Frosts in New York 533 



Plateau division, and the frost records for observation stations in the East- 

 em Plateau appear on pages 536-539 for spring and fall, respectively. 



Explanation of table 



The record given in each case covers the entire period for which authen- 

 tic information is available. When no specific reference to killing frost 

 was made in the record, the last occurrence of a freezing temperature in 

 the spring and the first in the fall were considered as equivalent to killing 

 frosts. 



As an example, take the spring frosts at Auburn, Cayuga county, in 

 the Central Lakes division (page 536). The data in the first four columns 

 need no explanation. In the columns that follow, the figures represent 

 the number of years in which the last killing frost in spring occurred 

 within the dates given at the head of each column, which are five-day 

 periods. In the column headed "15 or before " under April, appears 

 the figure i, which means that there was one season in the 14 years of 

 record when the last killing frost occurred on or before April 15. Under 

 May, in the column headed " 1-5," is the figure 4, which means that 

 in 4 out of the 14 seasons the last killing frost occurred between May i 

 and 5, inclusive. The sum of the figures in the columns under May is 

 10, which means that in 10 out of 14 seasons killing frosts occurred after 

 May I. 



The record of the first killing frosts in the fall is tabulated in a similar 

 manner. The figure i, in the column headed " 21-25 " under September, 

 means that killing frost occurred within the five days, September 21 to 

 25 inclusive, in i out of 14 seasons. The sum of the figures in the columns 

 preceding any date gives the number of seasons out of the years of record 

 in which killing frosts occurred previous to that date. Thus, the sum of 

 the figures preceding October 11 is 7, which means that killing frosts 

 occurred before October 11 in 7 out of 14 seasons. 



Practical application oj frost records 



The character of the seasons at a given place as shown by actual records 

 of past years, is the best guide we have to the character of the seasons 

 at that place in future years. 



Consider, for example, the frost record of Wedgwood, Schuyler county 

 (pages 536-537), with a view to detemiining the most favorable time for 

 planting com in that vicinity. By consulting the record of frosts at 

 Wedgwood it is found that killing frosts occurred after May 20 in 5 out 

 of 22 seasons, after May 25 in i out of 22 seasons, and that there were 

 no killing frosts after June i. If, during the 22 years, com had been 

 planted so as to be up on May 20 it would have been frosted in 5 out of 



