64 ROYAL SOCIETY OF CANADA 



The above method of obtaining the range in these dirt'erences is the 

 fairest that can be used. It would not have given as fair a comparison, 

 if several of the highest and lowest values were omitted, as was done for 

 the Lower St. Lawrence and the Strait of Belle Isle, where the observ- 

 ations extend over a whole year, and thus include the winter season ; 

 and where the wind influence is so great. It is also to be noted that the 

 average of a larger number of high and low différences, or any other 

 reasonable method of determining the limiting values, would only change 

 the result by a few minutes ; as the differences range fairly between the 

 above limits. 



It is evident from these comparisons that Pictou is the best port of 

 reference for Northumberland strait. The range in the difference with 

 both Halifax and St. Paul island is there much less ; and a much greater 

 proportion of the range is due to diurnal inequality than in the case of 

 Charlottetown ; and as this can be reduced to law, the outstanding errora 

 will be less. Pictou is also more centrally situated, and the variation in 

 the diurnal inequality along the strait will therefore be better divided, 

 and leave less residual error, if the differences are taken in the two direc- 

 tions from there. It will probably be found also to stand in better rela- 

 tion to the tidal currents in the strait, when these come to be examined 

 systematicall}'. 



The tidal stations selected, enable comi)arisons to be made with 

 Pictou for the two ends of the strait, at Souris and Cape Tormentine, or 

 as far as the tides have a marked range. The differences given are the 

 averages of the three highest and three lowest values, and are in 

 absolute time as before. The observations at Souris extend from July 

 8th to September 2nd, and from September 16th to November 15th, 

 comprising in all 192 tides. At Cape Tormentine the cbmpainsons 

 extend from July 26th to 31st, August 4th to 0th, and September 5th to 

 30th ; giving in all comparisons for 70 tides. The results are as follow.s, 

 Charlottetown being repeated for convenience in the comparison : — 



Souris tu Pictou. Distance (io miles. 



Difference in the time of H. W. ranges from Oli. 16m. to 2h. ()8ni. 



Range in the difference. Ih. n2m. Aveiage difTerence = lh. 14m. 



Pictou to Cape Tormentinf-. Distance 70 miles. 

 Difference in time of H. W. from 2.5m. earlier, to .57m. later. 

 Range in the difference, Ih. 22m. Average difference=Oh. 24in. 



Pictou to Charlottetoivn. Distance 50 miles. 



Difference in time of H. W. from Oh. 19m. earlier, to 2h. 48m. later. 



Range in the difference, .Sh. 07m. Average difference=lh. 02m. 



The maximum outstanding error along the strait will not exceed 

 the semi-range in the above differences which may occur at certain times 

 during the month. In the case of Charlottetown, this error is the 

 greatest ; and the irregularity in so short a distance may be due to some 



