246 
Thomas and Ferguson . — On the 
that the observed and calculated values of E are in satisfactory agreement, 
the percentage error being shown in the last column. 
It is not necessary to give full details of any of the other experiments 
made. A number of the graphs of log E and log a are given in Fig. 1 
below, and it will be seen that in all cases the graphs are fairly accurately 
linear, whilst the variation of n under different circumstances is clearly shown 
by the varying slope of the different lines. 
The letters appended to the various lines in Fig. 1 are identical with 
those in the first column of Table II. A reference to this table will there- 
fore show at once the conditions under which the experiment was made. 
In Table II below we have a synoptic view of the results of the whole 
series of experiments. Those indexed A were carried out under ‘dark’ 
room conditions; experiments under the index B were made in a room 
typical of an ordinary laboratory, and those under the index C were made 
in the open air. The seventh column gives the distance d of the level of the 
evaporating water surface below the rim of the containing vessel, while the 
eighth and ninth columns give the values of k and of 11 in 
E = ka n 
as deduced from the graphs of the logarithms of the observed values of E 
and of a. (It should be noted that in experiments A 3 and A 4 a maximum 
and minimum thermometer was not used, and the mean temperatures given 
are the means of several readings of an ordinary thermometer.) 
4. Discussion of Results. 
The preceding table brings out several points of primary importance 
in connexion with the main problem under discussion— the calibration of 
atmometers. In the first place it is to be noted that under no circumstances 
