VISCOMETRY 



with the properties of an elastic soHd. These bonds could be long- 

 range intermolecular forces, ordinary chemical bonds at points of 

 contact between dispersed particles, or forces of some unknown nature. 

 The only important factors are that they be relatively few and that 

 they have very high energies. This condition will result in a structure 

 which will undergo only elastic deformation under low-velocity gra- 

 dients, but which will flow when sufficiently high shearing forces are 

 applied. The whole solution, consisting of the Newtonian component 

 and the network, will exhibit anomalous viscosity. Eyring has shown 

 that flow curves resembling those actually obtained for many types of 

 systems exhibiting structural viscosity can be constructed on the basis 

 of such assumptions. 



The difficulty presented by this situation is obvious. If the 

 anomaly is due solely to hydrodynamic efTects, the problem can be 

 solved by extrapolating the viscosity data to zero velocity gradient; 

 but, if the anomaly is due in part to structure, this procedure is de- 

 feated because the contribution of the structure — the unwanted factor 

 for our purposes — is at a maximum at zero velocity gradient. 



The solution to the difficulty is very simple in theory. Hydro- 

 dynamic anomalous viscosity in dilute solutions should be independent 

 of concentration, that is, the ratio of the apparent specific viscosities 

 at any two values of velocity gradient should not vary with the con- 

 centration of solute. On the other hand, the structural effect should 

 tend to vanish as the concentration of solute approaches zero. Values 

 of the velocity gradient should exist for which the ratio of the apparent 

 specific viscosities changes with a change in the concentration. Thus, 

 it is theoretically possible to differentiate between structural and 

 hydrodynamic anomalous viscosities and to find experimental condi- 

 tions under which the latter vanishes. Practically, however, two 

 difficulties remain: first of all, these theoretical considerations have 

 not yet been investigated experimentally; and, second, it might be 

 necessary to work with extremely dilute solutions to realize the condi- 

 tions postulated. This might not be possible until after the develop- 

 ment of much greater precision than has heretofore been required in 

 viscometry. 



It has been the purpose of this discussion to present some im- 

 pression of the simplicity of the viscometry technique, of the meaning 

 of viscosity data in terms of the shape of suspended or dissolved particles, 



