4 Mr. C. Chree on the 



ideas derived from Prof. Tyndall's experiments on radiant 

 heat, and seemed ready to attach at least sufficient importance 

 to this agency. He assured himself, however, that the inser- 

 tion in the liquid of a thin film of Swedish filtering-paper, 

 which would have a great influence on radiant heat, produced 

 scarcely any effect in his experiment. 



The other possible disturbing cause is diffusion. Guthrie, 

 however, painted the base of his upper cone with a soluble 

 aniline dye, but observed no trace of colour in the liquid near 

 the lower cone, and thence concluded that no sensible part of 

 the phenomenon was due to this agency. These results are 

 of fundamental importance from their bearing on the methods 

 of subsequent observers. 



Herr Lundquist*, in a paper' which I havejbeen unable to 

 meet with, obtained, by the same method as Angstrom em- 

 ployed for metals, absolute values for the conductivity, k, of 

 several liquids. The following results, in which the units are 

 centimetre, gramme, and minute, are given in a paper (pre- 

 sently to be considered) by Herr Weber, who considers Lund- 

 quist's method very accurate but very tedious. 



Temperature, k, k, 



Centigrade. Lundquist. Weber. 



Water 40-8 -0937 -0953 



Salt-solution, of") 



43-9 -0897 -0901 



density 1*178, 



of U diity°l-382;} 45 ' 2 -0952 '° 872 



For the last liquid Lundquist took as the specific heat the 

 value -77, while the correct value according to Weber is '697, 

 and the corrected value of k would be '0862. Weber deter- 

 mined k at only two temperatures, the highest being 23°, and 

 deduced the third column of the preceding Table by assuming 

 the conductivity to vary uniformly with the temperature. 

 Thus the agreement of his results and Lundquist's does not, 

 as will more fully appear subsequently, form so convincing a 

 proof of their accuracy as might be thought at first sight. 



Herr Winkelmann J has investigated the conductivity of a 

 considerable number of liquids. His apparatus consisted 

 essentially of two brass cylinders, one enclosing the other. 

 The inner was completely closed, save for a small hole at 

 the centre of its upper surface in which was fastened a glass 

 tube. This tube passed through a corresponding hole in the 



* Upsala Universitets Arsskrift, 1869, p. 1. 

 + Pogg. Ann. cliii. p. 481. 



