68 



EFFECTS OF WINDS AND OF 

 TABLE No. 14. 



Each least-square solution gives two values for the lag in each barometric 

 effect, E-W and N-S, as noted on page 22. The discrepancy in each pair 

 between the two computed values of the lag is a test of the accuracy of the 

 adopted value of the lag. Table No. 15 gives the discrepancies of that 

 character arising from the final solutions for barometric effects. The lags 

 were computed by use of table No. 2, as indicated on page 22. 



In table No. 15, of the 10 possible discrepancies, 3 did not develop, since 

 there was a failure to get a determination of the lag, as indicated in the 

 footnote. 



Of the remaining 7 discrepancies, 5 range from 2.0 hours to 7.0 hours, and 

 2 are 0.2 hour and 0.1 hour, for the lag in C n at Milwaukee and Buffalo, 

 respectively. 



Note that by comparison with table No. 7, page 33, the two pairs 

 of determinations of lag which gave the very small discrepancies noted at 

 Milwaukee and Buffalo were for very large values of C n , +8.44 and 15.62, 

 respectively; also, that in the case in which both possible determinations of 

 lag failed for C n at Harbor Beach, the value of C n involved is very small, 

 -0.44. 



In general, then, it appears that the lag in the barometric effects is 

 apparently determined within less than an hour for the largest effects, is 

 subject to a possible uncertainty of several hours for the very smallest 

 effects, and for the effects of intermediate size the chances are about even 

 that the error in the lag is only one or two hours. 



When one notes how slowly the barometric effects change from hour 

 to hour it appears that such determinations of lag represent a rather high 

 degree of accuracy in the determination of barometric effects. At Buffalo, 

 where C n is 15.62, the largest value of the kind encountered in this in- 

 vestigation, the maximum change in the N-S barometric effect in one hour 

 found in the computations was 0.07 foot. The discrepancy between the 



