On Nifjht-Frosts. 39 



of tlie heating are seeii fioin Curves III„, III,, and III„ Pl. IV. The iippev 

 curves for T,, are every nearly ideiitical, and the lower also lesemble each 

 otlier. About midniglit the température is lowest, and then a rising occurs, 

 very slow in the heginning, but afterwards more rapid. 



On account of a sliglit breath of wind the experiment was tinished on the 

 iC" June at 1 a. m. 



On comparing Experiment I with Experiment III, we shall easily tind 

 that the cold current of air which passed through the valley and was of such 

 marked effect in Experiment I, is no longer perceptible in Experiment III. 

 The uninterrrupted course of Curves III prove this beyond all doubt. 



Of course, the objection might be made that the number of torches, in 

 proportion to the area, was nnich larger in the latter case than in the former, 

 hut this objection will lose in importance when we remember that a larger 

 nnmber of torches was placed on the soutli border of the field, which border 

 was about 4Ü m in length, so there Avas almost 1 toi'ch to every meter, while 

 there remained only 30 torches for the rest of the üeld. As will be sliowu 

 below, a smaller lield will require, in proportion, a larger number of torches, 

 and consequently, the two experiments must be considered fully equal to 

 eacli other. 



As all the lower thermometers in the field began to rise directly after 

 midnight, though the température of the surrounding field was falling slowly 

 and no formation of ice could be observed in the sheltered field, [though appear- 

 ing in the surrounding fields at midnight already] it was not considered ne- 

 cessaiy to continue the experiment after 1 a. m. From the tolerably rapid 

 rising of the theimometers in the field of experiment, it is evidently under- 

 stood that this would have continued until sun-rise. 



EXPERIMKNT IV. 



June 22"''. 



The weather which, had been calm and summer-like, changed on the 17"" 

 and become windy, with a falling thermometer which continued until the 18"' 

 and 19"' June. On the last-named day, at sun-set 9,25 p. m., the wind feil almost 

 entirely, and, before long, the sky was covered with a tliin veil of clouds. 

 The température, which at 4 p. m. had been + 10", did not reach 0" tili 10. 

 45. p. m., thus, 1 hour 20 minutes after sunset, and, constantly lisiiig and 

 falling, it remained + 1° about this point. This was caused by a faint NW. 



