[CALLENDAR] OBSERVATIONS OF SOIL TEMPERATURES 65 
composition to a depth of about 8 feet 6 inches. Below this, to a depth 
of about 30 feet below the surface, extends a bed of stiff blue clay. 
Water is always found in the sand for a certain distance above the blue 
clay. It is intended to sink a tube in the neighbourhood of the ther- 
mometers, with a float to indicate the variations of water level, from time 
to time. 
A similar thermometer is also exposed on the surface of the grass 
for the observation of surface temperature, and another is placed in a 
screen of the usual pattern, four feet above the ground, to give the air 
temperature, 
DESCRIPTION oF THE ISOTHERMAL DrAGRAm. (Fia. I.) 
The results so far obtained are recorded in the accompanying dia- 
gram. The abscissa of the diagram represents time, on the scale of one 
- centimetre to one week. The ordinate represents depth, on the scale of 
one millimetre to one inch. The temperature, at any depth, at any time, 
is shown by drawing the isothermal lines for each degree Fahr. This 
is probably the most convenient way of showing the whole at one glance. 
The isothermal lines are plotted by drawing, from the observations for 
each day, the curve showing the variations of temperature with depth. 
From these curves it is easy to find the depth corresponding to each 
degree of temperature. 
A curve giving the variations of daily mean temperature of the air 
above the ground, at the rate of one millimetre to 1° F., is also shown 
on the diagram, as well as the depth of snow lying (1 mm. to 1 inch), 
and the rainfall (1 em. to 1 inch).’ 
From the 21st of October to the 3rd of November, the mean air 
temperature varied very little from 50° F.; the temperature of the soil, 
to a depth of 9 feet, was also nearly uniform, and lay, for the most part, 
between 49° and 50° F. 
A heavy fall of rain (85 inch) on November 3rd was followed by a 
rapid decrease of temperature, marking the onset of winter, on Novem- 
ber 4th. The temperature of the surface soil fell very rapidly till 
November 10th, when the rate of fall was checked by a snowfall of 4 
inches, which lay for three or four days. The ground was first frozen, 
at a depth of 4 inches, by a severe frost, on November 19th. From this 
date onwards the 4” thermometer never rose above the freezing point 
till the end of the winter. The air temperature rose to nearly 40° F. for 
the latter part of this week, but the fall of temperature near the surface 
was accelerated in a marked manner by the fall of rain on November 
25rd percolating through the half-frozen ground. 


1 For convenience of reproduction, the figures have been reduced to about half 
the scale of the originals. 
Sec. III., 1895. 5. 
