38 NORTH AMERICAN FAUNA. [no. 27. 



vious week, and on the' 26th and 27th it became still warmer and the 

 ice practically ceased running. On October 28, however, the weather 

 became colder and the ice again appeared. From this date it contin- 

 ued to run and the snow steadily accumulated. The thermometer did 

 not rise above the freezing point, and consequently there was no 

 thawing, except to a very slight extent in sheltered spots directly 

 exposed to the sun. While the river remained open its expanse of 

 ice-laden water added greatly to the intensity of the cold. From the 

 same cause the trees and shrubs were nightly loaded with ice crystals. 

 On the night of November 7 the minimum temperature recorded was 

 — 10°, and on the night of November 15 a temperature of — 27° was 

 registered. 



During the night of November 18 the river finally set fast. In 

 this process the drifting ice accumulates until it has so filled the river 

 that it jams in some place where it has partially bridged the channel. 

 Against this barrier the oncoming fioes, laden with saturated snow, 

 are pushed by the current, and becoming lodged in all possible posi- 

 tions are almost instantly cemented together by the intense cold. 

 This process continues upstream, usually without interruption, until 

 the whole river is closed. As a result the surface of the river becomes 

 extremely rough and is almost impassable until the drifting snow 

 fills the interstices and the projecting points are somewhat rounded 

 off b}r evaporation. The time of the closing of a river depends greatly 

 on the height of water. 



During November and December the cold steadily increased in 

 intensity and the snow gradually accumulated. When no thaw 

 occurs the snow remains so light and powdery that it does not settle 

 appreciably. On the shortest days the sun rose in the southeast about 

 9.30 o'clock and, after describing a low arc over the tree tops, set 

 about 2.30. Even at midday its heat was scarcely appreciable. After 

 the middle of December the thermometer scarcely ever rose above zero. 

 From January 1 to March 12, 1904, it rose above zero on only 

 eight occasions, as follows: January 6, 3°; January 7, 1°; January 

 28, 2° ; February 24, 2° ; February 25, 5° ; February 27, 4° ; March 4, 

 4°; March 5, 4°. The lowest temperature recorded was — 54°, on 

 January 20 and 21. During the third week in January the average 

 daily maximum was — 30°; the average daily minimum — 45°. 



A grave which was dug on February 20, 1904, afforded an oppor- 

 tunity to ascertain the depth to which frost had penetrated. The 

 location was a sandy knoll somewhat sheltered on the north by a 

 thick growth of young trees and open to the south. Snow lay to a 

 depth of about 3 feet. The frost had reached a depth of only 20 



