorine in the salt gave 18.87 per cent. the formula C,HCl( NH,) 2 0 4 requiring 19.24. this salt' I am indebted to Herr W. Mushmann. For the following measurements of the crystals of Ammonium Ch lorofum arate.Monosymmetrical. a b c = 1.3892 1 1.0059. ,G = 71.12". Crystalline forms (100) mPm; (110) wP; (001) OP; (ioi) +gm and (010) mFw. The last plane rarely seen and quite small. Observed. Calculated. (100) (110) X.52" 45' -(100) (001) *71 12 -(loo) (ioi) ~ 6 7 o -(110) (001) 79 0 78" 49' (iio) (ioi) 76 21 76 19 Cleavage very perfect in the plane of symmetry ; colourless ; trans-parent . The opbical axial plane is vertical to the plane of symmetry and almost coincident with the base (001). The axes pass through the plane of cleavage towards (OlO) but out of the field of vision; they are perceptible only in Schneider's apparatus. With moderately dilute solutions of alkaline chlorofumnrates neither barium calcium nor mercuric chloride give a precipitate but wit 700 PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS.strong solutions calcium chloride gives a white crystalline precipitate sparingly soluble in hot water. The colour of a solution of copper sulphate becomes darker on the addition of a chlorofumarate. Like the fumarates chlorofumarates give with ferric chloride an amorphous, light brown or buff precipitate. Chlorofumttric acid boiled with copper carbonate produces a tur-quoise-blue powder very little soluble in water. A solution of copper acetate boiled for some time with this acid deposits a dark turquoise-blue powder seen under the microscope to consist of small crystalline nodules. Ethyl CIz lorofurnarate. This was at first prepared by gradually adding chlorofumaric chloride to alcohol washing the resulting ethyl-derivative with water and sodium carbonate drying it over potassium carbonate and distilling undei.reduced pressure (210 mm.). A little hydrogen chloride usually came off towards the end of the distillation the boiling point rising a little. Two specimens were analysed one prepared with the chloride from tartaric acid b. p. 200-202", 210 mm. and the other with the chloride from racemic acid b. p. 199-201*5" 210 mm. They gave-I. 11. Theory for Ether from Ether from C8H,,C104. tartaric acid. racemic acid. Carbon . 46.5 45-54 46.18 Hydrogen . . . . 5.32 5.38 5-39 These numbers show that the chloride from racemic acid gave the purer product. Their density determinations gave for I d- 1.2036 ; 15" 15" 1 15 for 11 a3- 1.19845.The magnetic rotations were determined but as the ethers were not considered quite pure the details are not given. No. I gave 31. rot. 11.225 and No. 11 11.300. A quantity of this ether was prepared from pure chlorofumaric acid obtained from the chloride made from tartaric acid and alcohol. On saturating the alcoholic solution with hydrogen chloride a good deal of acid was deposited but on standing redissolved and etherified; the ether after being washed treated with sodium car-bonate and dried boiled a t 202-203" at 210 mm. A little hydrogen chloride was formed therefore the product was left over a little potassium carbonate until this was removed and then filtered for use. A portion of this ether when distilled under a pressure of 760 mm.boiled a t 250" constantly with but little decomposition PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. 701 t. Henry (AnfluZen 156 178) prepared this ether by acting with phos-phorus pentachloride on ethyl tartrate. The boiling point he ob-tained was 243-245" at 735 mm. or about 245.5-247.5' at 760 mm. The density determinations of the ethyl chlorofumarate obtained from the pure chlorofumaric acid and alcohol gave-Sp. rotation. 4" 4 d- 1.2048, 10" 10" a- 1.1990, 15" 15" d- 1.19372, 20" 20 d- 1.1893, 2 5" 25" d- 1.1849. It will be seen that these numbers are lower than those obtained with the ether prepared from the chlorofumaric chloride. It may be as well to mention here that the ether obtained from the chloride pre-pared from t a r t a h acid had a slight permanent rotation to the right.One specimen in which the pentachloride of phosphorus was used in excess for the preparation of the chloride gave 15' for column 102 mm. long; another in which less pentachloride wa.s used gave 28' for the same column. The ether prepared from the pure acid had a permanent rotation of 5' 30" f o r column 102 mm. long. I t is difficult to say whether this rotation is due to the pure ether or not but it appears to be so. bers :-The magnetic rotations of this ether gave the following num-I-- -19 -0" 19.3 19 -6 16 *? 16'6 16 -5 Average 18.0 1 *1785 1 -1606 1 *1794 1 -1833 1 * 1838 1.1'798 1 *18@9 --Mol. rotation. 11 -365 11 '387 11 ' 3 i 7 11.369 11 -392 11 -352 11 -3'77 Ethyl chlorofumarate has a vinous ethereal odour ; cooled in ice and hydrochloric acid it does not solidify.It has rather an irritating action on the skin. Heated with aniline to boiling chemical action sets in and the product on cooling solidifies to a crystalline mass from which a neutral substance crystallising from alcohol can be obtained 702 PERKlK CHLOROFUMARIC AND CHLOROMALEIC ACIDS. Ethyl Chlorofumarate and Anznionba. When experimenting on the subject of the action of ammonia on ethyl chlorofumarate I was not aware of the results which had been obtained by Clam and Voeller (Ber. 14 151) in this direction but as I have operated in a different manner t o them it may be useful if I briefly give the results of my experiments.Claus and Voeller used the same ether as I did but called it chloromaleic ether and their compounds are therefore called ma leic-derivatives but as the ether is now known to be ethyl chlorofumarate, their nomenclature requires altering accordingly. They employed alcoholic ammonia of a known strength and in definite quantities. My experiments were made with aqueous ammonia. On agitating ethyl chlorofumarate with ammonia of 0.880 diluted with about three times its volume of water i t gradually dissolved, and after a time a crystalline product separated this was washed with cold water and examined by the microscope when it was found t o be a mixture of two substances. If dissolved in hot water acidified with hydrochloric acid and allowed to cool beautiful colourless trans-parent crystals separated which appeared to be rhombohedrons, sometimes they appeared as flat prisms with bevelled edges.They con-tain chlorine. h nitrogen determination gave the following result :-02289 gram of substance gave 14 C.C. nitrogen a t 5" and 762 mm. = 7.56 per cent. nitrogen ; the formula C6H,C1NOs requires 7-88 per cent. This is the same substance as that obtained by Claus and Voeller, and has the constitution-COOC,H, c c1 I CH CONH2 The fusing point of my preparation was lOO" that previously found was 102". This substance has been called ethyl chloro-maleamate but should be the chlorof umaramate. If this substance be dissolved and strong aqueous ammonia added to the solution decomposition takes place and on cooling yellowish needles are deposited.The same product is obtained direct from ethyl chlorof umarate by the continued action of concentrated aqueous ammonia. The product may be purified by crystallisation from water ; but even when animal charcoal is used its yellow colour cannot be altered. t = 12.5". It gave the following numbers on analysis :-0.1995 gram of substance gave 52.2 C.C. of nitrogen. Bar 766.5 PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. 'TO3 0.1618 gram oE substance gave 0.2228 gram GO and 0*0808 gram OH,. Theory for C,H7N,O2. Pound. Carbon 37.21 37.48 Hydrogen 5.42 5.54 Nitrogen. 32.55 32.34 This is the same formula as Claus and Voeller's diamide of amido-maleic acid or diamide of amidofumaric acid :-CONH, ~ N H , CH I CONH, But Claus and Voeller say they obtained their product in colourless scales melting at 122".My product crystallises in pale-yellow needles and when heated strongly becomes dark in colour and decomposes with evolution of ammonia ; no definite fusing point being obtainable. Its decomposing point is apparently above 180". This discrepancy is not easy t o explain. The diamide of amidofumnric acid is sparingly soluble in cold wat,er moderately soluble in hot water ; it also dissolves in hydro-chloric acid. On gently heating this diamide of amidofumaric acid with dilute soda it dissolves with evolution of ammonia and on making a silver salt from the acid produced it gave on analysis 62.34 per cent. of silver ; the formula of silver amidofumnrate C4H3NAg,O4 requiring 62.60 per cent.This agrees with the results of Claus and Voeller, who obtained in the same way from their diamide of amidofumaric acid an acid the silver salt of which gave them 62.93 per cent. of silver. Before accepting these results as proving that the acid is amidofumaric acid it would be well to determine the nitrogen in this product. Chloromaleic Anhydride. When chlorof umaric acid is gently heated with chlorofumaric chloride in an oil-bath molecular proportions being used hydrogen chloride begins to be freely evolved when the mixture has attained a temperature of about 1 2 5 O ; in the course of about an hour's heating at this temperature the product becomes quite liquid and on raising the temperature it distils over freely as an oil at about 192" the operation being almost complete by the time the thermometer has reached 195".This distillation should not be conducted slowly. O '704 PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. refractionating the oil thus obtained most of it comes over between 193-195" corr. and consists of chloromaleic anhydride. The preparation of this anhydride was also tried by acting on chlorofumaric acid with acetic chloride. On heating a mixture of these substances-the acetic chloride used containing acetic acid-at 100" in a sealed tube a clear liquid was obtained which deposited a large quantity of brilliant flat prismatic crystals on cooling. On opening the tube very little hydrogen chloride was given off. The crystals were dried first on a porous plate and then over sulphuric acid in 8 vacuum ; they then became somewhat opaque.On drying at loo", one preparation lost 26 per cent. and another 32 per cent. chiefly of acetic acid though there appeared to be a small quantity of acetic chloride as the vapour passed through silver nitrate gave a small amount of precipitate of silver chloride. The dried product proved to be unchanged chlorofumaric acid. From direct experiment acetic acid does appear to combine with chlorofumaric acid. Pure acetic chloride was next used and a higher tlemperature employed viz, 150-160° the chlorofumaric acid dissolved and a slightly brown solution was produced. On opening the sealed tube hydrogen chloride was given off in abundance and on distilling the product the boiling point gradually rose to 190-193" the latter being the temperature a t which it mostly came over; the product was chloromaleic an-hydride.When chlorofumaric acid is gently boiled in a retort it gives off hydrogen chloride carbonic anhydride and carbonic a3id as already mentioned ; but if heated more vigorously at first an aqueous liquid comes over and after a time oily drops appear rendering the distillate milky. This oily product soon increases in quantity coming over alone as a clear oil which should be collected separately ; when recti-fied it boils regularly at about 194" and is chloromaleic anhydride. A specimen of this anhydride obtained by the first process given gave the following numbers on analysis :-0.1927 gram of substance gave 0.2529 gram of CO and 0-0141 gram of H,O.Theory for C4HClOp Found. Carbon 36.22 35.79 Hydrogen 0.75 o*so The average of seven determinations of the boiling point of chloro-maleic anhydride mostly made with different preparations gave 196.3" corr. ; those taken under reduced pressure 210 rnm. gave 150-151O corr. The method of preparing it with acetic chloride appears to be tjhe best of the three given above PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS 705 This anhydride was for a long time supposed to be an oil as i t re-mained liquid for several months but on cooling i t with ice and hydro-chloric acid and riibbing it with a glass rod it was obtaiued in a crystallised condition fusing at about 0". Afterwards it suddenly commenced to produce hard crystals the temperature rising to about +30".These were then very strongly pressed between dried calico ; in this way a small quantity of an oily product was removed and the anhydride obtained as a dry white product resembling maleic anhy-dride. It fused at 34*5" and gave on analysis the following num-bers :-0.2489 gram of substance gave 0.3320 gram of CO and 0.0186 gram oE H,O. Theory for CqHC 10 Found. Carbon 36.22 36.57 Hydrogen 0.75 0.83. After standing for about a month during which time it remained perfectly fluid it was cooled with ice and hydrochloric acid and well rubbed with a glass rod. It then became a crystalline mass but remelted again before it reached 0". On touching it with a crystal of the anhydride melting at 34*5" the temperature immediately rose and it became a hard mass of crystals fusing at 34.5".I have obtained this remarkable result on several occasions so that the anhydride appears t o have two distinct fusing points lying wide apart. The higher fusing point is the one most usually obtained. The following density determinations of three specimens of pressed anhydride gave-This product was fused and placed in a stoppered bottle. I. 11. In. Average. 1.5682 1.3642 1,5664 4" 4 O d- 1,5670 10" 10" d- 1.5589 1.5601 1.5561 1.5584 1.5543 1.5503 1.5526 15" 15 d-" 1.5531 1.5488 1.5454 1.5473 20° 20" d- 1.5478 1.5438 1.5400 1.5421 25O 25" d- 1.5424 The following numbers were obtained for its magnetic rota-tion :-VOL. LlII. 3 501; PERKIN CHLOROFUMARIC AR'D CHLOROMALEIC ACIDS. 15' 15 I.Product with density d- = 1.5531. 1 *2'7444 1 *2765 1 -2754 t. 6 -066 6 -074 6 483 21° 21 21 Average 21 ______ t. 22 -5' 21 *7 21 -0 16 *8 16 -8 16 ' 8 Average 19.3 sp. rotation. 1 ~ 0 1 . rotation. Sp. rotation. 1 '2'7'74 1 -2786 1 -2744 1.2823 1 -2833 1 * 2793 1 -2809 ~ - - - ---15" 15 I1 Product with density d L = 1.5503. Mol. rotation. 6 *096 6 '098 6 -076 6 -099 6 -101 6.082 6 -092 -.-__. ~~ The mean of these results is 6.083. Fumaric Acid and Fuinaric Chloride. Fumaric chloride when boiled with fumaric acid molecular propor-tions being used reacts with it hydrogen chloride being evolved and the mixture gradually becoming liquid ; on distilling malic anhy-dride is obtained in considerable quantity; the yield being 77 per cent.of the theoretical. Chloromaleic Acid. C hloromaleic anhydride if left in contact with water gradually dis-solves forming a strongly acid solution ; with hot water this change takes place quickly. The solution of chloromaleic acid obtained in this way after being evaporated to a nearly syrupy condition on the water-bath and allowed to stand deposits the acid in tufts of crystals ; these when spread out and examined with a lens appear as trans-parent plates ; when viewed between Nicol prisms they give a beautiful play of colour. A solution of chloromaleic acid when evaporated down with strong hydrochloric acid is converted into chlorofumaric acid PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. 707 Chloromaleic acid readily combines with bromine if heated with it in a sealed tube to 100".Chloronzaleates. Hydrogen Sodiutm Chlornma1ente.-This salt which is easily soluble in water crystallises in beaiitif ul rosettes of transparent crystals with well-defined and brilliant faces. A specimen dried at. 100" lost 23.68 per cent. of water. A sodium determination gave 10.38 per cent. sodium on the salt before drying a.t 100". Theory for C4H,NaC1O4,3H,O requires 23.84 per cent. water and 10.155 of sodium. Hydrogen Potassium Ch1oromnleate.-This is a beautiful salt crgs-tallising i n large brilliant well-formed crystals. It is easily soluble in water nearly 10 times as much so as the corresponding chloro-fumarate 100 parts of its solution at 15" containing 29.2 parts of salt.T t does not lose weight when heated at 100". The following analyses of this compound were made :-I. 0.2514 gram of substance gave 0.2325 gram of CO and 0.0282 gram of OH,. 11. 0.5029 gram of substance gave 0.1408 gram of K2S04. Theory for C,H,KC10,. Found. Carbon 25.40 25-21 Hydrogen 1.05 1.24 Potassium 20.73 20.87 For the following crystallographic measurements of this salt I am indebted to Professor Haushofer. Bydrogen Potassium Chloromnleate. CrFqtal system rhombic. a b c = 0-4841 1 0.3479. Small colourless crystals. The combination P (111) = 0. OP(OO1) = c ; 2Pm(OZl) = p. The surfaces c&m(010) are sub-ordinate and occur but seldom as small truncations of the angle q/q 708 PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. Observed.Calculated. c 0 = (001) (111) *141" 24' -0 0 = (111) (111) xi48 27 -q = (021) (oai) 69 40 69" 40' 0 q = (111) (021) 142 40 142 49 The neutral potassium salt crystallises freely but is extremely solnble and deliquescent. Silver Chloroma1eate.-This is obtained as a bulky white precipitate on adding silver nitrate to a neutral chloromaleate. The analysis of a salt prepared in this way gave 58.97 per cent. of silver the formula C,HCAg,O requires 59.25 If an acid chloromaleate be used this salt comes down in a crystalline condition but from the low percentage of silver it gives on analysis it evidently contains some quantity of an acid salt. One preparation made in this way gave 56.4 per cent. of silver. Neutral chloromaleates do not give a precipitate wit>h ferric chloride but produce a dark brownish-red liquid so that in this particular they behave like neutral maleates.Ethyl Chloromaleate. To prepare this about 27 grams of silver chloromaleate were treated with an excess of ethyl iodide. The reaction did not set in to any large extent in the cold but when the mixture was gently heated it took place vigorously so that it was necessary to cool it. After the reaction had quieted down the mixture was warmed on the water-bath and then allowed to remain for some time. The product was thrown on a filter and the silver iodide formed washed several times with ethyl iodide. The filtrate was then distilled on the water-bath to remove +he excess of ethyl iodide and subsequently heated up to about 180-190" the thermometer being in the liquid.On being fractioned under reduced pressure it nearly all came over a t 189.5-190*5" corr. at 210 mm. At 760 mm. pressure it came over a t 235" only slight decomposition taking place. The former is about 12.5" lower than ethyl chlorofumarate which when distilled in the same apparatus came over a t 202" a t 210 mm. Ethyl chlorofumarate is a colourless oil having a pleasant odonr. The specimen examined had a permanent rotation of 6' to the right for a column 102 m m long. Boiled with alcoholic potash it gives an abundant precipitate of potassium chloride. It gave the following numbers on analysis :-0.3366 gram of subshnce gave 0.1971 gram of CO and 0.2003 gram of OH, PERKIK CHLOROFUNBRIC AK'D CHLOROMALEIC ACIDS. 70:) Theory for C&r&lO'p Pound.Carbon 46.5 46.5 7 Hydrogen 5-32 5.65 Its density determinations gave-10" 10 15" 15" a- 1 ~ 2 , d L 1.1821, 20" 80 25" d- 1.1780, d, 1.1740. da The following results were obtained for its magnetic rotation -t. 22 -5O 22 -5 22 -5 22.6 23 -0 21 *!5 21 -0 18 -0 18 -0 18 '0 Sp. rotation. 1 * 1180 1'1156 1.1159 3 -1155 1-1171 1 *1193 1 '1212 1 *I250 1 *1244! 1.1246 -Mol. rotation. a0 -908 10 *885 10 * 888 10 * 882 10 -903 10 -925 10 *939 10 -943 a0 -937 10 -940 Average 20.9 1 1.1196 1 10.915 ~ ~~ The results of this inquiry show that the chloride obtained by the action of phosphorus chloride on tartaric acid is fumaric chloride and not maleic chloride and the acid obtained from it chlorofumaric and not chloromnleic acid ; also that chlorofumaric acid yields chloro-maleic anhydride by processes similar to those by which fumaric acid yields maleic anhydride and further that the chloromaleic acid ob-tained from this is reconverted into chlorofumaric acid by heating with hydrochloric acid just as maleic acid by the same means is changed into fumaric acid.Chlorofumaric and chloromaleic acids also behave with bromine in a way analogous to fumaric and maleic acids the former not com-bining with bromine so easily as the latter. The reaction with ferric salts is also similar chlorofumarates giving pale-brown or buff precipi-tates whilst chloromaleates give dark brown-red solutions. With respect to other aci.ds called chlorofumaric and chloromaleic acids the chlorofumaric acid obtained by Bandrowski (Ber.15, 2695) by the action of fuming hydrochloric acid on acetylenedicarb-oxylic acid appears to be most like the one described in this paper, but the fusing point given for it is 178" which is considerably lowe '710 PERKIN CHLOROFUMRRIC AND CHLOROMALEIC ACIDS. than that of chlorofumaric acid which is 191". Unless there is some mistake on this point of course they cannot be identical. The acid obtained from benzene by Carius does not appear to be either the chlorofumaric or chloromaleic acid described in this paper as it fuses a.t lil-l72" and its acid potassium salt crystallises with 1 mol. H,O moreover 100 parts of its solution a t 16" contain 6.13 of dry salt, whereas the corresponding chlorofumarates and chloroma,leat,es are anhydrous salts and the former gives a solution containing 3.8 in the 100 parts and the latter 29.2 a t 15".Observations on Boiling Points. I n a paper recently published (Trans. 1888 SOO) I drew attention t o the great difference between the saturated and unsaturated com-pounds in relation to their boiling points similar variations in composition causing opposite differences in boiling point ; the com-parison being between the ethers and the anhydrides the latter in the saturated series being higher in boiling point whereas in the unsaturated series they are lower. It may be of interest here to show how apparently inconsistent the boiling points of the unsaturated products examined in this paper and a few other allied compounds are ; the densities are also given.B. p. 1 5" d-1 5 O ' Ethyl maleate*. . . . . . 823.0" 1.0r40 Ethyl fumarate . . . . 218.5 1.0625 - 4.5" 0.01 15 - -15" d@' B. p. 210 mm. B. p. 760 mm. Ethyl chloromaleate . . 190*0" 235" 1.1822 Ethyl chlorofumarate . . 202.5 250 1.1937 i- 12-5 + 15 + 0-0115 _c_ -I n the case of the non-chlorinated ethers the boiling point and densities are lower for the fumaric compounds but in the chlorinated compounds they are greater. It is curious to notice that the differ-ence is smaller for the boiling points taken a t 210 mm. If a comparison be made between the non-chlorinated compounds and the chlorinated the influence of chlorination on the boiling point will be seen.* This comparison of the boiling points of maleic and fumaric ethers is given in my previous paper (Trans. 1888 594) but R wrong set of numbers were copied by mistake; the above are those that should have been given PERKIN CHLOROFUNARIC AND CHLORO MALEIC ACIDS. 7 1 1 15" B. p. 210 mm. B. p 760 mm. '- 15"' Ethyl chlorofumarate. . 202.5" 250.0" 1 * 193 7 Ethyl fumarate 175.0 218.5 1 - 0 63 7 27.5 31-5 0-7 300 Chlorofumaric chloride. - 186.0 1.5i-31 - 162.5 1.4202 Furnaric 7 * 23.5 0.1529 Chloronialeic anhydride 195.3 Maleic anhydride. . 200.0 - - ---+ 4.7 Here we find how differently chlorination affects the boiling point, in the case of the anhydride actually lowering it so that mitleic anhydride has a higher boiling point than the chlorinated derivative.I n the comparison of ethyl chlorof umarate with ethyl fumarate we again find that the difference betweon the boiling points varies with the pressure. The differences between the densities although not consistent are not very wide apart except that they vary i n the reverse way to the boiling points. If the boiling points of some of the bromine-derivatives of maleic compounds be compared in a similar way we get-B. p. E thy1 bromomaleate 256" Ethyl maleate 225 31 Bromomaleic anhydride. . '215" Maleic anhydride. 200 15 --In this case the results go much in the same direction the bromine increasing the boiling point of the ether considerably but influencing that of the anhydride to a much smaller extent. It is however, remarkable that ethyl bromomaleate differs in boiling point from ethyl malente only to the same extent as ethyl chlorofurnarate differs from ethyl fumarate.Observations on Magnetic Rotations. I n considering the magnetic rotations of the substances examined in this paper it will first be convenient to compare the molecula ‘i 12 PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. Xthyl chlorofumarate 11.377 Ethyl fumarate. 10.112 Influence of C1 replacing H 1.265 Chlorofumaric chloride 10.044 Fumaric chloride. . 8.747 Influence of Cl replacing H 1.297 These results again are quite as consistent as could be expected ; the influence of the chlorine replacing hydrogen is however smaller than has previously been found in any other monochloro-compound and is iiearer that found when replacing a third hydrogen.The following c xamples will explain this :-M. rot. Chloroform . 5.55 9 Methylene chloride. 4.313 -Influence of C1 replacing K 1.246 Chlorethylene chloride. . 6.796 Ethylene chloride 5.485 -_-Influence of C1 replacing H. . 1.311 and as the bodies under consideration represent saturated compounds less HO some kind of analogy can be imagined to exist between their replacement. In a previous paper (Trans. 1888 598) it was shown that the magnetic rotations of both maleic and citraconic anhydrides were anomalous. That of chloromaleic anhydride also does not come ou PERKIN CHLOROFUMARIC AND CHLOROMALEIC ACIDS. 7 13 consistently. Thus if we take the replacement of hydrogen by chlorine we get-Chloromaleic anhydride 6.082 Naleic anhydride 4.545 Influence of C1 replacing H 1.537 This number is higher than that found for all other allied compounds ; it seems however unlikely that in this compound the chlorine should have a larger value but would lead rather to the infer-ence that this replacement has affected the molecule of maleic anhy-dride making its abnormally low rotation somewhat higher and thus seemingly increasing the value of the chlorine it contains ; even then it is lower than it apparently should be.These results again go to show that there is some peculiarity in these anhydrides which is not clear at present and this is borne out by the remarkable character of their boiling points. The following are some of .the more important results of this inquiry, The chloride obtained by the action of phosphorus pentachloride on tartaric and rncemic acids is chlorofumaric chloride.Chlorofumaric acid when distilled or treated with its chloride or with acetic chloride yields chloromaleic anhydride from which chloro-maleic acid and its derivatives can be obtained. Chloromaleic acid is converted into chlorofumaric acid by heat,ing i t with hydrochloric acid just as maleic is converted into fumaric . -acid The chloromaleic acid prepared from benzene by Carius does not agree in its poperties either with chlorofumaric or chloromaleic acids. The boiling points of the halogen-compounds of maleic and fumaric acids do not show any regular relationship to each other those of maleic anhydride showing the greatest variations. The magnetic rotations of chlorofumaric and chloromaleic ethers have to each other practically the same relationship as exists between those of fumaric and maleic ethers. The effect of the replacement of hydrogen by chlorine in maleic and fumaric ethers and fumaric chloride on the magnetic rotation of these compounds is to increase them practically to the same extent, viz. by about 1.284. The magnetic rotation of chloromaleic anhydride although ab-normal is not so far from the calculated number as that of maleic anhydride

