54 THE FRESH- WATER LOCHS OF SCOTLAND 
expected a priori that these would not greatly affect the periods, 
at least those of lower nodality ; but it is obvious from the Hydro- 
dynamic Theory itself that such deviations near to nodes might 
considerably affect their position. 
To give an idea of the accuracy that is possible under the most 
favourable circumstances in determining the periods of a lake, I 
subjoin some of the tables that were used in determining the first 
three periods of Loch Earn. The station, the instrument, and instru- 
mental adjustment are the same in each table, but not the same in 
any two tables for the same period. In the second column of each 
table is given the height of the surface of the lake above a certain 
arbitrarily fixed level. In calculating the mean from each table the 
EiG. 19. 
weight of each value of the period was taken to be proportional to 
the number of oscillations used in calculating that value. 
Naturally the period most accurately determined is the uninodaL 
It will be seen that the mean of the results from Tables L, II., and 
III. is 14*524, which differs by less than '005 from any one of the 
three. Probably, therefore, the determination of the uninodal period 
of Loch Earn differs from 14 '52 by less than 1 in 1452, say '07 per 
cent. The extreme accuracy of this determination of the uninodal 
period of Loch Earn, and incidentally also of the binodal period, is 
due to the great regularity and persistence of the seiches of this lake, 
and to the fact that the ratio T^^/T^ is very nearly equal to 9/5 ; so 
that there is a well-marked " configuration period " in the dicrote 
of the uninodal and binodal seiches, which can be utilised in the 
determinations. 
