M. Melloni on the Theory of Dew, 133 



or accumulation of dew on metals ? In other words, is the 

 feeble radiation of metals the only and true cause on account 

 of which these substances are never moistened with dew % 



The following experiment appears to me to decide this 

 question completely, and it demonstrates, at the same time, 

 the erroneousness of the hypothesis of the rising or falling 

 of dew, as well as the truth of Wells' principle. 



On a disc of tin, of a single piece as broad and thin as pos- 

 sible, I traced a concentric circle of a radius equal to one- 

 third of that of the disc, and covered it with a thick coat of 

 varnish. I then cut another sheet of tinned iron, a second 

 disc of 10 millimetres narrower than the varnished circle ; 

 and having soldered to its centre, and perpendicularly to its 

 surface, the extremity of a wire 2 millimetres thick, and from 

 2 to 3 decimetres long, terminating in a point at the free 

 extremity, I perforated the large disc with a small central 

 opening, and introduced the wire into it on the side of the 

 varnished surface. The moveable piece was then pushed 

 along this wire till its distance from the small disc was re* 

 duced to about 5 millimetres ; it was then fixed in this posi- 

 tion by means of some drops of solder. 



The two plates, thus united into a single system, are con- 

 veyed in the evening to the middle of the fields, and left for 

 a few minutes to themselves in a horizontal position, and far 

 from the contact of any other body. If the night be calm 

 and clear, we then easily and distinctly see dew produced 

 on the surface of the large disc. 



It is sufficient, indeed, to remember that in the position we 

 are supposing the two discs to occupy, the smallest is upper- 

 most, and as its radius is 5 millimetres less than the varnished 

 circle of the large inferior disc, itTollows that an annular 

 band of this circle, 5 millimetres broad, will stand out h-round 

 the vertical projection of the roof formed over it by the small 

 metallic disc above. It is clear that this band will radiate 

 towards the sky, will diminish its temperature, and become 

 covered with dew, and will by degrees propagate the cold 

 and dew consecutively from the side to the centre, and from 

 the side to the circumference. The propagation, however, 

 will go much further in the latter case, for the points cooled 



