138 METEOROLGY. 



its superincumbent air, below that of the ocean, so that the pressure in 

 the temperate now conspires with that in the frigid zone to produce cur- 

 rents towards the ocean. These currents will be nearly at right angles 

 to our sea-coast, and hence our autumnal winds are mostly from the 

 north-west. When these winds begin to blow, the previous season of 

 repose is broken up, and the Indian Summer is over. It indicates the 

 advance of winter from the frozen regions of the north. 



We must, however, not forget to ascribe a most prominent influence, 

 in the production of the Indian Summer, in the prevalence and direction 

 of winds, and in the whole meteorology of the Atlantic states, to the 

 Gulf Stream. This current, flowing from a comparatively warm cli- 

 mate, and bringing with it its warmed waters, is constantly giving out 

 heat and aftbrding vapor to the air immediately above it. The result 

 being a comparatively inferior density and pressure, winds must be di- 

 rected towards it from the land with greater or less regularity, until the 

 temperature of the air over both again becomes the same, that is, until 

 about the middle of April. 



About this time, another season of comparative equilibrium between 

 ocean and land, and another " Indian Summer," in all but the name, 

 again occurs. The atmospheric changes necessary to produce this sea- 

 son of repose, are essentially the same as those just explained, only 

 that they occur in a reverse order. The increasing altitude of the sun 

 and length of day together produce a rapid increase in the temperature 

 of the surface of the earth in northern latitudes. The air, which is 

 thus warmed and dilated, is at first forced southward by the greater 

 pressure from the North, producing the INIarch winds. After these are 

 over, the temperature of the land rapidly approaches an equality with 

 that of the ocean, and an equilibrium takes place. This occurs from 

 one to six weeks after the equinox; frequently being apparently divided 

 into two stages-, the fii-st, during the last week in March, was the case 

 in the spring of the last and present year ; the second, about the middle 

 of April, or a little later. About that time we may with almost as much 

 regularity look for a season of fine, pleasant and smoky weather as 

 during October. 



Secondly, The similar seasons, mentioned as existing about mid- 

 winter, and mid-summer, or a little later, are not to be regarded so much 

 as seasons of equilibrium of pressure between land and water, as be- 

 tween the Northern and Southern hemispheres. 



During the rapid cooling of the Northern hemisphere, tiirough the 

 months of November, December and January, when the daily loss of 

 heat is greater than the gain, and the rapid heating of the opposite hem- 



