SEICHES AND OTHER OSCILLATIONS 
61 
hand and the uninodal on the other are scarcely noticeable, but the 
seiche in our observations was hardly ever either purely uninodal or 
purely binodal. In these seiches the 5-9- configuration period caused 
by the interference of the uninodal and binodal components is usually 
reproduced with the most beautiful regularity, sometimes for a whole 
day or even longer. For example, in the seiche observed at Lochearn- 
head from 16th to 22nd October 1905, which lasted about 6^ days, 
say for 127 configuration periods, only six of these periods were 
found too short by one uninodal, and three too long by the same 
amount. It is probable that the gradual change of phase ac- 
He' 
T77 
^ dock "^a^vo-u M 4yrv(rcUj ^Z- 9 OS'. 
hci^vrv. 2/- q OS? 
M 
At 
in 
Fig. 20. 
companying the rise and fall of the amplitudes of the components 
more than compensated for the fact that 9/5 is not so close an 
approximation to T-^/T2 as is the sixth convergent, 70/39. 
In times of storm, or even moderate wind, there is, of course, a 
strong embroidery of various kinds, but usually the UB-dicrote 
configuration can be seen through all the confusion, and it soon 
becomes the prominent feature when the weather begins to settle. 
At this point we may indicate how the lake can be made to 
analyse its own seiches. Fig. 20 shows three simultaneous limno- 
grams, the lowest one taken at Picnic Point, about 480 yards from 
the eastern end of the lake, the middle one taken near the binode, 
