On Mght-Frosts. 59 



SECOND METHOD. 



Just as certainlv as a repression of radiation of beat by smoke and danipness 

 breatbof will prevent frost, so will a motion in tbe layers of air, even a sligbt 

 breatb of wind, be suffident to prevent its ravages. As we bave already de- 

 monstrated, tbe air-particles of wind move obli(iuely towards tbe surface of tbe 

 eartb, causing a blending of tbe layers of air and resulting in a liigber tempe- 

 l'ature. A borizontal motion would most probably be resultless. 



As layers of warm air are to be found immediately above tbe cooled 

 layers nearest to tbe suiface of tbe eartb, my attention was fii'st directed to 

 bringing tliis beat down for tbe benefit of vegetation. Tliis metbod appears so 

 natural and so easily available tbat a series of trials was undertaken, tbeir 

 object being to pump tbis air down by means of a borizontal fan of large di- 

 mensions. 



Tbese experiments served only to sbow tbe superiority of tbe torcb- me- 

 thod, botb as regards easc in applying tbe measures of protection, and tbe 

 cost. But tbe torcbes may also serve tbe same pm-pose, i. e. tbey make tbe 

 layers of air blend, altbougli less directly. 



All tbe experiments undertaken by me bave resulted in a sligbt breeze 

 and a blending of tbe layers of air, and tbe conséquence bas been a some- 

 wbat rapid rise in tbe température. If it sbould be objected that tbis breeze 

 bad a natural cause, and was not tbe resuit of our expei'iments. we may say 

 tbat some effect, sucb as tbis produced by torcbes, is so probable tbat it can 

 be foretold tbeoretically. 



We will take a field of e. g. 500 bectares and above it a layer of air, 

 witb a temperatiu-e of 6 " at the siuface of tbe eartb rising steadily upward 

 until. at tbe heigbt of 2 m it will stand at + 5 "; in summer tbe température 

 of the air rises witb its beigbt and probably reacbes 10" to 12". ïbe pre- 

 cise beigbt at wbicb tbis température is reacbed bas not been examined, nor 

 tbe point determined at wbicb it begins to fall again, neitber bas any law for 

 tbis decrease near the earth been discovered. We only know witb certainty 

 that during a frosty night we bave a warmer layer of air between tivo 

 cooler; one ratber tbin near the surface of the eartb, the otlier deeper, above 

 tbe intermediate warm layer. We will assume tbat the air is perfectly calm, 

 but how high upward tbis calm extends we do not know witb certainty [See 

 NOTE VII]. 



If now, in tbis field we choose 5 symetrically situated surfaces of 10 

 hectares each, and set up at each surface 2,000 torches, wbicb are to be 



