14 



VARIATION OF TEMPERATURE 



as would be di suable, yet every precaution Beems to have been taken by those who had 

 the general direction of them, to secure accurate results. Each obsen er is required to con- 

 form io ilu' rules laid down, and to certify under oath to the accuracy of his observations. 

 I have attempted to deduce the law from these data ; and (hough sonic anomalies may 

 be noticed, yet the resull on the whole is as satisfactory as could be expected. I have 

 included also a series of observations taken al Williams College, just out of the limits of 

 the State, do! more than two or three miles from the line, [n prosecuting the investiga- 

 tion, I have compared places t\w> by two having nearly the same latitude but different 

 elevations. In some cases where the latitudes differed too much, I have compared one 

 place wuli the mean between two or three others. For example, I compared the tempe- 

 rature of Canajoharie, whose latitude is 42° 53', with the mean of the temperatures of 

 . Bridgwater and Hamilton, whose mean latitude is also 42° 53'. In making 

 the comparison, I have uniformly employed the mean temperature of those years only in 

 which they were reported from both the places compared. The following table shows 

 the result : 



l.OHZR 



STATION. 



i rri:R 

 STATION. 



Difference 

 of lerel. 



Difference 

 fi-r io. 





658ft. 



750 





1051 



730 



910 



063 



712 

 544 



349 

 154 

 206 



664 





550 



276 



No of years 

 compared. 



Kinderl k . . . . 







) 



igburgh ) 



-li . . 

 . . . . 



Canaj"liani' 



tJtica .. 

 Auburn 

 ■ 

 Ithaca, 

 Aurora 

 Belleville 



Oxford 



Il.irt.v ick 



Williams College* 



Cherry-vallej 



I y-vallej .... 



Bridgwater, 

 Hamilton, 

 Fairfield . . . 



Pompey . . . 

 I 



Homer . . 



■ ille 



12 



1) 



13 

 3 



3, 2 & 3 



13 



II 



•1 



- .-. 5 



6 



9 



• Elevation estimated. t The year 1337 omitted, on account of a sopposed error in the record. 



The table shows very clearly that elevation exerts a perceptible influence on the tempe- 

 rature, though with considerable apparent irregularity. Perfect uniformity could not be 

 expected, and perhaps the deviation from a regular law is not <;rcaier than would natu- 

 rally result from the different exposure of the thermometers at the different localities, and 

 other accidental circun In no instance where the difference m the level of two 



places amounted to 300 feet or more, and where the latitude of both was nearly the same, 

 I found the mean temperature of the lower station to be less than that of the upper. 

 I'tica. as compared with Fairfield, was an exception during the years 1831 to 1837, but not 

 for the whole thirteen years embraced in the table. 



