41 
the temperature of the springs, and also to show that some 
of them have, to a certain extent, the character of warm 
springs. 
As is well known, the diurnal and annual variations in 
the temperature of the air only influence that of the ground 
and subjacent rocks to a moderate depth. Near the surface 
the temperature of the soil of course varies much from 
winter to summer, but in penetrating to a greater depth the 
amount of this difference gradually decreases, so that at a 
depth of from 50 to 100 feet the temperature is the same at 
all seasons of the year, and in this part of the globe 
corresponds very closely with the mean annual temperature 
of the locality. On proceeding to greater depths the tem- 
perature continues uniform at all seasons of the year, but 
gradually increases with the depth, on account of our 
approaching the more highly heated interior of the globe. 
Such being the facts of the case, it necessarily follows that, 
if the water of a spring has only percolated through the mere 
surface of the ground, its winter and summer temperature 
would differ considerably. If, however, it proceeds from a 
moderate depth, it would have nearly the same temperature 
at all seasons of the year, only a small variation being caused 
by the passage of the water to the surface. A careful deter- 
mination of the temperature of springs may therefore furnish 
us with a very good approximation to the mean annual tem- 
perature of the localities where they occur. It is, however, 
necessary to proceed with caution. If the temperature of 
any spring in winter differs considerably from that in summer, 
it would not be safe to place much reliance on the mean of 
the two, since the same circumstances which cause this dif- 
ference might also cause each of the detenninations to differ 
much from the truth. On the contrary, if the temperature 
after exposure to the cold weather of winter corresponds 
within a degree or two with that after the hot weather of 
