VARIATION OF MOLECULAR SURFACE-ENERGY WITH TEMPERATURE. 665 
the liquid, and kaowiug from our experiments that the relation between capillary 
ascent and temperature is a nearly linear one, the rise in the tube T could be calcu¬ 
lated approximately enough for our purpose by constructing’ a diagram in which the 
rise at atmospheric temperature was made one end of a straight line connecting it 
with the critical temperature, where the rise is zero. At intermediate temperatures 
the rise in the tube T was read off from the diagram ; it may be taken as practically 
accurate ; for any divergence from truth is much less than the limits of error in 
reading the ascent in the narrow capillary tube. This rise was therefore added to 
the apparent rise in the capillary tube, and the sum was taken to represent the 
total rise. 
No correction has been made for the increase in internal diameter of the capillary 
tube due to rise of temperatui’e. The surface-tension is directly proportional to the 
diameter of the tube ; but as the latter is increased by only 0‘5 per cent, for a rise of 
temperature of nearly 400°, the correction was neglected. 
Nor has any correction been introduced for the form of the meniscus. The total 
correction should have amounted to 0’05 millim. But readings were taken with the 
lower edge of the cross-line of the cathetometer on the lower surface of the meniscus, 
so that the minute error was thereby somewhat compensated. 
In calculating the results of experiments so as to ascertain the surface-tension, it 
has been assumed that there is no angle of contact between the liquid and the glass. 
The reasons for this assumption are as follows:— 
A tube of capillary bore (0'65 millim.) contained ether free from air. It was con¬ 
nected with the pressure-apparatus, and its temperature was raised. Bubbles were 
made to appear in the tube, and it was found possible, at any given temperature, to 
keep them stationary ; they did not ascend, or more correctly, the rate of ascent was 
exceedingly slow, so that a bubble might be kept for an hour, without rising a 
millimetre. A bubble in such a capillary tube is confined on its lower and upper 
surfaces by cups of liquid ; the walls of the tube between these cups is wmt, it is true, 
but by an exceedingly thin film of liquid; so thin that the transference of liquid from 
the upper to the lower part of the tube is very slow, and hence the almost stationary 
position of the bubble. 
By decreasing the volume of the bubble on screwing in the plunger, the cups 
approached each other, the lower cup rising nearer to the dome confining the bubble 
of vapour on its iq^per side. The bubble began to ascend rapidly only when its 
vertical diameter had been made equal to its horizontal diameter ; fie., when it was 
approximately spherical. Above 160° C. a bubble of ether refused to remain stationary 
in the tube, but at once ascended ; and with ethyl alcohol, the ascent commenced 
above 220°, These temperatures are evidently conditioned by the stability of the 
* Some of these experiments were made in conjunction with Mr. R. W. Stewaet, B.Sc., with whom 
one of the authors made a preliminary survey of the field in 1891 and to whom he desires to express his 
acknowled gm ent s. 
MDCCCXCIII.—A. 4 Q 
