12 
Proceedings of the Koyal Society of Edinburgh. [Sess. 
The quantities required for the forecast equation are : — 
It 
(1) The value of ^ from minimum to maximum; this can be ob- 
K 0 
tained by about 5 to 6 p.m. (see Tables I, II, III). 
(2) The lag on the day in question ; this may be observed by about 
It 
5 to 6 p.m., or may be found from the value of — (see Table IV). 
K 0 
(3) The estimated relative humidity of the coming night ; this requires 
a certain amount of practice and knowledge of local conditions, 
but fortunately a considerable error does not involve a correspond- 
ingly large error in the forecasted minimum. 
(4) The number of degrees ( 0 ) which the surface can fall below the 
4-in. temperature before the upward conduction balances the 
Ir- 
radiation ; this depends on ~ and the relative humidity (see 
Table V). 
(5) The probable difference between the air minimum over open-soil 
and the surface-soil minimum (see Table VI). 
Before discussing the forecast equation, I propose to deal with these 
in turn. The values given in Sections II, III, IV, V are all from my 
own observations during the last five months — March-July 1919, — and 
of course are probably only strictly accurate for the particular location 
and soil in which the observations were made. 
Though considerable alterations might be found in these values on 
moving from loam to clay, chalk, or sandy soils, I do not believe more 
than slight differences would be found to occur from place to place in 
the same type of soil. This is a question that requires further investiga- 
tion, and I propose to make a series of observations in different types 
of soil this autumn ; from a few observations made in September 1918 
R 
and August 1919, it appears that while the values of ^ for loam may 
H'o 
vary between ‘44 and -28, the values for sand may lie between *60 and 
*25, and for clay between *41 and ’35 under similar conditions of weather. 
R 
II. The Values of ^ in Soil of Varying Degrees of Wetness. 
R, 
R 
The value of ~ is so essential a part of the forecast equation that 
Jl-o 
it is important to know it exactly. This entails reading the minimum 
as well as the maximum at both the surface and 4-in. depth, a matter 