ISSN:0368-1645    

DOI:10.1039/CT8885300695

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

714 LV.-On a New Method for the Prepvation of Nixed Tevtiary Phosphines. By NORMAN COLLIE PhD. F.R.S.E. SlNCE the announcement of the discovery of the phosphorus bases by Hofmann and the subsequent publication by Cahours and himself of the classical work on this subject now nearly 30 years ago little has been added to our knowledge of this interesting group of substances. Michaelis has indeed investigated the aromatic phosphorus compounds, but those of the fatty series remain almost untouched. The great difficnlty experienced in their preparatioii has doubtless acted as an obstacle to their study and preparation ; still when they have once been prepared the simplicity and energy which characterises their reactions and the theoretical interest attaching to many of their combinations amply repays any time spent on the investigation of their properties.I have already called attention (Collie Phil. Hug. July 1887 27), t o the decomposition which some of the organic ptiosphonium com-pounds suffer when subjected to the action of heat and in a paper by Professor Letts and myself (Letts and Collie Phil. Aiag. August, 1886 183) we first noticed the decomposition which took place when chlorides of the phosphoniums were destructively distilled :-(C2H5)IPCl = (C2H,),PHCl + CZH,. This decomposition suggested a further examination of the action of heat on phosphonium chlorides which had been found to be such an excellent practical method for obtaining a free tertiary phosphine from a quaternary compound. Hofmann has already shown how from a primary phosphine it is possible by successive additions of a hydrocarbon iodide (and subsequent liberation of the free phosphine with potash) to build up secondary tertiary and finally quaternary phosphonium compounds.But if we start a t the other end of the series and seek for some method by which we may retrace our steps to the original primary phosphine we a t once find the task is one by no means so easy of accomplishment. In fact any one who has worked with these quaternary phosphoninm compounds is at once struck with the great difficulty which surrounds even the first step in the backward progress. This decomposition of tetrethylphosphonium chloride bas the advantage of being nearly a quantitative method for accomplishing this first step.Some of the oxygenated salts of tetrethylphos COLLIE PREPARATION OF MIXED TERTIARY PHOSPHINES. 715 phonium yield it is true triethylphosphine when heated hut the chief product of their decomposition is always the oxide of triethyZp1Lo.s-plLinp a substance whose stability is so great that to prepare from it either triethylphosphine or even compounds which might yield the secondary or primary phosphines is a problem of the greatest difficulby. As these phosphonium chlorides seemed to be the only salts from which the tertiary phosphine could be obtained in any quantity I have investigated the action of heat on several of them, chiefly in order to ascertain whether this mode of decomposition was general and I have also used it as a new method for the preparation of mixed tertiary phosphines.I had wished to make the list of phos-phonium chlorides experimented upon as numerous as possible but owing to the great difficulty experienced in the manufacture of the necessary material I have had to content myself with the salts obtain-able from triethylphosphine and a very small quantity of trimethyl-phosphine. All attempts to prepare tripropylphosphine were un-successful as the thick glass tubes which contained the alcohols and the phosphonium iodide were usually shattered by the enormous pressure of the gases produced during the experiment. When this was not the case only minute quantities of the phosphuretted hydrogen had entered into combination with the hydrocarbon radicles ; attempts also to get the triethylphosphine to unite with chlorobenzene were without success.These two substances were heated to as high a temperature as 250° and yet they did not combine with one another. In the following experiment it is shown that in the case of a large number of phosphoniuni chlorides the decomposition which they suffer when heated is always the same. Ethylene or an unsaturated hydrocarbon i s formed together with a hydrochloride of a tertiary yhvsphirhe. When there is more than one ethyl-group present in the phos-phonium chloride ethylene is formed-but when the phosphonium chloride contains only one ethyl-group, then although ethylene is still formed another decomposition goes on at t,he same time and a mixture of two hydrochlorides results :-e(C2H.5) (CH,),PCl = 2(C2H5) (CH3)ZPHCI + CZH,.(CzH,) (CH3)sPCl = (CH,),PHCl + CZH,. It is remarkable that if we compare the decomposition by heat of the phosphonium chZorides with the decomposition of any of the ammonium salts it must be with the hydroxides and not with the corresponding chlorides 716 COLLIE A NEW METHOD FOR THE Vhilst the phosphonium hydroxides resemble in their mode of decomposition the corresponding ammonium chlorides. (C,H,),(CH,)NCl = (C2H6),N + CH3C1. (CZH~)Z(CH,)NC~ = (C2H,)2(C&)N + CH3Cl.t In the first case it is the ethyl-group which splits away from the molecule whilst in the last it is any hydrocarbon radicle rather than ethyl which parts from the nitrogen or phosphorus. If the hydrochlorides of the tertiary phosphines are heated they probably dissociate when in the gaseous condition into hydrogen chloride and the tertiary phosphine; but the acid and the phosphine recombine again on cooling.I n some cases the hydrochlorides of the tertiary amines behat.e in a similar manner the hydrochloride of triethylamine can be distilled without decomposition but the hydro-chloride of trimethylamine decomposes when heated-. 3(CH3),NRC1 = 2(CH3)3N + (CH,)H,NCl + 2CH3Cl.t By this reaction a primary amine is produced. Up to the present this kind of decomposition has not been noticed amongst the phosphine Compounds ; no method for the preparation of a primary or secondary from a tertiary phosphine is known and all attempts made in that direction have been without success. I n the phosphonium and ammonium compounds it is usualIy supposed that the valency of the phosphorus and the nitrogen is pentad.This belief receives support from several facts which have been noticed during this research, (1.) The decomposition of the mixed phosphonium chlorides by which a mixed tertiary phosphine is produced from an ordinary tertiary phosphine. (CzH.=,),P + CTHVC1 = (C,R5)3( C7HT)PCl. (CzH,)S(C,H,)PCl = (CZH5),(C,H,)PHCl + CZH,. X Hofmann Annulen 78 283. t V. Meyer and Lecco Annulen 188 177. 5 Vincent Bull. SOC. Chim. 30 18'7 PREPARATION OF MIXED TERTL4RY PHOSPHINES. 717 (2.) The same mixed phosphonium salt may be prepared bJ two methods. (C2H5)(CH3)?P t CH31 = (C,H5)(CH,),PI. (CH3)3P + CZH5I = (CzH5) (CH,),PI. This kind of reaction V. Meyer has already noticed in the nitrogen series :-(CaH5)zNH + ZCHJ = (CZH,)2(CH3)2NI.(CH3)zNH + 2C2HbI = (C2H5)z(CH3)2NI. These facts are more easily explained on the supposition that phosphorus is pentad for a molecular change would have to occur if the valency be taken as triad. The new phosphines that have been prepared by this method are as follows :-B. p. 83 - 8 5 O Dime t h y le thy1 p hos phine Diethylmethylphosphine 110-1 12 Diethylpropylphasphine 146-149 D iethylisoam ylphosphine 185-4 8 7 Diethyl benzyl phosp hine 252-2 55 Dibenzylethylphosphine 320-330 The diff erenee in boiling point between the dimethylethylphosphine and the trimethylphosphine is considerably greater than between any other of the phosphines in the ethyl methyl series ; i'f these boiling points be compared with those of the corresponding mehhylethyl-methanes a; similar difference will be also noticed :-B.p. (CH3)SP 40- 42" (CH,),(CZH,)F 83- 85 (CzH,)z(CH,)P 110-112 (CZHJsP 127 B. p. (CH,),CH. - 17 (CZH5) (CH3)2CH 3f ) (C2H5)2( CH3)CH 60 (CzHs),CH 90 Unfortunately comparison with the corresponding amines is im-possible as only two out of the four have been obtained. Action of Heat on Ethyltrimetl~y1phosphonium Chloride. The corresponding iodide was prepared by two methods :-(1.) By the action of iodide of ethyl on trimethylphosphine. (2.) By the action of iodide of methyl on dimethylethylphoaphine. Prepared by the first metrhod in ethereal solution i t remained as a white crystalline and deliquescent solid after the ether was eva-parat ed 718 COLLIE A KEW METHOD FOR THE 0.351 gram gave 0.356 gram AgCl = 25-99 per cent.Cl. Calculated for (C,H,)(CH,),PCl C1 = 25.09 per cent. When heated it did not melt but a t a very high temperature-con-siderably above 300"-it decomposed yielding a crystalline distillate and some gaseous products ; 5 grams of salt were heated and 600 C.C. of gas were evolved. The gas was proved to be pure ethylene for it was completely absorbed by bromine and gave dibromethylene boiling at 130-134". The distillate was acid to litmus-paper and soluble in water. When this aqueous solution was treated with caustic soda an oil separated ; this was very volatile and had all the properties of a tertiary phosphine. It was dried and distilled ; i t then boiled between 40" and 60".As there was not enough of it to separate by fractional distillation it was treated with methyl iodide with which i t a t once combined and the resulting iodide was converted into the chloride. With chloride of platinum a platinochloride was obtained, and on analysis gave the following numbers :-I. 0.420 gram salt gave 0.1350 gram Pt = 32.14 per cent. Pt. 11. 0.620 , , 0.1996 , = 32.19 ,, The calculated percentage of Pt in (EtMe,P),PtCl is 31.55 per cent. Pt while that required for (MelP),PtCI is 33.05 per cent. Pt. The salt is therefore probably a mixture of the two platinochlorides: hence the chloride of trimethylethylphosphonium decomposes in two ways when heated :-(CHs)s( C,H,)PC1 = (CH,),PHCl + C,H4 B(CH3)3(C,H,)PCl = 2(CH,),(C2H,)PHC1 + CZH4. This is probable for several reasons.The gas obtained was consider-ably less in volume than the theoretical amount required by the first equation (600 C.C. obtained; theory 800 c.c.). Also the tertiary phosphine which was produced had no fixed boiling point (SO-6O0), whilst the boiling point of the trimethylphosphine is given as 4 0 4 2 " . This kind of decomposition was also noticed in the benzyl series during the experiments on the action of heat on ethyltribenzyl-phosphonium chloride. Moreover the same results were again obtained from trimethylethylphosphonium chloride prepared from the product of the action of iodide of metbyl on dimethylethylphosphine (p. 720) No further experiments were made with trimethylphosphine 2s the starting point for phosphoniuiri chloride ; the remaining salts were all produced from triethylphosphine PREPARATIOS OF ;CITI;ED TERTIARY PHOSPHISES.719 Action of Heat on Triethylmethylpphosphonium Chloride. 50 grams of triethylphospliine were converted into triethylmethyl-phosphonium iodide. The reaction was extremely violent even when the two substances were diluted with ether. The resulting iodide was converted into the chloride from which salt a small quantity of the platinochloride was prepared for analysis. I. 0.535 gram gave 0.152 Pt = 28.41 per cent. Pt. 11. 0.227 , 0.203 CO and 0.115 H,O. Found. Calculated for r--,h-- 7 (MeEt,P) ,PtCI,. I. TI. C 24.92 - 24.35 H 5.41 - 5-62 Pt 28.93 28.41 -6 grams of the chloride were heated; it decomposed above 300" without any charring and gave a solid crystalline distillate and 760 C.C.of ethylene mixed with 40 C.C. of 5t gas which was not absorbed by bromine but which burnt with a greenish flame. The theoretical yield of ethylene supposing the reaction to be as follows-(C,H,),(CH,)PC1 = (CJ&)z(CH,)PHCl + CzHd, would be almost 800 C.C. The distillate was dissolved in water and treated with caustic soda ; it then gave at once a free tertiary phosphine-diethylmethyl-phosphine. The remainder of the triethylmethylphosphonium chlo-ride was heated and the same phenomena were again observed. The whole of the diethylmethylphosphine obtained was dried over solid caustic soda and fractionally distilled ; nearly the whole boiled between 110-112". Its odour was similar to that of triethylphos-phine ; it united with sulphur and oxygen to form addition products, and with bisulphide of carbon gave red crystals.0.332 gram gave 0.700 CO and 0.3795 H,O. Calculated for (C2HMCHdP Found. C . . . 57.69 57-53 H 12.50 12-70 The remainder of the diethylmethylphosphine was converted into iodide of diet hyldime thy 1 phosphonium '720 COLLIE A NEW METHOD FOR THE Action of Heat on Dimet?~yldiethyl~?iosphoni~?n Chloride. This salt was obtained from the iodide. An analysis of the platino-0.450 gram gave 0.135 Pt and 0,5885 AgCl = 30.00 per cent. Pt Calculated' for (Et2MeJ?)2PtC16 Pt = 30.19 per cent. C1 = 32.97 The remainder of the ehloride of dimethyldiethylphosphonium was heated. 9 grams of salt in one experiment gave 1250 C.C. of ethylene and a solid crystalline distillate ; the theoretical amount should he 1300 C.C.of ethylene if the decomposition takes place in the following manner :-chloride was made. and 32.30 per cent. C1. per cent. The distillate which was obtained from this and other experi-ments when treated with caustic soda gave dimethylethylphosphine, which after drying and fractionally distilling boiled between 83" and 85". As there were only a few grams of this phosphine no experiments were made with it but the whole was converted into the iodide of trimethylethylphoaphonium. This salt has already been mentioned (p. 717) but it had then been prepared by another method. The whole of it was converted into the chloride of trimethylethyl-phosphine. Some of the platinochloride was prepared :-0.551 gram gave 0.1725 Pt = 31.21 per cent.Pt. Calculated for (Me,EtP),PtCl, Pt = 31.55 per cent. This platinochloride exactly resembled the salt which had been pre-pared before. On heating the chloride of trimethylethylphosphonium, it decomposed also in a precisely similar manner to the chloride previously mentioned. The free phosphine obtained from the distil-late boiled between 40" and 65" and after treatment with methyl iodide and conversion into the platinochloride it gave numbers identical with the former ones. The separation of this platinum salt into the two platinochlorides which were supposed to be present in it was attempted but1 without any satisfactory results. The amount of t,he substance was very small and both the platinochlorides seemed to be equally sparingly soluble in water.Action. of Heat on Propy ltrieth ylphosphoniurn Chloride. This sa,lt was prepared by heating triethylphosphine for several The hours with excess of chloride of propyl in a sealed tube at. 130" PREPARATION OF MIXED TERTIARY PHOSPHINES. 721 resulting white crystalline mass was dissolved in water and separated from the excess of propyl chloride. The aqueous ,solution was evaporated to dryness over a water-bath. A small quantity of the plntinochloride was prepared for analysis :-I. 0.460 gram gave 0.120 Pt = 26.08 per cent. Pt. IT. 0.352 , 0.377 CO and 0,1995 H,O. Found. Calcuiated for I-- 7 (Et3PrP) 2 Pt Cl6. I. 11. C . . . . 29-51 - .2923 H . 6.03 Pt 26-64! 26.08 L 6.29 -" . 7 The rest of the chloride was heated.In one experiment 2.5 grams of salt gave 300 C.C. of ethylene and a solid distillate; 2.5 grams should give 290 C.C. of ethylene according to the following equa-tion :-(C2H5)3(C3H7)PC1 = (CJL)z(CsH,)PHC1 + Cz& The distillate together with another quantity obtained by distil-ling the remainder of the chloride boiled between 225" and 228", and gave diethylpropylphosphine when treated with caustic soda solution. After drying and distilling this phosphine was obtained in a pure state boiling between 146" and 149". It united at once with ethyl iodide to form a phosphonium iodide which was not attacked by caustic soda. Converted into the platinochloride-0.3615 gram gave 0.0950 P t = 26.28 per cent. Calculated for (Et3PrP),PtC16 Pt = 26.64 per cent.This platinochloride wits exactly similar in appearance and crystal-line form to one prepared from the product of the reaction of propyl on triethylphosphine. Actio?t of Heat on Triethylisoamylphosphonium Chloride. When excess of isoa8myl chloride (b. p. 99") was heated at 130" in a sealed tube with triethylphosphine a crystalline mass consisting of the triethylisoamylphosphonium chloride was obtained. This salt is very deliquescent and difficult to obtain in the form of crystals ; but by heating some of it at 200" in an oil-bath a dry salt fit for analysis was obtained. 0.414 gram took 18.0 C.C. decinormal A&"3 solution = 15.43 per Calculated for (C2H,)3(COH11)PC1 C1 = 15.81 per cent. VOL. LIII. 3 c cent. C1 722 COLLIE A NEW METHOD FOR THE Some of the platiiiochloride was also prepared ; it is moderately soluble in water and crystallises in thick needle-shaped crystals.I. 0.360 gram gave 0.0385 Pt = 23-75 per cent. Pt. 11. 0.2455 gram gave 0.298 COz and 0.150 HzO. 111. 0.4925 gram gave 0.123 Pt and 0.534 AgCl = 24.97 per cent. Pt and 26-82 per cent. C1. Found. 7 Calculated for r--h--(Et,AmP),PtC16. I. 11. 111. c - 33.10 - 33.59 H - 6.78 - 6.61 Pt 24.87 23.75 - 24.97 C I . . 27.03 - 26.82 -The remainder of the chloride was heated. It decomposed at a temperature above 300° into the hydrochloride of diethylisoamyl-phosphine and ethylene. The hydrochloride after redistillation boiled between 270" and 271". took 14.5 C.C. decinormal AgN03 = 18.85 per On analysis-I. 0.273 gram cent. Cl.11. 0-306 gram gave 0.609 COz and 0.320 H,O. Found rI Calculated for (CZH,) (CsHiJPHCl. I. 54.28 C 54.96 - 11.61 H 11.19 I Cl., 18.06 18.85 -This hydrochloride was very deliquescent and soluble in water. Treated with caustic soda it gave diethylisoamyIphosphine. This phosphine is a colourless slightly viscid liquid which does not oxidise so readily on exposure to air as the triethylphosphine and has an odour which distinctly recalls that of fuse1 oil. It boils between 185" and 187". 0.223 gram gave 0.550 C02 and 0.261 H,O. Calculated for (C2HJ2 (Call) P. Found. C 67.50 67.26 H 13.12 13.00 The decomposition which the chloride of triethylisoamylphos-phouium undergoes when heated is as follows :-( C ~ L ) ~ ( C ~ H I ~ ) P C ~ = (CJ&),(C5Hii)PHCl + C2H4 PREPARATIOX OF MIXED TERTIARY PHOSPHENES.72 3 Action of Heat o n T r i e t l ~ y l b e n z o y ~ h o s ~ l i o ~ ~ ~ ~ m Chloride. I n order to obtain this salt triethylphosphine was warmed with a large excess of benzyl chloride. The chloride was at once formed, with evolution of a very considerable amount of heat. It was easily separated from the excess of benzyl chloride by treatment with water, and the aqueous solution on evaporation yielded this salt in the form of needle-shaped crystals. Hofmann (Am. Spl. I 323) has prepared this salt by the action of triethylphoaphine on benzylidene chloride but i t was in aqueous solution and he only analysed the platinochlorido. This chloride can be easily crystallised from an aqueous solution but is better obtained from an alcoholic solution by precipitation with dry ether.It crystallkes with 1 mol. of water of crystallisation and melt>s at 178". 2.041 grams dried in a vacuum over sulphuric acid lost 0.137 gram = 6.71 per cent. Calculated for 1 mol. of water of crystallisation H20 = 6-85 per cent. I. 0.446 gram dry salt took 19.0 C.C. of decinormal AgNOs = 11. 0.283 gram dry salt took 11.8 C.C. of decinormal AgNO = 15-12 per cent. C1. 14.80 per cent. C1. 111. 0.363 gram dry salt gave 0.842 CO and 0.2915 H,O. Found. Calculated for r-- 7 (C,H,),(C jH7) YC1. I. 11. 111. C 63-80 - - 63.26 - 8.92 H . . - 900 Cl - 14.52 15-12 14.8 When heated this salt decomposed at a temperature considerably above 300" yielding gaseous and solid products.The gas was pure ethylene and the distillate hydrochloride of diethylbenzylphosphine. This hydrochloride boiled without decomposition between 325" and 3SOo and when dissolved in water and treated with caustic soda gave benzyldiethylphospkine. An analysis of the hydrochloride was made :-0.622 gram took 29.9 C.C. of decinormal AgNO = 17.06 per Calculated for (C,H,),(C,H,)PHCl Cl = 16.39 per cent. The decomposition of the phosphonium chloride can be expressed cent. C1. by the following equation :-(CLH ),(C,H,)PCl= (C2H,)z(C,H,) PHCl + CZH,. 3 c 724 COLLIE A SEK METIIOD FOR THE A consider;tble quantity of tliethylbenzyl phosphine was prepared On fractionation it boiled between 2SO" and 25;". (During this fracb-tioiiat ion traces of triethylphosphine mere noticed.) 0.232 gram gave 0.627 GO and 0.1'32 H20.Cn1ciil;itcd for (c211s)2( C;H;) P. Found. c 73.33 73.70 H 9.44 9.19 0 1 1 exposure to the air this phosphine fumed strongly being easily oxidised. It was converted by the action of nitric acid into the oxide of diethylbenzylphosphine but when prepared by this method the oxide was found to be mixed with traces of nitro-compounds. It was, however obtained pure by another process namely. the action of heat on the hydroxide of diethyldibenz-jlphosphine :-(C2H,),(C7H7)P*OII = (CZHb)z( C7H;)PO + C7HS. Thils prepared it boiled between 328" and 330° and cq~stallised in 0.478 gram gczve 1.1785 COz and 0.3675 H,O. long needles. Calculated for (czI15)2(ci€Ii) PO. Found. c . . 67.34 67.24 I3 8-67 8-54 Sodium acted on this oxide in ,z remarkable manner setting free diethylbenzylphosphine.In this respect it differs from either triethyl- or trimethyl-phos-pliine oxide neither of which is acted on by sodium. The sulphide of diethylbenzylpliosptiine was also prepared. Sul-phur was added to an etlierenl solution of the phosphine. After evaporation the sulphicle remained as a crystalline mass. It is not soluhlc in water but when the water is boiled i t melts. It boils at a temperature between 300 and 310' with partial decomposition and melts a t 94-95'. When heated with sodium a violent reaction takes place and a free phosphine is produced. The remainder o f tlle dietliSlberizylphosp~iine was converted into diethyldibenzylphos-phonium chloride by heating it witti chloride of benzyl.Action of n e a t o n DiethyI,libeizzl/lp?iosphoni~Lm Chloride. This salt was separated from the excess of benzyl chloride and after the aqueous solution had been evaporated it yielded crystals of this salt on cooling PREPARSTJON OF MIXED TERTIARY PHOSPHINES. 7 25 I. 0.7435 gram took 25.0 C.C. decinormal AgN03 = 11.33 per cent. C1. 11. C.311 gram gave 0.799 CO and 0.224 H20. Found. Calculated for r--A- 7 (C2Hd 2 (CjH j ) 2PC1. 1- 11. C . . 70.47 - 70.06 H . . 7.83 - 8.01 C l . . 11.58 11.93 -The platinochloride was prepared. It consists of slender needles almost insoluble in cold water but more soluble i n alcohol. I. 1.0434 gram gave 02145 Pt = 20.55 per cent. Pt. 11. 0.2325 , 0.3835 C 0 2 a d 0.1046 H,O. Pound. Calculated for 7-- 7 (Et2B~$)2PtCl, I.11. c 45.47 - 44.98 H . . 5-05 4.99 Pt 20.52 20.55 -This chloride of dibenzyldiethylphosphonium decomposed when heated in a manner exactly similar t o the other chlorides-(C2H&(C7H7)2PC1 = (C2H5j(C7J%),PHC1 + C J L , but there was a small quantity of toluene formed when the solid distillate was dissolved in water. The tertiary phosphine which was obtained from the distillate on drying and distilling boiled at 320-330". It was a somewhat oily liquid which when warmed fumed on exposure to the air and had a pleasant penetrating odour. 0.232 gram gave 0.627 C02 and 0.192 H20. Calculated 'for (C2Hd (C,Hd,P. Found. c 73-70 73.33 H 9.19 9 44 A c t i o n of H e a t o n Tribenzylethy 1p~oosllhoniuwL Chloride. Dibenzplethylphosphine was heated with excess of benzyl chloride in order to obtain tribenzylethylphosphonium chlorida.This salt can be purified by recrystallisation from water It crystallises with 1 mol. H,O. 0.315 gram salt lost at 110" 0.015 gram H,O = 4.76 per cent. Hz@ = 4-65 per cent. Calculated for (CzH5)(C7H7)3PC1H20 '7 26 PREPARATION OF RIIXED TERTIART PHOSPHTSES. A chlorine analysis was also made of the dry salt-0.811 gram salt took 22.5 C.C. decinormal AgNO solution = 9.84 per cent. C1. Calculated for (C2H5)(C1H1)PC1. C1 = 9.63 per cent. This salt is very Some of the platinochloride was also prepared. insoluble in cold water but is more soluble in alcohol and hot water. I. 0,2715 gram gave 0.04SEi Pt = 17.86 per cent. Pt. TI. 0.1915 gram gave 0.350 Ft = 18.27 per cent.Pt and 0.152 AgCl = 19.64 per cent. C1. 111. 0.186 gram gave 0.3478 C02 and 0*0837 H,O. Found. Calculsted for t-- 7 (Et Bz3Y),PtC16. I. 11. 111. C 51.40 - - 50.99 H . . 4.84 - - 5.00 Pt 18.15 17-86 18.27 Cl 19.80 - 19.64 -- The chloride of ethyltribenzylphosphonium when heated did not decompose in at all a simple manner and from the small quantity which was used (4 grams) no very definite results could be obtained. Ethylene was evolved but in small quantity only; hFdrogen chlo-ride stilbene and a resin were also produced. The distillate was washed with water and on adding caustic soda t o the aqueous solu-t ion some diethyl benzylphosphine was liberated. Whether the resin was tribenzylphosphine or not esuld nok be determined for even if the amount had been larger it would have been impossible to purify it.Some years ago in a paper published in conjunction wit,h Profesqor Letts (Letts and Collie Trans. Roy. Soc. Edin. 30 Part I 2131 the difficulty surrounding the preparation of tribenzylphosphine was pointed out; a resin was obtained where i t was expected that tri-benzglphosphine would be found and when analysed this resin gave numbers agreeing fairly well with those required by tribenzyl-phosphine. The decomposition of the chloride of ethyltribenzylphosphine may possibly be as follows :-(CJL)(C7Hi),PC1 = (C7HT)SP + C2Ha + HCl qC,H,)(C,H,),Pcl = 2(C7H,),(CzH,)PHCl + C14H12. The resin was heated with chloride of benzyl but not even a trace of te trabenzyl phosp honi urn chloride could be obtained. The decomposition of tribenzylethylphosphonium chloride is similar t o the change which takes place when ti~imethylethylphosphonium chloride is heated (p. 718)

