362 BOTANICAL GAZETTE [NOVEMBER 
mates those which have so far resulted from most physiological 
studies of temperature relations. It seems that this value is more 
likely to be too low than too high for ranges of temperature com- 
monly met with in nature, but the present status of the problem 
does not warrant any attempt at a closer approximation. Indeed, 
it seems almost certain that the magnitude of the temperature 
coefficient will be found to vary, not only with different plant forms 
and with different stages of development of the same form, but also 
with the values of the temperatures’ considered. It is readily 
conceivable that the relation which we are seeking may be deter- 
mined satisfactorily only by the use of a temperature coefficient 
which is itself a variable, changing in value with the progress of the 
organism through its life cycle and with the annual march of the 
Seasons, as well as with variations in the temperature itself. The 
time is not yet ripe, however, for even an a priori discussion of 
this matter. 
Having tentatively established the temperature coefficient which 
is to be taken as a measure of the effectiveness of temperature 
in advancing growth (the intensity factor), we must make a similar 
assumption in case of the duration factor. For what period of 
time should we apply our assumed temperature coefficient ? This 
question is precisely the same as the one met with in connection with 
direct temperature summations, and its answer, as in that case, 
must involve the relation of time to the developmental stages In 
tie “gani.«is concerned. Since it is here desired to deal with the 
who. “ter,in the broadest and most general way, it is requisite 
to fix upon a time period, which will, as nearly as possible, approxl- 
mate the period of active growth in the majority of higher. plants. 
As one of the authors? has emphasized, the controlling climatic 
conditions are primarily effective for most plants only during the 
season of active growth. This growing season may be approxi- 
mated, for phytogeographical purposes, as the average or normal 
length of the frostless season, the number of days which intervene 
between the average date of the last killing frost in spring and the 
first in autumn. This has been adopted as the time factor in the 
present study. 
*3 Proc. Amer. Phil, Soc, 1913; as already cited, 

