AFFINITP. CONSTANTS OF AMINO-ACIDS. 153 XV.-The Afinity Constants o f AminocarboxyZic and Aminosulphonic Acids as cletem-nined By the aid oj’ Met hy 1 - orange. By VICTOR HERBERT VELEY. Introductory. IN a former paper (Zeit. physikal. Chem., 1906, 57, 147) on the above subject it was established that the affinity factors of organic acids experimentally found by a tintometer method were in complete accordance with those deduced by Ostwnld and his co-workers by the electric conductivity method, in accordance with the well-known general equation : +(k) = u2/(1 - a) V a = p/p (1) The acids formerly investigated, namely the carboxylic, and certain VOL. XCI. M154 VELEP : THE AFFINITY CONSTANTS OF hydroxy-, nitro-, and chloro-derivatives of the aliphatic and aromatic series, conformed to two general types, expressible by the equations : 1.j Y = k X ( y = k x - b (ib) 11. logy=logk+x loga (ii) namely those of straight lines and logarithmic curves. I n the above equations y is the variable height of a column of a methyl-orange solution, x unit masses or volumes of acids added to a fixed similar column of the same methyl-orange solution, k the affinity factor. The constant b in equation (ib) depends on the conditions of each set of observations, and the constant x in equation (ii) is the number of unit masses of acid added less one. I n the present communication this simple line of investigation is extended more particularly to the aminocarboxylic acids, generally regarded as true amphoteric electro- lytes, and the aminosulphonic acids, a class of substances s0mewha.t neglected from the physico-chemical standpoint, although herein of more importance, as the indicator methyl-orange belongs to them.Method of Experiment .-This was precisely similar to that previously described, and consisted in adding successive portions of 0.1 C.C. of the several acid solutions to 20 c,c. of a N/40,000 methyl-orange solution contained in one tube of a tintometer, and varying the height of a column of the same solution contained in the other tube of the tinto- meter until the colour tints were equally matched. Improvements in the method of working might doubtless have been introduced, but as the investigation had been commenced somewhat hurriedly for the Sixth International Congress of Applied Chemistry a t Rome, it was thought best t o continue the work on the previous simple lines.Greater accuracy might have resulted from such improvements, such as a reduction of error from 5 to 3 per cent. or even less. However that may be, the results herein detailed conform in every way to t,hose obtained in the past work, and even though the experimental error may be regarded as excessive, yet it is not greater than those deduced from electric conductivity experiments, wherein observational errors of a first power magnitude become, by the process of calculation, of a second power magnitude. It will be understood that if 0.1 C.C. of a x/!N-solution of an acid is added to 20 C.C. of the standard methyl- orange solution, the concentration a t the first observation is xIZc)ON, and at the nth observation is x / - X The reciprocals of these numbers, designated V, are expressed in eqiiivalent acidic and not in molecular concentrations.Samples of Acids.-Some of the samples were purchased from well- known firms specially for the investigation; others were kindly supplied 200 7%AMINOCARBOXYLIC AND AMINOSuLPHONIC ACIDS, 155 from the laboratory collections of the University of Oxford and of Mag- dalen College, and by individual friends. I have to express my obliga- tions for kindness shown, and herein especially to the Directors of the Badieche Anilin- und Soda-Fabrik, who kindly presented me with a collection of aniline- and naphthylamine-sulphonic acids ; my only regret is that some of the acids were found, owing to their sparing solubility, to be not very suitable for the purpose of the investigation.Aminocccrboxylic Acids-Amphoteric Electrolytes. Within the last few years especial attention has been paid t o this class of substances; it is only necessary to refer to the work of Winkelblech (Zeit. plqsikal. Cli,ern., 1901, 36, 546), Bredig (Zeit. ElAtrochem., 1899, 6, 34), and Walker (Proc. Roy. Xoc., 1904,73, 155, and 1904,74, 271) on the amino-acids (aliphatic and aromatic), and of Johnson (Yroc. Roy. Soc., 1906, 78, 82 et sep.) and others on the methyl derivatives of the latter. (The special case of cacodylic acid, classed with the amphoteric electrolytes, will be considered separately.) I n the above-mentioned investigations both the acidic and basic con- stants have been determined by various methods ; in the present work only the former are studied, either as regards the acids themselves or their hydrochlorides.Aliphatic Aminocadoxylic Acids.-A rninoacotic acid (glycine), a sample of which was originally purchased from Kahlbauni and re- crystallised subsequently, gave no acid reaction, even with a solution of original concentration N/10 (compare Imbert and Astruc, Compt. rend9 1900, 130, 37) ; aminopropionic acid (alanine), originally purchased from Schuchardt, gave under the same conditions an acidity change too faint for accurate measurement ; hydroxyphenylaminopropionic acid (tyrosine) also gave no reaction with a concentration approximately iV/50, namely, at about the limit of its solubility in warm water, Thus, as regards their reaction with methyl-orange, the acidic and basic constants compensate one another.Hydrochlorides of the above Acids.-As preliminary experiments showed that these substances reacted with methyl-orange, mainly a3 hydrochloric acid, the aminocarboxylic acid only producing a slight positive or negative effect according t o its specific nature, results were obtained with hydrochloric acid itself for the purpose of comparison. Hydrochloyic Acid.