XXII ROYAL SOCIETY OF CANADA 
results of a satisfactory degree of accuracy can be secured, and the mul- 
tiplication of the more expensive permanent tidal stations may be 
avoided. As another example, it has been found in Northumberland 
Strait that the leading variation in the difference follows the declination 
of the moon, or the period of the nodal month. The variations in 
terms of the phases and the distance of the moon are relatively small. 
Again, in the Bay of Fundy, the moon’s distance has as much effect on 
the height of the tide as its phases; that is to say, the difference in the 
range of spring tides at Apogee and at Perigee is as great as the differ- 
ence between the mean range of neap tides and the mean range of spring 
tides, the amount in each case being close upon 11 feet. These varia- 
tions are detailed in the present report. It is thus evident that each 
region must be investigated for itself, to determine the period in which 
the dominant variation takes place. 
By this general method, the characteristics of the tide in any par- 
ticular region are first allowed for, in the tables calculated astronomi- 
cally for the principal station. The tides for other localities in the 
region are then calculated by means of a variable difference in time, 
from this principal station; and it has usually been found that a varia- 
tion in some one period includes so large a part of the whole, as to give 
a result sufficiently accurate for praetical purposes. In one instance 
above cited, however, a further plus and minus correction was used in 
a second period over-running the first, to allow for another variation 
next in importance. With such variable differences, a result that is 
practically correct is secured, which would not be the case with the old 
method of using a constant difference. Formerly, the discrepancy was 
often large, as in Northumberland Strait, where the time of the tide 
as found by a constant difference from an Atlantic port was as much as 
14 hours early or late at certain parts of the month. 
The tide at Quebec was also computed formerly by a constant dif- 
ference in time from London, England. This may serve to indicate the 
substantial improvements in accuracy already secured by this Survey. 
The total expenditure on this Survey during the fiscal year from 
June 30th, 1900, to June 30th, 1901, was $7,060.20. This total expen- 
diture is classified as follows :— 
(1) General expenses; maintenance of the seven principal tidal 
stations, with repairs, heating, and supplies; salaries of observers and 
assistants ; office work and travelling expenses, $2,910.35. 
(2) Summer tidal stations, on the Lower St. Lawrence in 1900, 
and in Northumberland Strait in the early part of the season (up to 
June 30th) 1901; erection of gauges, salaries of observers, and inspec- 
tion, $1,503.25. 
