'mancy of Air and ITydroyen. 423 



the absorption of a layer of gas increases with the thickness 

 as with its density ; but for air and aqueous vapour these limits 

 must be drawn closelj. 



Melloni had shown by experiments * that a layer of air of 

 5-6 metres thickness exercises no perceptible absorption on 

 rays proceeding from a body heated to 100° C. Tyndall and 

 Wild have accepted this fact^ and indeed based upon it the dis- 

 position of their apparatus. 



But Melloni's experiments in this direction were performed 

 in an open space. The air employed, like that of any space 

 to Avhich the atmosphere has free access, must have contained 

 aqueous vapour, though this is not expressly mentioned. If the 

 absorption of moist air increased for such great distances, and 

 as considerably as Wild supposed, the experiments of Melloni 

 necessarily would have led to a different result. I conclude 

 from this that the absorption of both, dry as well as moist air, 

 very rapidly diminishes, and must soon become imperceptible 

 in a layer the thickness of which may be left out of considera- 

 tion when compared with the magnitude of the distance at 

 which Melloni had carried out the lirst of his comparative ex- 

 periments. 



The experimental part of the present investigation had 

 been completed Avhen my attention was directed to the analo- 

 gous researches ofHoorwegt. The observations of Magnus 

 and Tyndall are here elaborately and impartially compared, 

 and critically elucidated, chiefly with regard to the absorptive 

 capacity of moist air. Hoorweg's own experiments led him 

 finally to the same result which Magnus had attained, and at 

 which I, too, had arrived by my experiments performed in a 

 perfectly different manner, viz. that the absorption of moist 

 air differs but little from that of dry. Hoorweg's researches, 

 however, extend only to those rays which, before examination, 

 had been deprived of the portion for which air is atherma- 

 nous. The disproportion of Tyndall's numbers therefore natu- 

 rally escaped him. 



In contradiction to Melloni's experience, previously men- 

 tioned, Hoorweg finds that the absorption in moist air increases 

 with the length of the layer. This result of his experiments, 

 however, is open to objections. He filled the space between 

 the source of heat and the thermopile with moisture by intro- 

 ducing air which had been forced through a box containing 

 moistened fragments of quartz. It is possible that in this 

 way, along with aqueous vapour, finely divided particles of 

 liquid water had been carried along into the space. 



I pass over the conclusions regarding the thermal conditions 

 * Thennochrose, S. 136. f ^^o^^. Ann. Bd. civ. S. 385. 