-A standard solution, N/10, of this acid was prepared and its value ascertained by standard alkali ; this solution was diluted to concentrations N/400, N/300, and iV/200 respectively. The results obtained are given in Table I ; P is the equivalent con- centrations a t the first observation, x the units of 0.1 C.C. added (V/x being the correspondiag equivalent concentrations), and y the heights M 2156 VELEY: THE AFFINITY CONSTANTS OF of the variable tintometer column expressed in centimetres.(In successive tables the x column will be omitted and taken to be under- stood.) X. 1 2 3 4 5 6 7 8 9 10 11 12 Y= 8 x lo4. P y (found). y (calc.). - - - 3 ‘9 - 5.1 5.1 6-6 6.6 8.1 8.4 10.2 9-6 11.4 11.1 12.9 12’6 14.1 14.1 15.3 15% yi) TABLE I. Y=ci x 104. y (found). y (calc.). 8‘3 8 ‘3 10.5 10.7 12.9 13.1 15% 15 -5 18’0 17.9 20’9 20.3 - - Y= 4 x 104. ‘y (found). (y calc.). > - ;:g} - 5 -1 5-2 8 ’4 8-3 11’4 11.4 14-4 145 1 7 -7 17% 21 ’0 20.7 24’0 23 ‘8 26.7 26.9 - - - - If the results are set out graphically it appears that the first few observations (as bracketed) lie on a logarithmic curve, which does not pass through the origin of co-ordinates, but has the x axis as an asymptote (see I, Fig.1, for series 111). The remaining observa- tions lie on straight lines of general equation y = kx - b (cf. supra), the constants of which are for series I k= 1 5 , 6 = 2.4, for series I1 k = 2-4, 6 = 1.3, for series I11 k= 3.1, b = 4.1. The constants k are thus in the ratios 1.5 : 2.4 : 3.1, whereas if referred to their original concentrations and expressed in terms of the first of them, taken as correct the ratios are 1.5 : 2.25 : 3.0. On comparing the found values of y with those calculated from the general equation, substituting the second constants, it is evident that, with one or two exceptions, the differences are well within the 5 per cent. admitted limit.I n no case was the reaction pushed to its extreme limit, as it has been found previously that the red ion of methyl-orange formed by addition of excess of mineral acid could not be matched by a variable column of methyl-orange solution, which contains only the orange or orange-red ion. As regards the curved portion the equations to the logarithmic curves are, so far as they can be determined by obser- vations, few in number; for series I y=(O.23)2%, for series I1 y = (0.525)2~, for series 111 y = (0*65)2”. The curved portion corresponds to some induction period j it may be idle to speculate on the matter, but a possible cause is the presence of ammoniacal compounds in the water used which set up an opposing reaction to the methyl-orange. I n a former paper (Proc.Roy. Xoc., 1901, 69, 87) attention has been drawn to the persistent retention of some ammoniacal compound or compounds in distilled water, and more recently Burgess and Chapman (Trans., 1906, 89, 1414 e t sep.) have found the well-known induction period ofAMINOCARBOXYLIC AND AMINOSULPHONIC ACIDS. 157 hydrogen and chlorine to be caused, iizter alia, by the presence of ammonia or more probably complex ammoniacal compounds, albumin- oses or the like. However t h i s may be, since this induction period appeared in so many series of observations herein recorded, it is prob- ably due to some common cause. Glycine Hydrochloride.-A beautiful crystallinc specimen of this substance was presented to me by Mr. J. E. Marsh, F.R.S. ; some of Fro. 1. . 0 2 4 6 8 10 12 14 16 I6 Note as .to Plates.-The ordinates represent the heights in centimetres of the variable methyl-orange colnmn, the abscisase the unit masses, or volumes of the acids added ; the origin is shifted for the Graphs I1 and 111 (Fig. I), and for I1 (Fig.11), for the better comparison and to avoid overlapping. the crystals being of dimensions 14 x 8 mm. Mr. T. V. Barker, B.A., B.Sc., of the Mineralogical Department, Oxford, was kind enough to ascertain that the crystallographic axes and forms of the specimen, namely, orthorhomic, 1 : 1.1108 : 0.0309 ; forms, E P, oc p,, cc Poc , @ cfs , $P oc ; P ; hemihedral ; cleavage perfect, 1 leipa , were identical with those given by Schw:tbus in a somewhat obscure publication (Vienna, 1855), partly reproduced in Jahresber., 1854, 676.I58 VELEY: THE AFFINITY CONSTANTS OF Two solutions of N/150 and iV/200 coneentration respectively were made UP, which gave the results set out in Table 11.TABLE 11. y (found). y (calc.). y (found). y (calc.). Y= 4 x I 04. Y= 3 x 104. c L ,4 Y r 1 - 2 '8 6'3 4.1 4.0 10.3 9 9 6.3 6.8 13.5 15.5 9 '3 0 '6 16.8 17.1 12.3 12'4 19.8 20.7 15 '2 16.2 - - 18.0 18.0 - - 20 '4 20 *8 - - E} - - 6-3 2'6 E} - The results obtained are similar to those obtained from hydrochloric acid, the first few observations being upon a logarithmic curve, the remainder on a straight line ?=kx - b (compare I, Fig. 2, for series I). The values in the second and fourth columns are calculated by introducing constants 1%=3.6, b - 4 . 5 , and k=2*8, b = 7 * 2 . The constants i% are in the ratio 2.5 : 3.6, whereas if referred to original concentrations the ratio is 2.8 : 3.7, a difference within experimental error, The equations to the logarithmic curves are y=(O-3)P and y= (0.75)2%, so far as it is possible to ascertain them.Alumhe Hydrochloride.