April 21, 1892^ 



NATURE 



591 



such have become known, and still fewer have been considered in 

 this manner. In addition to the above-mentioned result of Ramsay 

 and Young, there claims attention one discovered by the Hun- 

 garian physicist Eotvos, according to which \}ciQ molecular surface 

 energy, as expressed by the product of the capillary constant 

 and the |rd power of the molecular volume, is shown to be 

 a linear function of the temperature. Since the surface energy 

 stands in closest connection with the energy of interaction, by 

 %irtue of which the substance of liquids, in contrast to that of 

 i;ases, assumes its own proper volume, and to which is accord- 

 ingly due the characteristic existence of the liquid condition, it 

 becomes at once evident that here certainly a means of access to 

 the theory of the latter is afforded. This means may be ex- 

 pected to lead more rapidly to the goal than the methods 

 hitherto almost exclusively tried, based upon a relation between 

 volume, temperature, and pressure. 



The stochiometry of the liquid organic compounds, founded 

 by Hermann Kopp, has enjoyed likewise a steady development. 

 While the question of the boiling-point seems to be essentially 

 postponed until the general theory of liquids becomes known, 

 yet that of the molecular volumes has reached a stage which 

 already assures the prospect of a successful period of develop- 

 ment. The additive scheme, proposed by Kopp as a first 

 approximation, according to which the molecular volume is the 

 sum of the atomic volumes — a scheme whose insufficiency Kopp 

 himself had shown in the case of oxygen — determines only the 

 roughest outlines of the phenomenon in question. Other factors 

 make themselves everywhere felt ; as was shown by Kopp 

 for oxygen — that the portion of the molecular volume due to 

 it can assume different values according to the function of this 

 element in the compound, i.e. according to the constitution 

 of the molecule — so the same holds for the other elements. An 

 essential difference between univalent and bivalent elements is, in 

 this respect, not present : ethylene and ethylidene chlorides 

 have different molecular volumes, although they both contain 

 saturated carbon atoms, and, in addition, only univalent 

 elements. 



We must, accordingly, more than ever before, recognize the 

 molecular volume as a constitutive property. This recognition 

 removes at once the firm barrier to which the additive scheme, 

 greatly against the will of its originator, had hardened. In vain 

 had been for so long striven to force the facts into this form ; 

 ever and again their living body would not fit upon the wooden 

 cross. Now we see that this undertaking was itecessanly in 

 vain : we begin to comprehend that methyl alcohol must be 

 more different from ethyl alcohol than ethyl alcohol from propyl 

 alcohol ; and that these two, again, must stand in a different 

 relation than do propyl alcohol and butyl alcohol, although each 

 time the "same " difference of CH.j is at hand — that there are, 

 in short, no two pairs of compounds whose differences are entirely 

 the same. 



Now, it is quite dependent upon the nature of the property 

 considered, in what relation the additive foundation stands with 

 the modifying effect of constitution. With the molecular volume 

 the first is comparatively superior ; with the boiling-paints, how- 

 ever, the latter make themselves to the most superficial observation 

 so energetically felt that, since the attempts of Schroder, Lowig, 

 and others, which over fifty years ago failed to carry through the 

 additive scheme for the boiling-points of organic compounds, 

 this line of effort has been definitely given up. The other pro- 

 perties which have been studied fall between these two limits. 



This holds especially for the molecular refraction. Just as 

 Buff had earlier shown that "double bound" carbon possesses 

 a greater molecular volume than does saturated carbon, it 

 has been demonstrated by Briihl that a similar relation holds 

 for the molecular refraction. This influence of constitution is, 

 however, not the only one ; a similar inference has been shown 

 for oxygen and likewise for chlorine, and it has been repeatedly 

 shown that, even if approximately additive laws be followed 

 among the higher members of homologous series, yet these do 

 not apply for the first members. This is necessarily so, as has 

 already been shown in discussing molecular volumes. 



The magnetic rotation is a property of much more strongly 

 marked constitutive character than are molecular volume and 

 molecular refraction. We possess here most excellent inves- 

 tigations by Perkin, which have often been found of service in 

 determining questions of constitution. 



In relation to the connection between the different properties 

 of substances a fruitful line of thought has been carried out by 

 Philippe-Guye. •-- ' ^' ....... , ^ . 



