414 THE FIGUEES OF EQUILIBRIUM OF A LIQUID MASS 



the second place, tlie beat of the hand slightly increases the temperature of the 

 liquid of the flask, and hence diminishes the density, whence there results a ten- 

 dency for the bubble to rise, •which, occasioning it to shift to one side or another, 

 and thus altering the symmetry of action, also leads to a rupture. To remove 

 these two causes of miscarriage, I have had a flask constructed of brass, fur- 

 nished with a tap and with feet, so that, when placed on the plate of glass 

 which serves as a cover to the vessel, the orifice of the tap reaches a little higher 

 than the edge of the funnel ; into this flask I introduce the liquid necessary to 

 form the bubble, and then let it flow by the tap into the funnel with a velocity 

 which can be graduated at will, while any influence of the heat of the hand is 

 at the same time avoided.* 



§ 7. By means of the system of apparatus which I have just described^we 

 obtain without difficulty results greatly developed. By giving to the little mass 

 of oil attached to the orifice of the tube a diameter of about three centimetres, 

 I have often realized bubbles 12 centimetres in diameter, and might, doubtless, 

 have realized still larger had the vessel been of greater capacity. When the 

 dimensions just stated have been attained, if the funnel be raised by a quick 

 and dexterous movement, the bubble remains behind, and the film of which it is 

 formed being elongated by its adhesion to the orifice of the tube, a sort of train 

 is constituted, which rapidly narrows and separates into two parts ; of these the 

 lower one serves to close and complete the bubble, which is thus left entirely 

 isolated in the liquid of the vessel. In this state it continues for a longer or 

 shorter time, amounting sometimes to more than an hour, and then bursts spon- 

 taneously. Experience will soon indicate the rate of quickness with which the 

 funnel should be withdrawn : if this be too great, the bubble will burst ; if too 

 small, the bubble rises with the tube and bursts in like manner when the orifice 

 of the latter quits the surface of the alcoholic liquid. Calculation gives, as the 

 mean thickness of the film which foiTus the bubble in the above case, 0"™.3,, 

 that is, less than the third of a millimetre ; I say the mean thickness, for this is 

 not uniform, and must, at certain points, be much thinner than 0™™.3. 



It may perhaps be asked why, when such a bubble is isolated from the tube, 

 it does not last indefinitely ; nor do we see in effect in the capillary action any 

 Cixuse which should induce its rupture. It is necessary, I think, to seek that 

 cause in a remainder of chemical action exerted by the alcoholic liquid on the 

 oil. This liquid, no doubt, dissolves the film little by little, so that at the point 

 where it is thinnest it is eventually quite absorbed for a small part of its* extent. 

 As in the experiments here reported I was not able to carry the expansion of 

 the bubbles to its absolute limit, I reduced the small initial mass to a diameter 

 of 2 centimetres. The bubbles then usually broke between the diametres of 7 

 and 11 centimetres. Still I have sometimes succeeded in augmenting the di- 

 ameter to as much as 12 centimetres, which would assign to the film a mean 

 thickness of O^'^.OO, that is to say, less than the tenth of a millimetre ; but I have 

 never been able to isolate the bubbles thus attenuated; some would burst spon- 

 taneously before the funnel had been withdrawn, others while it was being 

 withdrawn. 



§ 8. In § 12 of the first series of these researches, after having described the 



* I will here mention quite a curious circumstance. I had at first employed a flask of tin 

 furnished with a tap of hou, but when thealcolioUc liquid contained in this tlask accidentally 

 included small spheres of oil, these, on leaving the tap, sometimes broucjht over oxide of n'on, and 

 thus becoming heavy, fell with much rapidity to the bottom of the bubble of oil ; now, when this 

 happened, however minute the ferraginous spherule might be, the film of oil was seen, after 

 some seconds, to become suddenly thin at the place where this spherule rested, the attenua- 

 tion being propagated by a retreat of the oil to a small distance around the point of con- 

 t;ict, and the bubble burst almost simultaneously at this place. The retreat of a part of the 

 oil at the contact of the ferruginous spherule is doubtless, as we may say in passing, a phe- 

 nomenou of the kind described by M. Dutrochet in his RechercJies physiques sur la force epipo' 

 Uque, 



