A NEW EVAPORATION FORMULA 25 
the corrected elevation (elevation corrected for wind and barometric effects), unit 
weight being assigned to the Buffalo gage on Lake Erie for which the probable error 
of a corrected elevation of the surface of Lake Erie was ±0.036 foot. (See Publi- 
cation No. 317, Table 28, page 110.) 
EXAMPLES OF COMPUTATION OF KNOWN QUANTITIES IN THE OBSERVATION 
EQUATION, EQUATION (1) 
In the preceding pages (beginning with page 8), the final form of observation 
equation used in this investigation for determining the evaporation has been 
stated, equation (1), and the definitions of the various quantities and the methods of 
computing them have been set forth in general terms. In order to remove any 
ambiguity which might exist in those general definitions and statements of method, 
it is now proposed to give a specific example of the computations of each of the 
quantities e, w and J. The examples will be taken from the computations on Lake 
Michigan-Huron, inasmuch as the observations on that lake had by far the greater 
weight in fixing the final numerical values of the unknowns in the evaporation 
formula derived in this investigation. 
EXAMPLE OF COMPUTATION OF e. LAKE MICHIGAN-HURON 
The definition of e is given on page (8), and the meteorological stations at 
which the weather elements of vapor-pressure and temperature were observed are 
listed in Table 1 and shown on Plate 1. The computation of e for Lake Michigan- 
Huron for the dates September 1 to October 10 inclusive, 1910, is shown in Table 
10. In this table, for each station in the order in which they appear in Table 1, 
there are three columns headed, respectively, "Mean Temp.", "Sat. Pres." and 
"Vap. Pres.". In the first of the three columns is recorded for each day the mean 
air temperature as observed at the station during the two days ending at midnight 
on the current day. In obtaining this mean, the mean temperature for each day was 
first computed by taking the mean of the maximum and minimum for that day. 
The mean of two such means, one for the preceding and one for the current day, 
was then taken as the mean temperature as recorded on the current day. In the 
second column is recorded the saturation vapor-pressure corresponding to the 
mean temperature for the two days ending at midnight on the current day. In 
the third column is recorded the vapor-pressure as observed at the station. The 
vapor-pressure on the current day was taken as the mean between the vapor- 
pressure as observed at 8 p. m. of the preceding day and 8 a. m. of the current day, 
75th meridian time. The temperature is in degrees Fahrenheit, and the vapor- 
pressures are in units of 0.01 inch of mercury. 
On the right-hand side of the table are shown the values of e. These values 
are obtained for each station by subtracting the observed vapor-pressure from the 
saturation vapor-pressure for that station, and are in 0.01 inch of mercury. The 
mean e for the lake computed from the stations on the United States side is shown 
in the last column but 4. The mean e computed from the two Canadian stations, 
Parry Sound and Saugeen, is shown in the last column but one. The final mean e 
for the lake shown in the last column is the arithmetic mean at all the ten stations. 
This is the e used in the observation equation, (1). 
Note that the observation equation for any day covers a two-day time interval 
ending at midnight of the current day, or day to which the equation is assigned in 
listing the equations. All of the observed quantities in the equation, therefore, 
