Observations on Earth Temperatures. 177 
the warm summer of 1893 seems to have been stored up in the 
deeper layers of the soil, for the 4 ft. temperature in the following 
winter—Jan. and Feb., 1894—did not fall so low by 3° as it had 
done in either of the preceding years. In consequence, perhaps, 
partly of this underground store of heat the early spring of 1894 
was remarkably forward up to the middle of April; but then a 
cold and wet period set in, marked by a concavity in the curve 
of ascent of the earth temperature. The highest temperature 
attained at 4ft. depth was 60-1, on Aug. 8th and 9th. Late 
autumn was mild, and the curve of 1894 then again rose above 
those of previous years. 
Ihave spoken of the storage of heat in the deeper layers of the 
soil, i.e. below the 4 ft. level. That such storage takes place is 
shown by the circumstance that in winter time, when the tem- 
perature at 4 ft. depth is higher than the mean temperature of 
the air, and also than the temperature at 1 ft. depth, daily 
observations show that a rise in the 1 ft. temperature is followed 
by a rise in the 4 ft. temperature, even when the latter is already 
higher than the former. Under such circumstances the 4 ft. 
temperature marks the fluctuating level of the resultant between 
the supply of heat from below and the loss of heat at the surface ; 
when the loss is lessened the level of the ground temperature 
rises, just as the level of the water rises in a well when pump- 
ing, though not discontinued, is reduced below the yield of the 
springs. 
Another point shown by daily observations is that the tempera- 
ture of the ground at 1 ft., and, still more, that at 4 ft. in depth, 
not only have a less range and fewer fluctuations than the mean 
daily temperature of the air, but also that in their movements 
the ground temperatures lag behind the latter. Any notable 
change in the daily mean temperature of the air is followed one 
or two days later by a change, similar in direction but less in 
amount, in the 1 ft. thermometer, and in about four days by 
@ corresponding but still smaller movement of the 4 ft. ther- 
mometer. 
Daily observations of the 1 ft. and 4 ft. thermometers during 
the frosts of February, 1894, and of January and February, 
1895, have illustrated how the loss of heat in the superficial 
layers of the soil is retarded by a covering of snow. During the 
frost of 1894 the ground was unprotected by snow, but during 
that through which we have been recently passing it has been 
covered with a layer of loose powdery snow about 3 in. deep; 
and although the cold period this year has been much more pro- 
longed than that of last year, the fall of the temperature of the 
ground at 1 ft. and 4 ft. depth has been markedly slower. 
To give figures. During the six days of frost (Feb. 17th—22nd, 
1894), the mean temperature of the air averaged 8°26° daily 
