Tension of Liquid Films. 199 



curvature*, and consequently are not curves of equilibrium of 

 tension. In order to verify this result by direct observation, I 

 use the two rings described in my former note; they are 70 

 millims. in diameter; the upper one is provided with a suspender, 

 the lower one resting on three feet. Suppose, for instance, the 

 distance of these rings is 40 millims., I seek the length of the 

 portion of the asymptotic line drawn from one to the other ; for 

 this purpose the portion of the meridian catenary comprised be- 

 tween the same rings may be divided by the cosine of 45°t; 

 I thus obtain for the desired length nearly 63 millims. On the 

 other hand, it is easily found that the meridian planes passing 

 through the extremities of this portion of the asymptotic lines 

 form an angle of about 88°. 



This being granted, the following is the way in which the ex- 

 periment is made. I fix the ring with a suspender to a support, 

 and I make it quite horizontal ; below it I place the ring with 

 feet on a horizontal glass plate, so that the vertical distance of 

 the rings is 40 millims., this distance being reckoned from the 

 middle of the thickness of each wire. Then by means of a piece 

 of wax I attach one end of a cocoon-thread, 63 millims. in 

 length, moistened with glycerine solution, to a point in the 

 upper ring, and the other to a point in the lower ; I then fix 

 the two ends of a second cocoon-thread of the same length 

 to two points at a distance of 10 millims. from the first, and 

 both situated on the same side of them. This being done, I 

 bring the lower ring- almost in contact with the upper one, 

 and fill the intervening space with the glycerine liquid ; I then 

 carefully lower the tripod ring, taking care that the threads 

 are completely immersed in the catenoiclal film which is formed ; 

 when this ring is on the glass plate, I measure the diameter 

 of the circle of the gorge of the film by means of a hori- 

 zontal cathetometer. I then move the tripod Y\<?. 3. 

 ring 20 millims. from the upper ring, burst 

 thefilm comprised between the cocoon-threads, 

 and then very slowly lower the lower ring to 

 the plate; I turn it about the axis of figure 

 so as to establish an angular distance of 86° 

 between the azimuths of the points of attach- 

 ment of each of the two cocoon-threads. At 

 this moment these form two curves of the same 

 form, but arranged in inverse directions (fig. 3); 

 these curves, as far as the eye can judge, 



* In the report already mentioned. M. Lamarle gives as the value of this 

 radius of curvature T V 2, T being the part of the tangent to the meridian 

 from the point under consideration to the axis of revolution. 



\ See the above-cited memoir of M, Aoust. 



