368 Prof. Bischof mi the Temperature of 



longer felt at the depth of three feet, as Munke's* observations 

 seem to shew, still heat would be imparted to it by the rivulets 

 through which it passes. It is therefore very possible that the 

 mean temperature of the spring was estimated as much too high 

 as that of the air was. 



I must here call attention to another circumstance, which is 

 certainly worthy of notice. Observations on the temperature of 

 the air are generall}- made in places protected from the direct 

 influence of the sun''s rays, as well as from radiated heat ; but the 

 soil is exposed to these. The yearly mean temperature of the 

 air must then, if ascertained in such a situation, be lower than 

 the mean of the superficial crust of the earth ; but variable 

 springs receive their heat from that external crust, so that their 

 temperature must be higher than that of the atmosphere. Single 

 observations on the temperature of the soil and of the air have 

 given very considerable differences. Thus, Humboldt* found 

 the temperature of the air on the Orinoco at two o"*clock to be 

 86**; that of a coarse moveable granitic sand, 140°. 45 ; that of 

 a similar white, but close-grained, fine sand, IJ^G^'.S ; and that 

 of the granite rocks, 117°.725. An hour after sunset, the coarse 

 sand showed 89°.6, and the granite rock, 101 f°. As sand ab- 

 sorbs the heat of the sun in a greater degree than other earths, 

 and as the meteoric waters, filtering slowly through it, easily 

 assume the temperature of the sand, it is very possible that the 

 mean temperature of springs which have their origin in a sandy 

 soil may rise higher than the mean temperature of the air. This 

 may, perhaps, be the case with the springs near Berlin. For 

 that reason, it would be very desirable to combine observations 

 on the temperature of the superficial strata with those made at 

 the same time on the temperature of springs and of the air. 



The thermometrical observations at Tiibingen were made in 

 a well of running water in the botanical garden. A comparison 

 of the yearly mean of these with the yearly mean temperature of 

 the air, speaks in favour of the opinion that the former is higher 

 than the latter, for they give— - 



• Gehlers neues Phjsikal. Worterhuch, vol. iii. p. 988. Fourier's " TWo-* 

 rie de la Chaleur," j)laces this limit at about 3 yards. 

 \ Voyage, toI. vii. p. 203. 