NO. I 



As is known. Maxwell had derived a definite 

 73, VOL. 45] 



relation between the coefficient of refiaciion and the dielectric 

 constant, from his wide-reaching speculative investigations, 

 which latter had yielded a complete analogy of the mathematical 

 expressions for electrodynamical and optical action at a distance, 

 together with an approximate equality of the fundamental con- 

 stants, and which have been finally made fruitful by the brilliant 

 experimental investigations of Hertz. This dielectric constant 

 is in turn, according to an expression due to Clausius, a simple 

 function of that fraction of the total volume of a dielectric which 

 is occupied by the actual material substance (considered as con- 

 ducting). But this so called true molecular volume is, finally, 

 nothing but the co-volume in the equation of Van der Waals. 

 There is accordingly to be expected a close connection between 

 the critical constants and the molecular refraction, and Guyehas 

 shown that the expected connection actually exists. 



Although spectrum analysis, with its manifold applications, has 

 for years had almost no rational development, it has recently taken 

 a quite promising start in the stochiometric direction. The theo- 

 retical and experimental researches of Balmer, Deslandres, Julius, 

 Rydberg, Kayser and Range, and others, indicate already that 

 the time is not far distant when there shall be simple and intelligible 

 regularities in this field, which until now has been so overgrown 

 with unfruitful hypotheses. Only upon one point I wish at this 

 opportunity, as a chemist, to direct the attention of the physicists. 

 It is held as an undoubted dogma that at the highest tempera- 

 tures, as, for example, in the electric light arc, all compounds 

 must be dissociated into their elements. This view is certainly 

 not justified. What we do know about the stability of com- 

 pounds is, on the contrary, that all compounds which are formed 

 with absorption of heat become tnore stable with rising tempera- 

 ture, and the reverse. Because the majority of the compounds 

 known to us are formed from the elements with evolution of 

 heat, and correspondingly become more unstable with rising 

 temperature, the conclusion has been drawn that this is in 

 general the case. But if we reflect that cyanogen and acetylene, 

 two compounds formed with great absorption of energy, are 

 readily formed in quantity, at the highest temperatures, in the 

 blast furnace and in the Davy arc light, we become conscious 

 that the spectra occurring at high temperatures may, under 

 proper conditions, belong to compounds which, formed^ with 

 great absorption of energy, may have a fleeting existence con- 

 fined to those temperatures only. From this point of view, 

 many difficult facts of spectroscopy and spectrometry would have 

 some prospect of a proper interpretation. 



At the extreme boundary of the optical properties, towards 

 the side of the constitutive character, stand finally colour and 

 rotation of the plane of polarized light. Although the first 

 property is decisive for one of the most important branches of 

 technical chemistry, the dye-stuff industry, still but little is 

 known as yet about the connection of colour with composition 

 and constitution. The investigations of Kriiss, Liebermann, 

 and more recently Vogel, all indicate that the property is in 

 great measure constitutive, becoming additive only within the 

 narrowest limits of closely-related compounds. This renders 

 correspondingly difficult a recognition of the connections at 

 hand. Some time later, on the contrary, directly on account of 

 this marked constitutive character, the colour will be an im- 

 portant aid in the determination of constitution ; at the same 

 time, when we shall have learned to recognize this connection 

 with some certainty, the discovery of new dyes with definite 

 properties will be no longer a matter of a lucky hand and of an 

 unconscious feeling for this connection, but will rest upon just 

 as broad a basis as, for example, the technic of the metallurgical 

 processes. 



The constitutivecharacter of the rotation of the plane of polari- 

 zation has been always known and recognized. Since van 't Hoff 

 and Le Bel, twelve years ago, pointed out the connection 

 between this property and the presence of an "asymmetrical" 

 carbon atom, i.e. one joined with four different elements or 

 groups, this idea has, at first slowly, then more and more 

 rapidly, had an important development. For the "optical 

 symmetry " shown by Pasteur in the tartaric acids, the ex- 

 amples have become more and more numerous ; the researches 

 of Wallach on the ethereal oils have especially furnished 

 valuable material. The presence of optical activity is now held 

 as an entirely undoubted proof for the presence of asymmetrical 

 carbon, and Le Bel has just announ.ed that he has succeeded 

 in the preparation of optically active nitrogen compounds con- 

 taining an asymmetrical nitrogen atom. 



The investigator whom we have already mentioned, Philippe- 