ISSN:0368-1645    

DOI:10.1039/CT8885300714

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

LVL-The Chemical Actions of some iWicro-organisms. By R. WARTNGTON. THE present research was commenced with the intention of isolating, if possible the nitrifying organism contained in soil. I n the autumn of 1886 I spent through Dr. Klein’s kindness some time in the laborat,ory of the Brown Institntion learning under his instruction, the methods usually employed for the separation of different species of bacteria. On returning to the Rothamsted laboratory I com-menced the study of five organisms-two isolated from visible growths in some of my nitrifying solutions; two obtained from soil at the Brown Institution ; and one the well-known Bacillus subtilis obtained from hay. A short time afterwards I became acquainted with the paper by W. Heraens (Zeit. f. Hygiene 1886 193).In this communication he names seven well-known organisms :-LW. prodigiosus Staphylococcus citreus Finkler’s bacterium the spirillum of cheem the bacteria of anthrax and typhoid fever and the “ Wurzelformige Bacterien,” as capable of nitrifying urine and concludes that the power of producing nitrification is by no means unusual among bacteria. In April 1887, D c. Klein kindly supplied me with pure cultures of 12 bacteria includ-ing several of tliose specifically named by Heraeus. The study of these organisms afforded results of considerable interest and in consequence others were from time to time supplied by Dr. Klein. The investiga-tion thus deviated from its original intention and became to a con-siderable extent a stmdy of some of the chemical propertiea of a variety of oipnisms many of them well known as pathogenic.The action of these organisms has been tested in four particulars :-1. The hydrolysis of urea. 2. Action on milk 3. Capacity for reducing nitrates. 4. Power of producing nitrification. It is proposed after publishing the results at present obtained to return to the investiga-tion originally intended and recommence the study of the micro-organisms in soil. THE MICRO-ORGANISMS STUDIED. The following were received in the state of pure cultures from Dr. Klein :-1. Bacillus of swine fever. 2. , of typhoid fever. 3. , of infantile diarrhoea 7%8 WARIKGTON THE CHENICAL ACTIOXS 4. Bacillus of anthrax. 5. , of septicaemia (mouse). 7. , subtilis (jequirity extract). 8. 7 ) , (scurf of scarlatina).9. , JEz~orescens. 6. 99 7’ (guinea-pig). 10. , $uorescens liguescens. 11. ,) intesthi.” 12. Spirillum Koch’s cholera asiatica. 14. , Deneke’s cheese. 16. Micrococci~s aureus. 17. Staphylococcus luteus. 13. 7 Finkler’s cholera nostra. 15. ,> Liagard’s noma. 18. 7 7 candidus. 19. 9 candidus liguescens. 2Q. Streptococcus s c a r l a t i n a h o r n Dr. W. R. Smith was received :-21. Micrococcus urece. The following five were isolated under Dr. Klein’s superin-tendence :-22. Bacillus su&ilis (hay). 24. , torulqormis. 25. , sulphureus. 26. , tardecrescens. 23 1 $occus. There mas also separated from specimens of M. prodigiosus-27. Nicrococcus gelatinrjsus. A considerable number of the organisms here enumerated are well In some other cases a reference to See Reports of Ned.Oflc. LOG. See same Reports 1886-7, It will be more fully described in the following known and need no description. the original description will suffice. Gov. Board 1877-8,169 ; 1686-7,446 P1. XXV. 447 P1. XXVII. Report. 1. Bacillus of swine fever (Klein). 3. Bacillus of i i f a n t i l e diarrhea (Klein). 5. Bacillus of septicmnia mouse (Klein). 6 . Bacillus of septicamia guinea-pig (Lingard). Ibid 1885-6 17@, Ibid. 447 P1. XXVI. Plate XIX. * When this paper was communicated to t.be Chemical Society the organism now called B. intestini was spoken of provisionally a8 Bacterium fepmo OF SOME MICRO-ORGIANISMS. 729 9 and 10. BaciZlus$uorescens. 11. BaciZEus iniestini. These two fluorescent bacilli will be described by Dr.Klein in.Med. O$ic. Report 1887-8. Obtained by Dr. Klein from the large intes-tine in a case of diarrhoea i n a rabbit. Grown in broth it appears as a short bacillus with rounded ends. The longer forms are 1.0-1.5 p in length and about 0.4 p in thickness, I n a plate culture it appears the first day as small translucent dots, which afterwards develop in the depth to opalescent spheres and on the surface produce a shining expanse of pearl-like lustre. I n a stabbed culture the growt,h at first is chiefly in the depth. The track of the needle is bounded by rows of opaque spheres. There is no liquefaction. In broth at 22" or 33" a dense white turbidity is produced with no film. 25. XpiriZZum of noma (Lingard). This non-liquefying spirillum is described in the Practitioizer 38 121 Figs.32-35. 18. StaphpZococcus candidiu (Klein). This non-liquefying organism was obtained from condensed milk. Hed. Ofic. Report 1886-7 386, No. 3. Obtained from the blood of a scarlatina patient. Ibid. 1886-1 374 Pls. I-XII. The organism used in the experiments described in this paper was obtained from the blood of a scarlatina patient. 21. d/li'crococcus urece (Smith). Quart. Jour. Micro. Science 1887, 371. 23. Bacillus ;Roccus.-Obtained from garden soil at the Brown Institution. A straight bacillus 3.0-7.5 p i n length and 1 . 0 ~ in thickness ; motile in broth forms long interlacing threads. In gelatin plate cultures at 22" it forms the first day faint colonies which under a low magnifying power resemble balls of cobweb; from these fine interlacing threads resembling the mycelium of a fungus spread in every direction.Liquefaction commences the first day and on the second extends over most of the surface. In stabbed gelatin cultures at 22" it forms a liquefied channel in the course of the needle the first day. From this channel numerous fine branches resembling knotted threads pass into the gelatin. The second day liquefaction has extended over the surface ; the channel has disappeared and the gelatin is filled with finger-like branches starting from the old tube. A soft, thick film forms on the surface ; below this are two layers of growth with clear spaces between. These two layers finally become one and sink to the bottom. The liquefied gelatin becomes brown near the surface.19. Staphy lococczrs caiididus liguescens (Klein). Ibid. 370 P1. VII. 20. Mreptococcus scarlatince (Klein). The fifth day liquefaction is completed 730 WARINGTON THE CHEMICAL ACTIONS On agar agar at 22" a grey finely crinkled film quickly spreads over the whole surface. I n diluted urine or in very Teak broth a t 22" it forms dense white flocks resembling those produced by anthrax the rest of the liquid remaining clear. In normal broth at 33" it forms after some days a thick soft film which falls as ribbons. 24. Bacillus toru1iformis.-Obtained from garden soil a t the Brown Icstitution. Maximum length 5.3 p a few 6.4 p ; thickness 1-0-1.6 p. The ends thickened and much ronnded. Many short, thick forms joined together.The general arrangement very crooked and irregular. In gelatin plate cultures i t forms the first day white dots, having a granular margin. The second day there are large white colonies many 2x or & inch in diameter. Under a low magnifying power they appear of granular structure. Liquefaction hw com-menced. I n stabbed gelatin the first day's growth is chiefly confined to the needle channel and consists of moderately opaque dots. The second day liquefaction commsnces. I t occurs principally a t the surface. A funnel of liquefied gelatin is formed with a short conical pipe and white organism in the pipe the upper liquid being clear without film. The gelatin at the bottom of the tube remains long un-liquefied. 011 agar agar a t 22" it grows rapidly forming a pasty moderately opaque white film.In broth at 33" it grows luxuriantly producing great turbidity an abundant fine deposit b u t no distinct film. 25. Bacillus su1phzweus.-This was isolated from a surface growth which had appeared on some solutions which had nitrified. In a deposib from dilute broth it had a leiigth of 1-1.5 p and a thickness of 0.3 p. Taken from the surface of the broth it appeared as a net-work of bacilli the sheaths faintly stained with well-stained oval dots within. The colonies were a t first milky points becoming afterwards translucent spherical masses. The colour of the colonies is at first yellowish-white. By exposure to light the colonies become opaque and of a bright sulphur colour. There is no liquefaction. In a stabbed culture there is scarcely any growth in the depth of the gelatin.In broth a t 2 2 O it produces turbidity. Afker some days small clots are deposited some strung together in ropes together with some firrer threads. Festoons of these ropes hang from the surface but, The bacillus is motile. In a plate culture no growth appeared for several days. Growth soon appears upon an inoculated surface OF SOXE XICRO-OXGAKISMS. 731 there is no film. light. The deposit in broth turns yellow on exposure to The bacillus does not grow at 35". 26. Bacillus tardecrescens.-This organism was isolated from floating gelatinous masses occurring in some solutions of ammonium carbonate which had nitrified. Grown in dilute broth it appears as a small oval bacillus 1.0 p in length and about 0.5 p in thickness.Grown on gelatin it is somewhat smaller and appears more like an oval coccus. It is stained by gentian-violet very slowly. The bacillus is under all circumstances B ~ O W in growth. In a plate culture the colonies became visible after a week or more ; they grow into small translucent droplets of a somewhat smoky tint. In a stabbed cultwe the growth was chiefly in the depths it con-sisted of strings of spherical coloniea. In broth growth is slow and only produces a very moderahe turbidity. 27. Microcoecus gelatinosus.-T wo specimens of M. prodigiosu.9 from different sources produeed a mixture of red and white growths on the surlace of agar agar. The red I did not succeed in obtaining pure; the white was easily isolated by cultivation on nsar agar at 35" when the white only was developed.The white organism resembles the red in microscopical characters but grows much better than the red at high temperatures and preserves its vitality for a longer period. Grown in broth it appears as an oval COCC';LS or short bacillus, 0-5-1.3 p in length and 0-3-1*0 p in thickness. I n plate cultures of gelatin it appears the first day as milk-white dots which the next day are increased to Q or 4 inch diameter and liquefied. In stabbed gelatin liquefaction in the channel commences the first day and has extended over the surface by the second day. A funnel is then formed with a wide pipe. On agar agar it grows rapidly at 22" less rapidly at' 35' forming a thick pasty whitish mass moderately opaque covering the whole surface .I n broth it produces an abundmt turbidity and a considerable amorphous deposit but no film. There was no liquefaction. There is no film. GENERAL METHOD. The methods of culture on gelatin and agar agar were those originated by Koch and now generally employed. The nutritive gelatin and agar agar were of the composition employed by Dr, Klein the former containing 19. per cent. of gelatin. The prepara ‘732 WARIXGTON THE CHEMICAL ACTIOSS tion of the various liquid mediums will be described under the head of each investigation. The micyoscope was one kindly placed a t my disposal by the Chemical Society for the purpose of this investigation. The powers used were Zeiss’ D and Leitz’s & oil immersion lens. The stained preparations were made with gen tian-vio le t.All inoculations of solutions were made by means of Klein’s capillary glass pipette or hollow needle the cotton- wool stopper not beiiig removed from the tube o r bottle containing the sterilised liquid. By means of the same pipette the acidity or alkalinity of a solution could be a t any time ascertained without removing the st’opper. I n examining a liquid for nitrites or nitrates a much larger pipetbe was employed the capillary tube of which was passed between the stopper and the glass ; sterilised wool was held against the pipette at the point where i t entered the bottle so that oiily filtered air might enter when suction commenced. The large pipettes were kept in boiling water till the moment when required.By proceeding in this way the cultures were commenced and brought to a close without once removing the cotton-wool stoppers. As a check on the results a culture was generally made on gelatin or agar agar a t the end of each experiment or series of experiments, made in a fluid medium the object being tlo prove that the action had proceeded without the entrance of any foreign organism. This check was at first employed in every case but after confidence in the mode of work had been attained it was made use OE only to check striking o r irregular results or to ascertain the growth of an organism in milk. The method was especially necessary for the latter purpose as successful inoculation with an organism which produced no striking change in the milk could in no other way be ascertained.THE HYDROLYSIS OF UREA. The hydrolysis of urea and its conversion into ammonium carbonate which takes place in putrefying urine is now universally acknowledged to be due to the action of micro-organisms. Pasteur has described a micrococcus which acts i n this manner. More recently (1F85) Leube and Graser (Yi~chow’s A ~ c h i v 100 555) have separated the organisms occurring in samples of ammoniacal urine, and have found four which were capable of converting urea into ammonium carbonate. The most active of these was a short bacillus, which they name Bacterium ureoe. The next in activity and the most abundant was a micrococcus. There were besides two bacilli, which exerted a much feebler action. The authors also name a fift OF SONE MICRO-ORGANISMS.7 33 organism " Lungen Sarcin," as capable of converting urea into ammo-nia. Heraeus (Zeif. f. Hygien,e 1886 215 221) isolated four bacilli which effected an ammoniacal fermentation in urine ; three of these liquefied gelatin and were therefore distinct from the organisms of Leube and Graser. During last year Dr. TV. R. Smith (Quart. Jour. Xicmscopicab Science 1887 371) has separated about 20 organisms from nmmoniacaJ urine and found among them only one which was capable of producing ammonia from urea. The active organism separated by Dr. Smith was a micrococcus ; it was apparently not identical with the micrococcus of Leube and Graser as it slowly liquefied gelatin whilst their micrococcus did not. I t appears there-fore that at least nine organisms are known which are capable of hydrolysing urea.Cultivations in diluted urine have been attempted with all the organisms studied during the present investigation. Nos. 13 14, 16 17 and 26 have not occasioned a distinct turbidity in the solu-tions; Nos. 19 and 20 have given only a slight turbidity; the remainder have afforded plain evidence of growth. The urine soh-tions were of two kinds ; the first was 8 1 per cent. solution contain-i n g gypsum used for the experiments on nitrification. The cultures in this series of solutions were tested with Nessler's reagent after being kept for two or three weeks a t 32" ; the presence of gypsum in this case preventing volntilisation of the ammonia. The tint given by Nessler's reagent in the various cultures was compared with that produced by the same test in similar solutions unseeded with organisms.The organisms Nos. 1 2 4 5 11 13 15 17 18 19 20, 22 23 24 25 and 26 were tried in this manner. I n one case only (No. 18) was any distinct increase of ammonia perceived over that present in the unseeded solution. The second urine solution contained 25 per cent. of urine and was prepared especially to determine the action of the organisms on urea. The test-tubes containing the seeded urine were kept a t the temperature of 22" for 7-10 days the alkaliiiity of each cultivation was thexi determined by means of standard acid and alkali and the result compared with that given by unseeded urine placed under the same conditions. The method did not admit of extreme accuracy as the indicators employed failed to determine the point of neutrality with great exactness.All of the organisms under investigation with the exception of Nos. 7 8 16 and 19 were tried by this mehhod; in the case of 13 14 and 17 however no quantitative result was attempted as there was no evidence of any growth in the solution. With the exception of two organisms to be presently mentioned, none proved capable of distinctly affecting the alkalinity of the urine, In the case of the Jficrococcus weCe (Smith) a very distinct effec 731 WARINGTON THE CHENICAL ACTIONS was produced on the urine. I n two experiments made with different solutions the urine acquired a distinct ammoniacal odour. In one case the alkalinityat the end of seven days was increased to an extent equivalent to 0.0066 gram of ammonia per 10 C.C.of solution while in the second experiment the increase of alkalinity in eight days was equal to 0.0043 gram of ammonia. A very distinct increase in alkalinity also occurred in the case of B. Jluorescem. In the first experiment the increase of alkalinity in eight days was equivalent to 0.0039 gram of ammonia in 10 C.C. of the solution I n the second experiment the increase in seven days was equal t o 0.0037 gram. It seems probable from this result that the B. jluorescens is identical with the bacillus giving a green fluorescence obtained by Heraeus from soil as this organism was found by him to convert urea into ammonia. The property of effecting the hydrolysis of urea is apparently but rarely met with among micro-organisms ; in the present case out of 24 organisms tried only two could certainly be shown to possess it.I am doubtful however if either of these organisms is the cause of strong ammoniacnl fermentation. At all events in two comparisons of the result produced by the M. urea? with that yielded under the same circumstances by the mixed organisms of arable soil the latter proved far more effective. The soil mas from heavy nnmanured, arable land. The fragment used was about the size of a large pin's head; it wa8 taken half an inch below the surface. The two com-parisons were made in different solutions. The amounts of ammonia found per 10 C.C. of diluted urine were as follows :-Soil. M. uvea. I. 0.0252 gram 0.0066 gram. 11. 