rent than a vacuum. " No imaginable source of error," they 

 affirm, " has here been left out of account." The arrangement 

 for filling with moist air was varied, the air allowed to stand 

 for a long time over the water in the gas-holder, and this moist 

 air then passed through several wash-bottles into the experi- 

 mental space, but with the same negative result. In common 

 with Magnus, MM. Lecher and Pernter ascribe my results to 

 the condensation of liquid films on the rock-salt plates, and on 

 the polished inner surface of my tube*. 



§ 4. Experiments resumed : Verifications. 



With a view to my own instruction and to the removal of 

 uncertainty from other minds, these researches on radiant 

 heat were resumed in November 1880. A brass experimental 

 tube 4 feet long, 2|- inches in diameter, and polished within, 

 was first employed"! - . Interposed between it and the source 

 was a "front chamber," through which, when exhausted, the 

 rays passed into the experimental tube. A plate of transpa- 

 rent rock-salt separated the tube from the chamber, while a 

 second plate of salt closed the distant end of the experimental 

 tube. The source of heat was at first a Leslie's cube contain- 

 ing water at 100° C, to one of the faces of which the end of 

 the front chamber was carefully soldered. The chamber also 

 passed air-tight through a copper cell in which a continuous 

 circulation of cold water was kept up. The heat which might 

 otherwise have reached the experimental tube by conduction 

 from the source was thus cut off. One face of a thermopile, 

 provided with a reflecting cone, received the rays which passed 

 through the experimental tube. The other face, also provided 

 with a cone, received the rays from a " compensating cube," 

 used, as formerly, to neutralize the radiation from the source, 

 and to bring the needle of the galvanometer to zero when the 

 experimental tube was exhausted. On the entrance, then, of 



* However I may otherwise differ from MM. Lecher and Pernter, 

 I agree with their opening remark, that few other questions of experi- 

 mental physics present difficulties so great as the one here under consi- 

 deration. Nor do I see reason to differ from their closing words, that " the 

 extraordinary difficulty of investigations of this sort would be richly repaid 

 hy the attainment of quantitative results ; whilst the corresponding optical 

 investigations (immeasurably easier) will always remain more of a quali- 

 tative nature." It is the difficulty here signalized that has caused so 

 many distinguished investigators to go astray in this field of inquiry. I 

 may state here that on the receipt of their paper I wrote to MM. Lecher 

 and Pernter, but my communication was returned from Vienna through 

 the dead-letter office. 



t The plate answering to this description will be found in the Philoso- 

 phical Transactions for 1861. 



