On Night-Frosts. 37 



-2> 

 showing, thus, a différence of température of 



In points b and d we hâve an average of 



0,45, 



the différence consequently being 

 or, in an average, in botli points 



3",57. 



Tliis rise of température may be considered as the effect of the torches, 

 which, besides, had prevented a formation of ice in the whole field and also 

 in a vast area round it, so this phenomenon of ice-formation was prevented 

 in at least 3,2 hectares. At l,2o a. m. already, the thermometers began 

 to rise, and about 2 a. m., as at the tiist experiment, a faint breath of air 

 arose which warmed up the tield at once. ïlie day had been relatively cahn 

 with sunshine and average heat. 



EXPERIMENT III. 



June 15"'. 



In order to satisfy myself as to the correctness of my argument about 

 the effect of the cold current of air in Experiment I, i. e. that considerably 

 cooled air poured down through the valley in tlie S, from the bog-lands situat- 

 ed in that quarter, a smaller experiment was made that day in 0, 125 hectares 

 of the tield, where the tirst experiment had taken place. On the map, this 

 field is marked by dotted lines, and indicated by a V c, which, at the 

 same time give the points Avhere the thermometers were placed, in the usnal 

 way. On the thermometer-table III a is =0, h'=h, c — c and cX — d. Point 

 ti was 40 m outside the tield, point e' 62 m S, facing the valley, // in the 

 vicinity of the barn at point /?, g near point 1Î and covered with a leaf. 



Seventy torches were placed in the field and in such a manner as to be 

 nearer each other on that side of the field which faced the valley, by which 

 arrangement the cold current of air from the bog-lands was eut off'. 