-This substance was prepared by dissolving aminopropionic acid in such a volume of concentrated hydrochloric acid that the latter contained a slight excess of that required for equimolecular combination. The solution was spontaneously evaporated, the crystalline residue washed with absolute alcohol, redissolved in water, and the solution spontaneously evaporated over sulphuric acid. The crystalline magma was dried on a porous tile, and the minute crystals, owing to their deliquescent nature, dissolved as quickly as possible in the required quantity of water.Only one set of experiments was conducted with this substance, as unfortunately within twenty-four hours of the preparation of an original solution a hypomycete had made a considerable growth therein.* * Though wholly foreign to the present inquiiy, yet it may be worthy of mention that these hydrochlorides of the amino-acids were found, from sad experience, to be most convenient media for the growth of such micro-organisms, doubtless as sup- plying carbon, amino-nitrogen, a i d chlorine ; so far as I m i aware the introductioq of tlrese snbstaiiccs in culture media has not been tried.ARIINOCARBOXYLIC AND AMINOSULPHONIC ACIDS. 159 TABLE 111. v= 4 x 104. I Y=4 x 104. h r- \ y (cnlc. ). - y (found).- 4.2 4.1 7.5 7‘4 y (found). y (cnlc.). 10-8 10.7 14’1 14.0 17’3 17’3 20’4 20.6 The figures in the second column are calculated from the equation y =kx - b, k- 3.3, b = 5.8 ; the results on the curved portion can be expressed by an equation 9 = (0*5)x2. FIG. 2. 14 12 10 8 6 4 2 0 0 2 4 6 8 I0 The behaviour of alanine hydrochloride is thus precisely similar to glycine hydrochloride. .Betchine Ifydmc?doyide.-A samplo of this substance was pur- chased from the Aktien-Gesellschaft f iir Anilin Fabrikation, Berlin, under the name of “Acidol,” being a preparation for the internal administration of hydrochloric acid in a convenient form for certain,160 VELEY : ?'HE AFFINITY CONSTANTS OF gastric complaints. The preparation was recrystallised from cold water by spontaneous evaporation and its chlorine contents determined by the Volhard method : Found, C1= 23.09.Solutions of original concentration N/400, i'V/300, and N/200, being equimoleculsr to those of hydrochloric acid (compare supra), were made up, and gave the results set out in Table IV. Calculated, C1= 23.1 per cent. TABLE IV. v= 8 x 104. 7- y (found). y (calc.) - ;::} - 2 *7 2'7 4'2 4.2 5.6 5.7 7-2 7'2 8 *7 8.7 10'2 10.2 V= 6 x lo4. - y (found). y (ca!c.). 0.6) - 1'2/ - 3.0 2 '9 4 '9 4.7 7.2 7.1 9.0 9.2 11.1 11 '3 13.2 13.4 Y= 4 x 104. y (found). y (calc.). 0 '6 - 3 '1 2.8 5.9 5.8 9.0 8.8 12.0 11.8 14'7 14.8 17.4 17.8 20'4 20.8 1.5) - The numbers in the second, fourth, and sixth columns are calculated from the straight line equation, the valws of k being taken as 1.5, 2.1, and 3, and those of b as 1*8,3*4, and 6.2 respectively (compare I, Fig.1, for series 111). T'yyosine Hydrochloride.-This substance was prepared according to the directions of Erlenmeyer and Lipp, and obtained in tufts of hard, glistening prisms, As it was further found, in accordance with the observations of these authors, that when excess of water is added to such crystals, tyrosine separates out, leaving a small portion, if any, of the salt dissolved in water, it was not possible to conduct any observa- tions with methyl-orange solution. General Conclusions.-In the following table the values of k for solutions of the same equivalent concentration are put together for the purpose of better comparison. TABLE V. V=8 x 104. V=6 x 10 V=4 x 10'. Y=3 x lo? - Hydrochloric acid .. . . . . 1 *5 2 '4 3 *1 Alanine hydrochloride . - - 3'3 - Betaine hydrochloride.. . 1 -5 2.1 3'0 - Glycine hydrochloride.. . - - 2.8 3 '6 Neglecting the result of alanine hydrochloride as possibly too high owing to the di6culty of obtaining this substance in a state of purity, the results of the remaining hydrochlorides of the amino-acids are very approximately equal to those of hydrochloric acid.ARIINOCARBOXFLI(T AND AMINOSULPHONIC ACIDS. 161 It would therefore appear that either (i) these hydrochlorides are hydrolysed completely, or nearly so, into the amino-acids and hydro- chloric acid, or (ii) the methyl-orange, as a disturbing factor, nearly completely displaces the former from the latter. The first hypothesis would seem a t first sight t o be in opposition to the experimental evidence of Bredig (Zoc.cit.) and Walker (Zoc. cit.), who found such hydrolysis to be partial and not complete. But the concentrations in the different methods of inquiry were widely different ; the most dilute solution used by the above observers equals V= 1024 (approximately lo”), whereas the most concentrated solution in my experiments equals V= 3 x lo4, or thirty times more dilute. Thus the discrepancy may only be apparent and not real; further electric conductivity measurements with such dilute solutions could only solve the question. The second hypothesis would involve the complete displacement of an aminosulphonic by an aminocarboxylic acid, although all results show that the former are more acidic and not more basic than the latter.Aspartic Acid.-A sample of this acid was purchased from Kahlbaurn and purified by recrystallisation. The following results were obtained : TABLE VI. Y= 3 x 104. Y= 4 x 104. Y=2 x 104. 1.05 2‘4 5 *1 2 -25 4‘2 7.8 3.9 6.3 11.1 5 5 5 8.1 14.7 The above results, though few in number as the reaction is soon complete, are similar in type to those of hydrochloric acid, in that the first few observations are in accordance with the logarithmic expression logy = k + xloga, the remainder with the straight line expression y = kx - b, the values for k being 1.6, 2-0, and 3.1 respectively. It appears remarkable that, firstly, aspartic acid, an amino- carboxylic acid, should behave as a strong mineral acid, and secondly, though the values of the initial period are less than those of succinic acid for the same concentration, yet in the corresponding straight line periods the value k = 1-7 of the former should be approximately three times greater than that of the latter, k = 0.6.Aspartic acid has been studied by P. Walden (Zeit. physikal. Chem., 1891, 8, 481) by the electric conductivity method, who found that as the values of V were increased in geometrical proportion, the values of162 VELEY: THE AFFINITY CONSTANTS OF k, instead of being constant, increased approximately in arithmetical proportion, and also that the value of k for solutions of the same con- centration were greater than those of succinic acid. This author obtained the following results. v. 32 6 1 128 258 512 1024 A?