INVESTIGATED BY THE METHOD OF JET VIBRATION. 343 



main reason of the great discrepancies the different determinations of the surface- 

 tension show in relation to each other is certainly to be found in the variable 

 condition of the tested surface. These irregularities could arise from impurities, for 

 example, fat, oil, or similar substances, even the smallest portion of which is able to 

 produce a great alteration in the surface-tension. Thus Lord RAYLEIOH* has proved 

 that a film of oil not thicker than 2xlO~* mm. reduces the surface-tension of 

 water about 28 per cent., and the same author has latert proved that an oil film of 

 even 1 x 10~* mm. produces a noticeable reduction in the capillary-tension of water. 

 W. C. RONTGEN'S* experiments show that even an oil film of only 0*5 x 10~* mm. is 

 able to appreciably alter the condition of the surface. A. OBERBECK has been able 

 to detect the existence of a film of oil that was only 0'3 x 10~* mm. thick. 



Apart from contamination, the surface can undergo different changes of a chemical 

 and physical nature. In this manner the fluid with which the surface is in contact 

 can cause a chemical change in it. If the liquid under examination is a mixture or a 

 solution the concentration at the surface will in many cases be different from that in 

 the interior. 



From the above it is clear that the surface-tension of a liquid is, as a rule, not 

 constant, but varies with the time that has elapsed from the formation of the surface. 

 The value of the capillary-tension, immediately after the formation of the surface, I 

 propose to call the " initial value," while the value of the capillary-tension, when the 

 surface is sufficiently old, is called the " stationary value." Most of the methods for 

 the determination of the surface-tension give values differing from these two limiting 

 values, but it is just these two limiting values that have the greatest interest. Of 

 these the stationary value is of importance in many practical cases, but from a 

 theoretical standpoint the initial value is, without doubt, of the greater interest, as it 

 must stand in a more simple relation to the properties of the liquid than the 

 stationary value, which is dependent upon alien conditions. 



With the other methods of measuring, attempts have also been made to work with 

 quite fresh surfaces. GBUNMACH, BKUMMER and LOEWENFELD have adopted a 

 method of obtaining pure liquid surfaces originally proposed by RONTGEN.|| The 

 principle of the method is, that the liquid is conducted from below through the neck 

 of a funnel over the upper horizontal edge of which the liquid flows. This method, 

 however, does not appear to be applicable to all cases (see GRUNMACH, ' Wiss. Abh. 

 d. K. Norm.-Aich.-Komm.,' Heft III., "Experiments with Mercury"), and has at 

 least two faults, firstly, that when the surface renewal takes place somewhat quickly, 



* Lord l:\\ i nun, 'Roy. Soc. Proc.,' 48, p. 364 ('Papers III.,' p. 345); 'Phil. Mag.,' 30, p. 386, 1890 

 (' Papers III.,' p. 383). 



t I.i.v-1 l; AYI.KK.H. ' Phil. Mag.,' 48, p. 321, 1899 ('Papers IV.,' p. 415). 

 J R.-NT..KA. ' WIED. Ann.,' 41, p. 321, 1890. 

 OBERHK K. \\ n n. Ann.,' 49, p. 366, 1893. 

 || RONTGEX, '\VIK.K Ann.,' 40, p. 152, 1892. 



