126 THE FRESH- WATER LOCHS OF SCOTLAND 
continuity is abrupt,^ the period of such a seiche is given by the 
formula 
V g(p-p') 
where t is the period of the seiche, I the length of the loch, p and p 
the densities of the cold and warm layers respectively, and h and Ti 
the depths of the cold and warm layers respectively. 
But even when the discontinuity is not very abrupt, a tempera- 
ture seiche may take place. Fig. 49 gives the observations made 
at Fort Augustus, on Loch Ness, at the surface and at depths of 
50, 100, 150, and 200 feet, during July and August 1904. During 
August observations were made every two hours. Time is measured 
along the abscissa axis, and temperature along the ordinates. The 
first curve represents observations made at the surface, the second at 
50 feet, the third at 100 feet, the fourth at 150 feet, and the fifth at 
200 feet. It will be seen that in the month of July changes in 
temperature which occur at the surface, due to changes in wind, 
etc., are traceable to all depths. But in August, when the discon- 
tinuity has become marked, the changes at and below 100 feet are 
independent of the changes at the surface, and have roughly a period 
of three days. A rough calculation of the period of a temperature 
seiche from the above formula gives a value for the period of the same 
order of magnitude. With a view to further test the theory of a 
temperature seiche, as explaining the oscillation in temperature 
observed below the discontinuity, observations were made simul- 
taneously at both ends of Loch Ness, and near its centre. The 
observations at the centre did not show oscillations of a period of 
three days, but they showed oscillations of a period of about one and 
a half days, which was probably due to a binodal seiche. The 
explanation of the absence of the longer-period oscillations is that 
the observations were made at or near a node. The observations at 
the two ends of the lake showed a rough opposition in phase, so that 
little room is left for doubt as to the nature of the oscillations. 
It was suggested by Halbfass that the temperature seiche might 
be peculiar to Loch Ness, or at least to very deep lakes. No 
observations existed in other lakes sufficient to show the presence or 
absence of such a seiche,^ but the observations in Loch Garry in 1908 
show that even in a lake with a mean depth of only 78 feet periodic 
temperature oscillations take place. F'ig. 50 gives observations made 
1 See Lamb's Hydrodynamics, p. 354. 
2 When making observations in Loch Garry I was not aware of Exner's 
observations in the Wolfgangsee. 
