;8o 



NA TURE 



[August i8, 1904 



differences, A, for an addition of CH, are given under the 

 following conditions ; ' — 



A. At 0° C. 



B. At the respective boiling points under i attn. pressure. 



C. At equal reduced temperatures (0-6396). 



D. At the respective boiling-points under equal reduced 

 pressures (0-02241}. 



E. At the respective critical points. 



It will be seen that in every case there is a decided rise 

 in the value of A as the series is ascended, but that the 

 rise is relatively smallest when the comparison is made at 

 the particular reduced temperature chosen. At higher re- 

 duced temperatures, however, it would be relatively much 

 greater, since it is very marked at the critical point, where 

 the reduced temperature = i. The rise is also comparatively 

 small at the common temperature 0°, but the comparison 

 would not be satisfactory if a higher common temperature, 

 sa.v 150°. ■^^■tre chosen, because the coefficients of expansion 

 differ considerably ; at 150° the values of A would be 87;;, 

 '3-45. and 15-38 respectively. 



In the case of nine of the lower esters the values of A 

 are by no ineans constant, whether the comparison be made 

 at o , at the boiling-point, or at the critical point. The 

 eleven values of A vary in the three methods between 16-34 

 and 1S.21, 20-84 and 23-42, .■;4-3 and 61-7 respectively; but 

 there is not a regular rise with increase of molecular weight. 



Both Kopp and Schroder compared the molecular volumes 

 of compounds at their boiling-points under normal pressure 

 but they deduced quite different values for the atomic- 

 volumes of carbon and hydrogen ; it is clear, however that 

 as A varies considerably no values whatever for C and H 

 could give accurate results, even in the case of true homo- 

 logues. 



Traube makes the comparison at a common temperature 

 usually IS and takes into consideration both the actual 

 volumes of the molecules and the co-volume, which he 

 assumes to have the same value, 24-s (n-a() where 

 n = 1/273, for all substances. He calculates definite values 

 or the atomic volumes of C and H at a given temperature ■ 

 bus, at IS = 09 and H=3.r, or CH, = i6-i, so that 



^houlri^h"" ! ''"^'■'™'^'^ ^°'' ^"= "' '-' ^'^'^^ temperature 

 sbould be constant. 



It does not appear to me that the problem has vet been 

 completely solved, although Traube 's method of ca culation 

 ichr'der'. ^"" """' '"""'' ''^^""^ "^"" '^"^ "^ '^PP ^"^ 



Comparison of Boilmg-poinls at a Scries of Equal Pressures. 

 The results of this comparison are often e.xceedinglv simple 

 . he two substances compared are verv closely related, and 

 t there is no molecular association in either ckse. Taking 

 or example, chlorobenzene and bromobenzene, it is found 

 that the ratio of the boiling-points (on the absolute scale 

 of temperature) under equal pressures is constant whatever 

 the pressure may be, or wi.aicvcr 



T'„ 



I 0590. 



A similar result is obtained with the other halogen deri- 

 '"■ I'^'^u" '^^"^'^"<'' ^^'ith ethyl bromide and ethyl iodide 

 w-ith ethyl acetate and propyl acetate, and some other pairs 

 of esters; but in some cases of close relationship-for ex- 

 ample, with ethyl formate and ethvl acetate— the ratio is 

 not quite constant, and the formula' becomes 



Ta^T'. 

 Tb T'b 



;Tb-T' 



where c has a very low value [0-0000417 for these two esters] 

 When there is no close relationship, but the molecules are 

 not associated, the value of c is usually larger— for example 

 0000118s for carbon disulphide and ethyl bromide. 



Lastly, when there is no close relationship and the 

 molecules of one or both substances are associated the 

 formula ' 



Ta^T', 



Tb T'_ 



1 The atomic weights [C = 11-97, H = i| 

 are retained. 



NO. 1816, VOL. 70] 



„-+c(Tb-T'b) 



mployed in the 



al papers 



may no longer hold, and a third term may be required, 

 thus : 



Ta^Ta 

 Ib T'l 



^4-<-(TB-T'B)-K/(TB-rB)=; 



or, in any case, the value of c becomes much higher, as- 

 with benzene and ethyl alcohol [1; = 0-0008030] or sulphur 

 and carbon disulphide [c = 0-0006845]. 



