1902 - 3 .] Determining Latent Heat of Evaporation. 455 
standard of e.m.f. being a Clark cell. The current is measured by 
similarly obtaining the potential difference across the terminals of 
a Crompton - 01 ohm through which the current passes. As the 
current shows a steady variation during the time of an experiment, 
(about forty minutes), readings have to be taken every minute, the 
times being simultaneously noted. In order to obtain as steady a 
current as possible three sets in parallel of two large 100-ampere stor- 
age cells in series have been used, and two variable carbon resist- 
ances in parallel inserted in the circuit. This was found better than 
having three in parallel, probably because the greater heating of the 
carbon with a negative resistance coefficient more nearly compen- 
sates the increase of resistance of the copper leads. The total 
variation of the current during an experiment is never more than 
1 per cent. 
Order of Experiment. 
Heating . — The current is turned on a little before the jacketing 
steam is expected to be supplied. Until the calorimeter vessel 
reaches the boiling point, steam is condensed on the outside of the 
vessel; and as some of this water may remain during the experiment, 
the catch water is provided in order to prevent it being lost except 
by evaporation. Boiling is recognised by the weight of the 
calorimeter commencing to decrease. 
Determination of Resistance . — Two turns of fine copper binding 
wire are passed round the electrodes as high as possible, and at 
observed distances from the lower ends. The other ends of the 
wire are connected to the potential leads. Using the same current 
as in the main experiment a series of readings is taken of the 
e.m.f. ’s across the potential leads and the *01 ohm alternately, in 
rapid succession. After the first two determinations these read- 
ings can be taken more and more rapidly, within intervals of ten 
seconds or less if the current is not varying irregularly. The 
resistance between the mid-points of the covered parts of the 
electrodes is then calculated by applying the correction already 
given. 
First Weighing. — The current being turned on to the spiral 
above the shield, the india-rubber closing the top is placed so as to, 
enclose the small aperture around the suspending wire. The ap>-- 
