326 Frederick Guthrie on certain 
success, by examining the size of a liquid drop. But I soon 
found that other factors, notably the shape of the solid bodies 
from which the dropping occurred, and the rate of dropping, 
introduced arbitrary conditions which removed the measure- 
ments from the class of simple physical constants. 
§ 16. The plan adopted in the following experiments is the 
endeavour to support a mass of liquid above a plane surface 
in such a way that no actual contact ensues, not even such as 
takes place between clean glass and mercury. If such can 
be done, it is clear that we shall have a circular flat slab with 
rounded edges, and into the shape of that slab the influence 
of adhesion by no means enters. If the thickness of the slab 
be found to be a constant, we shall have a constant as charac- 
teristic as density, and, like density, varying for the same 
mass only according to volume, such volume-change in our 
case being brought about by heat alone. Such slab-thickness 
has for its negative influence the action of gravity (density), 
for its positive the cohesion and surface-tension. 
§ 17. The actual measurements of the slab-thickness I have 
performed in two ways: — (1) by a spherimeter which, when 
used as such, gives results trustworthy to the xoooo °f an iuch. 
But the upper of the two surfaces whose distance has to be 
measured being liquid, and the lower one not very hard, the 
spherimeter cannot be used by the method of touch. Accord- 
ingly I have measured the slab-thickness indirectly. A known 
volume of the liquid is poured on the surface, and teased into 
the circular form if it shows any noticeable departures from it. 
Four or five diameters are measured by means of a small hori- 
zontal cathetometer. The mean being taken, an allowance 
has to be made for the meniscus. This reduces the shape to 
the cylindrical, from which the thickness h is deduced by 
means of the equation 
;,= * 
irr 2 
§ 18. In regard to the actual apparatus: — Upon a thick 
round slab of paraffin, a foot in diameter and 4 inches thick, 
a massive foot of plaster is cast. The whole is placed on 
a three-screw levelling support. The surface of the paraffin 
is scraped into a true plane. When water was being exa- 
mined, the surface of the paraffin was lightly powdered with 
lycopodium and the water poured on vertically from a fine 
opening. With some care a perfectly round slab of water 
6 inches in diameter can be formed, which is so free to move 
that the greatest nicety of adjustment in the levelling-screws 
is necessary. Precisely the same arrangement can be adopted 
