][g s. Nakamurd, and K. Honda : 



Of the agencies which cause a seiche in a lake, Ave may 

 mention, after Forel and others, atmospheric electricity, earth- 

 quakes, unequal local distribution of atmospheric pressure, strong 

 wind etc.. We had no experience of an earthquake during our 

 ■observations, but we several times met with pretty heavy thunder 

 storms, and almost always had a very significant seiche after 

 them. This may be due partly to the unequal distribution of 

 iitmospheric pressure existing in such cases, but certainly it must 

 be due also to the attraction between the water of the lake and the 

 electrified cloud over it. The water just below the cloud is heaped 

 up and then is let go when the electrification of the cloud is 

 ■discharged, and as a result the water is set in oscillation. The 

 position of the electrified cloud relative to the lake perhaps 

 determines the mode of oscillation; when it is just above or near a 

 loop of (jne of the natural modes of motion of the lake, then that 

 particular motion is sure to be excited. Let us now give some 

 examples of such seiches excited after thunderstorms, observed in 

 lake Biwa (see PL VI.). 



Here we met with many thunderstorms, the first of which 

 occurred on the afternoon of the 13 th. August. There are 14 

 provincial department offices scattered around the lake, where 

 the conditions of tlie weather are officially reported once every 

 ■day. From these reports, we see that the storm was restricted to 

 the southern part of the lake. It began to rain at about 13 h.. 

 and continued for about two hours. Though the wind was either 

 light or moderate everywhere lüowing mostly in tlie westerly 

 direction, it caused an anomalous seiche in the lake. The motion 

 was recorded l)y the limnimeters set at Ötsu, a station at the 

 south end of the lake, and at Iraazu, another station on the 

 nortliwostei'n side. The curve traced at Ötsu is reproduced in 



