INVESTIGATED BY THK MKTHOD OF .TET VIBRATION. 381 



do not appreciably alter the surface-tension, as far as quite fresh surfaces are 

 concerned. 



As the result of my experiments I fix the initial value of the surface-tension of 



water as 



T 15 = 74-30 dyne/cm. 



The surface-tension of water has been so often determined, and in so many manners, 

 that even a moderately exhaustive representation of the results is impossible and, 

 besides, without great interest, as many of the measurements have but little value. In 

 Table XIV. are shown the results of a few determinations by the capillary-tube method, 

 and Table XV. contains most of the results obtained by the method of capillary ripples. 



With respect to the values found by the capillary-tube method, it appears that, 

 with the exception of QUINCKE'S values, they are all smaller than those found here. 

 This is quite natural, because it is the stationary value of the surface-tension that is 

 measured by the capillary-tube method. Under the given conditions this value must 

 be smaller than the initial value. 



The values found by the method of capillary ripples in most cases agree well with 

 the value found here. They are as follows : Lord RAYLEIOH, 74*35 dyne/cm. ; 

 DORSEY, 7372 ; WATSON, 7476 ; and KALAHNE, 74'22 dyne/cm. The mean value 



for all four is 



T 1S = 74-26 dyne/cm. 



An exception from this, however, is made by GRUNMACH'S measurements (BRUMMEK 

 and LOEWENFELD, who worked exactly in the same manner as GRUNMACH, are not 

 mentioned here). 



GRUNMACH'S measurements divide themselves into two groups, the surface of the 

 liquid being either the same during the investigation or continually renewed. 

 Table XV. shows that GRUNMACH'S value for distilled water in the first case is 



T 1& = 78-41 dyne/cm., 



and in the other 



T 16 = 75-89 dyne/cm. 



These results are very extraordinary, for two reasons. Firstly, it would be expected 

 that the former value would agree with the values found with the same method by 

 other investigators. This, however, is far from being the case, as GRUNMACH'S value, 

 78*41 dyne/cm., is 5*6 per cent, larger than the corresponding mean value, 74'26 

 dyne/cm., of the other measurements. 



Secondly, it would be supposed that the surface-tension for the continually renewed 

 surface would be the greater. GRUNMACH, however, came to the opposite result and 

 found a value 3 '3 per cent, lower in this case. 



A satisfactory explanation of this circumstance will certainly demand fresh investi- 

 gations, and before these are finished it will be difficult to judge of the value of 



