21 
1919-20.] The Cooling of the Soil at Night. 
upholds the view that the temperature and conductivity of the soil play 
a most important part in determining the minimum temperature at night.* 
Thus in various places in his paper he states : — 
(1) In 1906 and 1907, on a newly sanded thinly vined section of the 
marsh, the 3-in. depth soil maximum averaged highest and the soil 
minimum averaged lowest, but the surface-soil minimum and air 
minimum averaged highest. 
(2) There is, of course, a direct relation between the soil temperature 
and the ensuing air temperature. An increase of heat during the day 
serves to raise the temperature of the soil, and this, in turn, prevents low 
minimum temperatures at night. When the soil temperature is high, more 
heat may be lost by conduction and radiation from the soil before the 
critical air temperature at or immediately above the soil is reached, even 
if the loss by radiation be at the same rate at all locations. 
(3) The relation between the temperature of the soil and the occurrence 
of frost is noticeable in that it is practically impossible for frost to occur 
in sanded sections of the bog on the first cold night after a warm spell ; 
but it is likely, if conditions are favourable, on the second night when the 
soil has become cold. 
(4) The average increase in minimum air temperature on clear nights 
during September 1906 at Berlin, Wisconsin, on a clean sanded location 
where earth was warm, over a clean peat location where earth was 
relatively cold, was : — 
(a) At surface 6*7° F., 
(b) At 5 in. height 23° F., 
(c) At 36 in. height • 5 ° F., 
showing that the warming effect of the heat of the soil reached even to 
a height of 36 inches. 
These statements are fully borne out by my own observations, the 
tendency for low minimum temperatures when the soil is cool at the 4-in. 
depth being at once apparent from Table VIII. But in addition we see 
that low minimum temperatures may be expected even with relatively 
high 4-in. depth temperatures if the ground is very dry, as the heat is 
then prevented from being readily conducted to the surface. 
If now we examine the results given in Table VIII a little more 
critically it will be noticed that these results may be divided into two 
groups : (a) when conditions were ideal, the sky being clear all night, 
with no wind; ( b ) when the conditions were only semi-ideal, the sky 
* “Frost and Temperature Conditions in the Cranberry Marshes of Wisconsin,” by 
H. J. Cox, U.S. Department of Agriculture, Weather Bureau, Bulletin T, 1910. 
