224 Wm. A Norton on the Variations 



To be able to determine how much the temperature of a given 

 depth of soil will be raised by a given amount of heat, we must 

 know the specific gravity as well as the specific heat of the soil. 

 Now the specific gravity of sand is stated to be 1-5, that of clay 

 to be 2-2, and that of common earth 2-0. It appears, therefore, 

 that the increased rise of temperature of one inch of soil, above 

 that of water, by reason of the less specific heat of the soil, 

 should be generally pretty nearly counterbalanced by the diminu- 

 tion consequent upon its greater specific gravity- But, as the 

 specific heat is doubtless less than 0*5, while the specific gravity 

 does not exceed 2, the augmentation will be somewhat greater 

 than the diminution, and therefore the heat given out in the de- 

 position of in 01 of dew will raise the temperature of one inch 

 of soil more than 11°. If we take the specific heat equal to 04 r 

 according to the before mentioned experimental determination^ 

 and the specific gravity equal to 2, the effect of this amount ol 

 heat we find to be 13° -7. Upon a perfectly dry and sandy soil 

 it would not, probably, be less than 36° (since the density would 

 be 1-5, and the specific heat about 0-2). On the other hand, upon 

 a clay soil saturated with moisture, it might be as low as 10°. 

 But I will suppose for the present, that the heating effect of ,n *0l 

 of dew is no more than 11° to one inch in depth of soil, and pro- 

 ceed to enquire what quantities of dew would suffice upon this 

 supposition, to reduce the loss of temperature due to nocturnal 

 radiation down to the actual losses observed in different seasons. 

 On referring to the table of annual variations of temperature at 

 various depths below the earth's surface, given on page 45, we 

 find that the variations at the depths, l m , 2 m , 3 m , &e M form pretty 

 nearly a geometrical progression, of which the ratio is about 

 (the variations continuing below 8 m , and becoming nearly imper- 

 ceptible at the depth of 18 m ). Assuming this law and ratio for 

 the nocturnal variations, at the depths of l in , 2 in , 3 in , &c, recol- 

 lecting that in a night of twelve hours the cooling extends to the 

 depth of about 18 in , and taking 12° as the fall of temperature at 

 the surface, which is about the average for the year, and forming 

 the progression, I find the sum of the different terms to be 36 ■ 

 This then is the actual loss of heat. Now the average fall ot 

 temperature at sundown is 2° per hour. If we suppose this to 

 be due entirely to radiation, then, but for the heat given out by 

 the dew, the entire decrease of surface temperature in the course 

 of the night would be 24°; and therefore the entire loss of heat 

 would be 2 x 36° or 72°. To reduce this to 36° the heat given 

 out by the dew must be 36° ; and therefore the amount of dew 

 must be about T ^ of an inch. We have then this result ; upon 





The 



furni 



I take f in preference to £ 

 remits nearer to the truto- 



