6 



Mr. C. Chree on the 



where n depends only on the nature of the liquid. From the 

 three double cylinders, for each of which p is known, he thus 

 gets three equations. Any two of these give a value for n, 

 and thus three values may be deduced for K. According to 

 Winkelman, these values agreed marvellously for nearly every 

 liquid examined. The following Table contains the apparent 

 conductivities k^ k 2 , k 3 deduced from the three apparatus, in 

 which the thicknesses of the liquid layers were *2, '26, and *5 



centim. respectively. K 

 cording to Winkelmann. 



denotes the true conductivity ac- 





*,. 



Jc 2 . 



k s . 



K. 



Water 



•0624 

 •0658 . 

 •0294 

 •0357 

 •0404 



•0697 

 •0869 

 •0359 

 •0446 

 •0413 



•0850 

 •1261 

 •0650 

 •0826 

 •0435 



•0924 

 •1605 

 •0904 

 •1186 

 •0449 



Salt-solution 



Alcohol 



Carbon bisulphide .., 





The units are the centimetre, gramme, and minute, and the 

 results apply to a temperature of from 10° to 18°. 



A theory in which the corrections required are so large can 

 hardly be considered satisfactory, and in Wiedemann's Annalcn, 

 Bd. x., it meets with a very severe criticism from Herr Weber. 

 He attributes the discrepancy in the results given by the three 

 apparatus to convection -currents, which he says would, in the 

 case of the liquids of least internal friction, become very pro- 

 minent unless the thickness of the liquid layer were small. 

 Be considers the results given by the first apparatus to be 

 moderately good, but those given by the other two and Win- 

 kelmann's corrections to be useless. In the same number 

 (p. 668) there is an elaborate reply by Winkelmann, who 

 shows that Weber's remarks as to the effects of internal fluid 

 friction are scarcely borne out by the facts ; but he does not 

 succeed in proving his method to be satisfactory. 



The next observer, Herr Beetz*, has dealt with a large 

 number of liquids, including various solutions of different 

 strengths. His apparatus consists of a long cylindrical glass 

 whose axis is vertical, enclosed in a coaxial glass of slightly 

 greater diameter fused to the inner glass near its mouth. The 



* "VVied. Ann. vii. p. 435 (1879); and Sitzungsberichte der math.-pJiys. 

 Classe der k. b. Acad, der Wissenschaften, Heft i. p. 86 (1879). 



