CLIMATIC CONDITIONS OF THE UNITED STATES. 319 



with the corresponding precipitation means and the precipitation- 

 evaporation ratios for the 15-week period, the latter to be considered 

 below. Where the 15-week mean had been obtained from only two 

 averages this is indicated by the figure 2 in parentheses. 



Scrutiny of the means for this summer season shows them to vary, 

 with the geographic position of the stations, in an apparently rational 

 manner, and there is no reason to doubt that they furnish an approxi- 

 mately correct picture of the distribution of intensities of evaporation 

 over the United States for the period in question. There are two 

 marked exceptions to the last statement— the means for Bozeman, 

 Montana, and Salt Lake City, Utah, appear to be far too low, though 

 no explanation is at hand to account for this. The data for these two 

 stations have been ignored in the preparation of our chart. 



Plate 56 is a chart of evaporation intensities constructed from the 

 means just considered. The relatively few stations for which data 

 are available makes it unadvisable to attempt any detailed study of 

 climatic zones in this case, and isoatmic lines are shown for only 150 

 and 300 c.c.^ 



The most significant features of the chart of plate 56 are the follow- 

 ing: 



(1) The Canadian region of low summer evaporation (less than 150 

 c.c. per week) extends, as a great lobe, south westwar d from Lake 

 Superior, as far as the valley of the Arkansas River. Another southern 

 extension of the northern area of low summer evaporation intensities 

 reaches south-southwestward from southern New England, occupying 

 the whole of the Appalachian Mountain system south of Massachu- 

 setts. This area broadens toward the south and embraces most of 

 South Carolina, Georgia, Alabama, Mississippi, and eastern Kentucky 

 and Tennessee. Miami, Florida, northern and western Washington, 

 and northwestern Oregon he in the region of low intensity, as must 

 also the high altitudes of the Rocky and Sierra Nevada Mountains 

 though our numerical data do not show this feature. 



(2) The main region of high evaporation intensity (over 300 c.c. 

 per week) extends northward from Mexico and occupies western Texas, 

 New Mexico, and Arizona south of the plateau, and the lower altitudes 

 of Nevada and of southeastern CaUfornia. This is obviously the 

 so-called desert or arid region of the United States, and corresponds 

 to the arid evaporation province as shown on plates 53 to 55, and in 



1 It is to be remembered that the numerical data are in terms of cubic centimeters of weekly loss 

 from the Li^'^ngston standard cylindrical porous-cup atmometer, exposed as were our instruments. 

 They are to be regarded merely as comparable indices of atmospheric evaporating power with 

 reference to this instrument, just as are numerical data in terms of depth of water-loss from 

 some specified water-surface exposed in a specified manner. Prevalent ideas in this connection 

 require repeated emphasis upon the fact that rates of water-loss from one form of atmometer 

 can not be mathematically deduced from those obtained with another form of instrument, ex- 

 cepting in a very general way. If this can be generally appreciated it will aid much toward 

 atmometric progress. 



