EXPEEIMENTS WITfL SOAP BUBBLES BOYS. 215 



to the eye cither when seen du-ectly or when projected upon a screen. 

 I have prepared a frame of brass wire upon which I can form a sheet 

 of soap water 6 feet long and one-quarter of an inch wide, but zigzag, 

 so that it occu])ies a surface of about 4 inches scjuare. Further, I 

 have so constructed it that the two ends of the'sheet are adjacent. 

 Then, on breaking the sheet at one end, we immediately see that 

 the bursting is ])rogressive, lasting about one-seventh of a second for 

 the entire course. If the sheet be broken as soon as formed, when it 

 is still thick, the motion of the edge is slower. If we let it draui until 

 it shows the white or the colors of the first order, the motion of the 

 edge is perceptibly more rapid. 



The soap bubble furnishes a convenient means of illustrating the 

 principle of stability. It Y^^ould suffice here to refer to Plateau's 

 labors upon this subject. However, I have arranged two experiments. 

 The fii-st is a variation of an experiment of Plateau's which showed 

 that a very light sphere of glass remamed in stable equilibrium upon 

 the lower extremity of a vertical ring of wire if a sheet of soap water 

 were spread over the ring. I have found that blown birds' eggs can 

 be employed for this experiment provided that they be no heavier 

 than those of the house sparrow, but the hole should be mended with 

 a fragment of tissue paper and celluloid varnish. 



The bhd's egg offers the advantage of introducing a second princi- 

 ple of stability. It can remam m equilibrium only if its greatest 

 cross section, i. e., its oval section, is in the plane of the liquid sheet. 

 If the ring be made to revolve slowly in its own plane, the egg begins 

 to roll or slip ; then, the speed increasing, it roUs and jumps, but never 

 leaves the liquid sheet. I have so arranged this experiment as to be 

 able to project it upon the screen. 



The second method of showing the conditions of stability depends 

 upon the employment of cylindrical bubbles whose length is nearly 

 t: times as great as their diameter. Beyond that length, as 

 Plateau has shown, a cylindrical bubble is no longer stable. As this 

 length is approached, the stability is diminished, or, in other words, 

 the bubble more easily loses its form. If we blow a spherical bubble 

 with oxygen between the poles of an electromagnet of moderate 

 power, we find that the action of the magnetism upon the oxygen is 

 not sufficient to make the bubble move appreciably when the exciting 

 current is closed. But if we make the bubble take the cylindrical 

 form with an apparatus so constructed as to render its length nearly 

 71 times as great as its diameter, the magnetic influence imme- 

 diately causes the separation of the bubble into two parts, the larger 

 attached to the nearer ring remaining between the poles of the 

 electromagnet and the smaller attached to the more distant ring. 



I shall now blow a large bubble, usmg my mouth as an injector 

 and employing my hands without any other apparatus. This method 