== rc ;, 102. Di flwcnc es.0-0067 - 0.00’79 1 2 0’0094 15 0.010D 15 0’0122 13 0’0137 15 The above results are expressible by a general formula, k = cc + blog V, or actually k = 0.0067 f (O.O014/log2) log;. Although it seems hardly possible to accept the hypothesis of Walden that this result is conditioned by the formation of an N H inner anhydride, CO,H*CH,*CH</,O , as this would, from the analogy of betaines, increase the basic function ; although also the abnormality of this acid might be due to the presence of an asymmetrical carbon atom associated with groupings which could not oompensate one another, yet it is a matter of special interest that the results obtained by the methyl-orange tintometer method proceed on parallel lines to those obtained by the electric conductivity telephone method.Cacodyli’c Acid, (CH,),AsO(OH).-This acid is here introduced as intermediate as regards its acidic function between the amino- carboxylio acids of the aliphatic and those of the aromatic series. During the past few years considerable discussion has taken place as to whether this acid can be truly classed among amphoteric electrolytes ; the literature h : ~ been collated by Johnson (Be?.., 1903, 37, 3625). It was shown by Imbert (Conzpt. TemZ., 1899, 129, 1244) that this acid is neutral towards alkalis with methyl-orange as an indicator, but it behaves as a monobasic acid with phenolphthalein a s indicator, an observation confirmed by Zawidski (Bey., 1904, 36, 3325). On the other hand, the valnes of k by the electric conductivity method obtained by different observers, although not concordant among themselves, nor even concordant with the same observer for different concentrations, yet so far agree in showing that the value is of an order corresponding to that of a true amphoteric electrolyte; this result is also confirmed by independent observations by the hydrolysis met hod.It seemed, therefore, of interest to study the behaviour of this acid by the methyl-orange method notwithstanding the above state- ment of Imbert, who, apparently, used the indicator according to the Usual practice of volumetric analysis,AMINOCARBOXYLIC AND AMINOSULPHOKIC ACIDS. 163 As the sample, purchased from Kahlbaum, was in the form of well- developed crystals, it was considered unnecessary t o purify it further. The following series of observations were made, of which two were conducted a t - one time with solutions from one stock, and the third after the interval of some weeks from another stock, TABLE VII.v = 2 x 103. Y=l x 103. Y= 5 x 103. 0.75 1.5 3 1-5 3 '3 6 2.25 4.8 9 -1 3.0 6 *3 11.9 3 * i 5 7'5 - 4 -5 - - It will be readily apparent that the results are all in accordance with the straight line formula y=kx, the values of k being 3, 1.55, and 0.75 respectively, namely, in the same ratio as their concentra- tions 4 : 2 : 1, and also by this method of experiment cacodylic acid has a higher acidity value than the aminocarboxylic acids of the aliphatic series, which gave no appreciable reaction with dilution v= 2 x lo3, but a lower value than the corresponding acids of the aromatic series (see below).Cacodylic acid would therefore be rightly classed among true amphoteric electrolytes. Amiiaocurhoxylic Acids (Aromatic). Aminobenxoic Acids.-As regards the reactions of these acids with methyl-orange, Imbert and Astruc (Zoc. cit.) observed that the 1 ; 2- and 1 : 3-acids are scarcely neutral, but the 1 : 4-acid is sensibly acid. All three isomerides were investigated; two of them the 1 : 3- and 1 : 4-acids were laboratory preparations, the third or 1 : 2-acid was purchased from Kahlbaum ; all were purified by recrystallisation. It was found that-these acids differed from the aminoacetic acids in possessing a distinct acid function, although the reaction soon came to a n end. The following results were obtained : TABLE Acid. Y= 1 x 104.3-1 1 : 2 6.2 c -- 1 : 3 J ;:; -- _- 1:4 I -- -.. VIII. Y= 2 x 104. 1.65 3'30 1'8 3 '6 5.4 2.2 4'6 6.5 - Y= 4 x 104. 0.85 1.7 2.53 0 '9 1'8 2'7 1 ' 2 2'4 3.6164 VELEY: THE AFFINITY CONSTANTS OF The above results, though few in number on account of the nature of the case, are all in accordance with the straight line formula y = kx, the values being the highest for the 1 : 4- and lowest for the 1 : 2-acid. The results are in general accordance, not only with the qualit'ative observations of Imbert and Astruc, but also with the conductivity measurements of Ostmald and Winkelblech, which have recently been discussed by Walker in connexion with his theory of amphoteric electrolytes. I t may be of interest t o compare the conductivity results of Winkelblech as set forth by Walker, and my results at the greatest dilutions only in each case : TABLE IX.Winkelblech. Vcley. Acid. KO x lo5 (Y=1024). K( V = 4 x lo4). 1:2 0.96 0.85 1:3 1 -07 0 '90 1 :4 1-17 1'2 The magnitudes are of a precisely similar order, the variations from a strict arithmetical ratio being such as might be expected by the application of methods so widely different. Oxanilk Acid.