66 Proceedings of the Royal Society of Edinburgh. [Sess. 
(b) The Seasonal Changes of Conductivity. 
If we combine the various effects of the agencies already considered 
R 
in a general review of the year, we should expect to find that ~ had the 
following values during the seasons : — 
(1) A high value in winter, due to percolation of water — except during 
periods of frost and lying snow, which will reduce the value in the late 
winter. 
(2) A rise in early spring after frost is over and when the longer 
sunny days with frequent showers have become a feature of the weather. 
(3) A fall in summer due to the effect of a dry surface mulch. 
(4) A second rise in autumn when damper conditions have set in and 
the sun is still strong. 
(5) A return to the normal winter value in November. 
R 
In Table VII are given the mean values of ~ for each month compared 
with the monthly rainfall ; there appears to be a very close connection 
R 
between the value of ^ and the rainfall, and perhaps an even closer 
connection between it and the days of precipitation ; this is particularly 
noticeable in January and October 1919, when 50 millimetres fell in both 
months, but on 24 days in the former and 10 days in the latter month — 
the corresponding values of 
^ being *40 and *34 respectively. 
±t 0 
T> 
Table VII.— Average Monthly Values of -Land the Rainfall. 
Rft 
1919. 
Month. 
R4 j 
R 0 
Rainfall in 
Millimetres. 
No. of Days of 
Precipitation. 
Remarks. 
January .... 
•40 
50 
24 
February 
•36 
50 
16 
Hard frost. 
March .... 
•30 
32 
22 
April .... 
•35 
70 
16 
May .... 
•34 
17 
10 
Sunny and warm. 
June .... 
•35 
40 
17 
July .... 
•32 
17 
8 
Sunny and dry. 
August .... 
•33 
40 
13 
Sunny. 
September 
•35 
52 
13 
October .... 
•34 
52 
10 
November 
•42 
115 
24 
Very wet. 
December 
•40 
95 
23 
Section V. — Heat Transfer in the Soil. 
With the evidence of the preceding sections before us, there seems no 
doubt that percolation plays a very large part in the transfer of heat in the 