Behaviour of Liquids when Mixed Togetlicr. 

 There are three points to consider when two liquids are 

 brought together — (i) their miscibility, whether infinite, 

 partial, or inappreciable; (2) the relative volumes of the 

 mi.xture and of the components; (3) the heat evolved or 

 absorbed. 



Liquids which are classed as non-miscible rarely, if ever, 

 bear any close chemical relationship. Thus water is 

 practically non-miscible with all hydrocarbons and with their* 

 halogen and many other derivatives ; again, mercury, so. 

 far as I know, is not miscible with any liquid compound, 

 organic or inorganic. It is true that the higher aliphatic 

 alcohols are almost insoluble in water, although there may 

 be said to be some chemical relationship between them, 

 inasmuch as an alcohol may be regarded as an alkyl deri- 

 vative of water. But the alcohols may also be looked upon 

 as hydro.xyl derivatives of the hydrocarbons, and, the higher 

 the formula weight of the alkyl group, the greater is its 

 influence, relatively to that of the hydroxy!, on the properties 

 of the alcohol. Thus, while the lower alcohols show con- 

 siderable resemblance to water — for example, in their be- 

 haviour with dehydrating agents, such as sodium, phos- 

 phoric anhydride, or lime, and in their power of uniting 

 with metallic salts to form crystalline alcoholates corre- 

 sponding to the hydrates — this resemblance diminishes as we 

 ascend the series, and is generally not observable with the 

 higher members. 



On the other hand, the higher the molecular weight of 

 the alcohol the closer is its resemblance to the hydrocarbon 

 from which it is derived. This, as already mentioned, is 

 well shown by the diminishing difference between the boil- 

 ing-points of the alcohol and paraffin as the series is 

 ascended ; it may also be noted that methane was long 

 classed as a permanent gas, while methyl alcohol is a liquid ; 

 whereas both hexdecane (C.^H,,) and cetvl alcohol 

 (C,8H,,OH) are solids, the former melting at 18° and the- 

 latter at 50°. 



It may, in fact, I think, be stated that the chemical 

 relationship between water and methyl alcohol is fairly close, 

 while that between water and cetyl alcohol is very distant. 

 So, also, two adjacent members of a homologous series, 

 such as methyl and ethyl alcohol, are more closely relatecl 

 than two members of widely different molecular weight, 

 such as methyl and cetyl alcohol. 



.■\dopting this view, it is, I believe, safe to state that 

 liquids which are chemically closely related to each other 

 are invariably miscible in all proportions. 



-As regards the relative volumes of a mixture and of its 

 components at the same temperature, it is well known that 

 inequality is the rule and equality the exception ; and, 

 further, that contraction is more frequently observed than 

 expansion on admixture. So far, however, as experimentat 

 evidence is available, it appears that when the liquids are 

 very closely related to each other the change of volume- 

 is exceedingly small. For example, with ethyl acetate and 

 propionate in equimolecular proportions, -(-0-015 per cent. ,- 

 toluene and ethyl benzene, —0-034 Per cent. ; »-hexane and 

 H-octane, -0-053 per cent.; methyl and ethyl alcohol, 

 -f 004 per cent. ; chlorobenzene and bromobenzene, no. 

 change. 



When the relationship is less close the changes are usually,, 

 but not invariably, larger, and are in some cases positive, 

 in others negative ; and it is rarely possible, in the present 

 state of our knowledge, to predict from the nature of the- 

 substances — unless one is basic and the other acidic in 

 character — whether contraction or expansion is to be ex- 

 pected. Thus, when methyl alcohol is mixed with water 

 considerable contraction occurs, although the relationship- 

 is less close than between methyl and ethyl alcohol, which 

 expand to a minute extent on mixing. 



.■\11 we can say with regard to the alcohols is that, the- 

 higher the molecular weight — or, if isomeric alcohols are 



