WITHDRAWN FROM THE ACTION OF GRAVITY. 633 



instability. We shall denominate this latter proportion the limit 

 of the stability of the cylinder. 



45. These conclusions however are liable to a well-founded 

 objection. Our liquid figure is complex, because its entire sur- 

 face is composed of a cylindrical portion and of two portions 

 which present a spherical curvature. Now we cannot atfirm 

 that these latter portions exert no influence upon the stability or 

 the instability of the intermediate portion, and consequently 

 upon the value of the proportion which constitutes the limit be- 

 tween these two states. To allow of the preceding conclusions 

 being rigorously applicable to the cylinder, it would be requisite 

 that the figui'e should present no other free surface than the 

 cylindrical surface, which is easily managed by replacing the 

 rings by entire discs. I effected this substitution by employing 

 discs of the same diameter as the preceding rings, but the results 

 were not changed ; the cylinder, 6 centims. in height, was well 

 formed, and was found to be stable ; whilst the figure 8 centims. 

 in height began to change before becoming perfectly cylindrical, 

 and was rapidly destroyed. The final result of this destruc- 

 tion did not however consist, as in the case of the rings, of a 

 perfect sphere and a double convex lens, but, as evidently 

 ought to have been the case, of two unequal portions of spheres, 

 respectively adherent to the two opposite solid surfaces. The 

 limit of the stability of the cylinder therefore really lies be- 

 tween 3 and 4. 



The experiments which we have just related are very delicate, 

 and require some skill. In this, as in all other cases of measure- 

 ments, the oil must be allowed to remain in the alcoholic mix- 

 ture for two or three days, then the pellicle must be removed from 

 it (note to p. 627) ; afterwards, when the mass, after having been 

 again introduced into the vessel, has been attached to the two 

 solid discs, some time must be allowed to elapse in order that 

 the two liquids may be exactly at the same temperature ; more- 

 over, it must be understood that the experiments should be 

 made in an apartment the temperature of which remains as con- 

 stant as possible. Lastly, it is scarcely necessary to add, that 

 ■when the alcoholic liquid is mixed, after having added small 

 quantities of pure alcohol or alcohol at 16°, the movements of 

 the spatula should be very slow, so as to avoid the communica- 

 tion of too much agitation to the mass of oil ; we are even some- 

 times compelled momentarily to depress the upper disc, so as to 



