XIV INTRODUCTION. 



There was but little cloud in winter ; wliat there was was usually thin stratus aud cirro- 

 stratus, and it did not appear to be at a high level. The S.W. wind was, however, attended 

 with high cirrus clouds. A smoky haze was frequent in the summei-, which was probably 

 due to forest fires to the south of the lake. 



Parhelia, paraselenfe, and haloes were of common occurrence. On two occasions parhelia 

 were obseiwed at sunset, between the observer and the opposite shore of the lake (distant 

 four or five miles). 



The prismatic colouring of cirrus and cirro-stratus clouds in the neighbourhood of the 

 Sun was frequently observed in the spring and summer, and was a phenomenon at times of 

 great beauty. The colouring was once noticed to extend to a distance of 30"^ 40' from 

 the Sun, 



Aurora was observed on every clear night throughout the winter, as will be seen from 

 the tables, pp. 98-109. The journal of auroras has beea printed in extenso, and the 

 readings of the magnetic instruments at the time have been added, either as specimens of 

 the disturbance that accompanies aurora, or where a marked change of reading has coincided 

 with some phase of the phenomenon ; but as only one observer was generally available, 

 simultaneous observations could not often be carried out. 



The height of the aurora appeared to vary greatly ; it was twice noticed between the 

 observer and a mass of cloud. 



It was not found possible to obtain ^photographs either of the aurora or of its spectrum. 

 Captain Abney suggests that this was probably due rather to the effect of the low 

 temperature on the sensibility of the plate than to the faintness of the light of the aurora. 



The first snow fell on the 27th September, but it was not until a month later that the lake 

 froze. The residents all agreed that the season was a very exceptional one, the winter 

 being unusually mild, and late in setting in. At the end of November the Mackenzie river 

 was still nearly free from ice, whereas it is usually full of drifting ice in October and 

 frozen over in November. There was also much less snow than usual. A party of Indians 

 who came in on the 16th January reported that the country 50 miles to the N.K.W. was quite 

 bare of snow, the ground being not even white. The winter was also unusually free from 

 storms, which from all accounts, and from the journal kept at the station, seem to be both 

 frequent and severe in ordinary years. 



The snow began to disappear about the middle of April, and on the 3rd June the ice began 

 to break up. By the 16th it had entirely disappeared from the neighbourhood of Fort Kae, 

 though it was visible for some time longer on the horizon in the direction of the main lake. 



The trees fi.rst showed signs of budding on the 16th May, and on the 1st June they were 

 in full leaf ; when the party left the place on the 1st September they were already yellow 

 and beginning to lose their leaves. 



The observations being concluded, the return journey was accomplished without difficulty, 

 and Eno'land was reached on the 20th November 1883. 



