184 



It is well to recollect that the results rest on the selection of the coeffi- 

 cient of friction, and that this depends upon the undeterminable 

 irregularities of the channel as actually excavated. Furthermore, the 

 actual currents will vary from day to day with the varying range and 

 even the varying form of the tides at the entrances. These variations 

 may be intensified by winds and storms, which may pile up the water 

 at one entrance and draw it away from the other. A precise deter- 

 mination of the currents and tides produced by a particular fluctuation 

 of the entrance tides is of academic interest only. The computations 

 call therefore for reliability rather than precision. 



353. Selection of representative entrance tides. — The computation of 

 the tides and currents in a canal is far too laborious to warrant repe- 

 tition for each successive tidal fluctuation at the entrances, even if 

 any useful purpose would be served thereby. Representative tidal 

 fluctuations at each entrance should therefore be selected from a study 

 of the actual tidal fluctuations during a month or more. 



If the tides at both entrances are of the semidiurnal type, with no 

 large variations between springs and neaps, and are not much de- 

 formed by overtides, the representative tide at each entrance may be 

 taken as a simple harmonic fluctuation with the speed of the M2 com- 

 ponent and the amplitude of the mean semirange of the tide. The 

 most convenient origin of time in this case is at a high water at the 

 initial entrance. The initial phase of the tide at this entrance is then 

 zero. The initial phase at the other entrance may be obtained from 

 the difference between the average lunitidal intervals at the two en- 

 trances, corrected, if necessary for the differences in longitude (par. 10). 

 The corrected difference, in solar hours, multiplied by the speed of 

 the M2 component, 28°98 per hour, gives the initial phase of the tide 

 at the far entrance. This phase is positive if the tide at the far en- 

 trance is the earlier, and negative if it is later than at the initial en- 

 trance. Published data on the lunitidal intervals at stations near the 

 ends of the canal may be based on such a limited number of observa- 

 tions as to have no great weight. If a reliable determination of the 

 lunitidal intervals is not available, the recorded differences in the 

 times at high water at the two entrances, and in the times of low 

 water, over a period of 29 days, or a multiple thereof, should be 

 averaged. A material discrepancj^ between the average difference in 

 the times of high water at the two entrances to the canal and the 

 average difference in the times of low water, indicates that overtides 

 are of sufficient importance to warrant consideration. 



354. If the tides at the entrances are of the same general type as 

 those considered in the preceding paragraph, but the daily tide curves 

 at one or both entrances are so distorted by overtides that they cannot 

 be represented satisfactorily by a simple harmonic fluctuation, either 

 average or composite tide curves may be prepared by the methods 



