TIDAL OBSERVATIONS. 11 
sheet of water analogous to the seiches by which the Swiss and other 
lakes are perturbed. 
These oscillations merely confuse the tidal phenomena, and it is 
desirable to eliminate them as far as possible. In order that it may 
be known to what extent they disturb the results at the chosen point 
of observation, it will be well that the series should begin with 
observations taken every four or five minutes during an hour or two, 
repeated at intervals twice or thrice during the first day or two. 
Since seiches always exist to some extent, each of the subsequent 
hourly observations should be the mean of several readings taken 
just before and just after the exact hour. It may be suggested that 
it would suffice to take the mean of five readings taken at intervals 
of two and a half minutes, from five minutes before the hour to five 
minutes after it. If, however, the preliminary examination suggested 
above should show that the seiches are apt to be considerable in 
amplitude, and to have a long period—such as a quarter of an hour to 
twenty minutes—it might be well to extend the period further before 
and after the hour. 
February 18, 1901. 