-This acid may conveniently be considered here ; the sample used was a laboratory preparation, which was purified by re- crystallisation. The folIowing results were obtained, and in the table the top figures in the third and fourth columns follow on from the bottom figure in the first and second columns respectively: I. Found. 1-5 2 -9 4.5 6'0 7.4 9'0 TABLE X. 11. Calc. 1.6 3-0 4.5 6.0 7 -5 9'0 v=4 x 104.111. Found. 10.8 12.3 18.8 15.6 17.7 19'8 IV. Calc. 10.5 12.0 13.5 15.0 16.5 18.0 The values given i n the second and fourth column are calculated from the straight line formula y = kx (k = 1.5) ; it will be observed that although at first the difference 9'- y is constant, namely, 1.5, yet towards the end there is a marked tendency for this difference to increase. Its behaviour is quite analogous to that of other acids, which likewise show an increase of electric conductivity factor (+)k with increase of dilution. Saccharin (BenzoicsuZphinimide).-Although this substance is notAMINOCARBOXYLIC: AND AMINOXULPHON~C ACIDS. 165 strictly an acid, but an imide, yet as it has been shown that it resembles acids in accelerating the decomposition of ammonium nitrite (Trans., 1903,83,747), being probably converted into the corresponding acid, SO,H*C,H;CO*NH,, on hydrolysis (a reaction which takes place on the digestive tract), it was thought worthy of interest to study its behaviour as containing a sulphonic and carboxylic grouping.The sample, originally purchased from Merck, was purified by recrystallisation, and the results obtained were as follows : TABLE XI. Y= 8 x lo-'. v=4 x 104. A A r \ /- 3 y (found). y (calc). y (found). y (calc.). 1.2 1.7 2.6 2.7 2.7 4.0 5'4 7% 10'8 2.7 4.1 5.4 7.6 10 -8 5.4 10 9 Y > I > Y > 5 '4 10'8 Y Y > > > * I n both cases the values of y a r e calculated from the logarithmic equation logy=logk+sloga, in which the values of k are 1.9 and 2.7, and of a are 1-41 and 2.0 respectively.It would appear from tho above results that saccharin on hydrolysis gives an acid with a sulphonic and not a carboxylic grouping, since towards methyl-orange it behaves when in solubion as an acid of high acidic function, resembling formic and oxalic acids. Aminoswlp?honic Acids. Hitherto the affinity constants of these substances has been deter- mined mainly by the electric conductivity method, partly by Ostwald (Zeit. physikal. Chevn., 1889, 3 , 406 et seq.), and more fully by Ebersbach (ibid., 1893, 11, 608), and it has been shown generally that the magnitudes of these constants are of a higher order than the aminocarboxylic acids ; the aminosulphonic acids are not therefore usually classed among the true amphoteric electrolytes, from which they differ in other important respects.Anilinemonosul~honic A cids.-Two out of the three isomeric modi- fications were investigated, namely, the 1 : 4 or sulphanilic acici (two specimens, one a laboratory sample, and the other purchased from Kahlbaum), both of which were purified by recrystallisation, and the 1 : 3 or metanilic acid, supplied by the Badische Anilin- und Soda- Fabrik, also recrystallised. Metanilic Acid.-The following results were obtained :166 VELEP: THE AFFINITY ONSTANTS OF v= 4 x 104. Y= 2 x l O J * 1.95 3.8 3.8 8.0 5 . i 12.0 TABLE XII. y=4 x 104. Y=2 x 10.' 7'4 16'3 9 .o 20 '8 10.8 I The results in both cases are in accordance with the straight line expression y = k x , the values of Ic being 1.85 and 4 respectively, the ratio of the numbers being that of their concentrations within the 5 per cent.error. SuZpha?dic Acid. obtained from one only of them are given : As the two samples, alluded to above, gave identical results, those TABLE XIII. Y= s 104, y (foluld). y (cnlc.). / A 7 0% 1 -05 1.6 1.6 2.3 2.2 3.3 3.2 4 '9 4'6 6.7 6.7 9'3 9.6 13.8 13.9 Y= 4 x 104. y (foLll~d). ?J (calc.). A f 7 1.2 1-65 3.2 3.3 6.6 6'6 13.2 13'2 > 7 7 ) The results given in the second and fourth columns are calculated from the logarithmic expression logy = logk + xloga, the values for k being taken as 1.05 and 1.65 and of a 1.45 and 2.0 respectively. It appears remarkable that of these isomerides one should conform to the straight line and the other to the logarithmic expression; the results for t h s same concentration are set out graphically in curve I1 (figs.1 and 2). Ostwald (Zoc. cit.) found that both the 1 : 3- and 1 : 4-acids gave regular residts, although the value of Ic for the former was about three times greater than for the latter (actually 0.0581 : 0.0181). However this may be, it mill bs shown in the sequel that such a difference in behaviour of isomeric aminosulphouic acids is not unique. Anilinedisulphonic Acids.-Only one of the possible isomeric modifi- cations was investigated, namely, the 1 : 2 : 4-acid) supplied by the Badische Anilin- und Suda-Fabrik. The sample was purified by dissolving in hot water, filtering through animal charcoal, crystallising, and drying the crystals on a porous tile; in this way a white specimen was obtained.As a sufficiently marked reaction was not produced with N/100 original solution (V being thereby 1 x l O 4 ) , a solution which gave Y= 0.75 x lo4 was used, and the following results mere obtained :AMINOCARR OXTLIC AND AMINOSULPHONIC ACIDS. 167 TABLE XIV. y (fOlUl~1). ?J (calc.). 1.5 1 -5 3'2 3.0 6.5 6.0 11.9 12'0 The values given in the second column are calculated from the formula logy=logr%+xlogtc (r%=1-5, a=2). It is evident that the introduction of a second S(13H grouping into the 1 : 4-anilinesulphonic acid decreases rather than increases the affinity constant or acidic function. As the same conclusion is arrived at, not only by electric conductivity results, but also by my results to be described in the sequel, it will not be necessary to consider further a matter so contrary to previously formed conceptions.T h e Ncq&hyZaminesulpA onic Acids.