580 REPOBT— 1886. 



densities of the saturated vapour, that of hydrogen under similar conditions of tem- 

 perature and pressure being taken as unity. At low temperatures the saturated 

 vapours of the alcohols and ether follow Boyle's and Gay Lussac's laws ; hut if the 

 temperature is raised the density of the saturated vapour is found to increase, slowly 

 at first, but more and more rapidly, luitil the critical point is reached, when the 

 specific gravity of the liquid is equal to that of its saturated vapour, and the devia- 

 tion from Boj^le's law is very great. Thus with ethyl alcohol the density of the 

 saturated vapour is normal at about 40°, and it remains normal at lower tempera- 

 tures. With acetic acid very different results are obtained. The constants for 

 this body have not been determined up to the critical point, the temperature being 

 too high for the method adopted. From about 140° to 280°, however, the density 

 of the saturated vapour is found to increase, and there can be no doubt that this 

 increase would go on up to the critical point. When the temperature is reduced 

 below 140° the density of the satui-ated vapour (which at this point is about 50, 

 the normal density being 30) does not continue to decrease but becomes nearly 

 constiint, and at still lower temperatures rises again more and more rapidly until 

 at 50° it is nearly 60. There is then a marked difference between such 

 substances as ethyl alcohol, in which no dissociation of any kind is known to occur, 

 and acetic acid, the vapour of which is believed by many chemists to consist, in 

 part, of molecules of the formula n(C.^H^O„), n being generallj^ supposed equal to 2. 

 It is very improbable that the increase of the density of the saturated vapour with 

 rise of temperature, which appears to be common to all bodies, is due to the union 

 of simple gaseous molecules to form complex groups, for dissociation is always in- 

 creased by rise of temperature ; on the other hand, owing to the high pressure, the 

 molecules are brought into clo.se proximity, and the abnormality may be caused by 

 a general attraction of the molecules for each other. At low temperatures, at 

 which the pressure is small, the molecules are far apart, and such a general attrac- 

 tion is out of the question. The deviation in the case of acetic acid at low 

 temperatures is therefore probably due to the combination of the simple gaseous 

 molecules to form complex groups of the formula n(0.,H,02). This is borne out by 

 the fact that with nitric peroxide, the vapour of which is known to dissociate 

 according to the equation N.^Oj = 2N0.,, a similar increase of the density of the 

 saturated vapour takes place. The densities of the unsaturated vapour at various 

 constant temperatures have been determined by the Natansons (' Wied. Ann.' July 

 1886) and the vapour pressures by ourselves ('Phil. Trans.' 1885). The junction 

 of the isothermals showing the relation of pressure to vapour density with the 

 horizontal lines indicating vapour pressures gives points on the cm've representing 

 the relation of the densities of the saturated vapour to pressure ; and it is thus 

 found that the density of the saturated vapour increases with fall of temperature 

 and pressure. 



It is extremely improbable that the increase of the density of the saturated 

 vapour of acetic acid at both high and low temperatures can be due to the same 

 cause, the conditions being so totally different ; and it may fairly be concluded that 

 at high temperatures the abnormality of this and of all other bodies is due to a 

 physical rather than a chemical attraction. The existence of complex molecular 

 groups in the vapours of normal substances, such as alcohol, is therefore very im- 

 likely, and it can hardly be supposed that they are formed at the moment of con- 

 densation ; hence, in all probabilit}', the molecules of ordinary liquids are simple 

 gaseous molecules in close proximity. 



Other facts observed in the course of this research lead to the same conclusion. 

 We shall discuss their bearing on the question in the ' Philosophical Magazine.' 



We have attempted to study the solubility of eosin in alcohol at temperatures 

 above the critical point, but the experimental difficulties are so great that no very 

 definite conclusions have been arrived at. It appears, however, most probable that 

 at temperatures but slightly higher than the critical point, the eosin remains in 

 solution for some time (as shown by its fluorescence), but that it is gradually 

 deposited. 



