40 Selim Lemström. 



wind which was continnallj^ sweeping over the field. The leaves of trees were 

 in a constant tliough feeble motion. From the expérience gained already it 

 appeared tliat a frost, in the proper sense of the word, was not to he expected 

 tliat evening, so the torches which had heen placed out, were not lighted. 



At uüdnight, véritable hoar-frost was ohserved on the grass nearest to 

 the trench and ahout 18 m up the field, yet on the ears of rye no ice-forma- 

 tion was to he seen, thongh, on the other hand, very little dew. 



Consequently, the thin veil of clouds and the faint hreath of wind were 

 enough to prevent a further fall of température. Abont 1 a. m. the breeze 

 became strenger and the température near the ground rose very rapidly 



to + 1",2. 



The 20"' June was warm, relatively calm and with a slight breeze in the 

 evening; this breeze grew in force during the 21"' June and was inthe eve- 

 ning of that day somewhat strong. 



On 22"'^ June, in the evening, another experiment was made. Two hec- 

 tares of rye-field A were measured out, but now in such a manner as to take 

 only the lower parts of the plots under shelter. On the map, this field is 

 marked with 



flo ^2 Cü and ^2 



which marks, at the same time, give the situation of the thermometers. 



The torches were now placed in a somewhat différent way. The whole 

 field was surrounded by torches at a distance of 3 meters (5 steps) from each 

 other and in the field itself, along the borders of the ditches, at a distance of 

 9 m (15 steps). The number of torches amounted to 250, or 125 torches to 

 the hectare. 



The placing ont of the torches was executed by two woi'kmen in half an 

 hour and the lighting took them another half-hour. The variations of the 

 thermometers was very nearly the same as during the preceding experiments. 

 In the field outside the shelterd area the température went down to 



-3",2 



and the différence of average température to about 



4°. 



Considei-ing that the thermometers were in a most unfavourable position 

 to the torches, the average heating may unmistakably be estimated at 



5°. 



