he streams
in which anchor frosts occur generally are those which contain
water of different temperatures--viz., surface-drainage and land
springs and main springs, the first being always colder than the
latter, in winter these never being less than 40 degrees, even in
severe frosts.

These colder globules being first frozen, float on the surface of
the water individually, being prevented from coalescing by the
intermediate main-spring-water, and where the water passes in a
shallow stream over the pebbles the crystals are intercepted by
the interstices of the stones, and then become heaped together in
thick beds.

The fact of the crystals of ice (which are specifically lighter
than the water) sinking below the surface, is a circumstance
requiring explanation. They do not sink from their specific
gravity, but in the commotion of the current they are occasionally
submerged, and while so are stopped by any obstruction, when they
commence and compose the aggregation.

Thinking this was an erroneous view of the matter, I replied as
follows:--

J. M., in his remarks on anchor frosts, appears to me to have
fallen into several errors in endeavouring to account for them
(they are called bottom frosts in Yorkshire); for, admitting that
main springs are of the temperature stated (40 degrees) when they
issue from the earth, I am by no means prepared to believe that
they keep that temperature long, or that the water issuing from
them does not mingle intimately and immediately with the water of
the river into which it flows; especially in the situations where
anchor frosts are most common, which are rough and rapid streams.

From J. M.'s statement it would appear that globules of water of
different temperatures mix together without the one imparting its
excess of caloric to the other, which is contrary to the
experience of everyone; it is true, that in still places there
will be different temperatures in the same body of water, but it
is not owing to the main springs of which J. M. speaks, but to the
peculiar way in which water is affected by cold. It is well known
that water increases in density down to 40 degrees, below which
temperature it begins to expand, and this expansion continues
until it reaches the freezing-point, so that in severe frosts
there will be strata of different temperatures from 32 degrees to
40 degrees. Again, he says that "the crystals of ice are
intercepted by the interstices of the stones, and then