M. A. LAUFFRR 



Electrolyte concentrations as low as 0.02 M are sufficient to reduce 

 the effect to a relatively small value. An even better, but much more 

 laborious, method of eliminatins^ the effect is to measure the viscosity 

 and the other variables in the Krasny-Ergen equation at various 

 electrolyte concentrations, and then extrapolate the results to zero 

 electrokinetic potential. Briggs et ol. have shown that this procedure 

 is entirely satisfactory. 



The error due to the variation of the apparent viscosity coeffi- 

 cient with the velocity gradient is more difficult to eliminate because 

 it can have more than one cau.se. One of the possible cau.ses of this 

 error, which will be referred to subsequently as anomalous viscosity, 

 is hydrodynamic. The Simha equations show that the intrinsic 

 viscosity of rodiike and platelike ellipsoids of revolution randomly 

 oriented in the flowing stream should be much greater than the 

 intrinsic viscosity for the same particles completely oriented parallel 

 to the direction of flow, as given by the treatments of Eisenschitz and 

 Peterlin. Thus, the intrinsic viscosity of rodlike or platelike particles 

 will vary with the velocity gradient for purely hydrodynamic reasons, 

 because the orientation of the particles depends upon the velocity 

 gradient. Were this the sole type of anomalous viscosity, the solution 

 of the problem would be simple, for all one would need to do is measure 

 intrinsic viscosity at various velocity gradients and extrapolate the 

 results to zero velocity gradient, where the particles are randomly 

 oriented and where, therefore, the Simha equations are valid. Robin- 

 son (21) has shown that the viscosity of tobacco mosaic virus actually 

 does vary with the degree of orientation of the particles as determined 

 by optical means, and that the data can apparently be extrapolated 

 to zero velocity gradient. 



However, the anomalous viscosity could be due in part to inter- 

 particle forces, which confer upon the solution the properties of a 

 weak gel. This effect will be called structural anomalous viscosity. 

 Structural anomalous viscosity can be understood in terms of Eyring's 

 concepts (24). Eyring has postulated that a solution which shows 

 structural anomalous viscosity is composed of two mechanical systems. 

 The solvent and perhaps some of the dissolved particles constitute a 

 medium which obeys Newton's laws of flow. In addition to this, 

 however, relatively few bonds of high activation energy are assumed 

 to exist between some of the dispersed particles, forming a network 



258 



