1902-3.] Determining Latent Heat of Evaporation. 453 
is to be made the calorimeter ves'sel must swing quite free of the 
mercury cups, even with an ordinary amplitude of vibration of the 
balance. In the second place, when the beam is lowered, the 
electrodes must dip well into the mercury cups. For this purpose 
the rest is removed from its ordinary place and set somewhat 
lower than its normal position. 
The shield consists of an outer square wooden cylinder and an 
inner of light tinned iron. This inner shield forms a circular 
cylinder capped by a cone with small apertures at top and bottom. 
For convenience, both parts are cut along a vertical plane through 
the centre (see plan) into a back and front j)ortion, which can be 
removed or replaced by a single movement. The lugs which hold 
these in position on the base also act as guides, which ensure that 
each part always comes into the same position. By this arrange- 
ment the calorimeter is shielded from falling drops as well as 
from radiation. 
The aperture cut in the wooden top is about | inch diameter 
but it can be entirely closed by a flat piece of -J-inch india-rubber cut 
partially along a line running across it. This grips the suspending 
wire tightly. For the weighing the rubber is slid along till at 
another part of its section it embraces the wire with a small 
circular aperture of about 1 mm. radius. As even with this small 
aperture or no aperture condensation takes place on the suspending 
wire, it has been found necessary to use the device adopted by Joly 
in his steam calorimeter — a small spiral of platinum wire surround- 
ing the suspending wire just above the aperture and brought to 
faint red heat by an electrical current. With this arrangement and 
the mode of jacketing, the weight becomes quite steady. 
The balance used is a simple 250 gm. balance by H. L. Becker, 
with a sensibility for load 150 gms. of 2£ divisions to a centigram. 
It would therefore be possible to weigh to a milligm., but with a 
total mass evaporated of 50 gms. a centigm. represents an error 
of only 1 in 5000. The balance is mounted on a deal board, 
which is cut away under one pan (as in elevation in diagram), 
and is completely enclosed in a suitable case with glass front (not 
indicated in the diagram). 
It was at first expected that the vapour produced by the 
evaporation from the glass vessel would be sufficient for jacketing, 
