THE PHILOSOPHY OF DEW. 



From observations made in 1838, lam led to believe that the differences in the radiating 

 power of the polished and the scratched side of the cube in Leslie's famous experiment, 

 is not owing, as was generally thought, to the diflerences in the mechanical state of the 

 two surfaces, but to a change of density resulting from the scratching. This appears, 

 moreover, to be confirmed by the three following facts : 1st. The variation in the emissive 

 power is only observable in metals; marble, jet, or ivory have the same radiating power 

 whether they are smooth or not. 2d. If silver be melted and slowly cooled in sand molds, 

 then burnished and afterwards scratched with a diamond, so that the bottom of the 

 scratches be compressed and condensed, its radiating power will be less after the scratch- 

 ing than it was before. 3d. This same piece of silver, melted and polished, has its radi- 

 ating power diminished by being hammered or laminated. 



Thermoscopic instruments similar to those described above, and having the cases cover- 

 ed with varnish, black lead, isinglass, sawdust, sand, dust and leaves, constantly indicat- 

 ed a very sensible fall of temperature before becoming moist with dew; the lapse of time 

 between the fall of temperature and the deposit of dew, sometimes amounted to several 

 hours; a fall of temperature, moreover, often occurred without any deposit of dew at all 

 during the night. This last phenomenon occurred the more frequently the higher the 

 thermometers were placed above the ground. By making your observations at a certain 

 height above the soil, you can delay or entirely prevent the deposition of dew on your in- 

 struments, and prove conclusively that it alwa}^s follows and never precedes the produc- 

 tion of cold. I have never seen the polished metallic cases of my thermometers covered 

 with condensed vapor during damp nights, provided there was no trace of fog in the at- 

 mosphere. 



Hence we see, that in order that a body may be covered by dew it must first cool to a 

 certain degree, and the reasons why metals do not become covered with dew, is that they 

 do not become sufficiently cold by radiation. But is there nothing else to hinder the pre- 

 cipitation of dew on metals? In other words, is the feeble radiating power of metals the 

 true and only reason why they are never wetted bj' dew ? 



The following experiment seems to me to answer this question, at the same time that it 

 refutes the theories that dew rises from the earth or falls from the clouds, and clearly 

 proves Well's principle: On a tin disc, as large and thin as possible, draw a concentric 

 circle with a radius equal to one-third of that of the disc, and cover it with a thick layer 

 of varnish. Then take another tin disc less by ten millimetres than the varnished circle; 

 and having soldered a pointed iron wire (2 millimetres large and 2 or 3 decimetres long,) 

 at its centre, and perpendicular to it, place the wire through a hole in the center of the great 

 disc on the varnished side. The great disc is to be puslied along the wire until the two 

 discs are about five millimetres from each other, at which distance they must be kept. 



The discs so joined are to be taken in the evening out in the fields, and left for a few mi- 

 nutes in a horizontal position, quite out of contact with any other bodies. If the night is 

 calm and fine, phenomena which one might easily foretell will soon be observed on the 

 surface of the larger disc. 



It suffices, indeed, to remember that in the position in which the instrument was left, 

 the small disc was ui)permost, and hence an annular band of the varnished part of the 

 lower disc will be exposed to the air. Now it is clear that this band will radiate heat, 

 cool, become covered with dew, and will propel cold and dew in consequence, from the side 

 next the center, and from.that next the circumference. This propagation will, however, 

 proceed much further in the latter than in the former direction, and for this reason 

 points cooled by contact will cool further by radiation, and will become covered with 



