HAEMOIiTIC AlfALYSIS AND PEEDICTION OF TIDES, 87 



In which F^, for brevity, is substituted for the coefficient 



24 a 15vh 



T sin ^^ and gives the relation of the average B ordinate in- 



TT po 2a ^ ^ 



eluded in the A grouping to the true B ordinate for the time t 

 represented by that group. The reciprocal of this coefficient will be 

 that part of the augmenting factor necessary to take account of this 

 primary grouping. If the primary summing has been for the com- 

 ponent S, this coefficient may be taken as unity since the original S 

 sums refer to the exact S hour. 



When the secondary stencils are applied to the component A group 

 sums, the groups applying to an exact component A hour at smj 

 time t and represented by that time, will be distributed over an inter- 



15z>^ 

 val of a component B hour, or — r^ solar hours. 



For an integral component B hour at any time t within the middle 

 day represented by a seven-day page of original tabulations the limits 



of this intei'val will he it — ^r- ) and lt + -^|- j • For the same page of 



tabulations, letting t represent the same time in the middle day, the 

 limits of the group interval for the day following the middle one, are 



+ 1, +2, +3, respectively, for the seven successive days represented 

 by a single page of oiiginal tabulations, the limits of the group interval 

 for any day of the page may be represented by 



^ ^ SQOpn I5f\ ^^^ /^3Q0pn^l5v^\ 



^)and( 



2& J-^^y^ a ' 2h J 



Formula (332) gives the mean value of the B ordinate for grouping 

 of the A summations. The mean value of (332) obtained by com- 

 bining the grouf)s falling in any particular day of page of tabulations 

 in the limits indicated above is 



360pw 15pi 

 26 



COS {ht + ^) dt 



^ FB P* 



15y^ ^ I , 300p 



_3in(6( + ^ + 3J^-lf)] 



= F,F,Bcos(j( + ^ + ?^25) • (333) 



if we put F^ = J sin —^ for brevity. 



