368 ACCOUNT OF THE ARCTIC REGIONS. 



derived at least from 14,600 observations, sup- 

 posing only two observations to bave been made 

 eacli day, 

 2c?, That in a period of years, the mean of the 

 extremes of temperature observed in any one month, 

 agrees with the mean temperature of that month, 

 generally, to the fraction of a degree. 



In Table A of meteorological results, No. II. of 

 the Appendix, the sum of the highest tem- 

 peratures observed in the month of 3Iay for 

 twelve years is 415°, and of the lowest 1212°; 

 the amount, 537, divided by 24, the number 

 of observations, gives a mean of 22.38, differ- 

 ing from the mean temperature of this month, 

 as derived from 956 observations, only 0.43 of 

 a degree. From the same Table, it appears, 

 that the sum of the highest temperatures ob- 

 served each year for ten years in June, is 405, 

 and of the lowest 214, the amount of which, 

 619, divided by 20, the number of observa- 

 tions, gives 30.95 for the mean, differing only 

 0.35 of a degree from the mean temperature 

 as derived from 831 observations. In the 

 same way the mean of the extremes of tem- 

 perature in April is 17.6, differing from that 

 derived from the mean of 370 observations 

 only 1^ degrees ; and of July, 37-86, differ- 

 ing from the mean of the month, by 548 

 observations, only 0.89 of a degree. 



