MANUAL OF TIDE OBSERVATIONS 71 



(b) Sum separately the times of high and low waters, limiting the 

 number of items in each case to two less than twice the number of 

 days in the month, following the method described for the summa- 

 tion of the moon's transits. From the sum of the time of the high 

 waters subtract the product of 24 hours by the sum of the numerals 

 indicating the days of the month on which only a single high water 

 occurred, and from the sum of the times of the low water subtract 

 the product of 24 hours by the sum of the numerals indicating the 

 days of the month on which only a single low water occurred. Des- 

 ignate these results by H and L, respectively. 



(c) Find the differences (H-T) and (L— T) and divide by the 

 number of items included in each summation, which should be twice 

 the number of days in the month less two. Such multiples of 12.42 

 hours may be applied or rejected from the means as may be necessary 

 to reduce them to positive values of less than 12.42 hours. The 

 results obtained should check very closely with unreduced means as 

 obtained directly from the individual intervals, although at times 

 there may be a small difference in the second decimal place. 



232. Example. — This method for checking the mean intervals is 

 illustrated below for Seattle, Wash., for the month of January 1928, 

 these intervals having been computed in the regular manner in fig- 

 ures 24 and 25. This being a 31-day month, the number of items 

 included in each summation must be 60, 



Sum of times of 60 transits, January 1928 727. 6 



24X (7-1-22), single transits occuring Jan. 7 and 22 696 



Difference (T) 31. 6 



Sum of times of 59 high waters, January 1928 665. 7 



Time of high water Dec. 31, 1927, for deficiency 23. 2 



Sum of times of 60 high waters 688. 9 



24 X (1+15-1-30), single high waters occurring Jan. 1, 15, and 30 1, 104 



Difference (H) - —415. 1 



Sum of times of 60 low waters, January 1928 745. 6 



24 X (9-|-24), single low waters occurring Jan. 9 and 24 792 



Difference (L) —46. 4 



(H— T)H-«0=— 446. 7H-60=— 7.44 hours for high- water interval. 

 (L—T) ^-60=— 78.0^60=— 1.30 hours for low-water interval. 



Applying the tidal period 12.42 hours to each of the above negative 

 intervals, we have 4.98 and 11.12 hours, respectively, for the unre- 

 duced high and low water intervals. It will be noted that these 

 results agree exactly with those obtained in figure 25 by the regular 

 computations. An exact agreement, however, is not always to be ex- 

 pected, but the results will usually agree within 1 or 2 hundredths of 

 an hour. 



233. Lunitidal intervals from Form 248. — For a short series of 

 observaticms the lunitidal intervals can be computed in Form 248 



