12 ROYAL SOCIETY OF CANADA 
These results were based partly on tidal information obtained during 
the Admiralty surveys and partly on observations arranged for by 
correspondence, before the Pacific coast was visited (15). 
In 1905, several stations were established along the coast; and on 
reducing the results, it was found possible to divide the whole coast 
into three regions; namely, the West coast of Vancouver island, the 
region of the Strait of Georgia, and the Northern coast from Vancouver 
island northward. Principal stations are maintained in each of these 
regions; in Clayoquot sound, at Sand Heads, and at Port Simpson. 
The cities on the coast are quite unsuitable as ports of reference; but 
tide gauges are maintained at Victoria, Vancouver and Prince Rupert, 
and tide tables are specially calculated for these harbours, on account 
of their intrinsic importance. 
As a rule, the tide on the Pacific coast is strongly influenced by 
the declination of the moon, and it is also subject to an annual vari- 
ation with the change in the declination of the sun. On the open coast 
the spring and neap tides are quite distinguishable, notwithstanding 
the other inequalities. In the Strait of Georgia the diurnal inequality 
becomes so large as to dominate every other feature of the tide. Next 
in importance is an annual variation, as the influence of the sun is 
very great relatively to the lunar effect. The turn of the current in 
the Passes is similarly affected, as the relation between the time of 
slack water and the tide shows amarked annual variation. To obtain 
good comparisons, it is thus desirable to have a full year of simultaneous 
observations. If this is not possible, at least six months are necessary; 
or else the comparisons with the reference station must be made about 
the time of the Equinox. 
Variable Tidal Differences.—The use of variable tidal differences 
is very valuable in extending as far as possible the region that can be 
referred to each of the principal stations, and thus avoiding the need 
for an additional reference station. As arule, the variation is so largely 
in some one period, that all others can be neglected; but even if there 
are two periods involved, their laws of variation can be determined 
for calculation purposes. 
On the Pacific coast generally, the annual variation in the tidal 
difference, as already explained, is the only one necessary to take into 
account. For, in most cases, the diurnal inequality is so much the 
same, both at the principal and secondary station, that it does not 
cause variation in the difference. 
In the Strait of Georgia, where high water and the half tides are 
so nearly at the same level, it is only the difference for lower low water 
that is affected by variation. For example, on the Fraser river, from 
Sand Heads at its mouth to New Westminster, the difference in time 
