326 ENVIRONMENTAL CONDITIONS. 



characterized as follows: The arid province extends northward from 

 Mexico and occupies the Great Basin. The semiarid province is shown 

 as a belt Ijang outside of the arid area, this belt not being wide enough 

 to reach the Pacific at any point, but extending into Canada at the 

 north and also extending eastward, from Nebraska and Oklahoma, 

 in the form of a large lobe that reaches nearly to the Appalachian 

 Mountains. The semihumid province occupies a narrow belt outside 

 of the semiarid one, but not reaching to the Pacific coast (on plates 

 53 and 55, at least). This belt broadens at the north, extends into 

 Canada, and lies just outside of the great eastern lobe of the semiarid 

 province. It includes much of the Gulf and southern Atlantic region, 

 a feature that is quite unlike the corresponding state of affairs on the 

 precipitation charts. The humid province occupies a small portion 

 of the northwest (reaches southward along the Pacific coast on plate 

 54, fig. 14), extends into Canada at the north, reenters the United States 

 west of the Great Lakes, and occupies most of Minnesota, Wisconsin, 

 Michigan, New York, and the three northern States of New England. 

 Another portion of the humid province lies along the Atlantic coast, 

 from Massachusetts (or southern New Jersey, see fig. 14) to North 

 Carolina, and a narrow strip appears on the eastern coast of Texas. 



C. RATIOS OF PRECIPITATION TO EVAPORATION. 



(1) Preliminary Considerations. 



Following the lead of Transeau (1905), we have employed the ratio 

 of precipitation to evaporation as the nearest approach that is as yet 

 possible toward an ideal index of the external moisture-relation of 

 plants. Transeau's introduction of this ratio marked a very definite 

 and important forward step in climatology, which will, no doubt, be- 

 come more thoroughly appreciated as data of evaporation become 

 available. Of course this ratio, like other intensity factors, may be 

 employed in connection with any duration factor or factors that may 

 seem desirable, and we have thus employed it in six different ways. 

 As has been said, however, it is not yet possible to obtain evaporation 

 normals, and all that can be done is to use Russell's data for the single 

 year, July 1887 to June 1888. Transeau obtained his ratio by divid- 

 ing the normal annual precipitation, for each station considered, by 

 the corresponding total evaporation for Russell's year of evaporation 

 data. Our various applications of the Transeau ratio will receive 

 attention below. 



(2) Ratios of Normal, Total Precipitation, for Period of Average Frostless Season, to 

 Total Evaporation as obtained by Russell for the same Period, for the Year 

 July 1887 to June 1888. PjE. (Table 11, Plate 57, Fio. 16.) 



The two terms of this ratio, for each of the stations included in our 

 list, are given in the second and sixth columns, respectively, of table 

 11, and the ratio values themselves (P/jE) are given in the eighth column 



