BAROMETRIC PRESSURES ON THE GREAT LAKES 

 The lag in C w was found thus: 



-0.90 



33 



-2.92 . n _ , B 

 = +0.67 and 



-4.34 



B, 



-1.78 



= +0.50 



From the first of these, according to table No. 2, the lag is 8.2 hours, 

 and from the second it is 6 . 1 hours. The mean (7 hours) was adopted as 

 the most probable value of the lag in C w . 



Similarly, the two values found for the lag in C n were 4.3 and 4. 1 hours. 

 The adopted value was taken as 4 hours. 



The values of C w computed from formulae (36) and (37) were, respectively, 

 -1.78-2. 92= -4. 70 and -4.34-0.90= -5.24. The mean -4.97 was 

 adopted as the most probable value of C w . 



The two values of C n computed from formulae (38) and (39), respectively, 

 were +8.33 and +8.56. The mean +8.44 was adopted as the most 

 probable value of C n . 



The computations of lag and of C w and C n were made for all five stations 

 from the values of B WQ , B^, . . . B^, and B n ,, shown in table No. 6, with 

 the results shown below in table No. 7. 



TABLE No. 7. 



*Note that these three values of the lag are marked with minus signs, 

 used. They are anticipations rather than lags. 



They are to be so 



From these values the barometric effect at each hour at each station may 

 be computed from the following formulse, each of which is formula (18), page 

 17, modified (a) to adapt it to the particular lake, (&) to take into account 

 the lags which are now known, and (c) to adapt it to the units used above in 

 table No. 7 and in the observation equations. 



In the following formulse the quantities (6-8), (5-7), (4-5), and (3-6) are 

 expressed in units of 0.01 inch. The computed effect, E\, is obtained from 

 the formulae in units of 0.001 foot. 



For Buffalo, Ei, for any hour=(6-8) (+4.72) + (5-7) (-15.62), in 

 which (6-8) must be taken for 1 hour later and (5-7) for 3 

 hours earlier than the hour for which the effect is being 

 computed. 



(41) 