0.0116 , 0.0043 ,, It would appear probable therefore that the soil contained a more powerfully hydrating organism than the micrococcus of Dr.Smith ; it possibly contained the active bacillus of Leube and Graser. ACTION ON MILK. Milk is an excellent medium for studying the chemical action of micro-organisms being a liquid susceptible of many changes. I n all the experiments I have made skim-milk has been employed; the inconvenience of a surface layer of cream was thus avoided. The first trials were made wit,h skim-milk kindly supplied by the Ayles-bury Dairy Company. This milk had been separated in the centri-fugal machine it was thus perfectly sweet and as nearly as possible free from fat. The milk had however been sterilised by heat before it reached me and had of course again to be heated after it had bee OF SOME JIICRO-ORGLIKISIW.735 transferred to test-tubes. Owing probably to this double heating of the milk it required a loriger time to curdle than the milk subse-quently employed. The milk afterwards used in the experiments was taken directly from the cow into a sterilised stoppered bottle; the portion collected was never the first or the last milk from the udder. The bottle of milk was stood in a cool place for 36 hours. The milk below the cream layer was then drawn off by a syphon and 10 C.C. measured into each test-tube The test-tubes closed of course with sterilised cotton-wool were then heated a t 100" for 20 minutes in a steamer. Only three imperfectly sterililred tubes were observed during the whole investigation. The precipitation of the caseiin which determines the curdling of milk may be brought about either by an acid or by a soluble ferment, as rennet ; the latter is capable of solidifying milk even when it has been made alkaline with sodium carbonate.The proportion of acid required to curdle milk is smaller the higher is the temperature; thus milk that is only slightly sour may be completely curdled if heated to boiling. Attempts were made to ascertain what was the smallest proportion of lactic acid that would suffice to curdle milk. Very dilute lactic acid was slowly dropped into 10 C.C. of sterilised skim-milk ; milk acidified with different pro-portions of acid w~ then heated and the effect observed. With 0.15 gram of' lactic acid per 100 C.C. of milk a tolerably complete solidification of the milk was obtained when the test- tube was plunged into boiling water.A considerably smaller proportion of lactic acid sufficed to produce R partial curdling. With 0.32 gram of lactic acid per 100 C.C. of milk the whole became nearly solid in two hours at 33". The proportion of lactic acid required to curdle at lower tem-peratures could not be ascertained by this method as it was found impossible to mix the necessary proportion of acid with the milk without a t once curdling it the acid being unavoidably present in great excess at the spot on which the drop of acid fell. It is difficult to determine the quantity of acid in milk with much accuracy the reaction with litmus-paper being far from sharp. Neutral litmus-paper was used as the indicator in all the experiments.To Hueppe ( M i t t h e i l . a. d . k. Reiclzsgesundheitsamt 2 309) we are indebted for the fullest study of the organisms affecting milk. He names five as capable of souring milk. One st bacterium or bacillus, which curdles with the evolution of carbonic acid gas; two cocci obtained from the mouth; M. prodigiosus; and the coccus of ostao-myelitis." * Marpmann (Arch. Pharm. 1886 243) states that he has obtained five organisms producing lactic acid from the milk supplied to Gottengen j his desirip-tions do not seem to be publiehed yet 736 WARIKGTON THE CHEMICAL ACTIONS Among the organisms which I worked with five have distinctly acidified milk ; most of these organisms are certainly different species from those enumerated by Hueppe.The acidification by these or-ganisms has been in very different degrees. The three first named act apparently simply by the production of acid ; in the case of the fourth and fifth the action is more complicated. StapA ylococcz~s candidus.-At 22" this organism distinctly acidified the milk. The cultnre was continued for 43 days without visible change in the milk save the production of a slight sediment. At 32" after eight days the acidity of the milk was equal to 0.13 gram of lactic acid per 100 C.C. The milk was not curdled and did not solidify when placed in boiling water. At 33" with another batch of milk the acidity after 17 days was 0.15 per cent. Bacillus i.lztcstini.-Grown in milk a t 2 3 O it produced in 11 days an acidity equal to 0.33 per cent.of lactic acid; in 21 days an acidity of 0.33 per cent. ; in 31 days an acidity of 0.33 per cent. The milk did not curdle at 23"; when the cultures were placed in boiling water they immediately solidified. At 34" the milk became solid in two days ; the acidity was then equal to about 0.35 per cent. of lactic acid. Rncillus of Infantile Diawhcea.-A culture at 22" has been quite fluid after 43 days but examined at 83 days was found to be solid. The following amounts of acidity were found in experiments made at different times:-At 10 days 0.33 per cent.; a t 11 days 0.34 psr cent. ; at 7 days 0.31 per cent. ; a t 14 days 0.35 per cent. ; at 21 days, 0-38 per cent. ; a t 31 days 0.44 per cent. of lactic acid. At 33" the milk became solid in two days; the acidity was then 0.32 per cent.A t 36" the milk curdled in one day and gas was distinctly produced. No peptone was found in old cultures either of the curdled or un-curdled milk. 2Cli'crococcus zcrece.-At 22" the milk begins to thicken in 12 days, and some days later becomes solid ; the curd is however always soft. The acidities determined were-at 10 days 0.17 per cent. ; a t 7 days, 0.16 per cent. ; a t 14 days 0.24 per cent. ; a t 21 days 0.22 per cent. The last-named culture was tested for peptone but only a slight re-action was obtained. The amounts of acidity in different experiments were-at 7 days 0.19 per cent. ; a t 8 days 0.24 per cent. Micrococcus gelatinosus.-At 10" the milk becomes solid in about 15 days. The curd afterwards very slowly shrinks or partly dissolves.Peptone was distinctly present in the whey of an old culture. At 33" a soft curdling takes place in 5 days OF SOXE MICRO-ORGANISMS. 737 A t 23" the milk solidifies in 2 days; the amount of acid a t that time is about 0.15 per cent. The curd very slowly becomes smaller ; the whey then contains peptone very distinctly. At 32" the milk curdles in 1 day ; the acidity is then about 0.14 per cent. On reviewing the results given by these five organisms some points of interest appear. The first three organisms apparently do little more than produce lactic acid. The amount of acid produced by the second and third is considerably larger than that yielded by the first, but with each organism the proportion of acid which can be formed seems to be nearly a fixed quantity ; a t least after a certain acidity is reached further increase takes place with extreme slowness so that apparently a culture of six weeks will not enable Staph.candidus to curdle milk at all nor will it enable Bacillus intestini and the Bacillus of infantile diarrhma to curdle milk a t 22". We may probably assume that when a certain proportion of lactic acid has been produced the further growth of the organism is checked and with this the pro-gress of chemical action. That the curdling which does take place is determined solely by the lactic acid produced appears very pro-bable €or the amount of acidity in cultures freshly curdled a t 33" of B. intestini and the B a c i l l ~ ~ of infantile diarrhea is practically the same as the amount found requisite to produce curdling in the pre-vious experiments with lactic acid.When we turn to the last two organisms M. urece and 174. gelatino-sus the circumstances are very different. 21.2 gelatinosus is far more active in curdling milk than either Bacillus intestini or the Bacillus of infa7atCle diarrhma for these fail entirely to curdle at 22" while M. gelatinosusnot only curdles in 2 days a t 22" but also after 15 days a t lo" and yet M. gelatinosus produces far less acidity than these two organisms. They fail to curdle milk a t 22" with an acidity of 0.38 per cent. ; J!. gelatinosus succeeds in curdling milk a t this temperature wit8h an acidity of only 0.15 per cent. ! It is clear then that in the case of Jf. cyelafiizosus some curdling agent other than lactic acid takes a considerable share in the reaction.The same may be said of the far less perfect curdling effected by 31. ureE as here too the amount of acidity produced is by itself insufficient to effect the curdling which takes place. Further light is thrown on this subject by the results afforded by the next) three organisms. BacillzLs.~zLorescens 1iquescens.-At 2 3 O milk seeded w i t h this organism thickens in 3 days and becomes solid in 5 days. The milk is neutral a t the time of curdling. The small quantity of whey a t the surface exhibits a slight bluish-green colour. I n a culture a month old the curd remained undissolved ; it had become distinctly acid. Tests for peptone in cultures a fortnight old showed none or a trace only.VOTI. LTII. 31 738 IVARISGTOS THE CHEMICAL ACTIOSS Spirilli~m of Asiatic Cholera (Koch).-At 22" the milk exhibited ft loose curdling in 4 days ; i t was then neutral. Milk made distinctly alkaline with sodium carbonate became solid in 3 days. After 25 days the milk was still solid ; it had then become distinctly acid. At 33" milk curdled firnily in 4 days ; in 11 days it was slightly acid. Alkaline milk became solid in 1 day and was then still alka-line. The curd in the last-named experiment shrank a good deal. The whey tested after 15 days was distinclly acid and p v e a strong reaction of peptonc. Peptone was also found i n smaller quantity i n cultures in the ordinary milk 11 days old." SpiriZZ.um of Cheese (Deneke).-This organism acts but slowly on milk.At 22" a bright yellow ring forms a t the surface of the milk in 4-6 days. A tough crust is then developed giving the milk the appearance of solidification. Later a t somewhere about 11 days the milk gelatinises. It is still neutral and has a butyric ociour. Still Iatci. the opaque jelly par ti all^ redissolves ; the whey becomes slightly acid and contains a distinct amount of peptonc. I n alkaline milk the gelatinisation begins in about 12 days and gradually increases in solidity. At 33" a similar course of change takes place but apparently more slowly. The milk after 21 days is found to be alkaline and little or no peptone is present. I n the case of the TI. j7.riorescens liquescens and the Spirillum of Asiatic cholera we have clear instances of the complete curdling of the casein without the formation of any acid ; the action is in fact similar to that produced by the ferment contained in reunet.It has been frequently assumed that when an organism does work exactly similar to that performed by a ferment that it does SO by the production of a ferment. There is a t least one undoubted instance of the production of a ferment by a micro-organism namely the formation of invertin by yeast. According to S. Lea ( J . l'hys. 6 136) the bacteria which concert urea into ammonia also act by the production of a ferment. Riicrophytes would t h u s appear to effect certain hydrolytic actions by the same agency as similar actions are accomplidred in the cell sap of plants of more complex organisstion. Facts and analogy thus seem to support the hypothesis that ferments are produced by bacteria; and until more certain light is thrown on the subject we may well assume with Hueppe and others that the rennet-like curdling of milk is due to the production of a ferment.With the information gained by the study of the last group of * Since writing tlic above Dr. Klcin has informed me that the f w u l t j of curdling milk lias not been hitlicrto rccogniscd RS belonging to Kocli's spirilluni. The cspcrimcnts liavc therefore been repeated but with thc same results. The gelatin cultures prcpared from the curciled milk BL] a cliecli gave normal growtlie OF SOME MICRO-ORGANISMS. 739 organisms we can now understand how M. gelatinosus and M. urem are able to curdle milk with the production of only a part of the acidity necessary to effect this purpose.They act in fact as ferment producers as well a8 acidifiers. The cheese spirillurn possesses only a feeble gelatinising power ; it appears to occupy an intermediate position between the group of rennet-like organisms and the group of peptonising organisms which we have next to consider. Hueppe names four bacilli and two cocci as capable from his own observation of gelatinising the casein and then redissolving it with conversion into peptone. According to him this is by no means an uncommon property of bacteria; he ascribes it to the presence of two ferments a rennet ferment and a trypsin ferment. Peptone has been looked for in the present experiments by the application of the usual test of sodium hydrate and copper sulphate ; and also by means of phosphotungstic acid a reagent considerably more delicate.The milk t o be tested was in every case placed in a small dialyser of parchment-paper ; after 24 hours the diffusate was examined for peptone. According to Vines and Green (Phil. Trans., 178 43) dialysis is needed to avoid a confusion between peptone and hemialbumose both of which give the same reaction with chemical tests whilst only the first will pass through a membrane. In the case of a large number of organisms small quantities of peptone have been found in the milk cultures especially Then they became old. The peptone is usually most abundant in cultures made at a low temperature (23") ; a.t a higher temperature (33") less pep-tone and more ammonia is found.Milk cultures have for the same reason more tendency to become alkaline at the higher temperature. The following five organisms do not produce a solidification of the milk which corresponds to typical curdling but the casein is more or less gelatinised and then redissolved. Bacillus subtiZis (hay ).-Milk cultures of this organism at 22" begin in two or three days to show a translucent space immediately below the surface. Each day this space increases and the milk is resolved into a nearly clear liquid above and a soft jelly below the latter steadily diminishing by re-solution. The clear liquid afterwards becomes turbid from bacterial growth. The reaction of the milk is neutral during the early stages of the reaction but becomes strongly acid later ; tho odour is then pungent and acids of the fatty series are clearly present.As soon as the separation into fluid and jelly is clearly established, peptone can be distinctly found and it becomes abundant when the re-solution of the casei'n has made some progress, At 33" the action commences earlier but the re-solution of the casein seems to proceed more slowly than at a lower temperature. 3 ~ 740 WVARINGTON THE CHEMICAL ACTIONS The bacilli from scarlatina scurf and from jeyuirity extract, peptonised as vigorously as the bacillus from hay. Bacillus anthracite.-The experiments were made a t 22". The action goes through the same course as that described in the case of B. subtilis the production of acids in the latter stage of the reaction did not apparently occur.B. Jlocczhs.-The peptonising action is apparently as vigorous as with the two preceding organisms. No production of acid was observed. SpiriZlum of Cholera nostra (Finkler).-The culture was made both a t 23" and a t 37". A t the higher temperature the entire gelatinisation of the milk preceded the formation of a clear solution. Peptonisatiou was more vigorous a t the lower temperature. This organism did not apparently act quite as rapidly as the three previously named. Bacillus tomZi)?ormis.-The experiment was made a t 22". The action was considerably slower than in any of the preceding cases. The milk did not gelatinise for rather more than a week and after 26 days one half of the jelly was still undissolved. The culture remained neutral.The remaining organisms produced little effect on milk although nearly all grew freely in it as was proved by making gelatin cultures from the milk many days after it had been seeded. Streptococcus scadatin@ is stated by Klein to curdle milk a t 37" in three days or a little later. I obtained no curdling with this organism. At 22" the milk became distinctly acid but at the end of 43 days was quite fluid. At 32" the acidity after eight days was equal to 0.07 gram of lactic acid per 100 C.C. ; the milk was quite fluid and did not curdle when placed in boiling water. A culture a t 37" continued for several weeks. was equally unsuccessful in producing curdling. Fresh cultures on gelatin were used for inoculation in all these experi rn ent s. On reporting these results to Dr.Klein he informed me that the cultures of Xtr. scarzatinae in his possession had also lost the power of curdling milk. Since then he has obtained a fresh supply of the organism and finds that as before it curdles milk completely a t 37', rendering it very acid. This loss of the power of producing lactic acid during a long series of cultures on gelatin is a fact of great interest. The organism I experimented with had been grown on gelatin €or a year and four months before the experiments in milk were made. The two BaciZZi of septicmnia make the milk distinctly alkaline a change which is visible to the eye as the milk loses much of it OF SOME MICRO-ORGANISMS. 