Breaks in Tide Curves from Recording Gauges 79 



less amplitude. By using some judgment in this matter, the system of 

 computation need only be used for certain selected tides; which will 

 save labour. 



Summary. — The general procedure in making the interpolation for 

 each missing day may be thus summarized: 



Trace the tide curves from the record obtained with the registering 

 tide guage, 29 days before and 29 days after the missing day; the two 

 curves to be placed on the same sheet, graduated for time and height. 

 Draw an interpolated curve precisely midway between these in respect 

 to both time and height. 



Before accepting this interpolated curve as correct for the missing 

 day, the points of high water and low water are to be checked. If the 

 diurnal inequality is large at the date dealt with, compute the time and 

 height of the Higher high water and the Lower low water by the system 

 described; basing the computation on readings from the recorded tide 

 curves. The points thus computed, to be plotted in position; and the 

 interpolated curve to be modified to pass through them, tangentially to 

 the horizontal. This procedure will show whether or not it was correct 

 to assume that the interpolated curve was midway between the original 

 pair, especially in places where the pair are somewhat widely separated. 



When it is suspected that any element in the tide still remains 

 unbalanced, because of the astronomical conditions at the time, the 

 Tide Table values at high and low water may be plotted ; and the similar 

 points on the interpolated curve can be compared with these, to see 

 that they have the same relative positions with respect to the high and 

 low water points on the recorded curves. 



When these modifications and comparisons are made, and found 

 satisfactory, the interpolated curve can be accepted as finally correct. 

 The mxthod here described, should apply equally well to other types of 

 tide;' and it should often enable a large part of a year of valuable tidal 

 record to be saved, when interruptions occur, and utilized for harmonic 

 analysis. 



Supplementary Note. — It has been assumed that the interruption 

 requiring interpolation occurs after Tide Tables are available from ana- 

 lysis of the first year or two of observation. But if the interruption 

 occurs a few months after observations are begun at a new locality, as 

 for example in the first winter after a good summer season, this method 

 of interpolation is still applicable. The only disadvantage is the loss 

 of the check for outstanding errors due to any astronomical conditions 

 that remain unbalanced, when these are not negligeable in amount. 



