228 



Meteorology of New Haven. 



November. 



December. 



These numbers being set off on the curves previously mentioned, 

 determined five points of the true diurnal curve for Xew Haven, and 

 indicated that the chief error in the numbers on page 199 consisted in 

 the range of the diurnal motion, and not in the law of the cnrve. 

 Assuming that the true diurnal curve was similar to that indicated 

 by the numbers on page 199, a curve was drawn through the five 

 points already detennined, and as nearly parallel as possible to the 

 one first constructed. The true curve has generally a somewhat 

 greater range than the provisional curve first assumed; but other- 

 wise the two curves bear a close resemblance, and afford a presump 

 tion that we have made a near approach to the true curve represent- 

 ing the diurnal change of temperature. The ordinates of these 

 curves have all been carefully measured, and are given in the follow- 

 ing Table. 



