110 THE FUESH-WATER LOCHS OF SCOTLAND 
also states that the water at the surface to a depth of 2 or 3 feet is 
of nearly uniform temperature. This is true after passing the first 
thin surface layer, but I believe the thin surface film, the tempera- 
ture of which is so difficult to measure, undergoes much greater 
variation than Forel indicates. There is often at the surface a very 
steep temperature gradient, and this is never better shown than when 
there is a thin covering of ice forming at the surface, showing the 
surface temperature to be very nearly 32° Fahr., whilst the reading 
obtained by a mercury thermometer, the bulb of which is immersed 
an inch or so below the surface, is seldom lower than 34° or 35° Fahr. 
The extent of the diurnal variation is greatly dependent on the 
size of the surface waves and the strength of the wind, so that obser- 
vations on different days are not comparable with one another. 
Another point, however, is of more importance than this, and that 
is the depth to which direct insolation — direct heating by the sun's 
rays — takes place. Rightly or wrongly, I distrust all deductions 
made by Richter and others as to the depth to which heating takes 
place by direct radiation. Convection and conduction currents 
determine the depth to which heating takes place, and these currents 
are dependent on other factors than the strength of the sun's rays 
and the angle at which they meet the surface. Only two factors 
need be mentioned — (1) the effect of wind and the action of the 
waves, to which I shall return ; and (2) the composition of the water 
and the quantity of matter in suspension, upon which the trans- 
parency of water depends. The diathermancy of water is small, and 
the heat rays are largely absorbed after passing through the first few 
millimetres of water, so that radiation into a loch is not appreciable 
at considerable depths ; and radiation out of a loch during the night 
is even slower, for Soret found that the diathermancy of water was less 
for dark rays than for light rays. In Melloni's ^ experiments on the 
diathermancy of liquids it was found that, when the source of heat 
was an Argand lamp with glass chimney, only 1 1 per cent, of the rays 
was transmitted through a layer of distilled water 9*21 mm. thick. 
Attempts have been made to measure the depth to which direct 
radiation is appreciable. Forel observed in the Lake of Geneva, by 
immersing a thermometer with blackened bulb at varying depths, 
that in midsummer at a depth of 1 metre the black-bulb thermo- 
meter read as much as 15° Fahr. higher than the ordinary thermo- 
meter. Interesting observations were also made in Loch Ness by 
means of the Callendar recorder, which was at times used to measure 
the difference in resistance between a bright and a blackened platinum 
wire immersed at various depths. When this bolometer was immersed 
at greater depths than 5 feet it was impossible to distinguish between 
1 AnnaL Ghimie et Physique, ser. 2, t. liii. p. 5, 1833 ; t. Iv. p. 337, 1834. 
