MISC. PUBLICATION SOS, U. S. DEPT. OF AGRICULTURE 
The limiting effect of temperature on plants was expressed in the 
same manner. The coefficient for temperature efficiency Thorn- 
thw^aite called the T-E ratio. The sum of the 12 monthly ratios for 
any station he called the T-E index. Determining by empirical 
means an equation that gave the poleward limit of tundra a T-E 
index of zero and the poleward limit of the tropical rain forest and 
savannah a T-E index of 128 (which was the same as the P-E index 
of the arid margin of the rain forest), six temperature provinces 
were defined as follows : 
A' (tropical), T-E index 128 and above. 
B' (mesothermal), T-E index 64-127. 
C (microthermal), T-E index 32-63. 
D' (taiga), T-E index 16-31. 
E' (tundra), T-E index 1-15. 
F' (frost), T-E index 0. 
The sixth province rarel}^ occurs in the United States, 
Where temperature efficiency is adequate, variations in precipita- 
tion effectiveness establish the primary climatic boundaries; other- 
wise temperature efficiency is the limiting factor and determines the 
boundaries. The combination of the two produces a climatic prov- 
ince, in which various combinations of temperature effectiveness, 
precipitation effectiveness, and seasonal distribution of precipitation 
are possible, as shown in figure 2. 
LENGTH OF THE GROWING SEASON 
The figures for the length of growing season are taken from the 
Atlas of American Agriculture (^5). Reference to the map showing 
average annual frost-free season will show any portion of the country 
to be included between lines that indicate the average length of 
growing season. Between any two such lines is an area wherein the 
average annual frost-free season changes with the distance from any 
designated line. It is obvious that an average or mean cannot vary, 
but it is equally plain that it may be different in one place from what 
it is in another. Therefore, in the characterizations under each 
region when the statement is made, for instance, that the average 
is from 10 to 20 in the northern portion and from 40 to 50 in the 
southern, it is meant that in the northern portion of the region in 
question the average of the extreme north may be 10, but at a short 
distance south it may be 20. In the southern portion likewise, the 
average may be 50 in the extreme south and a certain number of 
miles north only 40. But it should be clear from this example that 
the growing season in the northern portion, changing in length from 
one place to another, varies from 10 to 20 and that the growing sea- 
son of the southern portion varies likewise from 40 to 50. 
SNOW COVER 
The figures for number of days with snow cover are taken from 
the Atlas of American Agriculture (4^). Here, as in the discussion 
of the growing season, the average annual figure is said to vary. 
Again it is obvious that averages do not vary, but may change with 
distance. When it is said, for instance, that the average annual 
number of days with snow cover is from 1 to 10, it is to be under- 
stood that the average figure changes with distance ; that somewhere 
