126 C. Barus — Subsidence of fine Particles in Liquids. 



differences of J w and J e are no larger than the many sources 

 of error led me to anticipate, particularly in view of the fact 

 that the two samples for ether and for water may not have 

 "been absolutely identical. The concentrated ether used was 

 the same commercial reagent with which I obtained the sub- 

 sidence phenomena. In consequence of the high but normal 

 values of both J w and J e , I saw no need of specially purify- 

 ing it. I add finally that after calcination, the dry tripoli lost 

 1*2 per cent in weight, and the dry bole about 11/4 per cent. 

 In both cases this is probably water of constitution, the elimi- 

 nation of which was of course not permissible. In spite of 

 differences of chemical composition, the bole and tripoli parti- 

 cles are about equally suspended before and after calcination, 

 and the phenomena with ether dried over CaCl, are identical 

 with the above. 



4. These results show that the densities in the two cases 

 (sediment in water and in ether respectively), are not essen- 

 tially different. Moreover, the density of -tripoli is so nearly 

 that of quartz, and the density of bole so nearly that of kao- 

 linite, as to leave the hydration hypothesis very seriously in 

 the lurch, so far as favorable evidence is concerned. It is im- 

 probable that the addition of water to the dry powder is ac- 

 companied by sufficiently marked volume changes ; it is cer- 

 tain that the enormous variation of rates of subsidence actu- 

 ally observed when the particles descend in water, in solutions, 

 and in ether, must be referred to some general cause apart 

 from the density of the particles and the viscosity of the 

 liquids. 



5. This premised, the explanation of sedimentation may be 

 so made as to give emphasis to the following principle : If 

 particles of comminuted solid are shaken up in a liquid, the 

 distribution of parts after shaking will tend to take place in 

 such a way that the potential energy of the system of solid 

 particles and liquid, at every stage of subsidence, is the mini- 

 mum compatible with the given conditions. In the case of 

 solid particles and pure water the configuration answering this 

 condition, is schematically, 



Particle, water Particle, water. 



In the case of solid particles and ether, or of solid particles 

 and solutions, this configuration is schematically, 

 Particle, particles ether, ether. 



For the exemplification of this inference my paper contains 

 varied experimental evidence. The principle asserts that in 

 case of water the sediment is graded and the suspended mate- 

 rial granular, whereas in ether the sediment is apparently ho- 

 mogeneous, as I found ; that the bulk of sediment is necessa- 



