Absorption of Heat on the Formation of Dew. 113 



steam still existed in the middle of the heated air. If it mixed 

 with the air, it must, on the edge (where it came in contact with 

 colder air), have been condensed and converted into mist. Now 

 the question here simply is, whether there was any condensation, 

 whether mist was present ; for it has never been doubted that 

 mist absorbs heat very well, and therefore radiates it equally well. 

 It appeared desirable therefore to make the experiment in an 

 altered form, and to compare the radiation of dry air with that 

 of moist, as well as of some other gases and vapours. 



For this purpose the gases or vapours whose radiation was to 

 be determined were passed through a brass tube 15 millims. in 

 internal diameter, and placed in a horizontal position. This was 

 heated to redness by gas-flames. One end was bent upwards, 

 so that the heated air ascended vertically. At a distance of 400 

 millims. from this ascending current of air, the thermopile with 

 its two conical reflectors was placed. To protect this against 

 lateral radiation, it was placed in a box which was a metre long- 

 by 0-6 metre in height and the same in breadth. The part of 

 this box turned to the hot current of air was of polished sheet 

 zinc. In front of this there was a double metallic screen, to cut 

 off the heat radiated by the heated tube. In the zinc side, as 

 well as in the screen, was an aperture 50 millims. in height by 

 25 millims. in breadth, the middle of which was in the prolon- 

 gation of the axis of the thermopile. The bent part of the tube 

 from which the heated air issued was 45 millims. from the side 

 of the box, but was at such a depth below the aperture that no 

 heat-rays could pass from it to the pile. The horizontal part of 

 the brass tube was at an angle of 40° with the plane of the side 

 of the box; thus neither the flames which served for heating, nor 

 the products of combustion arising from them, especially the 

 carbonic acid, could radiate against the pile, and the heated part 

 of the tube was as far as possible from the side. In spite of this 

 distance of the tube and of the interposed screen, the front side 

 of the box became heated and radiated towards the pile. To 

 neutralize the current thus occasioned, a vessel was placed in 

 front of the other cone of the pile, in which water was kept boil- 

 ng by steam being passed into it. By a screen moved by means 

 of a screw, in a similar manner to that which Professor Tyndall 

 used in his experiments, and Professor Frankland in the experi- 

 ment mentioned, the cone was so far shaded that an equally 

 strong current in the opposite direction was produced in all 

 cases. 



Recourse must be had to this method of compensation in those 

 cases in which, as in this, a gradual heating of the pile is unavoid- 

 able; but it is far inferior to that of Melloni; for the action 

 of the pile depends on the difference between its temperature 



Phil. Mag. S. 4. Vol. 32. No. 214. August 1866. I 