--I was advised by Prof. Bernthsen, the Director of the Laboratory of the Badische Anilin- imd Soda-Fabrik, that although the preparations mere moderately pure, yet for the purpose of physico-chernic-il investigations they should be purified by recrystallisation. Unless otherwise stated these preparations mere purified by one or more recrystallisations from water. a - 21rc~pJ~tl~yZanzir~e~~o~ao,~u Iphonic A cicls. The 1 : Z-Acid.-As it was found that this sample contained some of its sodium salt, being doubtless derived from the corresponding salt of the 1 :$-acid from which it was prepared, and did not give homogeneous crystals by the process above described, the original material was digested with dilute hydrochloric acid for two days, the coloured liquid drained off, and the residue washed with cold water until the washings gave no precipitate of silver chloride on addition of solution of the nitrate.The residue was then treated by the usual process, and, after a considerable but unavoidable waste of material, pale pink crystals were obtained, which appeared quite homogeneous when examined under the microscope. The following results were obtained :168 VELEY: THE AFFTNITY CONSTANTS OF TABLE XV. Y= 8 x lo4. v= 4 x 104. ------7 ?J (found). y (talc.). y (found). y (calc.). 0.75 1.8 j - 1 -5 1 - 1.2 1 5 4 5 .j 5.1 4 3 8‘4 8.0 9.3 9 ‘4 - - 14.4 14.0 - - 18.0 18.6 3 .O 3-0 2.7 J The above results are less satisfactory than those obtained in any other set of observations, and it is thought possible that some secondary change might intervene at the outset as the tint produced, on addition of the successive portions of the acid t o the methyl-orange solution, only assumed its final shade after standing for some minutes, and not immediately as in other cases.But notwithstanding the imperEections, from whatever cause they may arise, the results are sufficient to show that this acid behaves as a strong acid (hydrochloric acid, for example), in that the first few results conform approximately to the logarithmic expression and the remainder t o the straight line expression y = kx - b. (The results given in the second and fourth columns are calculated from constants k = 2.5 and 4.6, b = 4.5 and 13.6.) The results are in accordance with those obtained by Ebersbach by the conductivity method, who found values for k varying from 2.23 (V= 64) to 1.09 ( Y = 2045) ; whether the former or the latter of these numbers or the mean thereof be taken, yet the value is five to fifty times greater than that found for any other naphthylaminesulphonic acid, and is of the order of magnitude corresponding to that of a nitro- aromatic acid, 1’he 1 : 4-acid was not sufficiently soluble in water for the purpose of this investigation.Z’?t,e 1 : 5-acid gave a pale pink solution with slight ++ blue fluorescence when dissolved in water, but owing to its sparing solubility i t was difficult to work with, and only one set of observations was made : The point will be further discussed in the sequel.* Wherever here, or i n the sequel, the word “slight” is applied to the fluorescence, it will be taken t o mean the appFcarmce of the solutions under the conditions of ordinary daylight; when the beam of an electric arc is projected through such solutions the eflect produced is quite magnificent.AMINOCARBOXYLIC AND AMINOSULPHONIC ACIDS. 169 TABLE XVI. V= 4 x lo4. Y= 104. I y (foulld). y (cslc.). 1 . 0 1.12 3 '4 2-25 4.5 4.5 The results in the second column are calculated from the expression The 1 : 6-mid dissolved in water to give a pale pink solution ; the log y = log 1.12 + x log 2. following results were obtained : TABLE XVII. Y= 8 s lo4. ~ = 4 x 104. v=2 x 104. v- /-- 7- y (found). y (calc.). y (found). y (calc.). y (foiiiid).y (calc.). 0.63 0.7 1.2 1.42 2 *8 2.8 1 '27 1 *4 2.8 2.85 6.0 5.6 2-65 2 *a 5'7 5 *7 11.1 11.2 5 *1 5.6 11 '4 11.4 19 -2 22 -4 The values in the second, fourth, and sixth columns are calculated from the expression logy = logk + xlog2 (k = 0.7, 1.42, and 2.8) ; the last observation in the 2 x lo4 series is rather low, but it was evident on repetition that the possible reaction was nearly complete. The 1 :7-acid dissolved in water to give a pale pink solution with a faint blue fluorescence ; the following results were obtained : TABLE XVIII. Y=8 x lo4. Y=4 x 104. --\ 7- y (found). y (calc.). y (foullcl). y (calc.). 0.6 0.6 0 ' 9 1.25 1 2 1.2 2.4 2.5 2-4 2.4 4 -95 5.0 - 9.9 10.0 - The values in the second and fourth columns are calculated as above (k = 0.6 and 1-25 respectively), and it is evident from the figures given that the 1 : 7-acid is slightly weaker than the 1 : 6-acid.Ebersbach's results showed that the 1 : 7-acid was slightly stronger than the 1 : 6-acid (k= 0.0227 and 0,0195 respectively), The 1 : &acid, although obtained in a perfectly homogeneous crystal, showed no acid function whatever, even when added in very consider- able excess to the methyl-orange solution. This result will be further discussed, but it is in accordance with Ebersbach's result, who obtained VOL. XCI. N130 VELEY: THE AFFlNITY COKSTANTS OF a value of k = 0.001 in round figures, one-twentieth of the values of the 1 : 6- or 1 : 7-acids. P-i~TccpiithyZamineszr Zphonic Acids, The 2 : I-mid was obtained in pale pink prismatic needles; its solutions gave the following results : TABLE iT= s x 1 0 4 .y (found). y (calc.). 0.3 0.5 0.9 1.0 2'25 2.0 4'2 4.0 8 -1 8.0 /-'-- XIX. Y= 4 x 104. -- 1.05 1 *15 2 *1 2 -3 4.8 4'6 9'6 9.2 18% 18.4 y (found). y (calc.). The values in the second and fourth columns are calculated as above ( k = 0.5 and' 1 *15, a = 2 respectively) ; the K constants are not widely different from those of the 1 : ?'