124 C. Barns- — Subsidence of fine Particles in Liquids. 



2. With these deductions as a point of departure, I then 

 attempted to find relations between rates of subsidence, the 

 viscous and capillary properties of the liquid, and its electrical 

 behavior, under analogous conditions of concentration and of 

 temperature. This general survey proved that the phenomenon 

 of sedimentation is unique ; that the frictional resistances en- 

 countered by the particles are apparently different from the 

 viscosities of the liquids in which subsidence takes place ; that 

 many of the occurrences observed are closely allied to the elec- 

 trolytic and the capillary properties of this liquid. Finally 

 utilizing Prof. Brewer's stratified subsidence, which I obtained 

 very clearly with certain kinds of tripoli, I commenced a series 

 of more rigorously quantitative measurements showing that 

 caet. pa7\, rate of subsidence is primarily dependent on the tur- 

 bidity of the mixture. 



In my experiments with tripoli, the observed rates of subsi- 

 dence (cm / (sec X 10 6 )) in ether, alcohol, water, glycerine were 

 7500, 1300, 3, 0*09, respectively ; but owing to the difference 

 in character of the divers precipitates, these figures have no 

 further signification than to emphasize the said difference in 

 character. §§ 5,6. Let water and ether be mixed so that there 

 shall be equal bulks of etherized water below, and aqueous 

 ether above, and then let the dust (bole) be added. If now the 

 mixture is violently shaken and then allowed to subside, the 

 ether is washed clean of particles in a few minutes whereas the 

 sediments remain suspended in water for weeks and even 

 months. Here however the separation and subsidence is pro- 

 moted by the surface energy of water. 



On the other hand if dry tripoli be added to ether dried over 

 CaCl 3 , in a test tube, and if the tube be held obliquely after 

 shaking, subsidence is so rapid that the upward current of ether 

 along the upper line of the tube is almost tempestuous. 



The close relation of the present phenomena to electrolytic 

 phenomena appears at once by observing that so little as a 

 single molecule of HC1 (for instance) added to 10,000 or even 

 50,000 molecules of H„0, produces appreciable increase of the 

 rate of subsidence. Remembering that the arrangement is in 

 three dimensions, and supposing the molecule HC1 as large as 

 the molecule H 2 0, the effect of the molecule HC1 must be 

 appreciable at a distance of at least 30 times t its own radius,* 

 and extend much beyond this asymptotically. Quincke's radius 



* The estimated diameter of H 2 (distance between centers of adjacent mole- 

 cules 2p = 40/10 9cm ) I take from Kohlrausch (Wied. Ann., vi, p. 209, 1879). In 

 how far the molecule of liquid water differs from H 2 remains to be seen. In- 

 deed the underlying cause of continuous and of discontinuous fusion and vapori- 

 zation, and the cause of allied phenomena such as retarded solidification, ebulli- 

 tion, etc., seems clearly to be some form of polymerization. Nevertheless the 

 atomic theory in its present stage of development fails to suggest a satisfactory 

 mechanism for these occurrences. 



