214 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1912. 



special influence of tiic dissociated molecules, which infliieiice Willard 

 Gibbs showed and Avhich tends to diminish the superficial tension. 

 This action of ammonia may be very simply shown by resting a 

 bubble upon a rmg of w;ire of somewhat less diameter. If we hold 

 above the bubble the stopper fi-om a bottle of diluted ammonia, the 

 bubble withdraws toward the lower side of the ring. If, on the other 

 hand, a glass containing ammonia be held beneath the bubble, the 

 action wiU be still more rapid, the bubble rising in opposition to 

 gravity and squeezmg itself through the ring. The motion occurs 

 in each case as if the bubble were inconvenienced by the smell of the 

 ammonia. If the bubble be too large to lise through the ring, a tear 

 is formed, indicating its distress. Naturally, the cause of these actions 

 is the increase of tension of the liquid sheet on the side of the ring to 

 which the ammonia is applied, and if the bubble is too large, this 

 increase of tension attracts a little of the liquid from the rest of the 

 bubble and from the wire. This it is which forms the tear. 



Dupre proved long ago that the speed with which soap bubbles 

 burst is determined by the equality of energy in the movement of the 

 little drops discharged at the speed of the retreating edge to that 

 v/hich is necessary to draw out the liquid sheet in opposition to its 

 own tension. This may be expressed in Newton's manner in the 

 following way: If the tension of a sheet of soap water be sufficient 

 to support the weight of a certain number of feet in a sheet of a certain 

 thickness or color, the speed v/ith which a sheet of this thickness or 

 color wlU break is the same as the speed acquired by a stone which 

 has fallen freely this same number of feet under the influence of 

 gravity. This is manifestly based on the suj^position that the liquid 

 is perfectly mobile. When the rigidity and viscosity increase, the 

 speed is reduced. For example, a solution of saponm has a surface 

 tension 50 per cent greater than that of a solution of soap. So a. 

 saponin bubble should burst more quickly than a soap bubble if 

 surface tension only were of importance. For the benefit of those 

 who are not familiar vnth saponin bubbles I will show you one as a 

 curiosity (pi. 1, fig. 1). I shall next make a froth of saponin and 

 gljxerin in a cell in the lantern which you see projected on the screen. 

 I may continue the operation until I obtain the ordinary appearance 

 of froth. But the cells which are formed soon begin to burst. You 

 may see the free edge retreat with a slow, iiTegular movement, which 

 is due to the fact that the liquid is far from being a perfect fluid. Hence 

 I conclude that a soap bubble will burst rather less quickly than the 

 calculation would indicate. Mr. Bull could easily show you this by 

 the aid of his very powerful micro-cinematograph. 



In marked contrast with the slow bursting of this particular bubble, 

 I can show you that the speed of a true soap bubble's bursting, which 

 may be as great as that of an express train, may be rendered visible 



