144 



THE ROYAL SOCIETY OF CANADA 



Figure 1 



in their production were re-obtained by calculation to within 30' of 

 the correct value, and A to within -05 cm. These errors could 

 obviously be reduced with a finished piece of apparatus. 



The question whether it is of interest to treat estuary tidal 

 curves by this method of analysis depends upon the closeness with 

 which a desired curve may be obtained, upon the number of constants 

 involved, and upon the character of the parent curve. It will be ob- 

 served in the subsequent figures that the lack of agreement usually 

 represents differences only of the same order as those which may be 

 the result of wind or local pressure disturbances. The long series of 

 constants which arise in the harmonic method, and which are quite 

 arbitrary in the case of estuary tides, are absent; and it is apparently 

 possible to find suitable parent curves. 



In some of the estuary tidal curves it is difhcult to determine a, 

 h and e. owing to a lack of symmetry, but if they are chosen so that 

 the high water and low water points are correctly derived, a fairly 

 satisfactory projection usually may be obtained. 



In the consideration of this method of analysis it should be noted 

 that no criticism of the ordinary harmonic method is implied for those 

 legitimate applications where each constant refers to a given physical 

 component, as in the case of open sea tides (or, for example, sound 

 waves) . 



