ATMOSPHEROLOGY. — TEMPERATURE. 383 



hibit in the ball where it teniiinated, any difference 

 between the state of the electricity of the ship or 

 sea, and that of the atmosphere. The test of electri- 

 city, was a Bennet's gold-leaf electrometer, brought 

 into contact with the ball ; but though trials were 

 made for several successive days, from latitude 68° 

 to 75°, during clear, cloudy, and showery weather, 

 not the least excitation was ever observed. That 

 the effect might be rendered more perceptible, the 

 electrometer was well dried and warmed immediate- 

 ly before each experiment, without which, indeed, 

 no excitation could be produced in it, either with 

 glass or sealing-wax. The nights being light, the 

 aurora borealis could not be seen ; but on the eve- 

 ning of the 20th of JNIay, an appearance was observ- 

 ed very much resembling the aurora borealis, yet no 

 signs of electricity were observed in the electrome- 

 ter applied to the conductor. 



SECT. V- 



Atmospheric Phenomena, dependent on Reflection 

 and Refraction. 



The production of ice-blinks, by the reflection of 

 the rays of light, has already been noticed, when 

 speaking of the ice. As a proof of the benefit that 

 might be derived from this phenomenon, I may here 



