TIDAL OBSERVATIONS. 9 
In harbours frequented by shipping, trustworthy tide-tables have 
a considerable commercial value, but in the case in point this element 
is, of course, entirely wanting, and the minuter details of the phe- 
nomena at the place of observation present but little interest ; what 
we wish to know, is the nature of the tidal oscillation affecting the 
whole ocean. 
The construction of accurate tidal predictions demands a con- 
tinuous series of observations extending over several years. Such 
extensive data are needed, not on account of the complexity of the 
tidal oscillations themselves, although they are pretty complicated, 
but because it is only possible to eliminate the incessant meteoro- 
logical perturbation of the tide by considering the average as 
estimated over a long period of time. In the case of the Antarctic 
Ocean these prolonged observations are neither practicable nor 
desirable. It must not, however, be concluded that very rough 
observations will suffice, for accuracy is needed for the determination 
of the main outline of the phenomena, and duration of observation is 
a necessary condition for even the approximate elimination of the 
varying effects of wind and barometric pressure. It should also be 
noted that observation extended over two months will give con- 
siderably more than twice the accuracy attainable by the results for 
one month. 
Tidal observations are necessarily conducted from the shore, and 
the existence of land exercises a perturbing influence on the oceanic 
tide which it is our object to examine. Hence, on this account alone, 
it would be highly desirable that the observations should not be 
confined to a single station. But, further, since the tides depend on 
latitude, it would be of especial value that observations should be 
made at such a place as Kerguelen Island, and also on the Antarctic 
Continent itself. The tides on the continent are more likely to 
suffer perturbation from the existence of land, than those at so 
isolated a station as Kerguelen Island. Hence it is eminently to be 
desired that the observations in the extreme south should be taken in 
two or more places. 
The multiplication of stations and the prolongation of the period 
of observation present conditions which are obviously incompatible, 
and their rival claims can only be decided as a compromise. It may, 
I think, be stated that the shortest period during which observations 
should be made is a fortnight, but that a month is a much more 
desirable period. The naval requirements of the Expedition will 
probably give the observers but little choice as to the number of 
stations and the periods of observation. I would then merely advise 
