50 On the Methods of obtaining Mean Temperatures. 



The water of our wells and springs is derived from extensive 

 strata of sand, lying above, and sometimes between, beds of clay. 

 These strata deviate but slightly from a horizontal position, and 

 as the surface of the country is undulating, the same bed will be 

 reached at different depths at places near together, so that wells 

 are often seen twenty or thirty feet deep, or springs rising from 

 hillsides, having water of the same temperature as wells in the 

 same neighborhood having a depth of eighty feet. When the 

 average depth of a stratum below the surface is seventy or eighty 

 feet, the variation of temperature of the water of wells through 

 the year is scarcely perceptible, and the annual mean is about 

 64.25°. But in wells bored to strata whose mean depth below 

 the surface is not more than twenty five or thirty feet, the varia- 

 tion is often several degrees, and the mean afforded not lower 

 than 67° or 68°. 



Art. VI. — Some Remarks on the methods in common use of 

 taining the Mean Temperature of Places , and on the suppt 

 difference between the Temperature of the Air and that of 



Earth ; bv Prof. W. M 





It is stated by Humboldt and others, that the mean tempera- 

 ture of the coldest springs in warm climates is often lower than 

 that of the air of the same places. If we examine those agencies 

 in which atmospheric temperature originates, and by which ter- 

 restrial temperature is modified, we shall perceive that such a 

 condition could not exist, and consequently, that the observations 

 on which such conclusions were based, were not accurate, or that 

 some unsuspected agency must modify the relations which should 

 otherwise be constant. In examining; the meteorological records 



vwv, »**w«.wx,*ww b 



of our own country and of other parts, and comparing the obser- 

 vations made by different persons resident at the same places, we 

 shall perceive that the results differ not more from the mean tern- 

 perature of the place, than from each other. These discrepancies, 

 or rather these departures from accurate results, are dependent on 

 many circumstances. In the first place, thermometers of very in- 

 ferior quality are in very general use, and they will often differ in 

 the results afforded by four or five degrees. In the next place, 

 sufficient importance is not attached to the position of the instru- 





