Marsh , and its Relations to Evaporation , &c. 313 
This time, 1908, I, was taken as the standard instrument. 
Then : — 
(1) Comparing the respective areas of the two evaporating surfaces, we 
have the ratio : — 
1908, I _ 141-20 sq. cm. _ 2-27 
1908 cryst. dish 62-07 sq. cm. i-oo 
(2) From the calibration data given above we find that the total 
evaporation for a given time, from the standard and the dish 
respectively, is in the ratio : — - 
From 1908, I _ 50-40 _ 3-25 
From 1908 cryst. dish 15-51 i-oo 
(3) Therefore, the different rates of evaporation per unit area from 
the two evaporating surfaces are according to the ratio : — 
Porous earthenware _ 3-25 _ 1-43 
Open water surface 2-27 i-oo 
It will be interesting to compare the rate of evaporation from the 
porous cylinders employed in my experiments with the rate in the case of 
Livingston’s instruments. The latter author used unglazed porcelain 
cylinders, with one end closed and rounded. 1 From the measurements he 
gives, the area of the evaporating surface of one of his cylinders may be 
calculated as about 86 sq. cm. The average of his calibration experiments 
showed that the evaporation from one cylinder was equivalent to that from 
99-26 sq. cm. of open water surface. Therefore the different rates of evapora- 
tion, per unit area, from the respective evaporating surfaces, were in the case 
of his cylinders according to the ratio : — 
Porcelain _ 99-26 _ 1-15 
Open water surface 86-oo i-oo 
That is, the rates of evaporation per unit area from the porous 
cylinders used in the three cases were as follows : — 
(a) From Livingston’s cylinders, 1-15 times as fast as from an open water 
surface. 
(b) From my 1907 cylinders, 1-12 times as fast as from open water. 
(c) From my 1908 cylinders, 1-43 times as fast as from open water. 
These differences in the rate of evaporation are probably due, as will 
be seen later, to differences in the grain of the earthenware of which the 
cylinders are composed. 
We may now compare the cylinders with each other. The calibration 
data obtained from the sum total of the laboratory readings enabled 
multiplication factors to be calculated by which all the field readings could 
1 Livingston (’06), p. 20 , 
