1881.] 



On the States of Matter. 



411 



that the cohesion or capillarity of a liquid depends upon temperature 

 alone, and that pressure has no effect upon it. 



Practically, it was found that the less soluble a gas is in a liquid 

 the less effect does its presence cause at any pressure, and it is certain 

 that had we a gas totally insoluble in a liquid, we might compress 

 that gas under any pressure over the surface, without effecting the 

 slightest change in the capillarity of the liquid. 



To return to Andrews' experiment. It was seen that to determine 

 the state of every quantity of matter, it was necessary to have a free 

 surface, but by Andrews' mode of experimenting at pressures greater 

 than the " critical," this was impossible, the tube being filled with a 

 homogeneous fluid ; but it is possible to have a free surface, and to 

 watch the alteration of the capillarity at any pressure, provided a gas 

 can be obtained which is insoluble in the liquid. We have seen that 

 when a gas dissolves in a liquid, the capillarity of the mixture is lower 

 than that of the liquid alone, so by these means we can tell whether 

 or not a gas is soluble. In this way it was found that hydrogen was 

 insoluble in several liquids, and thus the conditions were established 

 for examining loss of capillarity with rise of temperature at any 

 pressures, the hydrogen keeping a free surface without affecting the 

 liquid. It was found that the capillarity fell to zero at the same tem- 

 perature, independent of pressure ; in other words, the matter passed 

 from a state where it "retains a permanent limiting surface exhibiting 

 capillarity," to a state where it " assumes no surface nor definite 

 figure, and which is so extensible, that any quantity will distribute 

 itself throughout a space," as we defined the liquid and gaseous states. 



The state of affairs in those experiments is this: — The hydrogen 

 takes the place of a portion of the vapour of the liquid, but unlike the 

 molecules of the vapour, its molecules are unable to penetrate the 

 surface of tbe liquid, in any quantity, but merely act by striking 

 against the molecules which would otherwise pass out, arid driving 

 them back into the liquid. Could there be any greater proof than 

 this of the special properties of a liquid surface ? Let the temperature 

 be very near the critical, yet the mixture over the liquid retains a very 

 much lower density than the liquid, as the hydrogen does its work in 

 keeping back a portion of the vapour molecules. In this case the 

 upper does not become equal in density to the lower portion of the 

 fluid as in Cagniard de la Tour's experiment, yet when the critical 

 temperature is reached the meniscus disappears and diffusion occurs 

 between the upper and lower portions. This phenomenon is seen at 

 any pressure up to five times the " critical " pressure, showing that 

 the change of state is independent of pressure. Gaseity is dependent 

 upon molecular velocity (thermal activity), and to give it this velocity 

 a very strong cohesion must be overcome. The velocity is not in any 

 way dependent upon the pressure or upon the size of the vessel, and 



VOL. xxxii. 2 F 



