328 ENVIRONMENTAL CONDITIONS. 



on plate 53. There is also apparent here a small locahzed area with 

 ratio values of 1.00 or 1.03, including Topeka, Kansas, and Lamar and 

 Springfield, Missouri, but this deserves no special attention. It is 

 especially interesting to note, as will need to be done also in connec- 

 tion with a number of the following charts, that the line for value 1.00 

 apparently leaves the mainland in the middle of the New Jersey coast, 

 returns at the western edge of Connecticut, leaves it again south of 

 Boston, and finally reenters at the southwestern extremity of Maine. 

 This brings it about that the eastern portion of the humid zone is 

 divided into two parts, a northern one extending from Minnesota to 

 northern New England and an eastern and southern one extending 

 from about Boston or New York to the Rio Grande and beyond. This 

 feature appears to be an important one, and it will receive more atten- 

 tion later. 



(3) Ratios of Total Precipitation for Period op Average Frostless Season, for the 

 Year July 1887 to June 1888, to Total Evaporation as obtained by Russell 

 for the Same Period and Year (P/^. (Table 18, Plate 58.) 



The method of deriving the moisture ratios that has just been des- 

 cribed involves the use of normal precipitation data along with evapora- 

 tion data for the single year of Russell's study. It was thought that 

 this would give to the derived ratios somewhat more of the character of 

 normals than would be the case if both precipitation and evaporation 

 data had been taken for the single year in question, but the novelty 

 and great promise of this climatic ratio render it worth while to present 

 the values obtained by the latter method. These are given in table 

 18. To make comparison easier, the ratio values from table 1 1 are here 

 repeated, in the last column. The second column of table 18 gives the 

 total precipitation for the period of the average frostless season, 

 derived for the actual months of Russell's observations. His 12 months 

 have been treated as though they all pertained to the same calendar 

 year, and our usual method of approximating average frostless season 

 data from monthly data has been employed, the original monthly 

 data of precipitation for the year July 1887 to June 1888 being taken 

 from the Summary by Sections. We term this precipitation value p, 

 to distinguish it from P. The third column gives the corresponding 

 evaporation values (E) for the period of the average frostless season, 

 being taken from the sixth column of our table 11. The fourth column 

 gives the new ratio values (p/E). When the name of one station is 

 followed by that of another (the latter in parentheses), the evaporation 

 data are for the latter station and the precipitation data are for the 

 former. 



