148 



THE ROYAL SOCIETY OF CANADA 



tide at Cap à la Roche^ on the same day. (The angle d, in this last 

 case could only be obtained approximately, owing to a limitation in 

 the apparatus when large angles were used.) 



Other combinations of d and with a different scale of plotting, 

 were found also to give a good agreement, but this range in adjustment 

 has not yet been considered. It is, however, apparent that if we have 

 the tides at Levis, and also some previous sets for other stations, 

 together with the usual tidal information sufficient to fix an origin 

 with reference to time and height, that a method for predicting the 

 tides at the other stations and tracing the changes up the river should 

 be capable of development. 



Figure 5 



It was of interest to note whether any one or more of the constants 

 of projection could be associated with any physical tidaj constants, 

 and thus give this method a further advantage. The expression 

 y = f(ax+by) in relation to y = f(cx), is quite compatible with a 

 possible dynamical significance for the constants of projection in a 

 case such as that of our estuary tides. 



Figs. 5 and 6 show the results of one test which was performed 

 with this idea in view. The converse method of projection as in 

 Fig. 2, was, for convenience, adopted again. Six consecutive tides at 

 Quebec for June 7, 8 and 9, 1917, are given in Fig. 5. In Fig. 6, the 

 continuous line represents the projections of the successive tides at 

 the angles given below, and the broken line gives the corresponding 

 pure sinusoids in which r' varies from loop to loop. The following 

 table shows how adjustment can be made so that most of the vari- 

 ations can be associated mainly with a change in 0, and a change in 

 r'; A' and r' refer to the sinusoid produced by projection and are 

 analogous in meaning to A and r. , 



ipiaton, Richelieu and Cap à la Roche are respectively 30, 40 and 50 miles 

 approximately, above Levis. 