74 1 opacity. The bacillus from the mouse acted more energetically than that from the guinea-pig.A trace of peptone was found in both cases in cultures three weeks old. The Bacillus of swine fever, B. fiuorescens and B. tardecrescens also slowly produce an alkalinity in milk. The Typhoid bacillus and Staphylococcus luteus slowly render the milk slightly acid without producing further change. B. suZphureus grew without producing any appreciable change. The SpiriZZum of noma did not grow easily in milk. I obtained a culture in alkaline milk only. My stock of Staphylo-coccus candidus liquescens and Jli'crococcus aureus was unfortunately dead before the milk experiments commenced. It is of great interest when one property of an organism can be correlated with another. I venture to think that these experiments with milk enable us to do this.The whole of the organisms which fail to gelatinise milk are organisms that do not liquefy gelatin. The three organisms first mentioned in this section which apparently simply attack the milk-sugar and produce lactic acid are also non-liquefying. On the other hand the whole of the organisms which act on milk as ferments liquefy gelatin. This conclusion is quite in accordance with the view already taken by some investiga-tors that the liquefaction of gelatin is itself an action produced by a ferment. We may venture therefore to predict that every liquefying organism will be found capable of gelatinising the casein of milk. This precipitation of the casejin may take place with or without the formation of lactic acid and with or without a subsequent re-solution of the case'in as peptone.Some experiments were made to see how the mixed organisms of soil would attack milk ; the soil employed was a heavy loam from an arable field long unmanured. At lo" milk seeded with a small fragment of soil became gelatinised in nine days ; the milk was then slightly acid. I n 11 days gas began t o appear. Half the volume of the milk was then an opaque jelly. The odour was very bad a circumstance that had not occurred with any culture made with pure organisms. Peptone was distinctly present. At 22" the milk curdled in 2-3 days and became slightly acid. In 4 days the evolution of gas commenced and continued active for some days. The curd slowly redissolved and the solution contained peptone very distinctly. The reaction of the milk was for some time only slightly acid but at the end of 5 weeks the acidity had become very large.The odour passed through various unpleasant stages, which it is difficult to describe. Soil thus showed itself possessed of organisms acting like rennet It produced no apparent change 732 WARlNGTON THE CHEMICAL ACTIONS and trypsin ferments. It differed in action from the pure organisms tried first by the production of much gas and secondly by the formation of putrefactive products having a powerful odour. TEE REDUCTION OF NITRATES. The reduction of nitrates to nitrogen in sewage and in waters containing sewage seems to have been first noticed by Angus Smith (Mem. Lit. and Phil. Xoc. Manchester 1867 [3] 4 56). He sub-sequently made many experiments on the snbject which are described in his Reports to the Local Government Board on the Pollution of Rivers 1882 and 1884.Schloesing in 1868 (Compt. rend. 66 237) showed that during the fermentation of tobacco juice or of putrefying urine or during the lactic fermentation of sugar any nitrate that was originally present disappeared nitrous oxide nitric oxide and nitrogen gas being produced. Schloesing further showed in 1873 (Cornpt. r e d . 77 353) that a rigorous reduction of nitrates to nitrogen gas occurs in moist vege-table soil when a change of atmosphere is prevented. Muntz has shown (Ann. Chinz. Phys. 1887 11 125) that under the same cir-cumstances chlorates are reduced to chlorides bromates to bromides, and iodates to iodides. Some experiments were made by myself in the Rothamsted laboratory in 1880 (Jour.Roy. Agri. Soc. 1881 332) on the reduction of nitrates in soil. 7 lbs. of arable soil were placed in a percolator, forming a column 8 inches deep. The soil was saturated with water, and sodium nitrate in quantity equal to a large agricultural dressing, placed on the surface. After a week water was applied daily for nine days in quantity sufficient to keep the surface covered and the drainage was collected and analysed. At the end of this time nitrates and nitrites ceased to appear in the drainage water Only 21 per cent. of the nitrates applied were recovered as nitrates and nitrites in the drainage. Large transverse fissures were formed in the soil by the production of gas. The power of soil to reduce nitrates to nitrites and finally to destroy the latter nitrogen being probably evolved was further shown by later experiments not hitherto published.To 125 C.C. of a sterilised 40 per cent. solution of urine contained in a bottle closed by a cotton-wool stopper about 0.2 gram of arable soil was added and the bottle placed in a cupboard having a tempe-rature of about 10". In five days nitrites were distinctly present, resulting from the reduction of the nitrates naturally occurring in the urine. I n L2 days neither nitrites nor nitrates could be found, the solntion giving no reaction with diphenylamine OF SOME MICRO-ORGANISMS. 743 To 100 C.C. of a 213 per cent. solution of urine containing 1 gram of nitre per litre about 0.5 gram of arable soil was added the surface of the solution was covered with a thin layer of paraffin oil, and the bottle then placed in an incubator at 20".In two days a slight evolution of gas commenced and nitrites were distinctly present. Nitrites have per-manently remained in this solution. A similar experiment was made at the same time the only differ-ence being that 5 grams of nitre and 5 grams of glucose per litre were present in the solution. Gas was evolved as before but in much larger quantity. In 11 days all nitrites and nitrates had disappeared. In a duplicate. experiment to the preceding but conducted at a temperature of 35" gas was evolved in one day and ceased in four days by which time the nitrites first formed had disappeared and no nitrates could be found. The oxidisable organic matter contained in the soil aud urine is seen by the first twa experiments to be capable of reducing only a small amount of nitrate ; the addition of glucose in the last' two experiments determined the reduction af a much greater quantity of nitratre and the final disappearance of nitrite from the solution.That the reduction of nitrates is due to a reaction between the nitrates and the organic matter present has been recognised from the first by every investigator. Meusel in 1875,(Jozlr. pharm. [4],.22 430) was the first to prove that the reduction of nitrates to nitrites in natural waters is brought about by the agency of living organisms which he pronoiinced to be bacteria. Deh6rain and Maquenne in 1882 (Compt rend. 95 732) were the first to establish the same agency in the case of the reduction of nitrates in soil.It has in fact been abundantly shown that if sewage or soil is sterilised by the action of heat or antiseptics no reduction of nitrates will take place. The conditions necessary for the reduction of nitrates include a nourishing medium and temperature suitable for the growth of micro-organisms ; also the presence of organic matter capable af oxidation. The exclusion of air is favourable to reduction but not essential to it ; thorough asration is however fatal to the process. When the organism is capable of growing at a high temperature such a temperature is most favourable to reduction. The organic matter suitable for effecting reduction is very varied. According to different observers albuminoids sugar propyl alcohol ethyl alcohol fats, glycerol glycol acetates and tartrates are all active in this respect.The amount of reduction if other conditions are equal depends entirely on the quantity of oxidisable organic matter present. Thus Gayon and Dupetit found that in sewage seeded with putrid urine, In nine days evolution of gas ceased 744 WARINGTON THE CHEMICAL ACTIOXS not more than 01-0.2 gram of nitre per litre was reduced ; while in chicken broth seeded from the same source 50 grams of nitre per litre could be reduced. The reduction may be to nitrites to nitric oxide to nitrous oxide or to nif rogen. The formation of ammonia has been sometimes observed, but the evidence generally points t o its origin in the decomposition of nitrogenous organic matter rather than from the reduction of nitrates.The formation of nitric oxide has been frequently noticed during the fermentation of sugar-beet molasses when the solution is not kept sufficiently acid (Compt. rend. 66 171 237). The pro-duction of nitrous oxide has been observed by Deherain and Maquenne during the reduction of nitrates by soil (Cowpt. r e i d . 95, 691 854). The differences in the products of reduction are determined :-1. By the conditions of the experiment; 2. By the specific nature of the acting organism Till within the last few years all experiments have been made with mixtures of various organisms such as naturally occur in the air of the laboratory in sewage or in soil. Working with such natural mixtures one is easilyled to conclude that the character of the reaction depends entirely on the composition of the solution or on the other conditions of the experiment.Thus Munro (Trans. 1886 667) found that river-water readily oxidised ammonia to nitric acid but after the addition of a tartrate reduction set in and all nitrades disappeared from the solution. From such facts some have concluded that not only the extent of reduction but the whole difference between oxidation and reduction is simply deter-mined by the composition of the solution. These ideas become, however greatly modified when we become acquainted with the specific properties of the different organisms which together produce the actions in question it then becomes evident that in many cases one stage of the work accomplished is performed by one organism, while a second stage is effect'ed by another.Within the last few years considerable progress has been made towards ascertaining the reducing faculty of individual species of bacteria. I n 1882 Gayon and Dupetit (Compt. rend. 95 1365) pub-lished quantitative results showing the amount of nitrate reduced in the same time by seven distinct organisms. The most active was a small, mobile bacillus producing few spores anaerobic. This organism, cultivated in chicken broth at 35" with exclusion of air reduced 9.6 grams of nitre per litre to nitrite in one day. Several other organisms only reduced a t the rate of about 0.5 gram per day. I n 1886 the same authors published a splendid research upon the reduction of nitrates by bacteria (Ann.de la science agronornique 1885 1 226). The products of the reduction of nitrates are very varied OF SOME lIICRO-ORGAISIS?rlS. 745 A good abstract of this paper will be found in this Journal (Abstr., 1886 823). The authors found that the reduction of nitrates to nitrites was a very usual property of the bacteria examined ; they met with only one species (they notice 11) which could be cultivated in broth containing nit're without occasioning reduction. They state that a large class of bacteria reduce only t o nitrites no nitrogen gas being produced. They experimented with but did not succeed in isolating in a pure state the organism producing nitric oxide. The-y isolated two bacilli from sewage which they named Racteriuiiz denitri-ficans oc and /i? ; these reduced nitrates to nitrogen gas nitrites being formed under most circumstances as a stage in the reaction.The action of the a-organism was under favourable circumstances very energetic the liquid in which it is cultivated becoming covered with foam and evolving in one day its own volume of nitrogen the temperature at the same time considerably rising. The same orgmism when grown in an artificial solution containing nitre and asparagine, produced a considerable amount of nitrous oxide ; when the aspara-gine was omitted no nitrous oxide was formed. Heraeus (Zeit. f. Hygenie 1886 215) grew 10 bacteria isolated from water and soil in solutions containing a nitrate and sugar. Six grew well in this medium and of these t w o bacilli reduced the nitrate t o nitrite and ammonia.In the autumn of 1887 I communicated to the British Association (Report 1887 653) a preliminary account of the results obtained with 20 organisms part of those forming the subject of the present paper. In March of the present year a paper was read before our Society by Dr. Percy Frankland (Trans. 1888 373) in whicb the author describes the results he has obtained respecting the reduction of nitrates to nitrites with 32 species of bacteria separated by himself from the atmosphere or from natural waters. The main series of experiments were made in a weak solution containing as organic matter 0.25 gram of peptone and 0.3 gram of sugar per litre at the tempera-ture of 30". Of the organisms tried 15 or 16 were found incapable of reducing nitrates to nitrit,es; it is to be remarked however that seven of these organisms produced either no visible growth or a very slight turbidity.It would seem possible therefore that under condi-tions more favourable to growth the proportion of active organisms might be increased. In the experiments with pure organisms which I have now to describe the power of each organism to reduce nitrates has been tested in nearly every case in two solutions :-1. Beef broth contain-ing 5 grams of potassium nitrate per litre ; 2. A 20 per cent. urine solution containing 1 gram of nitre per litre. In some case 746 WARINGTON THE CHEMICAL ACTIONS peptone has beeu added to the broth and glucose to the urine solution ; but these were for additional trials outside the main series.For the broth 2 Ibs. of lean beef were taken cut small treated with cold water slowly heated with stirring and finally boiled for half an hour. The filtered broth was made slightly alkaline to neutral litmus-paper by the addition of sodium carbonate brought to boiling and again filtered. The perfectly clear filtrate then received the necessary amount of nitre and was diluted t o 1200 C.C. The urine solution was in some cases clarified by making it dis-tinctly alkaline with potassium or sodium carbonate boiling and filtering ; the filtrate was then neutralised with. phosphoric acid. Urine thus prepared remains perfectly clear when heated for sterili-sation and also when it becomes alkaline a fact of considerable advantage if the growth of an organism is to be determined by turbidity or by the formation of a deposit in the solution.Urine so clarified is not however so nutritive as urine which has not been filtered after boiling. The nitrated broth or urine was placed in small wide-mouthed bottles previously baked at 140" ; each bottle received 100 c.c. which nearly half filled it. The mouths were closed with plugs of sterilised cotton-wool a paper cap tied on and the contents of the bottles sterilised by heating for several hours at a temperature near 100". The trials were as a rule made a t two temperatures ; the lower one varied at different times from 20-23" the higher from 32-35'. The formation of nitrite was ascertained by two reag,entu 1. The zinc iodide and starch solution of Trommsdorf ; 2. Metaphenylene-diamine.The first reagent is so extremely delicate that one is apt to conclude from its indications that a large amount of nitrate has been reduced when in fact the amount of reduction has been extremely small ; the extent of reduction has therefore always been judged from the results of the second test. No strictly quantitative determina-tions of the nitrous acid formed have been made. The reducing power of each organism must be concluded mainly from its behaviour in the experiments with broth. The reduction of nitrate in the urine solutions was never large the proportion of nitrous nitrogen very seldom exceeding 1 per million of the solution. The results with urine also varied a good deal at different times, according probably to the varying nutritive character of the solution.With freely growing organisms the reduction of nitrate generally takes place speedily (if it occurs at all) both in broth and urine solutions and most speedily at the higher temperature if the organism will bear it. To avoid however repeated testing of the solations the cultures in broth at 32" to 35" were seldom examine OF SOME MICRO-ORGAh'ISMS. 747 till at least three days old and the cultures at 20-23" not before five days. The cultures of slowly growing organisms and many of the cultures in urine were allowed a longer time before examination. Looking first at the bacilli which form the largest class examined, I am disposed to place the following three first as possessing the greatest power of reducing nitrat'es to nitrites. Bacillus JcEoccus., jluorescens non- lipuescens. ? of swine fever. In the case of these organisms the culture in broth gave an exceed-ingly strong reaction with metaphenylenediamine a large precipita-tion of the colouring matter taking place immediately. Next in order stand the following eight bacilli the cultures of which in nitrated broth gave a very strong reaction with meta-phenylenediamine but with no immediate precipitate of colouring matter. Bacillus intestini. , of typhoid fever. , of infantile diarrhcea. , of septictemia (mouse). , anthracis. , 7 ) (guinea-pig). Spirillum of Asiatic cholera. , of cheese. Far removed from these in reducing power stands-Bacillus subtilis. The hay bacillus does not reduce nitrates in urine even when glucose is added.In broth ah 22" there is frequently no reduction for a week or more but generally a slight reduction afterwards occurs. l n broth at 35" a small reduction can be perceived after a few days. Reduction does not apparently occur till the broth becomes alkaline. The addition of peptone favours the reduction. The largest amount of nitrite noticed has probably not exceeded 5 parts of nitrous nitrogen per million of solution. B. subtilis from the scurf of a scarlatina patient gave the same amount of reduction as the bacillus prepared from hay. The bacillus from jequirity extract gave a somewhat greater reduction. Frank-land speaks of B. subtilis as giving no reduction; this is probably owing to the small amount of reduction it occasions when it feebly nutritive solution is employed.The following six bacilli have given no reduction of nitrates to nitrites : 748 WARINGTOW THE CHEMICAL ACTIOSS Bacillus Juorescens lipuescens. , toruliforinis. , sulyhureus. , tarclecrescens. Spirillum of cholera nostra. , of noina. 