GROUND OR ANCHOR ICE. 321 
40° F. down to the freezing-point, it expands, and therefore becomes lighter, 
and remains floating at the surface, and then, on freezing there, it expands still 
further, and therefore still more tends to float. In rapidly-moving river water, 
on the contrary, and especially at the foot of rapids, ice is often found to grow 
attaching itself to the rocks or stones forming the bed of the river, as a spongy 
or porous mass, which, seen in the aggregate and not examined minutely, pre- 
sents a general appearance not unlike the spawn of frogs. In large rivers in 
cold climates, as, for instance, in the St. Lawrence, immense quantities of this 
ice, called ground or anchor ice, are found to accumulate with astonishing 
rapidity. These accumulations of ice, by damming up the water, cause great 
floods, and by yielding to the force of the water, and moving down with the 
current, especially after they have become jammed and heaped up with other ice 
formed on the surface, act in producing very striking geological effects in dis- 
turbing the bottom and banks of the river, and in shoving along huge boulders 
which otherwise would remain immoveable. The ground and surface-ice also, 
by their shoving-action, introduce formidable difficulties and dangers in the 
construction of bridges or other engineering works requiring to be founded on 
the beds of rivers in cold climates. Jn the construction of the Great Victoria 
Bridge across the St. Lawrence at Montreal (the most costly bridge which has 
ever been executed), these difficulties have been successfully overcome, and a 
structure has been raised which is likely to stand secure against the much- 
dreaded forces of the ice. On account of the tendency both of water approach- 
ing to the freezing-point and of ice to float, it has long been regarded as rather 
4 singular circumstance that ice should ever be found growing at the bottom 
of a river. From among the many suggestions which have been offered at 
various times to account more or less completely for the phenomenon, the author . 
gets out by accepting as quite correct the view that the essential difference 
* between the circumstances of the freezing of lake and river water is, that in the 
former case the water is left undisturbed to the action of the cold, and is allowed 
to adjust itself in strata in which the coldest parts, being also the lightest, float 
to the top; while in rivers the whole water is, by mixing, due to its rapid flow, 
brought to an uniform temperature at the freezing-point from top to bottom, and 
is thus brought into a condition in which it is ready to freeze at any part where 
additional cold may be applied. He is not, however, satisfied with any of the 
numerous suggestions which have been offered to account for the growth of the 
masses of spongy ice at the bottom, rather than that the ice should be found at 
the top, or in a state of mixtnre with the water throughout its depth. Some, 
for instance, have thought that radiation from the bottom to a cold sky (see 
paper by the Rev. James. Farquharson, Philosophical Transactions, 1835) would 
cause ice to grow at the bottom of the river much in the same way as hoar-frost 
frows on land. Arago, having rejected the supposition of radiation being the 
cause, assigned two other reasons: first, that there might be expected to bea 
peculiar aptitude to the formation of crystals on the stones and asperities at the 
bottom, like as there is found to be a special readiness for the formation of 
crystals on rough bodies in saline solutions ; and secondly, he supposed that 
the existence of less motion of the water at the bottom would favour the growth 
of the crystals there. As against this view, the author of the present paper 
Vou. VIII. Y 
* 
