506 INFECTION AND RESISTANCE 



tion had been allowed to stand until it had reached equilibrium with 

 the gravitational forces. The result was that the same law of dis- 

 tribution was found to hold in this case as in the case of the atmos- 

 phere. The kinetic theory is thus shown to apply quantitatively as 

 w r ell as qualitatively to colloidal solutions. 



3. ELECTRICAL PROPERTIES. If a U-tube be filled with water, 

 electrodes placed in each arm, and these electrodes maintained at a 

 constant difference of potential either by a battery, dynamo, or other 

 source of direct current, it is noticed that there is a continual flow of 

 liquid in the tube, in one direction near the walls and in the opposite 

 direction in the interior of the tube. There is every reason to believe 

 that such currents will be set up in all cases whatever the nature of 

 the liquid or of the tube, although the current set up may in par- 

 ticular cases be very small and even very rarely approach or equal 

 zero. If we name the current along the walls simply the "current," 

 and that through the interior the "countercurrent," then in the case 

 of glass and water the direction of the current is from anode to 

 cathode, and that of the countercurrent is from ca^iode to anode. 

 This phenomenon is explained by the hypothesis that at the surface 

 of contact between the glass and the water there is established a dif- 

 ference in electrical potential, the glass becoming negatively and the 

 water positively electrified. If this assumption is valid it follows 

 that if a particle of glass placed in water be subjected to the influence 

 of two electrodes placed in the water, it will, being negatively 

 charged, be attracted by the positively charged electrode (the anode) 

 and repelled by the negatively charged electrode (the cathode). The 

 result would be a wandering of the particle of glass through the solu- 

 tion toward the cathode. This result is confirmed by ample experi- 

 ments. Furthermore, the phenomenon is common to all particles in 

 all liquids, so far as is known, so that any colloidal solution placed 

 in a potential gradient will show wandering of its particles in one 

 direction or the other. Thus in water, ferric hydroxid, chromium 

 hydroxid (and most hydroxids in the colloidal state), methyl violet, 

 and some other dyes wander to the cathode. All colloidal metal solu- 

 tions, sulphur, the halogen salts of silver, chlorophyll, rosin, mastic, 

 most dyes, and, in fact, the great majority of substances investigated 

 wander toward the anode. Albumin (and probably some other sub- 

 stances) wanders toward the cathode in acid solution, and toward 

 the anode in alkaline solution. As will be seen later, the hypothesis 

 of the existence of such electrical charges on colloid particles has 

 been of very great use in explaining many forms of conduct on the 

 part of dispersed systems. 



4. SURFACE TENSION. If a globule of mercury be divided into 

 two parts, these two parts will unite again if opportunity be given. 

 All the opportunity which is necessary, if the surfaces be clean, is to 

 bring the two parts into mechanical contact. The union of the sep- 