13. fluorescens liquescew and B. su@hureus have not been tested a t a temperature above 23" as they refused to grow in the hot incubator. H. toruliformis Finkler's spirillum and the spirillum of noma, have each been grown with an attempt at complete exclusion of air, the surface bf the culture-liquid being covered after seeding with a layer of paraffin oil ; this addition did not a t all prevent the growth of the organisms. The use of paraffin did not in any case determine a reduction of the nitrate.It was thought a t first that the layer of oil would effectually protect the liquid from contact with oxygen but a subsequent experiment in which a solution of cuprous oxide in ammonia was substituted for the culture-fluid showed that oxygen did in fact pass beneath the layer of oil. I hope a t some time to try the effect of an absolute exclusion of oxygen. Proceeding next to arrange the micrococci examined ascording to their reducing effect and classifying them as far as possible on the same principles as the bacilli we place in the first rank-Jficrococcus w e e . Xtaphylococcus candidus. , gel at irtosus. 7 luteus. These are apparently practically equal in efficacy to the three bacilli In the second rank stands one coccus-previously named as possessing the highest reducing power.Xtaphylocozcus candidus 1iquesceBs. Far removed from these comes-Streptococcu.s scarlatiim. This must for practical purposes be classed as without a reducing power for nitrates but in fact a trace of nitrous acid is shown 1.y the delicate iodide test when it is cultivated for some time in nitrated broth. Strept. scarlatince has been grown covered with a layer of paraffin oil without developing further reducing powers. One other micrococcus, Micrococcus aureus, has been grown in nitrated broth with and without a layer of paraffin on the surface without any production of nitrite OF SONE MICRO-ORG~AXISMS. 749 Among the organisms examined by Prankland it was the bacilli only which exhibited reducing properties ; among those now reported on the micrococci have proved to be in many cases as powerful as the bacilli.The table (p. 750) summarises the whole of the experiments on reduction. The letters R R r tr. 0 indicate the extent of reduction observed. The letters n. gr. signify “ no growth.” As already ment,ioned the results in urine varied a good deal a t different times the whole period of experiment being more than a year I n some cases single experiments showed a distinct amount of nitrite by the iodide test though none is mentioned in the table a majority of experiments made a t other times showing no nitrous acid. In several cases in which a large reduction occurred in broth a t R low temperature the experiment was not repeated a t a high one. It may be safely concluded that in each of these cases the reduction would have been equally great at the high temperature there being in none of these instances any ditficulty of growth at a high tempera-ture.Whenever an organism showed no or very little reduction a trial was always made at both temperatures. In several cases of small or no reduction the experiment was repeated with an addition t o the broth of 5 grams of peptone per litre; in no instance was the amount of nitrite increased by this addition the broth being in itself sufficiently rich in organic matter. The addition of glucose to the urine was chiefly tried with organisms which afterwards proved to be destitute of reducing power ; in one case in which the organism was able to reduce it,s action was dis-tinctly increased by the glucose.The whole number of organisms reported on is 25 ; of these 7 were entirely witlhout reducing power 1 only produced a mere trace of nitrite and 1 only a very small quantity. The remaining 16 reduced nitrates in broth with considerable vigour. Of those which failed to reduce one was of feeble growth produc-ing only a very limited turbidity in the broth but others in the same class were among the most vigorously growing organisms examined. The power of reducing nitrates is thus in no way determined by rapidity and vigour of growth. As far as I am able to judge the reduction which has occurred has been simply from nitrates to nitrites but this point can only be decided by strictly quantitative experiments. No production of gas has been observed in any of the broth cultures although the amount of nitrate present was quite sufficient to occasion visible gas if a reduction to nitrogen had taken place.B.$occzcs was grown in broth under a layer of paraffin with the especial object of observing any bubbles of gas that might be formed but without any such result Experiments on the Reduction of Nitrates to Nitrites by various i I n broth. I I 20-23". No. 32-3'7". -23 9 1 11 2 3 12 5 6 4 14 22 10 24 25 26 13 15 21 27 17 18 19 20 16 I Number of trials. --B.JEoccus 1 of swine fever 1 intestini . 1 of typhoid fever 1 of infantile diarrhea,. . 1 Sp. of Asiatic cholera 1 B. of septicaemia (mouse) 1 9 , (guinea-pig) . . .1 of anthrax 3 Sp. Deneke's 2 B. subtilzs 6 JEuorescens liiquescens 4 torulzyorrnis . 1 sulphureus 1 tardecrescens. 1 Sp. Finkler's . 1 of noma . 1 M. urea 1 gelatinosus. . 1 Htaph. EzLtezLs 2 candidus . 1 , , liquescens 1 Streptococcus scarlatina . 2 Juorescens non-lipuescens 1 M . aureus . 1 Reduo-tion. Reduc-tion. Number of trials. R R --I__-R R R n R R R R n It R tr. I) 0 0 0 0 0 R R R 0 0 R R -B -n R r 0 0 n. gr. 0 0 0 R R -R 0 0 2 1 -A - -1 I -2 2 2 3 2 1 1 4 2 1 2 1 5 3 - OF SOME MICHO-ORGANISMS. 751 this bacillus which was obtained from soil is thus not one of the clenitrifying bacteria of Gayon and Dupetit.Another means of jadgiug whet her reduction has proceeded further than the stage of nitrite is by studying the alkalinity of the solution. If the reduction of potassinm nitrate yields gaseous oxides of nitrogen, nitrogen or ammonia the solution must become alkaline from the formation of potassium carbonate. With the exception of the broth cultmes of Staph. candidus Ziquescens I have observations of the alka-linity of the broth cultures of each organism at the time when trial was made for nitrites; at this time 3-10 days from starting the cultures the reaction of the broth was slightly alkaline in cultivations of M. u r e a and B.$uorescens Ziquescens and decidedly alkaline in the case of B.$uorescens non-liquescens; in all other cases the broth was neutral or in a very few instances feebly acid.With the exception, possibly of the culture of B. fluor. non-Ziq. the reduction to nitrites would therefore appear to have occurred without the production of nitrogen oxides of nitrogen or ammonia. After a considerable time, most of the broth cultures became distinctly alkaline; but this in itself is no proof of the destruction of nitrites for many organisms render simple broth alkaline after growing in it fop some time the production of alkali being due to the destruction of the organic salts which the juice of meat contains. As far then as this imperfect evidence goes none of the organisms examined (with the possible exception of B. Jluor. non-Ziq.) possessed the power of reducing nitrates to nitrogen or to its gaseous oxides.That soil does actually contain organisms which are capable under certain circumstances of so reducing nitrates has been confirmed by my own experiments already described (pp. 742-3). Nitrites are not normally present in well aerated fertile soils or at least only to that minute extent in which they can be shown by delicate tests to occur everywhere. The drainage watera collected at Rothamsted only rarely contain sufficient nitrite to give a distinct reaction with the delicate iodide test. Misapprehension has some-times occurred on this point as some analysts are in the habit of reporting the quantity of nitrogen present as " nitrates and nitrites," without apparently ascertaining in each case the presence of nitrites in the water.NITRIFICATION. A good many investigations have been published during the last two years having for their object the discovery of nitrifying organisms or the farther elucidation of the process of nitrification : little success seems however to have attended these endeavours 752 WARI?;GTO?J THE CHXMICAL ACTIOSS Celli and Zuco (Gnzz. chim. Italians 17 99) found that a solution of ammonium chloride containing i+GG its weight of mercuric chloride passed and repassed two or three times a day through a column of sterilised sand and calcium carbonate gave after a few days a distinct reaction with diphenylamine. The amount of nitric or nitrous acid present was iiicreased when spongy platinum was sub-stituted for the sand. They conclude that the nitrification of ammonia in soil may take place without the intervention of an organism although organisms may assist in nitrification.Frank (Forschtingen auf clem Gebiete der Agdxlturphysik 10 56) expresses the opinion that nitrification is in greatest part an inorganic process. I n his hands soil which had been ignited was still capable of nitrifying ammonium chloride. Frank's experiments have been since repeated by Plath (Landw. Jahbiicker 16 89l) who comes to the opposite conclusion that soil when sterilised has no power of oxidising ammonia. We surely need not controvert a t length these conclusions of Celli and Frank contrary as they are to the accumulated evidence of many well-ascertained facts. I may perhaps refer to one of my own early experiments (Trans. 1878 44) which seems to give a distinct answer to the question at issue.Similar soil was placed in three tubes, through which air was daily drawn the air in one case containing a little chloroform vapour in another the vapour of carbon bisulphide. Where air alone was aspirated the nitric nitrogen in the soil increased from 8 to 50 parts per million ; where the air contained the vapour of chloroform o r carbon bisulphide no appreciable increase of nitric nitrogen took place. Although however we may strongly hold that the nitrifying power of soil is due to the action of a living organism we have still to explain the appearance of small quantities of nitric or nitrous acid in the experiments above mentioned. It is clear that in all such experi-ments the presence of traces of nitrites in the atmosphere must always be taken into account.I have already shown (Trans. 1881, 229) that distilled water cannot be freely exposed to air for any length of time without containing traces of nitrous acid and the presence of this acid is soon manifested if the exposure takes place in a room in TT-hich coal-gas is burnt. When therefore an experimenter on nitri-fication makes use of the diphcnylamine test which gives a blue coloration with 1 part of nitric or nitrous nitrogen in 10 millioiis of water or the far more delicate iodide and starch test which will indicate 1 part of nitrous nitrogen in 100 millions of watei' he will not improbably obtain a distinct reaction sooner or later without a living organism having had any share in the operation or indeed without any oxidation of ammonia having occurred in his solutions OF SOME MICRO-ORGANISMS.753 We turn next to the investigations with living organisms. Celli and Zuco isolated various organisms from the subsoil waters of Rome ; these organisms were then grown in a solution containing an ammonium salt. In every case the solution after a time gave a small reaction with diphenylamine. The amount of nitrous or nitric acid did not increase during several months even when a current of oxygen was passed through the solution. When the cultures were passed and repassed through a column of sand the reaction with diphenylamine was distinctly increased. All the organisms tried seem to have behaved in a similar manner. Adametz (Fomch. Agr.-P72ys. 1886,381) isolated 22 organisms from soil.At the end of five weeks an extremely small quantity of nitric acid was present. Heraeus (Zeit. f. Hygiene 1886 193) has conducted an investiga-tion on nitrification in Koch’s laboratory ; about 29 organisms were examined. He obtained a powerful nitrification when a large quantity of soil was placed in a solution of ammonium carbonate. The bactei-ial skin which formed on the surface of the solution in this experiment was also very active in setting up nitrification. I n 6 days the solution seeded with this skin oxidised 1.6 C.C. of indigo, and in 10 days 5.5 C.C. From this bacterial skin he isolated a bacillus, and a streptococcus ; these together with a bacillus and yeast isolated from fermenting urine were grown in an ammoniacal solution similar to that already employed.All the solutions gave a distinct reaction for nitrous acid with the iodide and starch test but the amount was tloo small to determine with indigo. About 13 organisms of well-known kinds were also grown in a 20 per cent. urine solution. Seven of these organisms produced in one day’s growth in the incubator suEcient nitrous acid to give a strong reaction with the iodide and starch test. Heraeus concludes that these seven organisms and the four previously mentioned possess oxidising pyoperties and me capable of forming nitric acid. Frank (loc. cit.) separated a number of organisms from soil but failed on trial to obtain nitrification with any of them. Leone (Atti d. R. Accademia (1. L i n c e i 1887 37) concludes from his experiments that all micro-organisms are more or less capable, under favourable conditions of producing nitric acid and that the same organisms in the presence of organic matter are capable of reducing nitrates.Muntz (Ann. Chim. Phys. 1887 11 128) also cautiously observes that organisms which appear identical with those which produce nitrification are capable of reducing nitrates when air is excluded. Manly Miles ( A g r i c u l t u r a l Science 1887 10‘2) writes as if he had FGur of these were grown in an ammoniacal solution. TOL. LIII. 3 754 THE CHEMICAL ACTIONS OF SOME MICRO-ORGANISMS. pure cultures of the nitrifying organism under experiment but he gives no description of the organism. P. Frankland (Trans. 1888 389) grew 32 orga,nisms obtained from air and water in a nutritive solution containing an ammonium salt for 40 days ; he obtained in no case more than a faint indication of nitrous acid.It seems very clea,r that not one of the investigators who have ex-perimented with isolated species of bacteria has obtained in his solu-tions more than a trace of nitrous or nitric acid; no one 72as obtained an amount that could be determined quantitatively. Another point which generally appears is that every organism tried gives nearly the same result. The failure to obtain the nitrifying organism in a separate form is most conspicuous in the case of Heraeus. He had in his hands a bacterial skin which possessed energetic nitrifying properties ; but the two organisms which he separated from this skin gave in a solu-tion of the same composition as that previously nitrified nothing more than the usual trace of nitrous acid.The statement of Heraens that seven of the organisms examined commenced the nitrification of a 20 per cent. urine solution in one day is. apparently due t o a mistake. My own experiments show that a urine solution of that strength cannot be nitrified by soil without the addi-tion of gypsum the commencement of nitrification in a strong solu-tion is also extremely slow (Trans. 1884 661). The nitrous acid which so speedily appeared in his solutions was doubtless due t o the reduction by the organisms of the nitrates naturally present in the urine (Trans. 1884 669). Celli and Zuco Leone and others apparently believe that the oxidising or reducing action of an organism is determined by the con-ditions in which it is placed and that the same organism is capable of discharging both functions.The capacity for exercising the entirely opposite functions of nitrification and the reduction of nitrates is indeed possessed by soil and by river water mediums which contain a multitude of different organisms but it cannot as yet be said to be proved of any single organism growing in a pure culture. The investigations of P. Frankland and myself have shown that all bacteria do not reduce nitrates even under specially favourable conditions and one that is capable of unmistakably oxidising ammonia has not apparently been at present isolated. My own results respecting the nitrifying power of isolated organisms admit of a very brief description.It has been already stated that two of the organisms B. sdphureus and B. tardecrescens had been separated from visible growths occurring in solutions which had nitrified. Two other organisms B. floccz~s and B. torulifornzis had been obtained from soil. The remaining organisms had not an origi SOME REACTIONS OF THE HALOGEN ACIDS. 755 that would suggest their possession of the nitrifying properties belonging to soil. The four organisms which appeared most likely to include one possessing nitrifying properties were submitted to several trials ; 20 other organisms were mostly tried but once. Four different solutions were employed the composition being varied with the view of obtaining growth with a minimum amount of carbon-aceous matter present. The examination of the solutions extended in every case over many weeks. When a fragment of arable soil was added to any of these solutions nitrification distinctly commenced within a fortnight, and in another fortnight all the ammonia (equal to about 25 of nitrogen per million of solution) had disappeared. None of the pure organisms experimented with gave any such result. A distinct reaction with diphenylamine was in some cases obtained but this did not appear to grow in amount although in such cases the examination was specially prolonged. The amount of nitric or nitrous nitrogen in the solutions did not apparently in any case exceed 1 per million and all of this could not be attributed to the action of the organism as the unseeded solutions in the incubator also gave some reaction with diphenylamine. When we have discounted the trace of nitrites probably obtained from the atmosphere there is clearly very little left that can be attributed to the action of the organism. The question whether any part of the nitrate or nitrite present was produced by the organism I am unable to decide; but it is quite clear that none of the organisms examined possessed any nitrifying power in any way comparable with that possessed by soil. An organism which nitrifies as soil nitrifies has yet to be isolated

