404 
Proceedings of the Royal Society 
of the water were reduced by about 8 fathoms, it would form three 
lakes, — the upper and largest extending from the head of the loch to 
Rowardennan, the middle one from Eowardennan to the chain of 
islands stretching from Luss to Balmaha, and the third, and 
shallowest, from these islands to Balloch. The ridges which 
separate these basins are covered in the present state of the lake by 
from 5 to 8 fathoms of water. The lowest, or Balloch basin, is of 
great extent and comparatively shallow, having a maximum depth 
of 13 fathoms. The middle, or Luss basin, is also of considerable 
extent, and has a maximum depth of 35 fathoms. The upper, or 
Tarbet basin, is long and narrow, and very deep, the maximum 
depth being 105 fathoms. At the upper end of this basin is a 
subsidiary one, which I call the Ardlui basin, with a maximum 
depth of 34 fathoms, and separated from the main basin by a ridge 
with a probable maximum depth of 17 fathoms. 
The general direction of the lake is north and south, so that the 
prevailing westerly and south-westerly winds blow across it, and, 
as is always the case in mountainous districts, they are diverted 
into squalls, which blow [sometimes up and sometimes down the 
lake. At Tarbet there is a deep rift in the mountains separating 
Loch Lomond from Loch Long, which gives access to the westerly 
winds to this part of the lake. On the whole, the geographical posi- 
tion of the lake tends to neutralise the effect of the prevailing winds. 
Extended temperature observations were made on the following 
days: — 18th August, 5th and 22nd September, 15th October, and 
14th November. On the 18th August observations were made only 
in the Tarbet basin, and only down to a depth of 30 fathoms. On 
the 5th September observations were made in the Luss basin, at 
four stations in the Tarbet basin, and at one station in the Ardlui 
basin. On the 22nd September observations were made in the Luss 
basin, and at two stations in the Tarbet basin. On the 15th Octo 
her observations were made at the same stations as on 5th Sep- 
tember, omitting Culness; and on 14th November observations were 
made in the Luss basin and at Inversnaid. 
The observations were made with an improved form of protected 
six’s thermometer, having a millimetre scale on the stem and a 
Eahrenheit’s scale on slips at the side. The average length of a 
» degree Fahrenheit was 3 millimetres, and all the thermometers had 
