LEOPOLD S ISLAND. 
235 
and sometimes fluctuating slowly to the northward. 
The chart opposite page 12 will show the capricious 
nature of this drift. 
But our general course was toward the south and 
east. On the 17th we were fairly in the sound. It 
welcomed us coldly. The mercury stood for a while 
at —19°, and sunk during the night to —27°. 
The next day, however, a shift of wind, gradually 
increasing in force, comhined with a tidal influence to 
drive us hack to our old position. The thermometer 
was at this time lower than we had ever seen it, and 
the sky seemed to sympathize with the temperature. 
The moon had a solid look, resting upon the snow- 
hills of Cape Riley, like a great viscid globe of illu- 
mination. In the morning the sky comhined all the 
tints of the spectrum in regular zones, a broad band of 
orange girding the horizon with an almost uniform in- 
tensity of color. The stars shone during the entire 
day. At daybreak on the 18th, Leopold's Island rose 
by refraction above the ice, standing with its unmis- 
takable outline clearly black against the orange sky ; 
but it went down as the sun neared the horizon, and 
passed to the south of his low circuit. My journal for 
the next two days shows the degree of illumination at 
the different hours. 
" November 20, Wednesday. The winds are unlike 
those encountered by Parry, our only predecessor in 
this region at this season of the year. It has been 
very providential, and very unexpected for us, this pre- 
dominance of breezes from the southward and east- 
ward. It has prevented our drifting into the dreaded 
sound, there to be carried, if it pleased Fortune, into 
Baffin's Bay by the easterly current. 
"We had a heavy gale from 2 P.M. of yesterday 
4 
