194 



NA TURE 



{June 27, 1889 



becomes gaseous independently of pressure and volume. 

 It is, in fact, that temperature which the researches of 

 Andrews have made us familiar with as the " critical- 

 point." In this paper Mendeleeff presents us for the first 

 time with a number of determinations of the critical- 

 temperatures of various substances, founded partly on 

 his own determinations, and partly on those of Cagniard 

 de la Tour, Wolff, and Drion. 



Other papers on physical chemistry relate to Contact 

 Action, to Fractional Distillation, and to the Heat of 

 Combustion of Organic Substances. In 1883 Mende- 

 leeff was made an honorary member of our Chemical 

 Society, and in the following year he contributed a 

 remarkable paper to the Journal of the Society (Trans- 

 actions of the Chemical Society, xlv. 126), in which he 

 developed an extremely simple general expression for the 

 expansion of liquids under constant pressure between 0° 

 and their boiling-points. This expression may be written 

 ijN t = I — kt, in which V/ is the volume at f (that at 

 o' being unity), and ^ is a quantity which varies with 

 different substances, but which may for any one sub- 

 stance be considered invariable between 0° C. and the 

 neighbourhood of the boiling-point. This formula is 

 analogous to that which expresses Gay Lussac's law of 

 the uniformity of expansion of gases. But just as Gay 

 Lussac's formula, V = \ -\- kt, applies only to a so-called 

 ideal gas, MendeleefPs expression is in like manner to be re- 

 garded only as a first approximation — that is, as applicable 

 only to ideal liquids. The deviations are not large in either 

 case ; they are, as might be expected, especially remarkable 

 near temperatures at which the states of the bodies change. 

 In the case of actual liquids the deviations from the ideal 

 form of expansion increase not only as the liquid ap- 

 proaches the point at which its state of aggregation is 

 changed, but also with diminishing density, increasing 

 cohesion, and diminishing molecular weight. This last 

 cause is especially noteworthy since Mendeleeff showed, 

 more than a dozen years ago {vide supra), that the devia- 

 tions from Gay Lussac's law were related to the molecular 

 weights of the gases. The well-known irregularities in the 

 expansion of water are, according to Mendeleeff, con- 

 nected with its small molecular weight, its high capillary 

 constant (which expresses its cohesion), and the compara- 

 tively small temperature-interval within which its state of 

 aggregation is unchanged. Subsequent observers, by apply- 

 ing Van der Waal's theory of the general relation between 

 the pressure, volume, and temperature of bodies to Mende- 

 leeff's expression for the thermal expansion of an ideal 

 liquid, have shown that the reciprocal of the constant k is 

 the number obtained by subtracting 273 from the product 

 of the critical temperature into a quantity which should be 

 the same for all substances. The value of this quantity is 

 approximately 2, and since the range of its variation is 

 apparently very small, the development of Mendeleeff's 

 formula affords a simple and ready method of calculating 

 the critical temperature of bodies from observations of 

 their expansions as liquids. 



Mendeleeff's skill in physical measurement is well 

 illustrated by his determinations of the Specific Gravities 

 of Aqueous Solutions of Alcohol. Such determinations 

 have been frequently made the subject of the most 

 rigorous experiment in this and other countries, inasmuch 

 as they constitute the basis of the methods of assessing 



the duty on spirits, which is so important a factor in the 

 national income of many States. Mendeleeff's work has 

 served to confirm and extend that of Drinkwater, Fownes, 

 and Squibb, and has been utilized by certain Continental 

 Governments [eg. that of Holland) for the purposes of 

 revenue. But it was not the utilitarian aspect of this sub- 

 ject which alone attracted Mendeleeff. In a paper com- 

 municated a couple of years ago to our Chemical Society 

 (Trans. Chem. Soc, li. 778), these determinations are 

 applied towards the elucidation of a theory of solution in 

 which it is sought to reconcile Dalton's doctrine of the 

 atomic constitution of matter with modern views respect- 

 ing dissociation and the dynamical equilibrium of mole- 

 cules. According to Mendeleeff, solutions are to be 

 regarded as strictly definite atomic chemical combina- 

 tions at temperatures higher than their. dissociation tem- 

 perature, and just as definite chemical substances may be 

 either formed or decomposed at temperatures which are 

 higher than those at which dissociation commences, sa 

 we may have the same phenomenon in solutions ; at 

 ordinary temperatures they can be either formed or 

 decomposed. In addition, the equilibrium between the 

 quantity of the definite compound and of its products of 

 dissociation is defined by the laws of chemical equili- 

 brium, which require a relation between equal volumes, 

 and their dependence on the mass of the active com- 

 ponent parts {loc. cit. p. 779). It follows from this hypo- 

 thesis that the specific gravities of solutions depend on 

 the extent to which active substances are produced, or 

 that the expression for the specific gravity, s, as a function 

 of the percentage composition,/, may be represented by 

 the general equation — 



.y = C -I- A/ + B/-. 



Between two definite compounds which exist in solu- 



ds 

 tion, the differential coefficient ^^ is a linear function of/ — 



ds 

 dp 



dp 

 = A + 2B/. 



By the application of this method to the case of aqueous 



solutions of ethyl alcohol, Mendeleeff infers the existence 



of three definite hydrates, viz. EtHO. i aHgO, EtHO.sHgO^ 



and 3EtH0 . HgO, the first two of which he has isolated 



by subjecting the mixture to low temperatures. The 



hypothesis respecting the linear character of the dif- 



ds 

 ferential coefficient ~ji has been proved to be correct for 

 dp 



solutions of many salts, of acids, and of ammonia. 



We have the consummation of this work on solu- 

 tion in the monograph published by Mendeleeff last 

 year. This volume, the fruit of many years of labour, is 

 unquestionably the most important contribution to the 

 theory of solution yet given to science. 



Much of Mendeleeff's scientific activity since 187 1 has 

 been absorbed in an extended work on the elasticity of 

 the gases, which he has executed in conjunction with his 

 pupils, Kirpitshoff, Hemilian, Bogusky, and Kajander. 

 Part only of the results have as yet appeared. The first 

 volume, published in Russian in 1875, contains details of 

 the modes of measurement, which involved many forms of 

 apparatus new to physical science. A summary of the 

 principal results obtained was published in the form of 

 a pamphlet in 1881. Regnault found that the product 



