1^ M. Melloni on the Theory of Dew, 



molecules of this same vapom' at the moment of their pas- 

 sage into a liquid state. 



To shew that these different hypotheses cannot be main- 

 tained, I first took three thermometers graduated on the stalk; 

 I then passed each tube through a cork stopper, and fixed it 

 at 5 or 6 millimetres from the bulb. This stopper served 

 as a point of support to two parts of the metallic apparatus, 

 in which I enclosed my thermometers intended for experi- 

 ments on nocturnal cold. The first is a small vessel of silver 

 r copper, very thin, resembling a thimble, and having its 

 surface smooth and polished, and of sufficient size to contain 

 the bulb of the instrument. The second is composed of a 

 cylinder of tin, closed at one extremity and open at the other, 

 forming an envelope to a graduated tube. These two me- 

 tallic pieces (which can be taken ofi^ and put on with the 

 greatest ease) are easily kept in their place by the pressure 

 and elasticity of the cork. 



Now imagine three wide recipients of tin, having a lateral 

 opening by which we may introduce, close to the bottom, the 

 armed bulbs of three thermometers, leaving their stalks and 

 4invelopes disposed horizontally on the outside ; imagine these 

 recipients supported by slender tubes of metal, provided 

 with covers of the same substance, the whole exposed to the 

 free air in a calm and clear night ; and remembering that one 

 of the thermometer's armature is blackened, the two others 

 in their natural state, and the recipients sometimes open and 

 sometimes closed, — you will have an idea of the experiments 

 J instituted to compare the nocturnal radiation of silver with 

 that of lamp-black. Let us suppose the recipients to be first 

 closed ; our three thermometers then indicate the same tem- 

 perature. Let us then leave one of the vessels containing 

 the metallic thermometers closed, and open the two others. 

 A^ry delicate instruments and very minute comparisons are 

 requisite to observe and measure the extremely slight sink- 

 ing produced in the metallic thermometer exposed to the 

 sky ; but the blackened one will visibly sink, and, after some 

 minutes, will indicate three or four degrees less than the 

 thermometer in the closed vessel, — an evident proof that this 

 difference is owing to the calorific radiation which the black- 



