Capillarity of Molten Bodies. 85 



may be neglected, and the weight of the portion of the drop 

 which falls may be treated as the W in equation (7). 



We may equally neglect the fact that new fluid comes down 

 at the time when the drop is separating, which tends to make 

 the drop too large. When this access of fluid is too great, 

 on the other hand, there is a thin jet of fluid which may readily 

 be resolved into smaller drops by taps from the outside. This 

 is the explanation of the fact that, in the case of many fluids, the 

 drops attain a maximum for a determined velocity in the supply 

 of the issuing fluid*. 



Although it thus appears that the process of the formation of 

 drops is exceedingly complicated, the application of equation (7) 

 would give us approximate values of the capillarity-constants a; 

 and this method has at least the recommendation that there is 

 no better, or none which is not complicated by too many expe- 

 rimental difficulties. 



5. The experiment is simplest for gold and silver. Vertical 

 threads of these metals were held by pincers and brought down 

 into a small gas-flame the dimensions of which were not greater 

 than 3 millims. diameter and 8 millims. height, so that the metal, 

 as soon as it was melted, formed in a drop at the lower end of the 

 thread. The drop increased in this way, and rose on the solid 

 thread, which was gradually lowered to the flame. When it 

 was too large it fell into a vessel filled with water, and was 

 immediately solidified, and afterwards dried and weighed. After 

 a little practice it became easy to avoid any shaking of the 

 threads, by which the drops were apt to be too soon detached. 



The molten metal was colder above than below ; and at the 

 upper part the temperature was only a little above that of the 

 melting-point of the substance. The weight of the drop in mil- 

 ligrammes, divided by the circumference of the wire in milli- 

 metres, gives us accordingly the constant of capillarity a for that 

 melting-point. 



The shorter the distance between the drop and the pincers 

 holding the wire, the larger the drops seemed to be. This was 

 due probably to the lower temperature of the drop, in conse- 

 quence of the abstraction of heat by the wire and pincers. 



Further, the drops from a gold wire melted over a common 

 gas-flame and over one fed with oxygen weighed nearly the same ; 

 so that the influence of temperature in these experiments may 

 be neglected. 



The diameter of the wires was measured by a microscope and 

 an eyepiece-micrometer which gave one hundred divisions. Each 

 single division (and tenths of a division could easily be esti- 

 mated) corresponded, therefore, according to the magnifying- 

 power used, to from '007 millim. to '02 millim. 

 * Compare Pogg, Ann. vol. cxxxi. p. 130. 



