1883.] On the Circumpolar Expedition to Fort Kae. Ill 



anemometers was on an island in the lake, in an entirely open 

 situation. 



The anemometers did not work quite satisfactorily, being at times 

 choked by ice ; but I hope by the comparison of the two satisfactory 

 results may be attained. 



The wind was usually either South-east or North-west ; and even 

 when it blew from the former quarter, the motion of the upper clouds 

 often showed the existence of a North-westerly current. 



The hair hygrometers were found to be useless out of doors in cold 

 weather, on account of the formation of ice on the hair. 



The earth thermometers were read every alternate day : the obser- 

 vations were interrupted by a carcajou, or other animal, which 

 extracted the thermometers from their tube for the sake of the fur in 

 which it had been found necessary to envelope them, and broke them all ; 

 other thermometers were, however, substituted, and the observations 

 continued. It was found impossible to obtain the temperature of the 

 soil at a greater depth than four feet on account of the rocky nature 

 of the ground. 



A series of observations of terrestrial radiation was made by means 

 of a thermometer placed on the surface of the snow, but the almost 

 continual wind detracts much from the value of these readings. 



I was told by the residents of the country that the year was an 

 unusually dry one, and certainly the rainfall is remarkably small; 

 they also said that the winter was particularly mild and free from 

 storms, which, from all accounts, and from the journals kept at the 

 fort, seem to be both frequent and severe; as it was, we only expe- 

 rienced one, in February. 



A plan is annexed, showing the position of the meteorological 

 instruments, and their relative heights. 



Astronomical Observations, 



My first determination of the longitude was made by means of 

 lunar distances, and time was found by the method of equal altitudes, 

 but after the observatory was finished both these points were deter- 

 mined by transits, and the first value of the longitude found to be 

 more than a minute in error. The latitude was determined by transit 

 observations in the prime vertical, and is probably within a few 

 seconds of the truth. The longitude may be ten seconds in error. 

 The time was generally correct to within three or four seconds. 



A more solidly constructed transit instrument would have been 

 desirable, as it was found that in the cold weather it required so 

 much force to move the telescope of the transit theodolite on its axis , 

 that there was great risk of disturbing the adjustments of this instru- 

 ment, composed as it is of so many parts. 



