WITHDRAWN FROM THE ACTION OF GRAVITY. 675 



with certain external circumstances, such as the presence of a 

 surrounding liquid, or the contact of the convex surface of the 

 cylinder with a solid plane. In all the subsequent statements, 

 we shall take the simplest case, i. e. that of the absence of ex- 

 ternal circumstances ; in other words, we shall always suppose 

 that the cylinders are produced in vacuo or in air, and that they 

 are free as regards their entire convex surface. 



6. Two cylinders of different diameters, but formed in the 

 same liquid, and the lengths of which are such that the divisions 

 assume in each of them their normal length, become subdivided 

 in the same manner, i. e. the respective normal lengths of the 

 divisions are to each other as the diameters of these cylinders. 

 In other words, when the nature of the liquid does not change, 

 the normal length of the divisions of a cylinder is proportional 

 to the diameter of the latter. 



The same consequently applies to the diameter of the isolated 

 spheres into which the normal divisions become converted, and 

 to the length of the intervals which separate these spheres. 



7. The proportion of the normal length of the divisions to the 

 diameter of the cylinder always exceeds the limit of stability. 



8. This proportion is greater as the liquid is more viscid and 

 as the configuring foi-ces in it are weaker. 



9. In the case of a cylinder of mercury, this proportion is 

 much less than 6, and we may admit that it is less than 4. 



In the case of a cylinder composed of any other very slightly 

 viscid liquid, such as water, alcohol, &c., it is very probable that 

 the proportion in question is very nearly 4. Hence, in the case 

 of the latter liquids, we have for the probable approximative 

 value of the proportion of the diameter of the isolated spheres 

 resulting from the transformation and the diameter of the cylin- 

 der, the number 1*82 ; and for that of the proportion of the 

 distance of two adjacent spheres to this same diameter, the 

 number 2'18. 



10. If mercury is the liquid, and the divisions have their nor- 

 mal length, the time which elapses between the origin of the 

 transformation and the instant of the rupture of the lines, is 

 exactly or apparently proportional to the diameter of the 

 cylinder. 



This law very probably applies also to each of the other very 

 slightly viscid liquids. 



This same law may possibly be general, i. e. it may be api)li- 



