ON THE SEICHES OF LOCH EARN. 475 



observed during August and September. The limnograms taken at the uninode and 

 Picnic Point are similar, except that the former shows merely a feeble binodal seiche, 

 while the latter has a well-marked trinodal superposed on the uninodal seiche. 



The wind on the 4th was light and easterly ; but a well-marked barometric depres- 

 sion, travelling with a velocity of about 18 (mile/hour), passed in a direction towards 

 N. 15° E., probably a little to the west of Loch Earn, the centre being nearest about 

 h 52 m on the 5th. 



The microbarograph at Ardtrostan shows a somewhat gradual drop of 2 mm., followed 

 by a sharp rise of 4 mm. between 15 h 44 m and 16 h 3 m . 



It does not appear that either the passage of the main depression or the minor 

 fluctuation attending it could have caused the sudden initial rise shown on the limno- 

 gram at 16 h 9'6 m . Both of these causes would indeed have worked, if at all, in the 

 opposite direction. 



We are therefore driven to the probable conclusion that the uninodal seiche was 

 caused by the flood. A glance at the map shows that the area — Glen Beich, Glen Ogle, 

 Glen Droma, Glen Ample, and Glen Voirlich — which drains into the western half of 

 Loch Earn, much exceeds that — Glen Tarken, A lit an Fionn, and Finglen — which drains 

 into the eastern half. It appears from the limnogram that for some time after the 

 flood commenced the level of the whole lake was rising at the rate of '32 mm. per minute. 

 In half the period of the uninodal seiche this would give us a rise of 2 '3 mm. If we 

 suppose this flood at the very beginning to be thrown only on the western half of the 

 lake, we have a disturbance equivalent to an increase of atmospheric pressure of 4 '6 mm. 

 of water. Acting during half the uninodal period, this, according to the calculation 

 given in Part V., p. 503, would produce uninodal and trinodal seiches having extreme 

 amplitudes of 6 '8 mm. and 2 '8 mm. If the first incidence of the flood were concentrated 

 on, say, the western quarter of the lake-surface, the resultant seiche would of course be 

 still greater. The rise shown at the binode was actually about 5 '5 mm., which corre- 

 sponds to an extreme amplitude for the uninodal seiche of 9*4 mm. It is therefore 

 quite possible that the seiche may have been wholly due to the sudden flood on the 

 western half of Loch Earn, and there appears to be no other way of accounting for it. 



4. Effect of Rainfall. — In order to obtain an idea of the effect of heavy rainfall in 

 causing a seiche, let us suppose a cloudburst to fall on the eastern half of Loch Earn 

 (idealised into a symmetric parabolic lake). If a- denote the rainfall in centimetres per 

 second, v the velocity of the rain-drops as they reach the surface of the lake, p the 

 pressure at time t after the shower begins, then we have 



p = <r(v + (jt) (dyne/cm. 2 ) 

 = av/g + <rt (gm./cm. 2 ) ; 



or, if we measure the pressure in millimetres of water, 



p =\0a-v/g+l Oat 

 = q + rt, say. 



