570 Proceedings of the Roycd Society 
It will be observed that these temperatures correspond almost 
exactly with such observations of the previous day as were made a 
mile and a half further north at the same depths, where the sound¬ 
ings were 618 fathoms. The bottom temperatures also corre¬ 
sponded with what I had observed with a different thermometer 
on September 21st, three weeks earlier. Using a cistern with 
proper valves, constructed by Mr Adie, for bringing up 96 ounces 
of water from the bottom, with a simple thermometer in it, I found 
that on September 21st, when the surface temperature was 54°, and 
also on October 11th, when it was 52°, the thermometer, on the 
instrument arriving at the surface, indicated 44° in the water 
brought up from the bottom, both in 87 and 103 fathoms of water. 
As the heating of the cistern in ascending must have been very 
nearly or altogether the same on both occasions, it follows that the 
corrected temperature at the bottom, as on 11th October, was 42° 
on 21st September. 
On 18th November I found it to be also the same. Cold weather 
had set in for a week before. The air was frosty, the ground dry 
and hard, the atmosphere very clear and perfectly still. Near the 
lower end of the loch, where the highway first touches it, the air 
temperature was 33° at half-past eleven. At Tarbet at one p.m., it 
was on land, but at the water’s edge, 37°; in the boat, in the middle 
of the loch, two feet above its surface, 42°; and in surface water, 
over 610 feet soundings, 46°. At the bottom, by a Casella’s thermo¬ 
meter, protected against pressure, and corresponding exactly in its 
graduation with the unprotected one previously used, the bottom 
temperature was again 42°. My design to make at the same time 
another complete series of observations, was prevented by unex¬ 
pected delays shortening my time very much, so that I had to con¬ 
fine myself to a single additional observation, for determining more 
nearly the upper limit of the cold substratum of water. At 250 
feet I obtained a temperature of 42°'2o, and consequently the 
upper limit of the water at 42° must have been as nearly as pos¬ 
sible at 270 feet in 610 feet soundings. 
Before drawing confident deductions from these observations, 
they require to be repeated at other seasons. But in the 
meanwhile it may be well to see what are likely to be the 
results, 
