242 Mean Dii'ection of the Wind. 1 



North prevailing more tlian lialf tlie day. It does not appear how 

 tliis fact can be exphiined, except by asciibing it to the influence of 

 the warmer water of the Gulf Stream. 



The northerly wind, which prevails during a considerable portion 

 of the day in the month of May, cannot be ascribed to the influence 

 of the neighboring water, but is easily explained by the higher tem- 

 perature of the Gulf Stream ; while during the principal part of the 

 day, the tempei-ature of the land rises so much above that of the 

 neighboring water, that a breeze springs up from the colder water 

 tOAvard the land. 



The northerly wind which prevails during a portion of the day in 

 the month of June, seems also to indicate the influence of the Gulf 

 Stream, while the southerly wind, wliich prevails dui'ing more than 

 half the day, is explained as in the month of May. 



In July the northerly wind almost entirely disappears, for now the 

 land is not only warmer than the neighboring ocean, but during a 

 considerable part of the day is warmer even than the Gulf Stream. 

 The sti'ong southerly wind which generally pi-evails, is a current flow 

 ing from the coolei- water toward the land. 



In the months of August and September the land is warmer than 

 the neighboring water during about half of the day, and colder during 

 the otlier half; and we And accordingly that the northerly current 

 prevails for about half of the day, and the southerly for the other 

 half 



In the month of October the circumstances are nearly the same as 

 in March, wliile in November and December they are nearly the same 

 as in January and F'ebruary. 



We find, then, that most of the observed tacts can be accounted 

 for from the unequal temperature of the land and the neighboring 

 Avater ; l)ut some of the fiets, especially those in A])ril, ^lay and June, 

 seem to indicate a decided influence of tlie Gulf Stream; and if the 

 influence of the Gulf Stream is appreciable during certain months of 

 the year, its influence must be exerted during the remaining months 

 of the year, altliough partly masked l)y being blended with other 

 causes. 



If the causes which we have here assigned ftn* the changes in the 

 wind's direction at Wallingford are correct, they ought to produce 

 similar efiects at other stations similarly sittuited; that is, at places 

 all along the Atlantic coast of the United States within the belt of 

 prevalent westerly winds. Observations at such places may then 

 afibrd a test of the accuracy of the explanation here given, j 



