MECHANICAL STRETCHING OF LIQUIDS. 
369 
General Remarks on the Experiments. 
It will be noticed that the manner of experimentation has the disadvantage that 
it does not permit of the whole of the liquid under tension being at the same 
temperature. The mean temperature was probably at no time more than 5° removed 
from 16° C. I am now endeavouring to make an apparatus in which it shall be 
possible to maintain the portion of the liquid, whose stretching is to be measured, at 
any desired constant temperature, whereby also the chief difficulty in the way of 
more exact measurements will be overcome. 
As regards the conclusion reached, that the changes of volume of a mass of liquid 
are equal for numerically equal increments of pressure, whether positive or negative, it 
may be justly observed that this was only to be expected. No one, however, could 
predict that the coefficient of extensibility would remain practically constant up to 
tensions of 17 atmospheres, and nothing but further experiment can decide what 
changes may take place in its value as the tension is increased. 
Bearing of the Observations on the Theory of Surface-Tension. 
The unequivocal proof that a liquid can exist in stable equilibrum, in a state of 
isotropic tensile strain, has a bearing on the theory of surface forces in fluids. For it 
can be shown to be necessary for equilibrium that a compressible liquid shall be, close 
to the free surface, less dense than in the interior ; in other words, the surface layers 
are in a condition to which interior liquid could be brought by stretching it, and are, 
therefore, a seat of energy in precisely the same way that stretched liquid is a seat of 
energy. A theory, such as that of Laplace, which assumes uniform density precludes 
us from admitting in the material itself any such modification correspondent to the 
surface energy, and drives us to seek it in the condition of the superjacent ether film. 
Note on a Curious Phenomenon of Adhesion between two Solids immersed in a 
Stretched Liquid. 
Desiring to ascertain whether an air-free liquid would adhere under tension, as well 
to a metal as to glass, I enclosed a small piece of folded sheet copper in a glass bulb, 
which was then filled with boiled-out, air-free alcohol. Experiments with this 
showed strong adhesion to the copper, as well as to the glass, provided the vessel was 
kept still, but any agitation at once caused the stretched liquid to let go its hold at 
the place of contact of the copper and the glass. Close attention showed that the 
copper seemed to “grow to the glass ” at the points of contact, when the surrounding 
liquid was in a state of tension. This led to experiments on bulbs, with smaller 
bulbs of glass inside, and in all cases the same phenomenon was observed : when the 
liquid was stretched, the loose bulb attached itself to the side of the vessel. The 
equilibrium was, however, very unstable. The release of the liquid took place on the 
MDCCCXCII.-A. 3 B 
