METEOROLOGY. 



351 



intensity at any latittitle north, is precisely the same as at an equal 

 latitude south of the equator. This immediately follows from the 

 formula, where the annual intensity is developed in a series of powers 

 of cos L, which is always positive, whether the latitude L he south or 

 north. 



Proceeding with the investigation, I have computed the annual 

 duration of sunlijrht^ according to the rising and setting of the sun's 

 centre, without regard to refraction. It is the half of 365.24 days, or 

 182.62' days, increased hy the quantities in the following table, for 

 the northern hemisphere^ and diminished hy the same for the southern 

 hemisphere : 



Annual Inequality of Sunlight, A. D. 1850. 



Having thus determined the duration of Sunlight, let us next con- 

 sider its increase hy Refraction and by Twilight. The mean horizon- 

 tal refraction, according to Mr. Lubbock's result, is 2075", or 34' 35"; 

 the barometer standing at 30 inches, and the thermometer at 50° F. 

 But as this is somewhat greater than what has been usually employed, 

 we shall adopt 34' as the mean value for determining the increase of 

 daylight by direct refraction. 



With respect to the duration of Twilight, A. Bravais, who has 

 made extensive observations upon the phenomenon, observes in the 

 Annuaire 31 eteorologique de la France for 1850, p. 34 : " The length 

 of twilight is an element useful to be known : by prolonging the day. 

 it permits the continuance of labor. Unfortunately, philosophers are 

 not agreed upon its duration. It depends on the angular quantity by 

 which the sun is depressed below the horizon ; but it is also modified 

 by several other circumstances, of which the principal is the degree of 

 serenity of the air. Immediately after the setting of the sun, the 

 curve which forms the separation between the atmospheric zone directly 

 illuminated by the sub, and that which is only illuminated second- 

 arily, or by reflection, receives the name of the crepuscular curve, or 

 Tiuilight Boio.* Some time after sunset, this bow, in traversing the 

 heavens from east to west, passes the zenith ; this epoch forms the 

 end of Civil Tivilight, and is the moment when planets and stars of 

 the first magnitude begin to be visible. The eastern half of the 

 heavens being then removed beyond solar illumination, night com- 

 mences to all persons in apartments whose windows open to the east. 

 Still later the Twilight bow itself disappears in the western horizon ; 

 it is then the end of Astronomic Ticilight ; it is closed night. We 



"^^ Tlie phenomenon is equally conspicuous in ihe west, before the rising of the sun, and in 

 cci tain slates of the atmosphere is scarcely less beautiful than tbo rainbow, for the symmetry 

 and vivid liuting of its colon. 



