86 ROYAL SOCIETY OF CANADA 



running and shallow stream, being all mixed together by the agitation 

 caused by the inequalities of the bottom, are all cooled down during 

 an intense frost to the freezing point, and that the stones there form 

 proper points of attachment to facilitate the formation of icy crystals. 

 This is, in fact, the same answer that was given to my theory, in 

 some of the newspapers, two years ago; but it is altogether inadequate, 

 for this plain reason, that, according to it, the phenomena of ground- 

 ice ought to appear in every hard frost, when the water reaches the 

 requisite temperature. But so far is this from being the case, that 

 in the hardest frosts which we have ever seen, not a particle of ground- 

 ice was found in the river. Take, for instance, the very severe frost 

 of 1813-14, when the Tay was frozen over for many weeks, yet no 

 ground-ice was to be seen. Some gentlemen present may remember 

 to have skated down the stream and through below the arches of the 

 bridge, whilst the ice everywhere was as clear as crystal, and the bed 

 of the stream entirely free from the white spongy ground-ice. 



Some now present will probably recollect that the theory which 

 T proposed, as a solution of these phenomena, was founded on inform- 

 ation which I had received from country people and others, whose 

 operations depended on water-wheels, and whose interests forced them 

 to attend to appearances, which might pass unheeded by others. The 

 sum of their information was, that the ground-ice was never formed 

 but after a heavy Hone, or hoar-frost. If this is the fact, the explan- 

 ation is obvious. The hoar-frost, which is congealed moisture, preci- 

 pitated from the atmosphere, and falling into the river when the water 

 is cooled down to the freezing-point, cannot be dissolved. It retains 

 in the water the very shape in which it descends from the air. Wlien 

 these small crystals fall on a deep unfrozen pool, the water being 

 above the freezing-point, the particles melt and are incorporated with 

 the water; but in the shallow and agitated stream, almost the whole 

 water is brought in succession, into contact with the intense frost, and 

 may thus be cooled down to the freezing-point to the very bottom of 

 the stream, before even a pellicle of ice is formed on the stagnant pool. 

 \11 the particles of hoar-frost, then, or frozen vapour which fall on 

 such a stream will remain unmelted ; and being tossed in all directions 

 by the agitations of the current, will be brought into contact Avith the 

 Tocks, or other substances projecting from the bottom, to which they 

 will readily adhere, and form a nucleus for that strange accumulation 

 calleà: ground-ice, which is found nowhere but in streams. 



I woidd not have brought forward this theory a second time, 

 had I not met with some facts collected by M. Arago, which afford 

 the strongest confirmation of the theory which I had advanced, though 



