CLIMATIC CONDITIONS OF THE UNITED STATES. 249 



rain-gage or atmometer tank, but these conditions have just begun 

 to attract attention in connection with the quantitatively dynamic 

 aspect of the study of plant activities, and we are not able to go nearly 

 so far with their treatment as is possible with temperature. Work 

 such as that of Koppen and of Lehenbauer for temperature influence 

 upon plants is greatly needed for the corresponding influence of the 

 moisture conditions, but this sort of work has not yet been attempted, 

 even if it may have occurred to anyone. When such work is accom- 

 phshed (which will be possible only with good equipment for the general 

 control of environmental conditions), then it will be time to consider 

 efficiency indices of the moisture conditions in somewhat the same 

 way as we have attempted to deal with the suggested indices of 

 temperature efficiency. 



(B) PRECIPITATION. 



(1) Introductory. 



Since rainfall is so remote from being the immediate environmental 

 condition controlling the water-supply to plants in nature, the measure- 

 ment of this climatic condition must not be expected to show very 

 definite relations to plant activity or distribution. As has been indi- 

 cated, we employ rainfall data not because they are desirable, but 

 because they are the nearest approach to what is desirable that the 

 present state of our knowledge affords. In this case, as in that of tem- 

 perature, we usually employ the length of the average frostless season 

 as our duration factor. Since the effect of precipitation is markedly 

 cumulative, we have also tentatively established a second annual 

 period, which may prove to be more satisfactory for this condition 

 than is the length of the average frostless season. This period is 

 obtained by adding to the average frostless season, at its beginning, a 

 period of 30 days. By this scheme the rain falling during the last 30 

 days of the frost season is considered as pertaining to the following 

 frostless season. Thus the snow and rain of March is frequently very 

 influential in determining the kind of plant growth that can occur in 

 the following month, especially if the latter is comparatively without 

 precipitation. The length of this added period is taken as 30 days 

 quite gratuitously; perhaps it should be longer or shorter and it prob- 

 ably should have different lengths, according to other climatic condi- 

 tions, for different localities. At any rate, it has seemed desirable to 

 make test of this modification. In the discussions that follow we shall 

 let P represent the normal total precipitation for the period of the 

 average frostless season, while tt will represent the corresponding nor- 

 mal total precipitation for the longer period just described. 



It is obviously not to the point at all to employ the sunmaed pre- 

 cipitation for a portion of the year as a measure of the water-supplying 

 power of the environment available for plant growth during that 

 period. Rather is it requisite to study the average water-supplying 



