On Night-Frosts. 51 



cause, and that the phenomenon takes place in the same way as it did in the 

 above described case of the bog-Iand. 



d) A phenomenon, according to our opinion hitherto inexplicable, has 

 been fuUy accounted for Ity the falling of tlie over-cooled niist. Dnring the 

 latter half of August and the beginning of September ve frequently see before 

 sunrise, after a dear niglit witli low température, an wnmistakahle formation 

 of ice up to the toj) of leafy trees, on hüls of moderate hcùjlit, in Valleys and 

 dells, almost everywhere. Wlien tirst observing tins phenomenon and studying 

 it, I felt quite dejected, for it seemed to put an end to the agreement with 

 naturels law which, until tlien, I had fouud prevailing in the whole phenome- 

 non of frost. 



According to this law, only the bottoms of Valleys and lower grounds 

 ought to be covered with hoar-frost, because cold air pours down from hills to 

 Valleys and is, on the former, supplieil by a warmer air. And now this 

 simple rule deduced from the law of gravitation could not be applied. Had 

 this réfrigération issued from the upper layers of the atmosphère, where, as is 

 known, the température is very low, and had it anything to do with the 

 motions prevailing in the liarometer-maximum and barometer-minimum? ^ 



This question has occupied me for several years and made me proceed in 

 my conclusions respecting the autumnfrosts with a certain doubtfulness. This, 

 has now been removed. This phenomenon has nothing whatever to do with 

 barometer-maximum and minimum, but is simply and naturally explained by 

 the falling over-cooled mist, or by the aq. vapour Condensed into intinitely 

 small drops of water. As soon as thèse drops touch the treetops or solid 

 objects on the earth in general, they turn into ice, and in this circumstance 

 the irreguler formation of hoar-frost is to be sought. 



We can quote an observation, which, besides proving that such cases of 

 rain of over-cooled water will occur early in spring, is a striking instance of 

 this phenomenon. 



About the middle of May 1891 I was staying at Breds estate, in the 

 South of Finland. The preceding evening had been calm, but the sky was 

 not qui te clear, so a severe frost was not expected. The vegetation was al- 

 ready so far advanced that cultivated meadows had begun to grow and the 

 cloverlields, in particular, were uncommonly advanced for the season. Half an 

 hour before sunrise (at 3, 20 a. m.) I walked across some fields sown with 

 clover and timothy, and found, to my utter astonishment, that hoar-frost had 



' In the central parts of a maximum the air moves downwanls, aud in a minimum, upwards. 