ISSN:0368-1645    

DOI:10.1039/CT8885300727

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

SOME REACTIONS OF THE HALOGEN ACIDS. 755 LV1I.-Some Reactions of the Halogen Acids. By G. H. BAILEY D.Sc. Ph.D. and G. J. FOWLER B.Sc. the Owens College. IT had been noticed by one of us that when phosphorus pentoxide was used for drying gases there was in certain cases an indication that a considerable amount of gas was taken up. It was found for instance on passing a stream of air through phosphorus pentoxide tubes which had been used for drying chlorine sulphur dioxide and hydrogen chloride respectively that these gases continued to be carried over by the air for several hours. This was thought to arise from a mere mechanical retention of the gases but as the quantities expelled were large it was desirable to determine the character and 3 3 756 BAILEY AND FOWLER SOME REACTIONS Time elapsed .~ - - _ _ _ 143 hours . . 167 ) ) 239 ) ~ 309 , 452 , . . ~ 572 ) extent of the absorption. This seemed especially necessary seeing that the problem of the reaction of dry gases on one another is one of great importance and phosphorus pentoxide has been relied upon in experiments of a delicate nature where perfect desiccation was required. For the purpose of determining the extent t o which this ret'ention of the gas took place dry chlorine hydrogen chloride and sulphur dioxide respectively were passed over a long layer of phosphorus pentoxide in a drying tube which was weighed from time t o time. It was found that although the experiment was continued several hours no marked increase of weight occurred. These observations appeared therefore t o be in contradiction t o those prei-iously made.Recently however we have been engaged in studying the relative values of desiccating agents and the results which were obtained, when working with phosphorus pentoxide in particular led us to reopen the question as to the action of this substance on gases. Absorp-tion. 112 .G 124.4 138.6 148.5 158'5 163-7 Action of Hydrogen Chloride on Phosphorus Pentoxide. Our method of experiment was to introduce into the gas over mercury a quantity of phosphorus pentoxide in a glass t,ube open at both ends so as to give free access to the gas. With oxygen hydrogen, sulphur dioxide carbon dioxide chlorine and sulphuretted hydrogen, no absorption was observed although the gases were left in contact with phosphorus pentoxide for several weeks.With hydrogen chloride however a considerable absorption was noticed and this went on from day to day commencing very slowly, rising to a maximum diminishing a p i n and finally ceasing. Observations of the amount of absorption were made at convenient intervals and the result's are embodied in the following table the volumes being in all cases reduced t o normal temperature and pressure. I. I 1 I . X Time elapsed. Ab sorp-tion. Rate in C.C. per hour. 27 hours. * . 48 )) . 70 ,) . 77 ) ) . 103 , . 11.9 , . 6.8 C.C. 18.3 ,, 35'0 ), 444.5 > ) 74.3 ,, 95.3 ,, 0 *36 0 -55 0 -76 1-35 1'16 1 '28 Rate in C.C. per hour. 0 -72 0 '57 0 -19 0 -14 0 -07 0 '04 ~~ * Readings in the second series of obsei*vations after the P,05 tube had been in a vacuum (see p.757) OF THE HALOGEN ACIDS. 757 The slow rate of absorption during the time of the first few obser-vations is worthy of remark ; though it may merely indicate that it was necessary for the hydrogen chloride t o expel the air mechanically retained by the phosphorus pentoxide before it could come into intimate contact with the latter. This would also explain why no marked increase in weight was observed when the gases were simply passed over a layer of phosphorus pentoxide. The next question was to determine what was the nature of the action which took place between the hydrogen chloride and the phos-phorus pentoxide. That this was something more than a mere mechanical retention of the gas was indicated by the fact that when the phosphorus pentoxide was taken out of the hydrogen chloride absorption tube and left in a Torricellian vacuum for some days no gas was given off.The phosphorus pentoxide was therefore again placed in an atmosphere of hydrogen chloride and as seen from Table 11, continued to take up more of the gas. The only similar case of reaction with phosphorus pentoxide is as far as we are aware that noticed by Gladstone (J. Chem. Xoc. 1866, 19 290 and 1868 21 64),* in which dry ammonia is taken up with formation of diamidopyrophosphoric wid. To determine quantitatively the amounts of phosphorus pentoxide and hydrogen chloride which enter into the reaction well-stoppered bottles were filled by displacement with hydrogen chloride and a quantity of phosphorus pentoxide quickly inserted from a weighed tube.The bottles were opened from time to time under mercury until there was no further absorption ; at this stage all the phosphorus pentoxide had become liquid. As an instance of the results obtained by this method the following numbers may be given :-Phosphorus pentoxide taken = 0.1255 gram, hydrogen chloride absorbed = 28.5 c.c. that is 1 gram of phosphorus pentoxide takes up 227 C.C. of hydrogen chloride. The amount of pentoxide added was thus known. The amount of the absorption was thus measured. These numbers agree very nearly with the equation-2PZOj + 3HC1 = POC1 + 3HP03. Calculating from this equation 0.1255 gram phosphorus pentoxide should take up 29.7 C.C.of hydrogen chloride ; a result which agrees * Professor Thorpe has called our attention to the fact that Mallet (Chem. News, 44 164) had observed that hydrogen fluoride was taken up by phosphorus pent-oxide from which he concluded that the pentaauoride of phosphorus was formed. Moissan has however shown (Alzm. Chim. Phys [ S ] 12 486) that phosphorus tri-fluoride combines readily with oxygen forming a stable oxyfluoride POF3 analogous to the oxychloride and it ;would thus seem very probable that the retention of hydrogen fluoride observed by Mallet was due to the format.ion of thie oxyfluoride 758 BAILEY AND FOWLER SOMX REACTIONS very nearly with that found especially when we take account of the small errors inherent in the method of experiment.I t was difficult in this way to obtain a sufficient quantity of the compound to determine its nature. An apparatus such as is shown in the figure was therefore constructed and as this form which our experiments showed to be most convenient may be found useful in cases where a regulated supply of gas free from air is required we describe it. A is a fractionating bulb into which a quantity of phosphorus pentoxide was introduced B is a three-way tap C a large bottle serving as a reservoir €or the hydrogen chloride and D a flask contain-ing sulphuric acid and common salt. After attaching the bulb A, containing the pentoxide the whole apparatus was filled by displace-ment with hydrogen chloride and the side tube of A sealed up. Any excess of hydrogen chloride may escape by the tube E dipping under mercury.The whole apparatus is thus sealed off from the air. The course of the reaction may be examined by allowing the vertical limb of the three-way cock B to dip under mercury which thus showed how the absorption was proceeding. By means of the three-way tap supplies could be drawn periodically from the reservoir to replace the gas taken up by the phosphorus pentoxide and the reservoir could be replenished by warming the flask D. When the reaction appeared complete the bulb A was detached, a thermometer introduced and the liquid distilled off precautions being taken to prevent ingress of moisture. The liquid boiled a t 108" and showed all the properties of phosphorus oxychloride. A residue of metaphosphoric acid was left in the bulb.From these experiments it appears that a reaction analogous to that of Kolbe and Lautemann (Annulen 113 240) for obtaining phos-phorus oxychloride occurs spontaneously when hydrogen chloride is left in contact with phosphorus pentoxide and is complete if sufficient time be allowed OF THE HALOGEN ACIDS. 759 Having established therefore the reaction in the case of hydrogen chloride it was interesting to observe what took place in the case of the other halogen acids. These were experimented on in the same way. In the case of hydrogen bromide absorption took place and the phosphorus pentoxide became liquid the reaction beginning slowly and increasing to a maximum as in the case of hydrogen chloride. Complete absorption however is effected much more slowly in the case of hydrogen bromide.With hydrogen iodide no absorption could be observed so that although the circumstances seem especially favourable for the produc-tion of an oxy-compound no such body is formed. Action of Hydrogen Chloride on Mercury in presence of Oxygen. Although it is known that hydrogen chloride has no action on mercury it had been noticed that in some of the experiments with this gas the mercury had been attacked and this occurred only where oxygen was present. Richardson has recently shown (Trans. 1887 Sol) that hydro-chloric acid when mixed with excess of oxygen and exposed to light, yields chlorine; but we found that even with small quantities of oxygen the mercury was attacked and became coated with a white salt.48 C.C. hydrogen chloride and 30 C.C. oxygen dried over calcium chloride were introduced along with a small quantity of mercury into a stoppered vessel and exposed to diffused daylight for three weeks to allow of the reaction reaching its limit. Only 5 C.C. of gas remained and the whole of the hydrogen chloride had disappeared as there was no further absorption on the introduction of water. The chlorine had manifestly been taken up by the mercury and oxygen equal to half the volume of the hydrogen chloride had disappeared, part of which only could have combined with the hydrogen. In order to obtain a quantity of the mercury compound a large cylinder was filled with a mixture of hydrogen chloride and oxygen, and a few globules of mercury introduced.The mercury was fre-quently shaken to renew the surface the charge of hydrogen chloride being renewed from time to time. When sufficient of the compound appeared to be formed it was collected and any globules of mercury adhering to it were removed as completely as possible by amalgama-tion with silver. Analysis however showed that the compound still contained free mercury. The analysis was performed by beating the substaace (prsviouslg The vessel was then opened under mercury 760 SOME REACTIONS OF THE HALOGEN ACIDS. dried over phosphorus pentoxide) with quicklime collecting and weighing the mercury and estimating the chlorine in the residue. 0.2890 gram substance yielded 0.2347 mercury. 0.0359 chlorine. The difference 0.0184 could consist only of oxygen or water.This evidence taken in conjunction with that derived from the volumes of the gases concerned in the reaction lead to the equation-2Hg + 2HC1 + 0 = HgrzOCI,,H,O. It seems therefore that in such an experiment there is formed an oxychloride resembling the oxycyanide Hg20 (CN),. Berthelot we found (Ann. Chim. Phys. [ 5 ] 23 ZOO) had already noticed that mercury was attacked in some cases by hydrogen chloride, and set down the compound formed as calomel but gives little experimental evidence in support of this. The reaction proceeds in the dark as well as when exposed to light, and indeed powerful sunlight seems if anything unfavourable to it. It was found by these experiments that if excess of oxygen is present all the hydrogen chloride is used up and vice veysd.I f , however a considerable amount of nitrogen is present as when air is mixed with the hydrogen chloride the reaction proceeds much more slowly and all the oxygen is not taken up even in presence of con-siderable excess of hydrogen chloride. The reaction of course goes on much more slowly if excess of hydrogen chloride is present. In one case a mixture containing only 9 per cent. of oxygen was l e f t in contact with mercury ; at the end of three weeks nearly half the oxygen remained the total amount of gas entering into combination being only 20 C.C. out of the 150 C.C. used. Action of Hydrogen Bromide and Hydrogen Iodide on Mercicry. A mixture containing 33 vols. of oxygen t.0 45 of hydrogen bromide was taken and kept in contact with mercury in the dark.The whole of the hydrogen bromide was found to have disappeared 37 C.C. of gas remaining. On testing this gas for oxygen with a glowing splinter a powcrful detonation occurred showing the presence of a considerable quantity of free hydrogen ; it appeared in fact that whereas in the case of hydrogen chloride the whole of the hydrogen combined with the oxygen with hydrogen bromide only part of it was thus transformed. Hydrogen iodide as is well known acts directly on mercury; the green mercurous iodide is first formed very rapidly and this is converted by excess of hydrogen iodide into the red iodide Thi ACTION OF POTASSIUM ON TETRALKYLAMMONIUM IODIDES. 761 takes place even if tlhe mercurous compound has been dried at 100”. The mercuric iodide is reconverted into the mercurous compound on adding excess of mercury.Chlorine if allowed t o act on mercury in the cold forms mercuric chloride even in presence of a considerable excess of mercury. Whilst the work was proceeding we communicated with Dr. Richardson t,hinking that he might have made some observations in the same direction and we have to acknowledge the great courtesy with which he placed such facts as he had noticed at our disposal, desiring that they should appear in conjunction with ours. We cannot do better than quote his words “ Mercury in contact with hydrobromic acid and hydriodic acid (in presence of oxygen) readily decompose in the light also however in the dark.” “Hydrobromic and bydriodic acids dry and free from oxygen were almost completely decomposed in the dark ; copper completely decomposes hydriodic acid.” The reactions observed open out an interesting field f o r studying the course of chemical reaction and one in which we look forward with interest for the results of Dr. Richardson’s investigations of the part played by light in these phenomena