-acid, although the reaction proceeds further. The electric conductivity constant has apparently not been determined. The 2 : 5-acid was obtained as a fine crystalline powder ; its solutions gave the following results : TABLE XX. y (foniid). 7J (calc.). y (found). y (cdc.). Y=S x 104. v= 4 x 104. ---- /-- 0.9 O*i5 1 *5 1.45 1.5 1 *5 3.0 2.9 2-85 3.0 6.0 5.8 - - 11'4 11'6 The values for k are taken as 0.75 and 1.45 respectively ; it mill be observed that though they are both higher than those of the 2 : 1-acid, yet the possible reaction sooner came to an end.The 2 : 6-acid was not sufficiently soluble for the purpose of investi- gation. The 2 :'l-cccid was obtained in pale pink needles; owing to its sparing solubility only one set of observations was made and the solu- tion kept warm for the purpose. The results are given below : TABLE XXI. i/-= 4 x 104. y (fouacl). y (calc.). 1 *2 1 - 1 6 2'4 2.3 4.5 4-6AMINOCARROXPLZC AND AhftNOSULPHONIC ACIDS. 171 The value of k=1*15 is identical with that of the 2 :l-acid at the same concentration, but all reaction sooner comes to a n end. Y'he 2 : Y-chcid was not sufliciently soluble for quantitative determina- tions, but it was proved that even on addition of a considerable quantity of a supersaturated solution to the niethyl-orange no change mag produced; the acid therefore resembles the 1 : 8-acid in showing no acidic function whatever.Dimeth~Z-2-na~l~thyZc~naine-8-su~~~o~ic Acid.--I am greatly indebted to Dr. C. Smith for a sample (about 1 gram) of the above acid pre- pared and described by him (Trans., 1906, 89, 1508); as sent it was in the form of crystalline (ncicular needles) powder. When dissolved in water at 25", so as to give a X / l O solution, the liquid was of a pale yellow colour, but more dilute solutions displayed a beautiful pale blue fluorewence, so that their appearance somewhat resembled a very dilute solution of copper sulphate.A8 the iVjl0 solution gave a n immediate acid reaction with t h e met'hyl-orange solution it mas further diluted, and the following results obtained : TABLE XXII. V= 8 x lo4. V S G x 104. Y= 4 x 104. v-2 x l o ? - 7- -7 - y (found). y (enlc.). y (found). y (cnle). y(founcl). y (calc.). y(fmiid). y (calc.,, 0 -7 0 -65 0.9 0.9 1 *2 1-35 2.7 2.75 1 '2 1 *3 1 '8 1'8 2.7 2.7 6'0 5-5 - - 3 -7 3.6 5'4 5 '4 10'8 11.0 Although the above acid differs from the other naphthylarnine- mono- and di-sulphonic acids (see below), mhich contain a n S0,H grouping in the 8-position, i n that it shows a n acid reaction with methyl-orange, and although also the values of k in the expression logy = logk + dog2 ( k = 0.65, 0.9, 1.35, 2.75) are between the values found for the 2 : 1- and 2 : 5-naphthylaminesulphonic acids, yet all possible reaction sooner comes to an end.The results for the most dilute solution (8 x lo4) are almost too small for accurate measure- ment; they have been quoted, not as of much import, but as of concordant arithmetical ratio with the remaining series. However, the results are in perfect concordance with those obtained by the electric conductivity method in the case of the benzenoid aminosulphonic acids, which have shown that the substitution of hydrogen by methyl in the NH, grouping increases the value of k. The following illustrative examples are taken from the papers of Ostwald and Ebersbach (Zoc. cit.) ; NH,*C6H,*S0,H (1 : 4) 0.0586 NHz4CC,;H',-SO3H (1 : 3) 0.0186 NHMe*C,H,*SO,H ,t 0'0666 NMe,*U,H,*SO,H ,, 0.037 k. k, N 2172 VELEY : THE AFFINITY CONSTANTS OF Opportunity was taken to determine certain other physical data of the 2 : 8-dimethylaminosulphonic acid, namely the density of the N/10 solution, the solubility in water, and the degree of fluorescence.The following results were obtained: D;; (solution a t or about its maximum saturation) = 1.0073 ; DZ = 1,0058 ; the weighings were corrected to a vacuum, and the thermometer corrected according to the Reichsanstalt methods, Solubility S (grams in 100 grams water at 25") =3*47. The blue fluorescence described by Dr. C. Smith is very remark- able ; in ordinary daylight, under certain conditions, it is visible with a solution of concentration of the order of N / 2 x 105, or nearly one part in a million.When a solution of the acid is allowed to evaporate spontaneously, well-formed transparent crystals separate, but at present they have not been obtained of sufficient dimensions for accurate crystallographic measurement. a-iVcqAtJbg Zaminedisu Zphonic Acids. Acids derived from tJLe 1 : 4-monosu~phonk Acid.-TAe 1 : 4 : 2-acid, obtained as a white, crystalline powder, dissolved in water giving a pale pink solution with a purple-blue fluorescence. Two series of observations were made : TAELE XXIII, v=i x 104. Y=O.G Y 104. 0.9 2.1 1.8 4.2 2 -7 6.1 3% 8.4 4.5 10.5 y Z 1 104. v=065 x 104. 5.4 12.6 6'3 14'7 7-2 16.5 8-1 - This disulphonic acid is remarkable in that, firstly, it was necessary to use more concentrated solutions to obtain observations, and, secondly, of all the mono - and di-naphthylaminesulphonic acids examined this is distinguished from the rest in giving results which conform to the straight line expression y =kx (k=0.9 and 2.1) ; in this respect it resembles metanilic acid.The introduction of a second sulphonic grouping in this as in other cases diminishes rather t)han increases the acidic function. The 1 : 4 : 6-acid was a white, crystalline powder, giving a nearly colourless solution with a pale blue fluorescence. The following results were obtained :ANINOCBRBOYY LIC AND AMINOSULPHONXC ACIDS. 17 3 TABLE XXIV. Y= 4 x 104. Y= 2 x 104. v=i.5 x 104. A - c > - y (found). y (calc.). y (found). y (calc.). y (found). y (calc.). 