October 23, 1890^ 



NATURE 



627 



each disputant departed holding precisely the same opinions 

 which he had brought with him. Since then, however, each 

 party has obtained, or thinks that he has obtained, positive 

 evidence in favour of his own views ; evidence which, if upheld, 

 must be accepted as conclusive, or which must be overthrown 

 before his opponents can claim the victory. The supporters of 

 the hydrate theory claim that the curved figures representing 

 the properties of solutions of various strengths show sudden 

 changes of curvature at certain points, which are the same what- 

 ever be the property examined, which correspond to the 

 composition of definite hydrates, and which, therefore, can 

 only be explained by the presence of these hydrates in the 

 solutions ; while the supporters of the physical theory, now 

 identified with the supporters of the osmotic pressure theory, 

 claim to have shown that, with weak solutions at any rate, the 

 dissolved substance obeys all the laws which are applicable to 

 gases, and that, therefore, its molecules must be uninfluenced 

 by, and uncombined with, those of the solvent. 



In another respect also I may notice that our position to-day 

 differs considerably from what it was four years ago ; for instead 

 of having to argue the matter out amongst ourselves, as we did 

 then, we are now favoured with the presence of some of those 

 whose work in this very subject has made their names familiar 

 household words with every physicist and chemist throughout 

 the scientific world. 



I propose in the first place to give a brief summary of the 

 evidence which has lately been adduced in favour of the hydrate 

 theory, and in the second place to inquire whether the 

 conclusions drawn from this evidence are invalidated by the 

 important facts elucidated by Raoult, van't Hoff, Arrhenius, and 

 Ostwald. 



In one respect the supporters of the hydrate theory start now 

 under a distinct advantage — namely, that their most active 

 opponents do not altogether deny the existence of hydrates in 

 solution, although it is only in the case of strong solutions that 

 they will admit their presence ; in such solutions, indeed, it is 

 difficult to see how their presence could possibly be denied. 

 The only means which we have of proving that a liquid is a 

 definite compound is by ascertaining whether its composition 

 remains unaltered by its passage through the gaseous or solid 

 condition — by fractionating it by means of distillation or crystal- 

 lization. With liquids of comparatively small stability, such as 

 hydrates, crystallization is the only method available ; the results 

 of crystallization have led us to conclude that the liquid represented 

 by H.jSOj is a definite compound, and precisely similar results 

 must force us to accept the definiteness of the liquids H2SO4SO3, 

 H.jSOjHjO, and H.^S()44H20 : in the case of each of them the 

 liquid freezes as a whole, and without change of composition ; the 

 temperature remains constant throughout the solidification, and 

 any excess of either water or sulphuric anhydride which may have 

 been added may be separated from the pure compound, which 

 alone crystallizes from the mixture. Thus, in the instance taken, 

 between the anhydride on the one hand and water on the other, 

 we have four definite compounds, all existing in the liquid 

 condition. 



It does not follow, however, that every hydrate which exists 

 in solution can necessarily be obtained in the solid condition ; 

 probably no solution, even when it possesses the exact composition 

 of some existing hydrate, consists of that hydrate only, but of a 

 mixture of it with the products of its dissociation (though the 

 amount of these may be very small) ; and whether the hydrate or 

 one of these dissociation products crystallizes out on cooling 

 must depend on the relative ease with which the bodies in 

 (juestion assume the solid condition ; when the hydrate does not 

 crystallize easily we can hope to obtain evidence of its presence 

 by indirect means only. 



Mendelceff's conclusions respecting the densities of solutions 

 of sulphuric acid and alcohol, ^ mistaken though I believe they 

 were, led to the discovery of the means whereby such evidence 

 might be obtained. 



I lestatedthat on plotting out the rate of change of the densities 

 with the percentage composition of the solution (the first differ- 

 ential coefficient) he got a series of straight lines, forming figures 

 with well-marked breaks at points corresponding to definite 

 molecular proportions ; but on plotting out the experimental 

 points which he said formed these figures, it is impossible to see 

 any justification for this statement ; in the case of sulphuric acid 

 the points and Mendeleeffs drawing of them have been given side 

 by side in the Trans. Chem. Soc, 1890, p. 81, and in the case of 



' Zeit.f. Phys. Chem., i. p. 275; Chem. Soc. Trans., 1887, p. 778. 



