the Beds of Running Streams. 343 



the water was 31°'9, air 35°. The ice consisted of small thin 

 plates interwoven with one another in every possible manner ; 

 where the water was not more than 2 feet deep, and the current 

 ran at 2^ miles an hour, the ice spread over the bed of the 

 stream to a depth of from 2 to 4 inches. At the edges on the 

 surface there was not much ice; in thickness it was about f of 

 an inch at the edge, and thinning off quickly to a shell edge 

 within 1 foot from the side of the bank. On the ditches in 

 the neighbourhood the ice was about 1 inch thick, which at once 

 broke when an attempt was made to stand upon it. In the 

 forenoon of the 13th the sun shone with considerable power ; 

 this had the effect of producing a sufficient increment of tempe- 

 rature to detach the ice from the bed of the stream, and when I 

 left the Brook the ground ice had begun to rise rapidly. I have 

 observed that changes of this kind are very sudden ; in half an 

 hour a large portion of the ground ice will often disappear ; it 

 rises to the surface and floats away with the current. It often 

 gets obstructed in its passage down, when the surface of the 

 stream becomes choked with loose floating ice. In some places 

 inundations follow, which render these accumulations a source 

 of anxiety to the neighbouring proprietors. 



I have at intervals of several years examined this locality for 

 ground ice when the state of the weather seemed to be favourable 

 for its formation, and rarely been disappointed. One or two 

 nights of sharp frost with wind appears to produce it. The view 

 that I take of the mode of its production is, that the wind acting 

 on the ditches in the exposed flat district soon cools them down 

 to 32°, and at the beginning of the frost forms scales of ice on 

 their surface, which are carried along by a slow current till they 

 reach the Brook ; here the velocity of the water increases, and 

 the light particles of ice are mixed up with the water by the 

 current and get entangled by asperities in the bed of the stream, 

 consisting chiefly of stones and portions of plants. In some 

 parts where the stream runs slowly there is no ground ice; but 

 if there be a short space of it with a more rapid current, there 

 the accumulation of ice is found in plenty. 



The frost continued from the 13th of February onward j some 

 days later I again examined the bed of the Brook, and found 

 nearly as much ground ice as on the first occasion ; it now wore 

 a gray aspect, from minute particles of clay and sand lodged 

 amongst it by the current of water. Under two light stone arches 

 over the rivulet I found it ; under one of them, where the bed of 

 the stream and the velocity of the water were favourable for the 

 accumulation of this kind of ice, it existed in the greatest possible 

 abundance. This observation recalled to my recollection a remark 

 made by Mr. Elliot, while discussing M, Arago's theory of the 



