64 THE FRESH-WATER LOCHS OF SCOTLAND 
our observations on Loch Earn to follow continuously the minute 
variations of the atmospheric pressure with an ease and certainty 
hitherto unattainable.^ Also, in Part V. of my " Report on the In- 
vestigation of the Seiches of Loch Earn by the Scottish Lake Sui'vey," 
the mathematical theory of the effect of pressure disturbances of 
various kinds on an ideal lake, of form not very remote from Loch 
Earn, has been worked out,^ so as to show that the usually assigned 
cause of seiches, viz. the minor local fluctuations of the barometric 
pressure, is in reality sufficient to cause the disturbances observed, 
and is not a negligible quantity on ordinary lakes, such as the tidal 
action of the moon can be shown to be.^ 
Regarding those causes of seiches which have never yet been 
proved to be other than accidental, it may be of interest to record the 
fact that during our observations, viz. on 21st September 1905, at 
^gh 3gm^ were favoured with what Dr C. Davison,^ in a paper on 
the Ochil earthquakes, calls a " principal earthquake."' The esti- 
mated duration of the shock was 3"4 sec. Some of the members of 
my family who heard it, took it for the rumble of a train passing at 
an unusual hour on the opposite side of Loch Earn. The centre of 
disturbance seems to have been about 19 miles S. 39° E. of St Fillans, 
and the normal to Dr Davison's isoseismal 4 makes an angle of about 
63° with the axis of Loch Earn. At the moment the waggon 
recorder was not working, but the converted Sarasin at the binode 
was running at high speed (158 mm. per hour) and giving a smooth 
trace. The circumstances were as favourable as could be conceived 
for showing any seiche disturbance due to the earthquake, but none 
can be identified. There is, of course, no reason to expect that the 
rapid oscillations of ordinary earthquakes could cause seiches. Still, 
negative evidence in special cases is not without value, because in ex- 
ceptional cases, such as the Lisbon earthquake of 1755, seiches have 
been produced, and we do not as yet know the reason why. 
Observers are now agreed that the development of seiches usually 
accompanies, local disturbances of the barometric pressure whose 
duration if they are transitory, or period if they are periodic, does not 
differ greatly from the period of the seiche in question. The observa- 
tions on Loch Earn fully bear out this conclusion. Disturbance on 
the microbarogram is always accompanied by disturbance on the 
limnogram, although the magnitudes do not always correspond. 
1 A separate account of the observations with the microbarograjDhs has been 
published in the Proc. Roij. Soc. Edin., vol. xxviii. p. 437, 1908. 
2 Trans. Roy. Soc. Edin., vol. xlvi. p. 499, 1908. 
3 Except in very large lakes, such as Erie. See Endros, Petermann's Mitt., 
Heft ii. p. 16, 1908. 
4 Quart. Journ. Geol. Soc, vol. Ixiii, p. 366, 1907. 
