ON TEMPERATURE OBSERVATIONS IN LOCH EARN. 649 
as will have been seen from the discussion of the temperature observations, the tempera- 
ture distribution at different points in the loch may be widely different, and the 
boundary may in reality not be a plane, but may be at different depths below the 
surface, its depth at any point depending on the temperature distribution at that point. 
CALCULATION OF THE PERIOD FoR Locu Harn. 
To apply the theory in its most elaborate form to the calculation of periods it 
would have been necessary to take into account all the local variations of distribu- 
Locn Earn 2% ducosr 1911. 15h. Locn Ewrn vucost 191. 
8 10 
“999800 600 400 200 f 000 DENS/TY 
015 014 “013 012 --Oll  W/SCOS/TY 
1 
10 
DENSITY 
cores ead VISCOSITY 
40m 4 
50m 
60™ 
' 
1 
i 
i 
i] 
i 
i 
70m 1 
1 
i 
i 
1 
80m 
Fic. 18.—Temperature-depth diagram, Fic. 19.—Density and viscosity diagram, 8th August, 1 p.m. 
8th August, 1 p.m. 
tion, but it was felt that this was a refinement which the approximate character of the 
theory, and of the data at our disposal, would not warrant. Accordingly an average 
distribution for the whole loch was obtained by taking the mean for several depths of 
all the observations made at each station at 1 p.m. on 8th August. This hour was 
chosen because, from an inspection of the observations, it appeared that the discontinuity 
was nearly horizontal at that time. In taking the means, the observations at Stations 
I. to V. were only given half the weight of those at other stations. In fig. 18 are 
shown temperature-depth curves for the five stations, the heavy line showing the mean 
distribution of temperature for the whole loch, arrived at in this way. In the diagram 
