become
heaped together in thick beds;" but if my observations are
correct, these depositions begin first round the large stones,
which are not likely to stop small spiculae any more than are the
water-gates of mills, where, he says, the accumulations also take
place.

Anchor frosts are most common in the rapid streams occurring below
deeps in rivers, and I have seen a weir on the river Wharfe which
had a wall of ice four feet high formed upon it in a single night
by a sharp north wind. In my opinion a sufficient explanation of
this freezing at the bottom of rivers is to be found in the fact
that water when kept still may be cooled down below the freezing-
point without being congealed; but if the vessel in which it is
kept be shaken, a portion of it will be converted into a porous,
spongy ice, and the temperature immediately rises to 32 degrees.
In the deeps of rivers the same cooling below the freezing-point
takes place without congelation, but as soon as this water reaches
the stream below, the agitation immediately converts a portion of
it into ice, which collects round the large stones at the bottom
in the same way that crystallization commences in a solution of
salt or sugar around a piece of thread or other substance which
may be suspended in it. If a severe frost is followed by a bright
day, thousands of these detached pieces of spongy ice may be seen
rising from the stones which have served as nuclei for them; which
proves that the detention of them is not merely mechanical, but
that precipitation (if I may be allowed to call it so) takes place
in the first instance, the stone serving as a nucleus, and that
this adhesion is destroyed by the action of the sun's rays.

I have never seen any attempt to explain the phenomenon of bottom-
frosts before this of J. M.'s, and I am not philosopher enough to
speak positively on the subject; but the above is the way in which
I have always endeavoured to account for it. Perhaps some of your
scientific readers may be able to give much better reasons for it
than have been offered either by J. M. or myself. (January 17th,
1832.)

Another writer (J. Carr, of Alnwick,) says that anchor frosts are
merely long and severe ones where long masses of ice are frozen to
the stones at the bottom of rapid streams, and this is simply
owing to these stones acquiring a degree of cold far below the
freezing-point, and the water in contact with them freezing and
spreading into large sheet