106 
and these differences again are important indirect factors in determining their rela- 
tive agricultural values. The larger pore space and coarser granulation provides 
gi eater capacity and better facility for the storing of the rain as rapidly as it falls, 
and as a consequence of this difference there is much less surface drainage, less sur- 
face washing of iields, and less loss of water-soluble salts in proportion to the rainfall 
from the northern and middle-west soils. The smaller pore space, the closer texture, 
and the feebler granulation, combined with the heavier rainfall of the South, all con- 
spire to produce the excessive surface washing of the fields so generally destructive 
in the South. Whenever a heavy rain falls there the close texture and feeble granu- 
lation of the soils result in the surface pores of the field becoming quickly so com- 
pletely closed that the soil air finds little opportunity for escape, and as a result of 
this the water can only enter by the slow process of capillarity, greatly opposed by 
the trapped air. The result is that the surface soil, after having lost much of its 
coarser granulation, is more easily taken up by the water held on the surface during 
rains, and is then carried with the water-soluble salts, which have accumulated by 
capillarity and evaporation, into the drainage channels, thus giving the muddy waters 
so characteristic of southern streams. 
What, then, shall be done to establish and maintain a deeper openness and a 
coarser and stronger granulation in the soils of the South, that less of the most valu- 
able surface soil, less of the rainfall, and less of the water-soluble fertility shall be 
directly and immediately borne away in the surface drainage with every heavy rain? 
The temperature of the soil at Goldsboro has averaged for the entire growing season 
79.4° at inches, 74.9° at 12 inches, 73.4° at 24 inches, and 72.3° Fat 36 inches between 
1 and 3 o'clock p. m. At Upper Marlboro it has averaged 7.2° colder at 6 inches, 5.2° 
colder at 12 inches, 5.5° colder at 24 inches, and 6.6° colder at 36 inches. At Lancaster 
the soil temperature has been 11.6° colder than at Goldsboro at 6 inches, 9.5° colder 
at 12 inches, 9.6° colder at 24 inches, and 10.5° colder at 36 inches. At Janesville 
the mean soil temperature has been 15.3° colder than at Goldsboro at 6 inches, 35.4° 
colder at 12 inches, 13.4° colder at 24 inches, and 14.8° colder at 36 inches. 
Differences in soil temperature such as these are enough in themselves to cause 
measurable differences in rates of nitrification in the different soils. They would be 
considered important differences in greenhouse work also, but just how far these 
differences, under field conditions, may have affected quantitively the yields or the 
quality of the product produced we can not at present say. It seems at least probable 
that the differences have been influential in determining the number of days required 
for maturing the crops. 
Vhen the two soil types in each locality are compared the differences in tempera- 
ture at Goldsboro and at Upper Marlboro are about 2° F. throughout the zone of 3 
feet, the more productive soils in each locality being the colder. In the case of the 
two soil types at Lancaster and the two at Janesville the differences in temperature 
have averaged a little less than 1°. As would be expected in all four regions the soils 
having the least water capacity, or those which are most sandy, have had the highest 
temperature. 
It is reasonable to expect that the temperatures above ground which are influential 
in determining plant growth are those of the interior of the plant itself rather than 
those of the atmosphere which surrounds the plant, and as these, especially during 
sunshine, are quite different the temperature of the air at the place can not be taken as 
indicating the temperature of the growing plant. It is not an easy matter to get w T ith 
accuracy comparable temperatures from the plant itself, and on this account we have 
set up in each of the four regions in the corn on one of the soil types a thermograph, 
inclosed in a cylindrical galvanized-iron shelter provided with a conical top, the 
shelter having the same form and dimensions at the four stations and all set at 4 feet 
above the surface of the ground, so as to record the temperature in the cornfields at 
about the mean height of the corn plant. 
The mean temperatures of these shelters for seventy days, beginning June 15, are: 
For Goldsboro, 104.3°; for Upper Marlboro, 96.5°; for Lancaster, 90.1°, andfor Janes- 
ville, 87.7°. While these temperatures are not to be supposed to be the temperatures 
of the corn plants in the four regions, they do probably represent their comparative 
temperatures and are probably nearer the actual temperatures of the plant than that 
of the atmosphere would give, the figures, however, beingprobably a little higher on 
account of the influence of evaporation in lowering the temperature not being felt in 
the shelters except when wet from rain or dew. 
The Upper Marlboro record shows an average temperature 7.8° below that at Golds- 
boro; the Lancaster record 14.27° lower, and the Janesville record 16.8° lower at 2 
o'clock p. m. At 6 o'clock a. m. the mean temperature at Goldsboro for the same 
period was (59.7°; at Upper Marlboro, 64°; at Lancaster, 64.6°, and at Janesville, 
58.°, making the mean range of temperature during this period about 35° at Golds- 
boro, 33° at Upper Marlboro, 26° at Lancaster, and 30° at Janesville. 
