592 



NA TURE 



[April 21, 1892 



Guye, has made a remarkable attempt to find laws for the 

 numerical values of the molecular rotation, by giving to the 

 asymmetry of the carbon atom a numerical measure de- 

 pendent upon the masses joined thereto, and, in cases of 

 analogous compounds, comparing this with the values for the 

 molecular compounds. While this attempt has been well 

 supported by a number of older measurements, especially those 

 of Pictet on the esters of the tartaric acids, yet his own deter- 

 minations on the active amyl derivatives have not indeed fur- 

 nished much very favourable evidence. He has not overcome 

 the difficulties of obtaining pure material, and certain facts were 

 observed contradicting the assumption that the sense of the 

 asymmetry is due to the masses added. It is not improbable 

 that this difficulty will be overcome by placing the optical 

 vioment, if I may be allowed the expression, not proportional 

 simply to the mass of the atom ; there is rather to be suspected 

 a connection with the atomic refraction. 



We turn now to a field whose development belongs entirely 

 to the last few years, to that of scbdions. If we call to mind 

 the old saying. Corpora non agnnt nisi Jluida, vel sohiia, we 

 perceive at once the very great importance of the field ; all 

 rational knowledge of chemical processes must be preceded by 

 a corresponding knowledge of the condition of dissolved 

 substances. ^ 



I do not need to remind you that van 't Hoff's discovery of 

 the identity of the laws of gases with those of dissolved sub- 

 stances, is to be characterized as the greatest step forward which 

 has been made in this direction. If we reflect that the de- 

 velopment of the molecular idea, which rules the chemistry of 

 to-day, is most decidedly based upon the laws of gases in their 

 simple form, we recognize at once that all the important rela- 

 tions which have been here found can be directly transferred to 

 the domain of solutions. The latter has, however, at the same 

 time, far more varied possibilities in the form of its phenomena. 

 While in the case of gases only two of the variables, pressure, 

 volume, and temperature, are independent, there is present for 

 solutions the manifold infinity of the non-miscible and partially 

 miscible solvents. To this is due the appearance of a great 

 number of new formal and numerical relations for solutions, 

 even under assumption of the simplest form of the governing 

 laws, whereby a rich field of inexhaustible fruitfulness is made 

 accessible to investigation. In fact, after this advance by van 't 

 Hoff, the theoretical investigations of Planck, Riecke, Lorenz, 

 van der Waals, and Boltzmann, as well as the progressive com- 

 bination of theory and experiment by Nernst, have shown how 

 varied and valuable are the results to be gained, results whose 

 details I am here compelled to omit. 



I wish, at this opportunity, to call attention to one particular 

 point. I have already mentioned that the way to a rational 

 theory of the liquid condition leads from the gases, through 

 their variation from the simple gas laws, and through the 

 critical point, whose constants express in especially simple form 

 the individual properties of the kind of matter in question. 

 Now it is to be expected from the theory of solutions, and it 

 has been demonstrated in detail by O. Masson and W. Ramsay, 

 that upon transition from a dilute to a concentrated solution we 

 observe entirely the same phenomena that appear when the 

 volume of a gas is diminished ; there is here also a critical 

 state with its corresponding constants. We thus have here a 

 second way to a theory of the condition of pure liquids, which, 

 by reason of the greater variety of the phenomena, is a far more 

 difficult one than is that first named, but which, however, may 

 in many cases be of assistance wheie the other fails. 



While the already discussed parts of the newly-opened terri- 

 tory are mainly those problems with which physicists have 

 concerned themselves, still its study has not been less fruitful 

 for chemistry in particular, especially for organic chemistry. 

 To the above-mentioned variety of the relations here present 

 corresponds an equal variety of the methods of determination of 

 that most important constant, the so-called molecular weight of 

 dissolved substances. Since the tireless Raoult had shown years 

 ago, in a purely empirical manner, the application of the proper- 

 ties of solutions to this purpose, it was reserved for the theory of 

 van 't Hoff to discover the rational basis of these relations, and 

 thus for the first time to give to wider circles of investigators a 

 feeling of security in the making of such molecular weight 

 determinations. Especial service has been rendered by E. 

 Beckmann in the technical development of these methods ; and 

 the Beckmann freezing apparatus and boiling apparatus form at 

 present just as necessary and much used a part of the equipment 



NO. II 73. VOL. 45] 



of a laboratory as formerly the Hofmann apparatus for deter- 

 mining vapour densities. 



