acceleration of the pendulum in different latitudes. 183 
donment of the house, before any satisfactory conclusions had 
been obtained. 
Towards the middle of June, the land was tolerably clear of 
snow in the neighbourhood of the harbour; the mean tempe- 
rature of the air had become but little less than that at which 
it was desired to carry on the experiments ; and the range of 
the thermometers in the course of the twenty-four hours had 
greatly diminished. An elevated and dry spot was now chosen, 
and the earth being removed for nearly two feet in depth (it 
having thawed above one foot), a foundation for the stands 
was made with as large stones as could be brought for the 
purpose, filled in with sand ; the clocks were then set up, and 
protected by a marquee, pitched, as at Hare Island, within a 
laboratory tent, a stove being placed at the door of the mar- 
quee with pipes leading through the tent. 
The clocks were going on the 18th of June; but a heavy 
gale of wind continuing through the two following days, 
forced the pegs and other fastenings of the tent on the wea- 
ther side (it being bad holding ground), and bore it down on 
the marquee, until relief was sent from the Hecla, when the 
tent was permanently secured by ice anchors. Whilst the 
marquee sustained the weight of the tent, the inner walls 
were unavoidably pressed in several places against the 
clocks and stands, which were shaken thereby ; one of the 
clocks was also stopped for the purpose of putting its pendu- 
lum in safety. This difficulty being passed, no other inter- 
ruption took place to the success of the experiments. The 
foundation subsided a little at first, but soon became suffici- 
ently firm and steady; a fire was generally required in the 
stove at night, but only occasionally during the day, the tern- 
