CLIMATIC CONDITIONS OF THE UNITED STATES. 383 



employed. At best, they are perhaps indications as to the nature of 

 better-grounded generahzations that are to be developed in the future, 

 from more satisfactory data, and by means of more adequate methods 

 than are now available. The conditions will be considered in their 

 order of presentation in the preceding sections. 



A. TEMPERATURE CONDITIONS. 



All of our temperature charts, whether for duration or intensity 

 of the temperature conditions, represent the area of the United States 

 as divided into temperature zones or provinces, these having a gener- 

 ally west-east direction, but being distorted more or less by mountain 

 systems and proximity of the ocean. These zones result from arbi- 

 trary divisions, depending on selected ranges or amplitudes of the 

 index values concerned, but, since the climatic characters are geo- 

 graphically continuous from province to province this is the only 

 method by which they may be profitably studied. For general pur- 

 poses it has seemed desirable to recognize five temperate temperature 

 provinces in the United States, which we have termed: very warm, 

 warm, medium, cool, and very cool. This terminology may be applied to 

 all our temperature charts, but it is necessary to name the temperature 

 index employed in each case and also the index amplitude representing 

 each province. Thus, the warm province based on length of the 

 average frostless season is not exactly coextensive with the province 

 of the same name based on the physiological summation index, etc. 



Special emphasis should be placed on the length of the average frost- 

 less season as an index of temperature duration. It has proved to be 

 of great value, not only as a temperature index per se, but also as a 

 duration factor for intensity indices of both temperature and moisture 

 conditions. This promises to be one of the most useful temperature 

 indices for use in ecological climatology, although it has not yet 

 attracted the attention that it deserves. 



Other duration indices of temperature that may prove valuable are 

 (1) the length of the period of high daily normals and (2) the length of 

 the period of low daily normals. 



The most promising intensity index of temperature conditions 

 appears to be that of the physiological summation for the duration of 

 the average frostless season, as devised by Livingston, but much more 

 physiological study will be required before this index can be regarded 

 as established. As here employed, however, this index has proved to 

 be very satisfactory in many ways. Absolute temperature minima 

 have also proved to be valuable as intensity indices. We have also 

 employed the average daily normal for the coldest 14 days of the year, 

 as well as Merriam's chart of the mean normal for the hottest 6 weeks. 

 The normal mean annual temperature is the most promising of the 

 various temperature indices for which values are directly available in 



