ATMOSPHERIC REFRACTION AT LOW TEMPERATURES. 685 



most interesting to investigate. The extreme difficulty attending 

 the use of the repeating circle during intense cold, rendering it 

 next to impossible to obtain with that instrument observations of 

 a star having quick motion with the minute accuracy requisite 

 for this purpose, we were led to adopt the simple method of ob- 

 serving the setting of a star behind a horizontal board fixed 

 edgewise on a neighbouring hill, the altitude of the board being 

 obtained at leisure by repeated observations with the circle. 

 These observations, the details of which are given in the Philo- 

 sophical Transactions^ make the refraction at low temperatures, 

 and from the altitude of 4J° to 7J°, as computed from the table in 

 the Nautical Almanack, considerably in defect. 



The instruments used were a theodolite and a small repeating 

 circle with two telescopes which were fixed at 122 feet above the 

 sea-level on casks filled with sand and firmly frozen to the ground. 



It Avill be seen from the observations made by Sir John Ross 

 that the accuracy of the Nautical Almanack tables cannot be 

 tested by the method adopted. 



3. Capt. Parry's Third Voyage by Spitzbergen, p. 99. 



Fog Boivs. — At half-past 5 p.m. on 23rd July 1827, in lat. 

 82° 43', a beautiful phenomenon, consisting of six fog bows, was 

 seen in a direction opposite to the sun. First one white bow 

 was seen, then it became tinged with prismatic colours and five 

 other complete coloured arches were formed within it, the interior 

 ones gradually narrowing. 



The larger bow and the one next within it had the red on the 

 outer or upper part of the circle, the others on the inner side. 

 Altitude of the outer arc, 20° 45', its extent at the horizon, 72 J°, 

 the altitude of the sun being 20° 40'. 



The smaller bows were seen for 20 minutes. 



In these travels the southerly drift of ice was more than four miles 

 a day for the last five days of the journey, so that after immense 

 labour in making a good 23 miles journey there vras found to be 

 only a gain of one mile northwards. 



Lat. 82°-45', long. 19°-25'E., dip, 82°-2r-6, and declination 

 18°- 10' W. No bottom at a depth of 500 fathoms ; the specific 

 gravity of water at that depth was 1 '034 at 37° F. 



At 6 fathoms the specific gravity of water was 1 * 0225. 



On July 27th, at 3 a.m., saw another fog bow, the sun's altitude 

 being 121°, that of the centre of the outer arch 28°, and its extent 

 at the horizon, 77^°. 



Three smaller fog bows were seen within the large one. 



Aug. 2nd.— In lat. 82° 6', and long. 17° 45' E., we met with 

 snow, tinged with some red colouring matter to the depth of 

 several inches, giving a rose tint and in some places a salmon 

 coloured tint to the snow when pressed. 



