through Dry and Humid Air. 45 



and Moist Air," a translation of which is printed in this 

 Number of the Philosophical Magazine. From it I learn that 

 the experiments on atmospheric vapour which struck him most 

 were those performed with a tube the ends of which were not 

 stopped by plates of rock-salt. The results obtained with this 

 tube were so opposed to those obtained in another way by 

 himself, that he returned to Berlin resolved to repeat my 

 experiments. The paper just referred to contains an account of 

 his researches, and an explanation of my results. 



Operating with an open tube, he displaced by means of a 

 pair of bellows dry air by moist and moist air by dry, and 

 obtained, though not always, deflections corresponding to mine. 

 But he was particularly surprised to find that the direction in 

 which the needle moved when moist air was blown into the 

 tube, indicated, not a withdrawal of heat from the thermo- 

 electric pile, but an augmentation of heat. When dry air was 

 forced into the tube, the deflection observed did not indicate 

 that a greater amount of heat fell upon the pile, but, on the 

 contrary, that the pile was chilled. He explains these effects by 

 reference to the absorption of aqueous vapour by the lampblack 

 which coated the face of his pile. This absorption, when moist 

 air was blown against the instrument, rendered heat free; 

 when dry air, on the other hand, was forced against it, the 

 evaporation of the condensed vapour chilled the pile, and the 

 deflection due to cold was observed. From all this it is to be 

 inferred that in my experiments I have mistaken cold for heat, 

 and heat for cold, and have ascribed to absorption effects which 

 are really due to the condensation and evaporation of aqueous 

 moisture at the surface of my thermo-electric pile. 



To commit such an error, and to persist in it so long, would, I 

 fear, leave me little claim to confidence as an experimenter. But 

 the truth is that some years have elapsed since I became 

 acquainted with the facts now urged against me by Professor 

 Magnus. Experimenting years ago on dry and moist air with 

 tubes which had been coated inside by lampblack or lined with 

 blackened paper, I found, when moist air was introduced, the 

 radiation from the interior surface so energetic as to compel me 

 to abandon the coating. The promptness and energy with 

 which these effects of condensation and evaporation are produced 

 are remarkable. Dry air urged against the face of my pile on a 

 day of average humidity drives the needle of my galvanometer 

 through an arc of 196 degrees, and keeps it for a time pointing 

 to nearly 90°, from which, while the air-current continues, it 

 gradually sinks to zero. On simply stopping the current of dry 

 air, the needle springs quickly to the other side of zero and 

 swings through an arc of 120 degrees, this large deflection 



