Formation of Ground- J ce. 9 



torn, and fewer radiating spicular masses ; and in very many- 

 places, even in situations where tlie water ran pretty rapidly, 

 and with a ripple, there were plates of ice underneath 

 the surface of the water, — sometimes two, sometimes three, 

 and more in succession, of perfect transparency, separated 

 from each other by intervening running water, and some- 

 times of wide extent, stretching quite across the stream. 

 They were attached in every instance to stones, and appeared 

 to be formed by shoots from them, in the manner of crystal- 

 lizations, extending horizontally. This was visible ; — but why 

 the spiculse took this direction in forming, I cannot even offer 

 a conjecture. In no instance was this unusual formation, 

 which, I am not aware, has ever been described before, asso- 

 ciated with the production of ice at the very bottom. When 

 there was only one horizontal layer, that one by checking ra- 

 diation from beneath, might prevent the water from freezing 

 there. When there were two or more layers, no doubt, the 

 lowest was foi-med first. The lowest was always firmest ; that 

 nearest the surface least so, and most crystalline, often in 

 disjointed laminae. There was also on both these mornings, 

 in the shallower parts of the stream, and at their margins, 

 a large quantity of white honey- comb ice, formed of delicate 

 thin plates, variously intersecting, which seemed to owe its 

 origin to the same cause as the grovmd-ice, and to I'ise from 

 beneath the surface of the water, resting and attached to 

 stones, and above its surface (which it was in some places to 

 the extent of two or three inches), owing to crystallization 

 and attraction of the capillary kind, the water being draAvn 

 up by the laminae as they formed, and congealing in the act, 

 much in the same manner as crystals are seen to form on 

 the necks of bottles holding certain saline solutions. 



After the 21st, the extraordinary cold weather terminated. 

 During the day the sky became overcast ; the temperature 

 rose to 41°, and the ground-ice rapidly vanished. During 

 the next thirty hours there was heavy rain, 2.94 inches fall- 

 ing, which converted the rivulets into torrents, and swept 

 away all traces of ice. On the 23d, in the afternoon, the 

 temperature of the air was 55° ; that of the streams was 

 singularly changed ; the tributaries very much swollen by 



