EXPERIMENT STATION BULtrETINS. 347 



tnnity o-f studying the behavior and note the number of snow crusts. 

 It was found that there Vv'as a constant melting of snow at the bottom 

 regardless of the temperature of the air; that nine distinct crusts had 

 formed during the warm Avinter days, and that the crusts rapidly dis- 

 integrated as they came nearer tlie surface of the ground, becoming 

 honey-combed at first, until when within about a foot of the ground 

 there was nothing left to show their fonner presence except a yellowish- 

 colored liorizontal streak. 



That the snow at the bottom melts quite constantly was further demon- 

 strated by another series of observations taken at the same time with a 

 galvanometer of Prof. Curtiss of the Michigan Agricultural College, these 

 observations showing that the temperature at the surface of the ground 

 remains quite constant during the winter and ranges from slightly be- 

 low to a little above 32" Fahrenheit. These observations show that the 

 l)lants are not likely to be damaged to any great extent, much less 

 killed, either by cold temperatures or by the interception of air circula- 

 tion, and that the theory of "smothering," whatever it may be else- 

 where, does not apply here. It was assumed that if the plants are 

 killed by a fungus which is capable of propagating under the snow, the 

 time of planting might be the most important question involved, and 

 the following tables of yields is arranged accordingly. When the snow 

 had all disappeared during spring, all tlje jjlots were found to be dam- 

 aged to some extent. The injured portions consisted either of long, nar- 

 row strips which followed the slope of the ground, or of solid patches 

 which were irregular to nearly round in shape. No crusts had formed 

 under the tar paper and boards which had been used for covering a 

 portion of two plots. Under this covering the snow was still four inches 

 deep when it had completely disappeared elsewhere, but there was noth- 

 ing to indicate either then or subsequently that the covering had affected 

 the plants either one way or another. The damaged portions of the 

 plants dried up within a few days after the snow had disappeared and 

 their surface Avas found covered with numerous sclerotium or fruit 

 bodies of a fiingus whose name or nature could not be ascertained. The 

 active stage of what was either the same or another fungus and which 

 was bright red colored, was found at the same time on two plots but 

 only upon three or four plants which were still healthy. The yields of 

 the plots are shown in the following table: 