It has naturally come to pass that, together with the suddenly 

 increased range of molecular weight determinations, our views 

 of the nature of this quantity and of the therewith connected 

 question of valence have undergone a corresponding change. 

 The conception had become gradually rather dogmatically rigid : 

 it was understood to require for each substance only a single 

 absolute molecular weight, the variations observed, for 

 example, in the case of acetic acid, being characterized as 

 anomalies. Molecular weight determinations in solutions have 

 shown that such variations are so extended, and, at the same 

 time, occur so regularly, that they may no longer be pushed aside 

 as anomalies. It is therefore at present generally recognized 

 that a substance may quite well have different molecular weights, 

 standing in the -ratio of simple multiples, the most important 

 weight for the chemist being of course the smallest of them. 



The consequences connected with van 't Hoff's discovery 

 being so important and wide-reaching, they have had in general 

 a friendly reception, although a few scientific men — not of the 

 highest rank — fearing the little plants cultivated by them to be en- 

 dangered by the flood of light falling upon them, have attenipted 

 a slight resistance. On the contrary, all the uneasiness which is 

 unavoidably connected with important revolutions has been 

 directed against a second idea, which, appearing somewhat later 

 than that of van 't Hoff, removed a fundamental difficulty in 

 the theory of solutions, which had until that time made its 

 acceptance impossible for me. This idea has at the same time 

 shown itself as an aid to investigation to be of unexampled 

 sweep and value. This is the theory of electrolytic dissociation, 

 of Arrhenius. 



It is certainly to be presumed that the fundamental idea of 

 this theory is generally known. In the aqueous solutions of the 

 electrolytes, the salts, acids, and bases, a greater or less propor- 

 tion of the dissolved molecules are regarded as split up into 

 electrically charged constituents or ions, which exist in the solu- 

 tion indepently of one another in the same manner as the partial 

 molecules of a dissociated gas. If the van 't Hoff theory be 

 admitted, it must be admitted that in a solution of sodium 

 chloride, for example, almost double as many individual par- 

 ticles or molecules are present as in a corresponding solution of 

 sugar or urea of the same formula weight. The experimental 

 connection of these variations with the fact and numerical 

 amount of the electrical conductivity, first discovered by 

 Arrhenius, and which cannot be denied, furnishes the basis 

 for the second part of the theory of Arrhenius, the assumption 

 of electric charges upon the separated molecular constituents or 

 ions. If now these fundamental ideas are accredited, the re- 

 mainder follows with directly evident necessity. 



The significance of these views becomes apparent upon con- 

 sidering the quite astonishing range of phenomena in the most 

 widely separated parts of physics and chemistry which have 

 received explanation from the theory of Arrhenius in connection 

 with that of van 't Hoff. It is simply impossible in the limits 

 of this address to even enumerate these single applications ; I 

 shall, as I think, do better by treating the question from a more 

 general standpoint, and, without speaking in particular of each 

 advance made, sketch in rough outline that field in which 

 both theories have brought or will bring decisive explanation. 



Let it be first called to mind that the laws of dissociation 

 were already earlier derived thermodynamically for gases. If, 

 then, in the field covered by Arrhenius, the question be one of 

 dissociation, and the laws of gases do, according to van 't Ploff, 

 hold for dissolved substances, it follows that the entire theory of 

 the chemical affinity of electrolytes must be yielded by ^ the applica- 

 tion of those laws of dissociation. This means nothing less than 

 that the problem of chemical affinity is in reality solved. 



The conception of chemical affinity is to be understood to 

 reach so far as to include all phenomena caused by the so-called 

 inner energy of bodies. It includes, then, not only the pro- 

 cesses especially termed chemical, but also those of vaporization, 

 and solution without exception as well. If it be wished in the 

 latter case to preserve the ever emphasized but still unclear dis- 

 tinction between "chemical" and "physical" processes, to the 

 former may be reckoned those processes in which electrolytic 

 dissociation comes into question, and to the latter those in which 

 this is not the case. Thus, the dissolving of oxygen in water is 

 in this sense a "physical," that of hydrochloric acid in water a 

 "chemical," process. But this distinction is secondary: it is 

 expressed only in the greater complication of the corresponding 



i 



