208 OBSERVATIONS ON THE MEAN TEMPEKATURE 



land is 8° ; and therefore calling it also 8°, which is far too 

 low for the latitude of 76" 45', we have for the mean tempera- 

 ture, uninfluenced by the ice, 33°.8 — 8".0 =: 25°.8 ; from 

 which subtracting the polar anomaly of 21", as computed by 

 Mr ScoRESBY, and we obtain 4''.8 for the mean temperature 

 of the Pole. 



In the preceding paragraphs, we have compared the results 

 of the formula with the temperatures of individual places, 

 which must often be influenced by local causes. We shall 

 therefore compare the formula with the temperatures of the 

 four parallels of 30", 40", 50", and 60", which Humboldt has 

 deduced from a great variety of observations, and which he 

 considers as well established. 



The differences between the observed and calculated tem- 

 peratures, both in this and the preceding Table, are frequently 

 owing to the circumstance of the thermometer having been 

 observed at two periods, the average of which does not give 

 the mean temperature of the day. The Reverend Mr Gor- 

 don has found, from a series of very accurate observa- 

 tions, that the mean temperature will be obtained most cor- 

 rectly in this country, when self-registering thermometers 

 are not used, by observing at 10 o'clock in the morning 

 and evening ; and it is highly to be desired that this prin- 

 ciple should be adopted in all our meteorological journals. 

 Another source of difference arises from local causes, which 



often 



