354 Mr. A. M. Worthinston on the 



e> 



the repulsion- and attraction-curves leads us to anticipate this 

 condition of things ; for we know that an increase of tempera- 

 ture, when the volume is kept constant, involves an increase 

 of pressure, which implies that the ordinates to the repulsion- 

 curve must be everywhere increased, but most to the left- 

 hand end of the curve, since the increase of pressure when the 

 volume is kept constant is far greater in a solid or liquid than 

 in a gas; thus an elevation of temperature implies a bodily shift- 

 ing of the whole curve upwards with an increase in the slope as 

 we approach the left-hand end. Now the shifting alone would 

 lift the curve quite above the curve of attractions (see fig. 4), 

 which would mean that above a certain temperature repulsion 

 would always exceed attraction, or that the substance could 

 only exist under pressure. There would still, however, be 

 the same region of instability as before, and consequently the 

 surface of discontinuity between liquid and vapour would still 

 exist ; but by a sufficient increase in the slope of the repulsion- 

 curve, such as we have seen will also take place, this region 

 of instability will diminish and finally disappear (see fig. 5) 

 when the slope of the repulsion- curve is everywhere greater 

 than that of the corresponding part of the attraction-curve. 

 At the temperature at which this takes place all discontinuity 

 will disappear, and the substance will be in the same condition 

 throughout. Up to this point the liquid will still exhibit 

 within the vessel in which it is contained the phenomena of a 

 surface-tension, though in reality the wall of the containing 

 vessel will experience, after the temperature has been reached 

 at which the repulsion-curve ceases to cut the attraction-curve, 

 only a pressure even where it cuts the liquid surface. But 

 the pressure will be less here than elsewhere, for our original 

 investigation always holds within the limits for which the 

 slope of the repulsion-curve is greater than that of the curve 

 of attractions. Thus it is probable that the meniscus observed 

 at the surface of liquefied oxygen or nitrogen does not corre- 

 spond to a true tension as in the case of water or mercury. 



14. If our method of dealing with the continuity of the 

 liquid and gaseous states be compared with that of Professor 

 James Thomson (as quoted on pages 124-126 of Maxwell's 

 1 Theory of Heat '), it will be seen that the ordinates to the iso- 

 thermal curves there given correspond to the difference of the 

 ordinates to the two curves in the above investigation — that, 

 in fact, the horizontal axis of coordinates in Maxwell's diagram 

 (which is reproduced in fig. 6) represents our attraction- 

 curve. The suggestion which Professor Thomson made thir- 

 teen years ago — that the portion A B C of the curve, which we 

 know from experimental evidence to represent the conditions 



