57 



Height of Water 

 above Sea, in 

 Feet. 



Distances from Weir above which Gronnd Ice was observed rising. 



101 . . 



Weir from which distances are counted. 



107 . . 



1825 feet, where crystals of Ice were found. 





3550 feet, Weir with Ice on stones. 



inc 



1/0 . . 



4375 feet, Kathfarnham bridge. 





1§ mile, a cascade. 



176 . . 



2J miles, Templeogue bridge. 



238 . . 



3g- miles, Weir at Firhouse. 



304 . . 



4 J miles, Oldbawn bridge. 



364 . . 



5 g miles, opposite site of parchment mill. 



474 . . 



1\ miles, near Ballynascorney Gap. 



If, in addition to the rough section which may be formed from these 

 numbers, the reader bears in mind that the bed of the Dodder contains 

 masses of gravel, granite boulders, and projecting rocks, he will be 

 satisfied that the conditions required for perfectly mingling the flowing 

 water are all abundantly present. The Dodder is usually shallow, and 

 it was in this state before the frost of January ; thus the water, in fall- 

 ing over the weirs and torrential parts of its coarse, presented a very 

 thin sheet of liquid to the refrigerating influences of the air, and losses 

 of heat by surface radiation. Wherever the river flows most rapidly, 

 it is also shalloAvest and most disturbed, and the water is therefore 

 exposed at such places to the full intensity of the refrigerating actions. 

 The colder particles at surface exchange their positions and tempera- 

 tures with the particles at bottom, and a forced convection is thus 

 brought about, which reduces the temperature of the entire mass below 

 the freezing point. Another feature in the structure of the bed of the 

 river now operates to bring about condition (2). This occurs whenever 

 the water reduced below the freezing temperature arrives at the back 

 of a weir or mill dam. Id this position, the water at surface partakes 

 both of coDditioDs (1) and (2) ; but, while it freely loses its heat, it still 

 retains a small velocity. The water at bottom is now almost perfectly 

 still, and conditions (2) and (3) are much better fulfilled than at the 

 surface. In this way ground ice and surface ice may both be formed 

 nearly in the same cross section of the river. 



It may be asked, why should not freezing take place in the water 

 flowing between the bottom ice and surface ice, as well as above and 

 below? This suggests the utility of attending more precisely to the 

 physical conditions of the growth of ice crystals. The general influence 

 of rough solid substances in promo tiug crystallization is well recognised, 

 and the familiar experiment of plunging a vessel containiug water into 

 a freezing mixture shows the tendency of ice to commence its forma- 

 tion from even the most minute projections on the inside of the vessel. 

 Some experiments recently described by M. Fred. Engelhardt bear 



K. i. a. proc. — vol. x. I 



