1887.] Capillary Action in some Chemical Decompositions. 67 
of capillary action. The same cause will also account for the con- 
densation of other vapours on solids and liquids when the space 
containing them is not saturated with those vapours; and there 
is no essential difference between condensible vapours and the 
so-called incondensible gases which could affect the result except 
in degree. We should expect all vapours and gases to be con- 
densed on the surfaces of solids and liquids more or less. 
No one can have observed the way in which bubbles of air form 
upon a piece of metal immersed in hot water, and compared it 
with the way in which globules of oil form on the surface of a 
glass plate which has been wetted with oil and then immersed in 
water, without being struck by the similarity of the appearances. 
The film of air in the one case, and of oil in the other, is rolled 
up into globules which adhere to the plate by a portion of their 
surface. The explanation given of the one will serve equally 
well for the other. In the case of the oiled glass the surface 
tension for water and glass is less than the difference of the 
tensions for oil and glass and for oil and water. So too for the 
metal plate and its film of air, the air is rolled up into bubbles 
because the surface tension between water and the metal is less 
than the difference of the surface tensions for air and metal and 
for air and water. The bubbles of air, and the globules of oil, 
remain adhering to the surface of the solid only when it has a 
tolerably large radius of curvature, but escape with great readiness 
from points, i.e. from surfaces of small radius of curvature. This 
agrees with what is observed in the case of drops. A large drop 
will hang from a large surface, such as the blunt end of a rod, 
if thoroughly wetted by it, but you cannot get a drop of oil to 
hang from the point of a needle or a drop of water from a fine 
pointed glass rod. The drop if it holds on to the 
solid always takes a form something like that shewn 
in the adjoining figure: the point sticks out through 
the drop. And if you want small drops to fall readily 
you drop them from a rod drawn out to a fine end. _—— 
However to form drops like that in the figure the 
surface tension between the solid and adhering liquid 
must be less than that between air and the solid. 
If the film of air on a pointed solid is rolled up 
to the point by the capillary action of the liquid 
wetting the solid it will have a small surface to 
adhere to and will be easily disengaged just as 
small drops fall from a pointed solid. But there is another reason 
which seems to me to conduce to this result. When the surface 
tension between the solid and liquid is small the surface of the 
drop is concave to the air at points a and b, the radius of 
curvature at a is less than that at 6 in the principal section, 
ra 
(ta 
Liss 
YW, 
il 
YV 
a 
<1) || 
g 
