Surron—Study of Evaporation from Water-Surfaces. 165 
high relative humidity, and, evidently, it is just when 6 is large 
that the effects of a decreased relative humidity are most strongly 
felt. ‘The essential fact that an accelerated evaporation corre- 
sponds to a diminished humidity holds good for the pipe just as 
it does for the open cup. Before leaving this part of the subject it 
is perhaps worth while suggesting that in this case also as in that 
of the hair hygrometer the reason why there is probably a 
humidity factor in the process of evaporation from a water-surface 
is that the greater average kinetic energy of the aqueous molecules 
at the higher temperature of the space is the more competent to 
break up the closed chains of molecules in the surface of the 
water. Thermal effusion is probably not concerned in the process 
excepting to a very small extent. 
‘In Tables 24 and 25 at the end of this paper will be found a 
summary of the results for different intervals and periods. A 
comparison between these and the smaller tables in the text 
suggests the following provisional conclusions: that while 
differences between the vapour-tensions at the water-surface and in 
the open air are competent to influence the rate of evaporation to 
a large extent, the intensity of the effect of vapour-tension 
differences is profoundly modified by the relation the temperature 
of the dew-point bears to the temperature of the air—or, in other 
words, is profoundly modified by the relative humidity. The 
water-temperatures are, as such, probably of no great importance, 
initially at any rate ; but when considered in conjunction with 
the temperature and relative humidity of the air, an influence 
becomes apparent which, so far as I know, has not hitherto 
received due recognition. It seems to me to be extremely probable 
that after the relative humidity of the open air and the differences 
of vapour-tension (7.e. 6 =p’ —p”) have been allowed for, much 
of the observed evaporation, from whatsoever form of water-surface 
or type of gauge, is due to convection currents. Thus, for example, 
in the case of the three metal pipes we see that the evaporation 
is not most rapid when the temperature of the water, or of the 
air, 1s highest, or when the relative humidity is least, or when 
the differences of vapour-tension are greatest, either, or all, but 
when, in addition to favourable values of these, the temperature of 
the air falls below that of the water-surfaces. According to the 
observations we get the maximum effect in the pipes between 
