16 
determined under controlled conditions of moisture supply and sun- 
light. 
3. Especially in connection with the preceding, and as integrating 
air temperature and sunlight influences, leaf temperatures should be 
measured as a more direct criterion of the temperature conditions 
regulating food production and growth. What is particularly 
sought, of course, is the relation between leaf temperatures, air tem- 
peratures, and sunlight, and whether or not this is essentially dif- 
ferent in different plants. It is probably necessary to determine gen- 
eral relations of this kind and to base observations of growth on long- 
term air temperature records. 
4. The maximum temperatures which may be tolerated without 
highly destructive reactions in the plant, leading to fatal results, have 
been investigated very little and apparently have received very little 
weight in considering problems of distribution, although a number 
of investigators have shown that growth-rate falls off rapidly beyond 
a certain optimum temperature. The difficulty of observations on 
this point lies in the extremely close connection which is likely to 
exist, under any natural conditions, between very high temperatures 
and excessive transpiration or positive drought in the soil's surface. 
In the general study of climatic or temperature zones affecting- 
plant distribution and life forms, Merriam's (30) 2 work has become 
classic. The more minute determination of forest zones may begin 
with comparison of mean temperatures or temperature sums above a 
minimum of about 40° F., or similar sums for the frostless period. 
Livingston (25), in a general survey of the temperatures of the 
United States, carried the matter one step farther by rating the chem- 
ical efficiency of temperatures above 40° F., according to the van't 
Hoff-Arrhenius principle of doubled activity for each 18° F. increase 
in temperature. These temperature efficiencies were then summed 
for the growing season, in place of the original temperatures. Samp- 
son (32), McLane (31), and Lehenbauer (24), have tried various 
modifications of the Merriam idea of temperature sums, all of which 
should be looked into. One will hardly escape the conviction that the 
consideration of any temperature term other than the mean tempera- 
ture will require the accumulation of hourly temperature records or, 
in other words, the use of the thermograph. 
In the more exact study of the rate of growth as influenced by 
temperature a greater number of technical problems are presented. 
The temperature coefficient can not be determined unless moisture and 
sunlight are under control. The actual measurements of growth 
rate are difficult, and necessitate first of all plants of uniform age 
and size for the various comparisons. It seems to be quite well estab- 
lished that growth of most plants begins at about 40° F., is very slow 
8 The figures in parenthesis refer to the bibliography at the end of this bulletin. 
