gß s. Xakamura and K. Honda. 



The soundings of the lake were made by the Naval Hydrogra- 

 phie Office, and the result was placed at our disposal. A model in 

 cement was constructed, and an experiment with it showed us two 

 uninodal oscillations. As the lake is circular, we can not 

 distinguish them by the terms, longitudinal and transversal. In 

 one of them, which we shall denote by t', the nodal line runs 

 E-W, and in the otlier t", it runs N-S (Plate XXII). In our 

 model, their periods were found to be 



t' — 0.67 sec. 



^'=0.76 sec, 

 which on reduction gave 



t' — 10.3 minutes 



r=i].8 ,, 



as the periods of seiches in the actual lake, and they are both 

 greater than the observed period. We see from this experiment, 

 that as these motions have nearly equal periods, when they coexist 

 we must have the phenomena of beat. Since the difference of t' 

 and t" is 1.5 minutes, we ought to have a maximum or a minimum 

 amplitude every 6, or 7 oscillations, which agrees with the actual 

 observations (Fig. 25). It is to be noted from Plate XXII, that as 

 the lake has a shallow on its south side, though the loops for t' are- 

 situated symmetrically with respect to the meridian axis of the 

 lake, the loops for t" are unsymmetrically situated, i. e. they are 

 much displaced to the south side of the lake, 



Calculating the mean depth /( of the lake by means of at 

 planimeter, we find 



li = 9.58x10' cm., 

 and miCasuring the diamètres of the lake along the directions- 

 suggested by the experiment with the model, we get 



//= 9.5x10' cm.. 



