ATMOSPHEROLOGV. — FOG. 443^ 



with tolerable accuracy, even upon a ship's deck ; 

 and upon a sheet of ice, with excellent effect. 

 Where, however, there is the least motion, this in- 

 strument cannot be made use of, though a very sim- 

 ple contrivance, which I adopted some years ago, 

 may often be substituted with considerable advan- 

 tage. The data for the latitude, it is well known, 

 are the sun's declination ; the angle subtended be- 

 tween the sun and the horizon ; and the height of 

 the eye of the observer. Obtaining the correct alti- 

 tude of the sun, therefore, is the principal object. 

 But, in fogs, the apparent horizon is brought much 

 nearer the observer than in clear weather ; and the 

 angle found between it and the sun, must be too 

 large, and the result therefore erroneous. The dis- 

 tance of the proper horizon, however, increasing with 

 the elevation of the observer, and decreasing as he 

 descends, he has it in his power, by taking a boat, 

 and placing his eye near the surface of the water, 

 to bring the horizon within less than half of the 

 distance at which it appears from a ship's deck. 

 At the height of 15 feet, the horizon is seen at the 

 distance of 4f miles ; but, at the height of 3 feet, 

 it is reduced to about two miles. So that, from the 

 latter situation, if the eye can penetrate near two 

 miles, a correct observation can be obtained. But, 

 if not, the error of observation will be much dimi- 

 nished. For, should the fog or land constitute an 

 apparent horizon, at the distance of a mile, the er- 



