INTRODUCTION. xi 



30 miles or so to tlio uortli and east the " barren lands " begin : a rocky country, destitute of 

 trees, though not of vegetation, extending to the coast of the Arctic ocean. This is the home 

 of the musk ox and the reindeer. It is the great hunting ground of the Indians and the source 

 of the food supply of the district. 



Nearer the lake the country is covered with birch, willow, and pine, as a rule small and 

 stunted, though in sheltered places the last-named trees sometimes attain a fair size. 



The peninsula of Nu-chie (the mountain island) as the Indians call it, projects from the 

 north-east shore, and is the only locality where limestone appears on that side of the bay. It 

 is almost an island, being only joined to the mainland by a small patch of swamp, and consists 

 of a crescent-shaped hill of the height of about 220 feet, precipitous on the outside and 

 sloping more gently to the lake on the inside (Fig. 2, p. x). At the south-west extremity 

 of this peninsula, at the foot of the hill, is a small extent of level ground. Here is the 

 Hudson's Bay Company's post cf Fort Rae, some half dozen log huts, with a large store for 

 provisions, furs, and goods, for trading with the Indians. 



The lake at this place is shallow, and there is a constant current from the north-west, 

 caused by two rivers that enter the head of the gulf. The gulf contains numerous islands, 

 especially along the north-east shore. 



It was 10 p.m. on the 30th August when Fort Rae was reached. The 31st was occupied in 

 unpacking the instruments and stores. The barometer, an anemometer, and the thermometer 

 screen, with wet and dry bulb thermometers, were at once placed in position so as to enable 

 observations to be commenced at midnight. There was most fortunately at the spot an un- 

 finished and unoccupied building, admitting of conversion into a Magnetic Observatory. It 

 was a log hut, built for a store, and a door and windows having been put in, a floor laid 

 down and a fireplace built, it answered its purpose very well. 



The instruments, on the whole, had suffered little from the journey, one of the barometers 

 and two thermometers were broken, a few screws had shaken loose from some of the 

 magnetic instruments, and a mirror required to be re-silvered. These and other similar small 

 repairs were executed whilst the Observatory was Ijeing prepared for their reception, and on 

 the 3rd Septemlier the declinometer, on the 4th the bifilar. and on the 6th the balance 

 magnetometer, were mounted in their places, and observations commenced therewith. 



The performance of the magnetic instruments was satisfactory, with the exception of the 

 balance magnetometer, as mentioned hereafter, p. 119. Metallic suspension would have been 

 preferable to silk for the bifilar magnet. 



These instruments were mounted on wooden pillars, sunk to a depth of more than three 

 feet in the ground. Stone pillars would have been better for the purpose, but the only stone 

 available would have required so much cutting that even had the necessary tools been at 

 hand, so much time would have been consumed in the preparation of the pillars that the 

 observations could not have been commenced until late in September. 



The latitude, longitude, and time were all determined with the transit theodolite. 



The longitude adopted is deduced from 10 observations of moon-culminating stars, the 

 latitude from the prime vertical transit of a Ursse Majoris. The observations were timed by 

 a chronometer watch Avhose going was frequently checked by the transit instrument, and 

 its rate was found to vary but little throughout the year. 



The hourly observations were commenced at midnight on the 31st August, the hours were 

 thus divided between the three observers : — A. was on duty from 6h. 30m. a.m. tc 6h. 30m. p.m., 



A 17420. • C 



