RESEARCH METHODS Dff STUDY OF FOREST ENVIRONMENT. 19 
When readings are taken in the morning, if there is no thermo- 
graph record by which the time of the maximum and minimum may 
be determined, the minimum then read should be tabulated on the 
form for "Air Temperature Record." as of the current day, and the 
maximum as of the preceding day. If readings are made in the 
afternoon, both maximum and minimum should be credited to the 
current day. The current temperature should, of course, be credited 
to the day on which taken. The instrumental corrections should be 
used when entering the data in the field, if cards therefor have been 
prepared, the card being tacked in a conspicuous place in the instru- 
ment shelter. 
The daily range — purely a computed quantity — in degrees and 
tenths should be the difference between maximum and minimum tem- 
peratures as tabulated for any calendar day. 
Hourly Temperatures. 
Where a thermograph is available the instrument should be set 
in the same shelter as the maximum and minimum thermometers, and 
hourly temperatures may be obtained therefrom. 3 Corrections for 
the thermograph trace should always be obtained from the readings 
of the maximum and minimum thermometer, as thermograph records 
are liable to considerable errors; but the hours to which these correc- 
tions are applied may well be a matter of judgment with the ob- 
server, depending on the shape of the temperature curve. 4 The tabu- 
lation of hourly temperatures when obtained will require the special 
form, " Hourly (Air, Soil, or Actinograph) Temperatures." Certain 
data therefrom will be entered on the "Air Temperature Record." 
For example, as a measure of conditions affecting growth rate, it mayr 
be desirable to know, besides the mean: 
3 In any ordinary comparison of the temperatures of plant habitats, hourly tempera- 
tures ;ire not likely to be used except to explain transient phenomena. However, the 
thermograph is an extremely valuable adjunct in determining' the maximum, minimum, 
and mean temperatures, not only helping to correct errors of observation but making 
possible the more exact determination of the extremes and temperature ranges for any 
period, such as the midnight-to-midnight day, which is the unit of time in most meteo- 
rological computations. 
* Various rules for applying corrections to thermograph traces are used by different 
students. It is obvious that errors may exist in the traces from two distinct causes: 
il t When the range of oscillation of the pen is too great or too small the thermograph 
may read correctly at medium temperatures but be high and low at the two extremes; 
(21 even if the pen is approximately correct in its possible range there is a lag due both 
to the lesser sensitiveness of the thermograph as compared with a mercurial thermometer 
and to the friction of the pen upon the paper, so that normally the pen does not quite 
reach to the extremes indicated by the thermometers. In the first case, it is essential 
that the error he distributed somewhat according to the temperatures : thus, if the pen 
read correctly at a temperature of 45, at all temperatures above 43 — the range of the pen 
heing too great — there would be a minus correction for the trace, and at all temperatures 
below 45 there would be a plus correction. On the other hand, if the instrument is 
properly adjusted, it is logical to apply a minus correction to all descending portions of 
the trace and a plus correction to all ascending portions, the amount of such correc- 
tions to be determined from the corrections at the minimum and maximum, respectively. 
