of a Liquid Mass devoid of Weight 4 291 



the liquid of which the film is formed, h the height to which 

 the same liquid rises in a capillary tube of 1 millimetre internal 

 diameter, d the diameter of the bubble, and lastly, let p. denote 

 the pressure which this bubble exerts, or, more precisely speak- 

 ing, the height of the column with which it would be in equili- 

 brium ; then this pressure is expressed by the formula 

 2hp 



The product lip is, as can easily be shown, proportional to the 

 cohesion of the liquid; the pressure exerted by a bubble upon 

 its enclosed air is consequently in direct ratio to the cohesion of 

 the liquid, and in inverse ratio to the diameter of the bubble. 



I verify my formula by the experiment with the glyceric liquid. 

 By means of my apparatus, which is merely Mr. Henry's slightly 

 modified, a bubble is inflated at the orifice of a small inverted 

 funnel which communicates with a water-manometer. The 

 difference of the level in the two branches of this instrument is 

 measured by means of a cathetometer ; and the latter is likewise 

 employed to measure the diameter of the bubble, for which pur- 

 pose it is placed in a horizontal position on suitable supports. 



The formula gives hd=2hp, 



which shows that the product of the pressure by the diameter 

 must be constant for the same liquid and at the same tempera- 

 ture, since under these conditions h and p do not vary :. it is this 

 constant which I first of all sought to verify. These measure- 

 ments have been made by means of ten bubbles, of which the 

 smallest had a diameter of 7*55 millimetres, and the longest a 

 diameter of 48*1 millimetres, and consequently within limits 

 which were to one another nearly as 1 to 6; the temperature 

 ranged between 18°- 5 and 20°. 



The mean of the ten values obtained for the product pd is 

 22*75. Except in the case of the two largest diameters, there 

 was very little difference from the general mean; and if the 

 results are arranged in order, with the diameters increasing, 

 it will be perceived that these small discrepancies are irregularly 

 distributed. The two values which form exceptions are 20*57 

 and 26*45, and it is seen that the first is under the mean, whilst 

 the second is over it. As the other eight values presented a 

 remarkable agreement, I have deemed it allowable to reject, as 

 spoiled by errors of accident, the two that I have mentioned, 

 and I thought I might take, in order to estimate the product 

 pd as regards the glyceric liquid, the mean of the eight agreeing 

 determinations, which mean is 22*56. 



It remains to compare the value of the product pd, thus de- 

 duced by experiment, with that which our formula gives ; and 



