SEICHES AND OTHER OSCILLATIOKS 
49 
wave length, the result is that the one entirely destroys the other. In 
No. vii. the periods are equal, and the resultant has the same period, but 
a phase different from either of the components. This explains why 
a physical cause disturbing an existing seiche in a lake may in certain 
cases have the effect of altering its amplitude or its phase without 
affecting its period, or may destroy the existing seiche altogether. 
Nos. iv. and v. show the effect of superposing two seiches of the 
same amplitude, but of slightly differing periods. The result is a 
dicrote seiche which presents to the eye the appearance of a seiche of 
a single definite period but of periodically varying amplitude — a 
phenomenon analogous to the beats caused by two musical notes 
which are nearly but not quite in unison. 
If the periods of two components approximate to a simple 
Fig. 17. 
numerical proportion, say 9 : 5, as in the case of the uninodal and 
binodal seiches of Loch Earn, the result is a limnogram with a 
periodically recurring configuration like a wall-paper, the individual 
waves of which approximate to the waves of one of the two compon- 
ents if the amplitude of that component preponderates, but which 
fluctuates if the two amplitudes are not very different. It will be seen 
that the thick curves in Nos. i., ii., and iii. imitate very closely the 
smooth dicrote seiches reproduced in Nos. 1 and 2 of fig. 14, which were 
drawn one fine day near St Fillans by the unguided hand of Loch Earn. 
Conversely, these principles may be used in the difficult process of 
analysing an actual limnogram, so as to discover the periods of the 
components of the seiche which it records. At the bottom of fig. 17 
is reproduced part of a fine limnogram obtained by Mr James 
Murray from Loch Earn by a series of half-minute observations with 
an index limnograph, which extended over eight hours. By count- 
4 
