22 



EFFECTS OF WINDS AND OF 



Similarly, by following through the reasoning corresponding to lines " 5 2 , 

 4 k Lag/ J "B z , 4 h Lag," and " B , 4 h Lag" on plate 3 it may be shown that 



T> _ '" ft /OQ\ 



B W-LW (33) 







<-* * 

 24 



7? = i 

 _. . 



(34) 



(35) 



when the lag is 4 hours. 



The various values of B Wl , B m , . . . corresponding to a lag of 4 hours 

 are shown in the seventh line of table No. 1. 



The tabular values are B m or B w etc., in terms of C w or C n expressed in 

 units of -2*4 C w or ^ C n . 



The various lines in table No. 1, for various assumed values of the lag, 

 have each been computed by the method illustrated by the examples which 

 have been given. 



Table No. 1 makes it clear that the values of B wa , B 



Wll 



D n .,. B r , 



depend upon the actual lags and upon the actual values of C w and C n . If 

 observation equations of the form shown in (30) are used and the values of 

 BU, BW> ' B B n , are derived from the least-square solution, the lag 

 may be determined from these values by the use of table No. 2, which has 



TABLE No. 2. 



