78 THE FRESH-WATER LOCHS OF SCOTLAND 
T = 15'6™, and there is an alteration of the UB from 2CJ = 3-7 mm., 
2B = l-7 mm. to mm., 2B = l-0 mm. The dicrote then 
remains steady until 22^^ on the 25th, when it undergoes a sudden 
disturbance, which rapidly destroys its configuration. This sudden 
disturbance and the almost total destruction of the seiche about 
5 hours later are difficult to explain by the meteorological conditions, 
unless they were due to variations of the wind. 
6. Effect of the Impact of Wind-Gusts. — Inasmuch as a wind 
velocity of 10 (mile/hour) is calculated to produce a pressure of about 
1*5 mm. (Aq.) by direct impact on a small area, it is reasonable to 
expect that the impact of wind-gusts, especially in the case of lakes 
enclosed by high hills, may at times cause seiches. There are, how- 
ever, various difficulties in obtaining data on the subject. It is 
difficult to determine the angle of impact of the wind-blasts. Then 
it is uncertain whether the wind ever falls at the same angle and at 
the same time over large parts of the surface of a lake. The appear- 
ance of the lake-surface on windy days very often suggests the contrary. 
What we frequently see are patches of wind disturbance progressing 
over the lake-surface with varying velocities. 
Then again it is difficult to separate the effect of wind impact 
from the disturbances of the ordinary barometric pressure which 
always accompany high winds. 
It has not been possible to deduce any definite results from the 
observations under the present head. 
7. Effect of Periodic Fluctuations of the Atmospheric Pres- 
sure. — The observations on Loch Earn afforded many examples of 
this cause of seiches. It must, however, be understood that strictly 
periodic fluctuations of the barometric pressure of short period rarely 
if ever occur. There are often, however, fluctuations extending over 
an hour or two in which the undulations are approximately of equal 
length, and still oftener two or three consecutive undulations of 
approximately the same length. Such fluctuations are described in 
what follows as periodic, and by the " period " is meant the average 
of the intervals between the passage of corresponding phases (say 
maxima) of two successive undulations at the same point. 
It follows from theory, and is confirmed by observation, that a 
periodic disturbing cause is most effective when its period is not very 
different from that of the seiche in question. In practice, however, 
the disturbing; effect is considerable even if there is considerable 
divergence between the two periods. It should also be noticed that, 
even theoretically, if we consider only one or a limited number of 
oscillations, and neglect the viscosity, the maximum effect does not 
correspond to exact equality of the two periods.^ 
1 See Trans. Roy. Soc. Edin., vol, xlvi. p. 514. 
