382 ACCOUNT OF THE ARCTIC REGIONS. 



degrees upwards, intensely dry. In consequence of 

 the uncommon dryness of the cabin of a ship in cold 

 weather, the wainscotting sometimes shrinks as 

 much as half an inch in a pannel of about fifteen 

 inches broad, being equal to one-thirtieth of its 

 breadth ; but on returning to Britain, the same pan- 

 nel expands again almost to its original dimen- 

 sions. 



Few observations, comparatively, seem to have 

 been made on the Electricity of the atmosphere, es- 

 pecially in high latitudes. Perhaps some trials that 

 I made in the spring of 1818, on this subject, were 

 the first that have been attempted within the Arctic 

 circle. On my passage towards Spitzbergen, when 

 in latitude 68°, I erected an insulated conductor, 

 eight feet above the main top-gallant~mast head, 

 connected by a copper wire, with a copper ball at- 

 tached by a silk string to the deck. The conduc- 

 tor consisted of a slender taperingtube of tinned iron, 

 terminated by a pointed brass wire. It was fixed in 

 an iron socket, supported by a large cylindrical piece 

 of glass, which glass, by means of another iron sock- 

 et, was secured to the top of a long pole, elevat- 

 ed several feet above the mast head. A tin cone 

 encompassed the bottom of the conductor, the mouth 

 of which being downward, defended the rod of glass 

 from getting wet, so as to injure its insulating pro- 

 perty. The conducting wire being kept carefully 

 clear of the rigging of tlic ship, v.as expected to ex- 



