made at Fort Franklin in 1825, 1826, and 1827. 243 



bulb with a thin film of silk paper, and then blackening it 

 thoroughly with china-ink and indigo. The silk paper was 

 used to overcome effectually the polish of the glass and pre- 

 vent the reflection of any of the sun's rays. From the com- 

 mencement of the observations till the end of April 1826, the 

 blackened-bulb thermometer was hung on the south side of 

 a rough deal shed used as an observatory; while the correspond- 

 ing thermometer with a clean bulb was secured on the north, 

 and consequently shady side. The black-bulb thermometer 

 was therefore sheltered from the winds that came from the 

 northern points of the compass only. In May 1826, and all 

 the following months, both the clean and black-bulb thermo- 

 meters were secured on the top of a slender detached post 

 rising three feet above the sandy soil. A square, thin, clear 

 glass bottle, four inches wide, placed on the top of the post, 

 enclosed the radiation thermometer, and protected it from the 

 wind. Its mouth was left open. The other thermometer was 

 secured on the same post, at the same height, and its bulb, 

 with the lower part of its scale, were enclosed in two concen- 

 tric brass cylinders, which permitted a free circulation of air, 

 but effectually intercepted the sun's rays. This was ascer- 

 tained by almost an hourly comparison with two other ther- 

 mometers, one inside the observatory, which was regularly 

 registered in connection with the magnetical observations, and 

 another hung in the open air on the north side of the obser- 

 vatory. The latter always felt the influence of the sun in 

 May and the summer months, both in the morning and even- 

 ing (owing to the high latitude), and being also unsheltered 

 from radiation of the sandy soil and of the deal observatory, 

 was scarcely ever lower than the thermometer enclosed in the 

 brass cylinders even at noon. The black-bulb thermometer 

 in the bottle was very sensitive, and has been noticed to fall 

 10 in the short space of time occupied by a cloud passing over 

 the face of the sun in a moderate breeze. In clear nights it 

 often shewed a lower temperature than any of the thermome- 

 ters with clean bulbs, the difference in some instances amount- 

 ing to 4°. During May 1826, a spirit black-bulb thermometer 

 unsheltered by glass was hung against the observatory and 

 shifted from side to side with the course of the sun. A re- 



