59 
of Edinburgh ,, Session 1878-79 
represented graphically by the curve. From it we see that the 
temperature rises very quickly in the first fathom, then very slowly 
for some distance, until, in the neighbourhood of the bottom, it 
again rises quickly, reaching 39 ’85° in the mud. Had the water of 
the loch been in the condition usually imagined to immediately 
precede freezing — that is, at the temperature at which its water 
attains a maximum density uniformly throughout its depth, we 
should expect to find in the distribution of temperature in the 
water after the formation of ice the remains of this condition. The 
condition referred to would be represented graphically by a straight 
line drawn parallel to the line of depths through the temperature 
of maximum density, and if there were no supply or removal of 
heat from any other quarter than the surface, the curve of tempera- 
tures at any subsequent time when the loch was covered with ice 
would tend to coincide with this line at a sufficient depth. The 
source of heat which these observations show to exist at the bottom 
of Linlithgow Loch would have a tendency to mask but not to ob- 
literate these remains. If the curve of Station Ho. 4 be studied in 
this light we discover the remains of a comparatively uniform tem- 
perature of approximately 37°, more than half the water being at a 
temperature between 36 ‘5° and 37' 5°. If we imagine the water to 
have been at a certain date uniformly at the temperature 37°, and 
the surface to have been suddenly covered with ice, and at the same 
time a source of heat to have been applied at the bottom, the dis- 
tribution of temperature shortly afterwards would, I think, be of 
the kind represented in the curve. It must be observed that, 
assuming the water to be pure, both the loss of heat from the sur- 
face and the supply of it from the bottom would affect the inter- 
mediate waters by conduction alone until the temperature of the 
bottom had been raised to 39*2°. In the present case this tempera- 
ture has just been passed, and, admitting the water to be pure, 
convection would be beginning to come into play. That it has 
begun to do so is shown by the flatness of the curve near the 
bottom, compared with its steepness near the ice. 
Having established the existence of this unexpected thermal state 
of the water, it was necessary to find an explanation. The first 
that occurred to me was to suppose that the water of Linlithgow 
Loch was not pure water, but contained dissolved ingredients 
