74 
Proceedings of the Royal Society of Edinburgh. [Seas.* 
If K and K g are the conductivities of unfrozen soil and snow 
respectively, 
Then 
K x 1 K- x 15 
10 
10 
K s = — K 
s 15 
= — x *004 
15 
= -00027. 
It is obvious that so long as the air temperature was above — 15° C. 
thawing must have been going on constantly at the bottom of the snow 
layer, this thawing being the more rapid the more the air temperature 
rose. From Table XII we see that between 6 p.m. November 12 and noon 
November 13, with an average air temperature of — 3° C., the snow 
diminished 2 inches in depth : but between 6 p.m. November 13 and noon 
November 14, with an average air temperature of — 12 0° C., only 1 inch of 
snow was melted. 
Section IX. — The Coming of Spring. 
Spring, considered as the beginning of the period of active growth of 
crops, depends in the main on the soil temperature at the depth at which 
these crops grow. At first sight it would appear a simple matter — with 
the meteorological data for any year before us — to say whether that year 
had enjoyed an early or late spring; but when we study the phenological 
returns for the last thirty years the question assumes a more complex 
appearance. 
If we plot the deviations from the normal of the date of first flowering 
of any plant against the corresponding deviations from normal for some 
chosen preceding period of the values of the mean maximum, minimum, or 
soil temperatures, we do not find the closely corresponding results that we 
may have expected. As, however, mean values cover a multitude of 
variations, this is not to be wondered at, and it is doubtful if any good 
result can be obtained from monthly mean values at all. For let us 
consider two extreme months — 
(1) With a very small range of temperature and a mean equal to 
the average. 
(2) With a spell of frost for, say, 15 days and very warm weather foi- 
ls days, and a mean above the average. 
Now, if the growing temperature for the plant considered was below 
the mean for the month, No. 1 would give nearly 30 growing days, and the 
plant would appear in flower earlier than in No. 2, which gives only 
