[CALLBNDAR A m'leod] OBSERVATIONS OF SOIL TEMPERATURES 117 



contrast is the value g-iven next in the table, for the winter months, when 

 the ground was covered with frozen snow, so that no watei* could possibly 

 percolate. It will be observed that the same value 0015, which is less 

 than half the yearly average, was obtained again in the beginning of 

 1896. The small diifusivity is doubtless due to the extreme dryness of 

 the soil, under these conditions. 



The variations of the ditfusivity with the seasons of the year, are 

 shown more clearly in the Thermal Dift'usivity Curve, Fig. IV. The 

 diti'usivity is greatest in the spring and fall, when the temperature of the 

 ground is changing most rapidly. In the summer when the soil is fairly 

 dry, and the percolation is much diminished by evaporation, values are 

 obtained which approach more nearly the annual avei'age. 



We have also made an attempt to investigate the differences in the 

 diffusivity at different depths on the same date. We find that the differ- 

 ences are 'such as would correspond generally speaking with the state of 

 wetness of the soil at different depths. In the summer when the surface 

 is dried by evaporation, the diffusivity is much less in the upper layers. 

 In the spring and fall, the reverse appears to be the case. 



In cases where the area included between the cvirves was too small, 

 or the drawing of the tangents too uncertain, to give reliable results, the 

 separate values of the diffusivit}' for the upper and lower strata have not 

 both been given in the table. 



An obvious defect of both the methods above described for deter- 

 mining the diffusivity, is that it is necessary to assume the diffusivit}' 

 constant throughout a certain stratum of the soil, although the earth at 

 different depths may be in veiy different conditions with regard to 

 wetness. 



On this account we are proposing shortly to put in operation a new 

 and very simple method of measuring the diffusivity at any particular 

 depth, irrespective of its value in other strata of the soil. 



By sending a measured electric current through any one of the 

 thermometers, heat is locally generated at any convenient rate. The 

 consequent rise in the temperature of the soil immediately surrounding 

 the thermometer depends on the diffusivity, which is therefore readily 

 measured at any time. Owing to the extreme delicacy of the ther- 

 mometers, which can be read to a thousandth of a degree if desired, the 

 current and the rise of temperature required for such a measurement are 

 very slight, and would not appreciabh" interfere with the general meas- 

 urements of soil temperatures. 