NO. 1095, VOL. 42] 



alcohol they will be found in the Zeit. f. Phys. Chem. VI. i. 10. 

 Crompton then showed ^ from an examination of Kohlrausch's 

 values for the electric conductivity of sulphuric acid solutions 

 that a second differentiation might in some cases be necessary 

 before rectilineal figures with breaks in them were obtained. 

 In my own work on various properties of solutions of the acid 

 I have made free use of this process of differentiation, but I 

 have combined it with, and now nearly entirely rely on, an 

 examination of the original curves with the help of a bent ruler. 



In the Phil, Mag., 1890, vol. i. p. 430, will be found rough 

 sketches of the figures representing the densities, contraction on 

 formation, electric conductivity, expansion by heat, heat of 

 dissolution, and heat capacity of the solutions, and in the Trans. 

 Chem. Soc, 1890, p. 338, that representing the freezing-points. 

 In some cases, such as the freezing-points of solutions near 

 58 and 100 per cent, strength, a mere inpection of the figure 

 enables us to locate the position of abrupt changes of curvature ; 

 in general, however, the recognition of such changes is more 

 difficult. On attempting to draw any of these figures with the 

 help of a bent ruler it was found that the whole figure could 

 only be drawn in several sections, and it was also found that 

 each section thus drawn consisted of a single curve of a parabolic 

 nature, although a ruler, when bent by the pressure exerted by 

 the two hands, by no means necessarily forms a parabola ; and 

 moreover — and this is the most important part of the evidence — 

 it was found that these figures, though differing so greatly in 

 their general appearance, all split up into the same number of 

 sections, indicating the existence of changes of curvature at the 

 same points ; and, further still, these points corresponded to 

 solutions of definite molecular composition in all cases where the 

 ratio of the acid to tha water was sufficiently large to render any 

 such comparison possible ; the average difference between the 

 composition indicated by the changes of curvature and that of 

 definite hydrates was only o'OSyHoO. With weak solutions it 

 is, of course, impossihle to assert that the changes occur at definite 

 molecular proportions, owing to. the smallness of the change in 

 percentage composition which would be caused by an additional 

 molecule of water to each H.,S04 ; but the changes with these 

 weak solutions are of precisely the same character as those with 

 strong solutions, and, unless some strong evidence to the 

 contrary be forthcoming, we must attribute them to the same 

 cause. 



To discuss fully the value of the evidence thus obtained would 

 take me more hours than I can now afford minutes ; but I think 

 that I may say that these results stand at present unquestioned 

 and uncontroverted, and that unless they can be controverted 

 we must accept the presence of hydrates in solution as having 

 been proved. I may also add that my results with sulphuric 

 acid solutions have been strengthened by obtaining analogous 

 results with solutions of several other substances : that one of 

 the hydrates indicated by them has been proved to exist by 

 isolating it in the crystalline condition : and lastly, that a law 

 governing the freezing-points of solutions has been formulated, 

 according to which we can calculate within experimental error 

 the freezing-point of any solution, whatever its strength may be, 

 provided we acknowledge the existence of every hydrate which 

 my work has indicated ; whereas, if we deny the existence of 

 these, the freezing-points calculated according to this or any 

 other law show such divergences from the found values that all 

 semblance of agreement disappears. I am indeed labouring 

 under no small disadvantage in attempting to support the hydrate 

 theory when the greater part of the evidence existing in favour 

 of it is as yet unpublished. 



Before proceeding to the second part of my subject I wish ta 

 draw attention to the great complexity of some of the hydrates 

 which my work has indicated, as well as to the fact that the 

 indications of sudden changes are nowhere more marked than they 

 are with these very weak solutions. The changes, which are 

 observed in the heat of dissolution curve from 5 per cent, down- 

 wards,- afford a good illustration of this latter fact ; or, again, 

 the freezing-points of weak solutions may be instanced,'' where 

 the rate of fall from o to 0*07 per cent, is a quarter as great 

 again as it is from 007 to I'O per cent. The complexity of the 

 hydrates indicated is so great that in the extreme cases they 

 must be represented as containing several thousand HjO mole- 

 cules, and the suggestion of such complexity will no doubt 

 prejudice many against my conclusions in general ; though oti 

 what grounds I know not, for we are entirely in ignorance at 



' Chem. Soc. Trans., 

 a Ibid., 1890, p. 107. 



p. 116. 



3 Ibid., p. 343. 



