92 The Structure of Protoplasm 



are dilatant in water, they are thixotropic in organic liquids such 

 as benzene, carbon tetrachloride, etc.^ It is possible to pass gradually 

 from one state to the other by using two miscible liquids. 



In these organic liquids, too, a small amount of a second sub- 

 stance may cause a marked change in behavior: A paste made from 

 a very fine iron powder and carbon tetrachloride is pronouncedly 

 plastic, a behavior which is practically always correlated with a 

 marked thixotropy.-'' On adding a small amount of oleic acid, the 

 paste is liquefied to a strongly dilatant suspension. The adsorption 

 of the oleic acid on the surface of the particles produces a state of 

 independence, and hence dilatancy. 



Suspensions of finely powdered solids in organic liquids, such 

 as mixtures of oils, have been used for a very long time as paints. 

 It is, therefore, no wonder that painters have been acquainted with 

 the manifold, anomalous mechanical properties of suspensions, 

 though without having defined such limiting cases as thixotropy and 

 dilatancy. But it can hardly be doubted that many pastes of paints, 

 described as stiff and plastic and as having a marked yield value, are 

 thixotropic (in our terminology) if the concentration of the solid 

 is suitably chosen, whereas if they were dispersed to a mobile 

 liquid by the right dispersing agent, they would show dilatancy, 

 again at a suitable concentration of the solid. Green,^" at a time 

 when the phenomenon of thixotropy was hardly known, pubhshed 

 striking photomicrographs of plastic pastes of ZnO formed by sus- 

 pension in kerosene, which showed the ZnO particles to be coagu- 

 lated. When poppy seed oil was added as a dispersing agent, the 

 same particles were independent and dispersed. They are shown in 

 Figures 2a and 2b and are good examples of the state of the particles 

 in a thixotropic and in a dilatant system. 



Our knowledge of the forces acting between the particles and 

 causing these phenomena is not yet sufficiently advanced to allow 

 me to state concisely, in the compass of a lecture, the processes 

 involved. Only a few points may be mentioned. The geloids in 

 thixotropic sols and gels are closely related to the factoids,^-' i. e., 

 double refracting groups of oriented particles formed spontaneously 

 in concentrated sols containing nonspherical particles,-""' and to the 

 coacervates as defined by Bungenberg de Jong,-'^ i. e., hquid coagu- 

 lates of hydrophilic sols containing one or more kinds of colloidal 

 particles. In geloids, tactoids, and coacervates, the colloidal particles 

 are very far apart,-^^ up to many yi. It is, therefore, improbable that 

 the attraction between the particles is due to van der Waals' forces, 