ISSN:0368-1645    

DOI:10.1039/CT8885300755

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

ACTION OF POTASSIUM ON TETRALKYLAMMONIUM IODIDES. 761 LV 111.- The A c t i o n of Potassium o n Tet r alk y lanzmoniunz Iodides. By C. M. THOMPSON and J. TUDOR CUNDALL. ACCORDING to Weyl (Wa,tts’s Dictionary 5 329) when potassium in solution in anhydrous liquefied ammonia acts on ammonium chloride ammonium in the free state is produced and dissolves in the ammonia forming a blue solution. This statement lacks con-firmation and has been contradicted by Seeley (Chem. News 23,169). It seemed possible that more decisive results might be obtained by using some tetralkylammonium-compound instead of ammonium chloride since just as the tetralkylammonium hydrates are easily obtainable whilst ammonium hydrate is not so the free tetralkyl-ammonium might be more stable that ammonium and in any case the products of the reaction would be more easily separated and examined.We experimented first with tetramethylammonium iodide. Some of this substance was inclosed with excess of metallic potassium in one limb of a sealed V-tube about 1 to 15 cm. in diameter the other lim 762 THOMPSON AND CUNDALL THE ACTION OF of which contained ammoniated silver chloride. In preparing this latter compound care was taken to exclude water as completely as possible. Nevertheless we found that some was present and i n order to dry the ammonia before condensation a quantity of sodium chips were placed between plugs of glass-wool immediately over the silver-compound. aaa. Plugs of glass wool. 6. Bent-up capillary tube. A. Ammoniated silver chloride. B.Sodium chips. C. Potassium and organic iodide. The tubes were left sealed up for two or three days before distilling over the ammonia so as t o allow time for the sodium to remove the traces of water. The potassium was placed in a separate small tube so that its solution in the ammonia did not at once come in contact with the iodide. On spilling some of the blue solution on the iodide no ap-parent change could be observed when the above precautions as to the removal of water were taken although in previous experiments when traces of water were present rapid decolorisation with e-rolution of gas took place. After the potassium solution and the iodide has been mixed the tube was left at the ordinary temperature for some hours in which time the ammonia was almost entirely reabsorbed by the silver chloride.On opening the tube the pressure observed was not very great and the odour of trimethylamine could be detected notwith -standing the strong smell of ammonia. The residue in the tube was treated with alcohol which would convert the excess of potassium and any trimethylammonium which might have been formed into alcoholates. The greatex- part of the tetramethylammonium iodide was left undissolved and quite un-changed by this treatment although excess of potassium was present. It was collected on a filter and in one case weighed 3.3 grams, 5 grams having been originally employed POTASSIUM ON TETRALKTLAMMONIUM IODIDES. 763 The strongly alkaline filtrate was neutralised with hydriodic acid and evaporated. In the solid residue we could detect nothing but potas-sium iodide.A portion of the gas contained in the tube was collected after passing through dilute hydrochloric acid. The results got on explosion with excess of oxygen and absorption of carbon dioxide by potash indicated that it consisted of a mixture of hydrogen ethane, and nitrogen. The ammoniated silver chloride left in the tube was heated and the gas evolved absorbed by dilute hydrochloric acid. This solution was mixed with the hydrochloric acid solutiou got in purifying the gas and the whole evaporated to dryness. The solid was extracted with alcohol the solution evaporated and the residue converted into platinochloride. The quantity of salt obtained from one tube was too small to divide into fractions.It contained 42.6 per cent. platinum whilst ammonium platinochloride contains 44 per cent. showing that tri-methylamine is probably present. Whilst therefore the greater portion of the tetramethylammonium iodide remains unchanged a part is acted on by the potassium giving potassium iodide ethane and trimethylamine. Tetramethylam-monium if formed at all is under the conditions of the experiment, unstable. As the results got with tetramethylammonium iodide were unsatis-factory we performed another series of experiments with phenyltri-methylammonium iodide. This substance was easily prepared by bringing together dimethylaniline and methyl iodide in molecular proportion and was purified by crystallising from water heating with aqueous sodium hydrate to decompose hydriodides washing well with alcohol and finally crystallising from alcohol.Since the iodide is insoluble in ether it might be better after washing out the sodium hydrate to wash thoroughly with ether. That the substance which we employed was pure was shown by the fact that it was odourless and that the aqueous solution gave no turbidity on warming with sodium hydrate. 5 grams of the carefully dried iodide were treated with excess of potassium in a sealed tube as before. On allowing the potassium solution to come in contact with the iodide the blue colour was a t once changed to brown. After the action was complete the tube contained two liquids one brown and wetting the glass the other showing a red metallic lustre like that of copper and running about in globules like mercury.The latter liquid is exactly like a strong solution of potassium in ammonia as described by Seeley and re-peatedly obtained by us. The tubes had to be left for some time before the ammonia wa 764 ACTION OF POTASSIUM ON TETRALKYLAMMONIUM IODIDES. reabsorbed for the brown liquid checked its evaporation. On open-ing they showed only a slight pressure and the escaping gas smelt strongly of trimethylamine. A portion of the gas was washed with dilute sulphuric acid trans-ferred to a eudiometer and exploded with excess of oxygen. Addi-tion of potash then caused only a slight diminution of volume. Calculating on the assumption that the whole of the carbon dioxide is derived from ethane the rest of the gas consisting of hydrogen and nitrogen we found the percentage of ethane in one case to be 8.15 in another 13.2.Since the total quantity of gas free from ammonia, which could be got from one tube containing 5 grams of iodide did not exceed 100 c.c. the quantity of hydrocarbons formed in the reaction is exceedingly small and may be the result of some secondary reaction. The end of the tube was cut off and the residual products of the reaction were extracted with dry ether. On evaporating the ether a liquid was left which boiled at 187-188" (uncorr.) and on heating with benzotrichloride and zinc chloride gave malachite-green. It was therefore dimethylaniline. I n one case from 5 grams phenyltrimethyl-ammonium iodide we obtained 1.1 gram though some was un-doubtedly lost during the evaporation of the ether.The residue after extracting with ether was treated with cold absolute alcohol. A crystalline substance was left which was found to be pure potassium iodide. The alcoholic solution was neutralised with sulphuric acid and the potassium sulphate which was free from organic matter was filtered off and the filtrate evaporated. It gave a residue which consisted of unchanged phenyltriniethyl-ammonium iodide mixed with a small quautity of resinous substance soluble in ether. By heating the silver compound contained in the tube a gas was given off which was absorbed by dilute hydrochloric acid. The solution got in this way from one tube containing 5 grams of the iodide was evaporated and the residue extracted with alcohol. The aqueous solution of the salts soluble in alcohol was precipitated by platinum chloride in five fractions.The first three on analysis gave numbers showing that they consisted of ammonium platinochloride. The fourth and fifth contained 37.2 and 37.9 per cent. of platinum respectively the calculated percentage for (NMe,)2,H,PtCl being 37.0. Whilst these fractions were being heated a strong smell of trimethyl-amine was observed so that there can be no doubt as to the nature of the substance. The two fractions weighed 0.778 gram containing 0.1'73 gram trimethylamine so that the latter is formed in considerable quantity if we take into account the facts that some of the phenyltri-methylammonium iodide is not acted on and that its molecular weight is very high THE VAPOUR-DENSITY OF HYDROFLUORIC ACID. 765 The following products of the reaction were therefore recognised with certainty :-Potassium iodide dimethylaniline and trimethyl-amine; whilst ethane or other hydrocarbon was not identified and cannot have been present in any considerable amount. I f phenyltrimethylanimonium were formed and afterwards decom-posed in all probability dimethylaniline and ethane would be produced. The maximum quantityof ethane which could have been formed from 5 grams iodide (13 c.c.) weighs about 0.017 gram while the amount calculated from the equation-2C6H5(CHs)3NI + 2K = 2KI + 2C6H,N(CH3) + CzH6 weighs 0.28 gram or assuming that one quarter of the iodide remains unchanged 0.21 gram. The formation of trimethylamine also shows that the above equation does not represent the main reaction and on the whole it appears t o us improbable that phenyltrimethylammonium is formed at all. University College Cardif

ISSN:0368-1645    

DOI:10.1039/CT8885300761

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

THE VAPOUR-DENSITY OF HYDROFLUORIC ACID. 765 LIX-The Vupour-density of HydroJuoric Acid. By T. E. TEORPE F.R.S. and F. J. HAMBLY. (Preliminary Notice.) GORE in his researches upon anhydrous hydrofluoric acid showed that by heating a known volume of hydrogen with a slight excess of silver fluoride the volume of hydrofluoric acid gas formed was approxi-mately twice that of the hydrogen taken if it were measured at about 100" C. but at lower temperatures it was considerably less than was demanded by the formula HF (PhiZ. Tg-ans. 1869 173). Mallet determined the density of hydrofluoric acid at 30.5" C. by weighing the vapour in a large glass flask coated internally with paraffin and obtained a value corresponding to the molecular weight 39.32 at that temperature (Amer.Chenz J. 1881 3 189). Although this result shows that hydrofluoric acid does not give a vapour-density corresponding tlo the formula HF at temperatures near its boiling point it cannot be considered as conclusive proof of the existence of the molecule H,F,. The alteration in vapour-density may in fact resemble the well-known case of acetic acid in which there is a gradual and progressive breaking down of a complex molecular grouping. We have therefore invostigated the subjec 766 THORPE ASD RODGER THIOPHOSPHORYL FLUORIDE. with the object of ascertaining whether the gas possesses a constitu-tion corresponding to the formula H2F2 through any appreciable range of temperature. By means of a large platinum apparatus provided with stopcocks of the same metal we have determined the vapour-density at tem-peratures varying from 26.4" t o 88.3". Pure anhydrous hydro-fluoric acid was prepared as required for each experiment from the acid potassium fluoride and was then redistilled through the platinum apparatus placed in a bath of glycerol and heated to the desired temperature. In all 14 experiments were made at short intervals of temperature between the points given and the values obtained correspond to molecular weights ranging from 51.19 at 26.4" to 20.58 at 88.3" the process of breaking up of the molecular grouping being analogous to that observed in the case of acetic acid. We hope t o be able to lay a detailed account of our experiments before the Society at an early date

ISSN:0368-1645    

DOI:10.1039/CT8885300765

出版商:RSC    

年代:1888

数据来源: RSC

摘要:

766 THORPE ASD RODGER THIOPHOSPHORYL FLUORIDE. LX.-Th.iop hosphory 1 Fluoride. By T. E. THORPE F.R.S. and J. W. RODGER. (Preliminary Notice.) WHEN phosphorus pentasulphide is heated with lead fluoride best in a leaden tube a gas is formed which analysis shows to be thiophos-phoryl fluoride PSF,. Bismuth fluoride heated with phosphorus pentasulphide forms the same gas but a higher temperature is required to bring about the reaction. We have also been able to prepare the new gas by heating a mixture of sulphur phosphorus and fluoride of lead using a considerable quantity of the last-named sub-stance ; the excess serving to moderate the otherwise violent reaction. An additional mode of formation consists in heating a mixture of arsenic trifluoride and thiophosphoryl chloride in a sealed tube at a temperature of 150" ; but the most convenient method of preparation, and the one giving the gas in a state of purity is that firs2 described.Thiophosphoryl fluoride is a transparent colourless gas which may be liquefied in the Cailletet apparatus. In contact with the air the gas if pure spontaneously ignites burning if it be issuing from a jet with a pale yellowish-green flame tipped with blue. If the experiment be so arranged that a considerable quantity of the gas is \jrought in contact with the air it ignites as before but produces i ACTION OF BROMINE ON POTASSIUM FERRICYAKIDE. 7 67 the first instance a beautiful blue flash of light subsequently followed by the yellowish-green flame easily observed in the case of the jet of gas.Thiophosphoryl fluoride is dissolved by water but not very rapidly. Under the action of heat o r the electric spark it is decomposed with comparative ease sulphur separating in the first instance. If a quantity of the gas be heated in a glass tube over mercury for some time the volume alters phos-phorus and sulphur are deposited on the sides of the tube and the resulting gas consists entirely of silicon tetrafluoride. This reaction may be taken advantage of in order to determine the amount of fluorine in the compound. Observations of its spectrum show that the gas is dissociated at the lowest temperature of the spark. Thio-phosphoryl fluoride is soluble to some extent in ether insoluble in alcohol and benzene. It is completely absorbed by peroxide of lead, and unites with ammonia to form a white solid. Passed over heated sodium the metalstakes fire and burns with a red flame the residual mass giving off spontaneously inflammable phosphuretted hydrogen on treatment with water. We reserve the details of our study of the properties of this gas for a communication which we hope to lay before the Society at an early date. It has no action on mercury

ISSN:0368-1645    

DOI:10.1039/CT8885300766

出版商:RSC    

年代:1888

数据来源: RSC