a St ee ee eee ee 
ON TEMPERATURE OBSERVATIONS IN LOCH EARN. 651 
H, calculated by method of least squares = 3°632 x 10°. 
Vv (om 1) (om UV Co 
a fa Unit Unit hae re Unit Unit ae = mie Unit Unit 
= MS} 48010. | 1-016 x 107. ae ™8] 48010. | 1:016 x 107, Line. | £8010. | 1-016 x 107. 
“2 sq. ft. cub. ft. aa sq. ft. cub. ft. +, sq. ft. cub. ft. 
0 515 ‘00000 10 0 04104 20 722 02728 
1 514 -00007 11 65 ‘03707 21 814 01780 
2 472 00839 12 154 03434 92 870 02118 
3 432 01178 13 238 02894 23 929 02522 
4 368 -01669 14 323 029380 24 995 -02030 
5 308 02289 15 373 02993 25 | 1047 ‘01567 
6 260 ‘02590 16 455 -03180 26 1111 00665 
7 195 02716 ly 528 02450 27 1149 00356 
8 132 02349 18 591 ‘02570 28 1168 00000 
9 87 03761 19 654 02704 
Proceeding in exactly the same way as in the previous Loch Harn calculations, a 
period of 14°99 hours was obtained, which is remarkably close to the observed period 
of 15°2 hours. 
A further calculation was made on the assumption that the boundary should have 
been taken at a depth of 17 metres, and this gave a period of 14°7 hours. The calcula- 
tion for a 15-metre boundary was not made, but inspection of the other calculations 
makes it probable that a figure of about 15°3 metres would be obtained, so that the 
period as calculated from the theory is well within the experimental limits of accuracy. 
As stated on page 640, the period of the oscillation on 15th August was about 14 
hours. The reason for this is, doubtless, to be found in the higher surface temperature, 
and a very rough calculation of the period to be expected with the temperature 
distribution at that time gave 13°8 hours. 
A calculation of the binodal period under the same conditions gives 8°44 hours, 
whereas the observed period was 8°2 hours. The observed ratio of the uninodal period 
to the binodal, = (z.e. 1°85), is greater than in the case of the ordinary seiche, when 
14°99 
8°44 
(ze. 1°78). The reason of the discrepancy is probably that, with the higher nodalities, 
the assumptions on which the theory is based are not so accurate as they are for the 
ordinary seiche. 
Before the theory was developed the period of the oscillations was calculated by 
the method used for the Madiisee, and which gave quite good results in that case. But 
the period which the calculation gave was about 16 hours, and it was this discrepancy 
which suggested the extension of the theory. The reason of the agreement, which was 
obtained in the case of the Madiisee, is that the discontinuity was much sharper than 
it was = (z.e. 1°79), though theoretically the ratio should be nearly the same, viz. 
in Loch Earn, and the conditions more nearly approximated to a sudden discontinuity 
in temperature. But in Loch Harn the discontinuity was seldom very sudden, and at 