0.6 0% 1.2 1 *35 1.9 3 '9 1.5 1 *2 2'7 2.7 4.2 3.8 2.4 2'4 5 -4 5.4 7.5 7.6 4.8 4 *3 10.9 10.8 - - The results are in accordance with the logarithmic expression ?'he 1 : 4 : 7-acid, obtained as a slightly deliquescent white powder, The results were as follows : (k = 0.6, 1.35, 1.9, C& = 2).became slightly discoloured when dried at SO". v=4 x 104. v=2 x 104. A c A 7 .-- \ y (found). y (calc.). y (found). y (calo.). 0.6 0-6 1 -2 1 *3 1.2 1 -2 2 -7 2.6 2 *1 2 4 5.4 5.2 The values found are practically identical with those of the 1 : 4 : 6- acid, although the reaction sooner comes to an end. The 1 : 4 : 8-acid was obtained in yellowish-pink tabular crystals giving a yellow solution with faint fluorescence. The acid reaction, even with original solution N/10, was almost inappreciable, and thus in terms of the other acids its effect may be considered nil.The 1 : 5 : 7(?)-acid (Badische Anilin- und Soda- Fabrik, D.R.-P. 69555) is prepared by sulphonating aceto-a-naphthalide or its 1 : 5 - sulphonic acid and hydrolysing the acetyl compound with dilute sulphuric acid. The 7-position of the sulphonic-grouping appears t o be doubtful. The sample sent was purified by allowing a cold- water solution to evaporate spontaneously ; pale yellow needles separated, which rapidly effloresced. A solution of original concentration Nj25 gave no acid reaction with methyl-orange even on addition of excess. Therefore this acid behaves as acids which contain a sulphonic grouping in the 8-position. But so long as the constitution of the acid remains a mcitter of doubt, it does not appear desirable t o draw any conclusion.174 VELEY : THE AFYINITY CONSTANTS OF The P-hTap11t11yZami32edisu123joonic Acids.T h e 2 : 3 : 6-acid was obtained partly as colourless, glassy plates, which rapidly effloresced, partly as a pale pink amorphous powder ; by a process of rapid evaporation it was obtained in the latter form only. The acid probably, therefore, occurs in n labile crystalline and a stable, amorphous modification, but the conditions under which one or the other is formed have not as yet been accurately ascertained. The solutions showed a purple-blue A uorescence. Two series of observa- tions were made with the following results : TABLE XXVI. y (foluld). y (calc.). 2/ (foluld). y (calc. ). 0.6 0.65 1.2 1 ‘25 1-5 1.3 2‘6 2 ’5 2’4 2% 5.1 5.0 - 9-9 10.0 v=4 x 104. v=2 x 104. A r- , .I The values for k are not widely different from those obtained for the 1 : 4 : 7-acid. The 2 : 6 : &-mid gave no reaction with an original solution N/50, even when added in excess ; hence relatively its effect may be regarded as nil. General Conclusions regarding the Nap~~tl~ylanaiiiesul~honic Acids. On reviewing the results obtained with these acids, two general facts come into prominence : namely (1) the positions 2 and to a less extent 7 afford cases of steric “furtherance,”” and (2) the position 8 affords a case of steric ‘‘ hindrance.” As regards the steric hindrance effected by the 8-position, my observations are quite in accordance with those of Hewitt and Mitchell, as also of C. Smith (Zoc. cit.), in their studies on the reaction of various naphthalene derivatives with diazonium salts.On the other hand, as regards the steric furtherance of the 2-position my observations are in accordance with the results of Ebersbach. A t present there do not appear to be many data as to the effect produced by the 7-position, relative to the positions 1 and 2. * As the phrase ‘< steric hiiidiance” has now come iiito chemical literature, I have ventured t o use the oltl Saxon word “ furtlierance ” as its opposite. It appears that the iionn has been user1 in this sense sirice 1440 ( I‘o7.k M y s t c ~ y Plny), ant1 thc verb so fa1 back as 888 (BZfrerE C‘hrm.). 1’eilial)s I iiiny raise n p l c : ~ for the use of Saxon ~vorc~s, rather than those derived from Grcc.1~ or Latin, sepafirtely, or even worse, con j oi 11 t 1 y.AMINOCABUOXYLIC: AND AMINOSULPIIOKIC ACIDS. 175 (i) The method of determining the affinity constants of acids by means of a dilute methyl-oiaange solution and a tintometer, formerly applied to the carboxylic acids and certain chloro- and hydroxy- derivatives, has been extended to the amino-carboxylic and sulphonic acids. It is shown that the same general mathematical expressions hold good, namely, those of straight lines, or logarithmic curves. (ii) Acids which show irregularities in the Ostwald electric con- ductivity expression +(k) = az/(l - a)V show similar irregularities in the methyl-orange method. (iii) The aliphatic aminocarboxylic acids act as neutral substances, but their hydrochlorides as hydrochloric acid only, hydrolysis being nearly complete a t the degree of dilution used. (iv) The special cases of aspartic and cacodylic acids have been investigated, (v) The aminobenzoic acids show a distinct acid function, and the factors obtained are in a similar arit.hmetica1 ratio to those deduced by Winkelblech a t a different degree of dilution of t h e electric condnc- tivity method. (vi) The two aminobenzenemonosulphonic acids studied, namely, sulphanilic (1 : 4) and metniiilic (1 : 3) acids, are remarkable in that the latter conforms to the straight line, but the former to the logarithmic expression. The introduction of a second sulphonic- grouping reduces rather than increases the acid function, (vii) The study of the naphthylaminemono- and di-sulphonic acids affords examples of steric furtherance as regards the positions 2 and 7 and of steric hindrance as regards the position 8. Lastly, I desire to express my thanks to Prof. Win. Esson for assistance in the mathematical portion, and again to my colleagues a t home and abroad for having kindly supplied me with such a wealth of material, without which this investigation could not have been completed.