22 



Mr. C. Chree on the 



The following table, mainly taken from Christiansen's 

 paper, gives a summary of his results. K denotes the ratio 

 of the conductivity of the liquid mentioned, at the tem- 

 perature i(T 2 + T 3 ), to that of air at £(T, + T 2 ) ; while K' is 

 the value found for the liquid by Weber in his experiments 

 at low temperatures, i. e. about 4°. I have calculated the last 

 column, assuming for the conductivity of air at 0° the value 

 •000049* in C.G.8. units, and taking 7 = '0018; k refers to 

 the temperature ^(T 2 + T 3 ), and belongs to the liquid men- 

 tioned opposite, the unit of time being one minute. 





KTx+T s ). 



i(T 2 +T 3 ). 



Z. 



Z/Z'. 



k. 



Water 



22-32 

 33-51 

 24-88 

 23-94 

 34-85 

 26-90 

 39-55 

 26-17 



11-67 

 15-78 

 13-65 

 11-57 

 15-46 

 13-71 

 1869 

 13-14 



21-09 



20-87 



7-82 



12-64 



12-49 



6-73 



6-66 



6-52 



283 



268 

 314 



288 



310 



•0645 

 •0650 

 •0240 

 •0387 

 •0390 

 •0207 

 •0210 

 •0201 





Alcohol 







Olive-oil 





Citron-oil 





It seems worth noticing that in each case the results are 

 consistent with an increase in the conductivity as the tempe- 

 rature rises. 



Herr Graetzf has carried out a considerable number of 

 experiments by a novel method. By means of a simple 

 arrangement the liquid is forced under a determinate constant 

 pressure through a straight horizontal pipe of very small 

 radius. The pipe is immersed in a bath, through which flows 

 water of a constant low temperature T . The ends of the pipe 

 are let into corks fitting in apertures in the sides of the bath. 

 The corks also fit into the mouths of two larger tubes, one of 

 which brings the liquid to the pipe, while the other carries it 

 away. The entering liquid is warmed in a bath, and has its 

 temperature T x determined by means of a fixed thermometer 

 just before it reaches the mouth of the pipe. A similar ther- 

 mometer gives the temperature U of the liquid just after 

 'leaving the pipe. The quantity of liquid traversing the pipe 

 per minute is also observed. 



Grraetz's two papers treat of two different series of experi- 

 ments, which differed, however, only in the pipes employed 

 and the velocity with which the liquids traversed them. In 

 the earlier series the pipe was of copper, whose dimensions are 



* Wullner, Wied. Ann. iv. p. 321, gives '000045 as the correct theo- 

 retical value ; values obtained by experiment seem mostly to exceed that 

 employed above. 



t Wied. Ann. xviii. p. 70, and xxv. p. 337. 



