f 



OF A NORTH-WEST PASSAGE. 155 



spring-tide, we began to think that the ice might one day remove us rather J( 1822 - 

 more forcibly than by the slow process just related, for we found at day- o*-v-J 

 light, to our great surprise, that a portion of the floe belonging to the bay, 

 and which we had considered as quite fixed for the winter, had now been 

 broken off and removed, leaving the clear water within four or five hundred 

 yards of our sterns. A cloud of frost-smoke was issuing from it, and a flock 

 of dovekies swimming about in it. 



On the 26th, we flew a kite to the height of three hundred and seventy- Sat. 26. 

 nine feet, (as obtained by geometrical measurement,) with a Six's register- 

 thermometer attached to it. This after it had been up a quarter of an hour 

 indicated a minimum of — 23§°, the temperature upon the ice, by the same 

 thermometer tried before and after being — 24§°. 



Mr. Pulfer the carpenter of the Hecla in taking a walk round the S.E. 

 point, on the 27th, was somewhat startled at suddenly observing a large Sun. 27. 

 bear at no great distance from him, and prudently retreated to the ships 

 before Bruin saw him. It is commonly believed by the Greenland sailors* 

 who have certainly the best opportunities of judging, that these animals are 

 not generally disposed to retreat from one man, though they invariably fly 

 from a party. 



On the 29th there was afresh breeze from the N.W., which on the fol-Tues. 29. 

 lowing day increased to a gale more to the westward. In this instance the Wed.30. 

 thermometer seemed to rise with the wind, namely, from — 26° in the morn- 

 ing to — 18° at midnight. The thermometer fell again on the 31st, at the'flmr.31. 

 same time with the wind ; but these two phenomena did not often appear 

 to have the same connexion as at Melville Island. The month of January 

 closed with cold though rather overcast weather, and we thought we had 

 escaped very favourably with a mean temperature of — 22° 96'. 



The appearances of the Aurora Borealis during January were generally 

 more distinguished for their frequency than their brilliancy, or for any ex- 

 traordinary forms which this phenomenon presented. Towards midnight, 

 on the 13th, the weather being clear, it appeared in a very bright arch from 

 south to N.E., being 10° to 15° higher in the centre. It afterwards assumed 

 a wavy or serpentine form, which constantly varied, and smaller streams of 

 light seemed to be continually meeting the larger, from near the zenith. 

 From midnight till two A.M. on the 24th, it continued very bright and 

 generally extended from east, where it was most brilliant, to W.N.W. The 

 following evening, an arch of the Aurora assumed the most perfect 



x 2 



