in the Magnetic Forces of the Earth. 223 



saturated with vapor, and all the vapor within eighty-six feet of the 

 surface of the earth would not condense into more than in *01 of 

 dew, at the temperature of 70°. If we suppose the rise and fall of 

 the dew point, occasioned by winds, to balance each other, then the 

 average fall will represent the excess of the fall attendant upon dew 

 and rain over the rise produced by nocturnal evaporation. The 

 average fall attendant upon dew alone may then exceed 2° ; and 

 the withdrawal of vapor extend to a proportionally less height. 

 We may confidently affirm, however, that on a clear night vapor 

 must fall to the earth's surface from a height of several hundred 

 feet. But we are not confined to mere inferential conclusions on 

 this point, for, direct observations upon the varying hygrometric 

 state of the air have been made upon the summits of the Righi 

 and Faulhorn in Switzerland, which have established that even 

 at such heights (4.530 feet and 7,240 feet), the quantity of vapor 

 decreases during the night no less regularly than in the valleys 

 below. But the fall of vapor doubtless occurs at a height much 

 greater than that of the point from which it descends during the 

 night as far as the earth's surface. 



What is the amount of heat given out in the deposition of this 

 quantity of dew ? The experiments of Watt, and of Clement 

 and Desormes have established that when a given weight of va- 

 por is condensed into water of a given temperature, say 32° F., 

 the quantity of latent heat disengaged is the same, whatever may 

 be the temperature of the vapor, and is sufficient to heat the same 

 ^eight of water, at the temperature of 32°, no less than 1157°. 

 This then is the amount of latent heat evolved in the deposition 

 of dew on a winter night. On a July night it will be 1119°. 

 We may accordingly assume that all seasons of the year it is 

 more than 1100° ; and may therefore conclude that in the depo- 

 sition of in -01 of dew, the heat evolved is sufficient to raise this 

 ° m '0l of water 1100°, and to raise the temperature of one inch 

 J1 °- I do not find the specific heat of mold or soil in any table 

 °f specific heats in my possession, but in the very extended table 

 given in Pouillet's Elements de Physique Experimentale, I find the 

 specific heat of water set down as much greater than that of every 

 solid in the list. It is therefore highly probable that the specific 

 h eat of the soil is much less than that of water. We may 

 lengthen this conclusion, and at the same time obtain some ap- 

 proximation to the element sought, by attending to the specific 

 Jjeats of the principal ingredients of the soil. These are, of silica 

 01 9, of alumina 0-20, and of carbonate of lime 0*21, that of 

 J^ater being 1. The specific heat of dry soil is therefore proba- 

 cy about 0-20. That of soil in its ordinary moist state must be 

 somewhat greater. A somewhat rough experiment gave me 0*41 

 *°r the specific heat of a mass of earth of medium moisture, com- 

 posed of sand with some clay, the density of which I found 

 to be 1-5. J 



