Liquids in contact with Different Gases. 407 



saturated, then it should be impossible to distinguish between 

 the surface-tension of an " air-free " liquid and one saturated 

 with air. And this is not the case, as Proctor Hall's ex- 

 periments * have shown a difference of *06 dyne per centi- 

 metre between the surface-tension of air-free water and 

 water saturated with air. 



It is probable, therefore, that the effect of solution of air 

 in water is small, but appreciable. 



Moreover, in the experiments here detailed considerable 

 trouble was at first taken to obtain gas-free liquids. But it 

 was found that, whether the liquids were gas-free or partially 

 or completely saturated with either or both of the gases 

 employed, the result was always the same, viz.: — the differ- 

 ence between the surface-tensions liquid-air and liquid-C0 2 

 remained the same, but the absolute values of the surface- 

 tensions increased slightly as the liquid became more and more 

 saturated, finally reaching a steady value. 



It must be remembered, therefore, that the figures quoted 

 below give the surface-tensions of liquids saturated with the 

 above gases. As the gases are but slightly soluble in the 

 liquids examined, these values will differ but slightly from 

 the values for gas-free liquids (cf. Proctor Hall's result 

 above). 



We now proceed to tabulate the experimental results. In 

 the table the symbols have the significance given at the 

 beginning of the paper, 1 is the temperature of the liquid 

 in the manometer and 6 the temperature of the liquid under 

 examination. Measurements of h x and 1i were made by 

 means of a Beck travelling microscope provided with a brass 

 scale divided into half-millimetres and an excellent screw, 

 which experience had shown to be dependable to within 

 2 ( y o of a millimetre or less. The radius of the tube em- 

 ployed was measured by the same microscope, ten readings 

 each being taken along four diameters increasing in orienta- 

 tion by steps of 45°. 



Observations for h x were taken when the bubbles were 

 being released very slowly, one bubble every two or three 

 seconds being the maximum rate employed. It was observed 

 that, for readings taken when the bubbles were released at a 

 faster rate than this, the value of h x was a function of the 

 speed of release, but that for values slower than this the 

 value of hi was quite constant, no difference being appre- 

 ciable between the value given when the rate of release was 

 one bubble every two or three seconds, and that given when 

 the rate was one bubble every ten or fifteen seconds. 

 * Phil. Mag. November 1893, p. 385. 



