486 



D. O. JORDAN 



? 



^o-'. 



o°,>j:.'-^'* " 



»9^< * 



.r^' 





t> O.OI9l«>/o 

 O O.OI4 7'>/o 

 • O.OI03»/o 



400 

 G 



Fig. 19. Streaming birefringence of solutions of sodium deoxypentose nucleate 

 in 10% sodium chloride (Schwander and Cerf'^^). 



decreases with ionic strength, but the rotational diffusion constant is ap- 

 parently independent of both ionic strength and nucleate concentration 

 (Fig. 19) and has a value of 36.6 sec.~'. From these results and the magni- 

 tude of the birefringence, Sadron^^^ has concluded that the particles are 

 very long and not capable of contraction. The experiments of Schwander 

 and Cerf'^^ in which the birefringence was measured using solutions of 

 sodium deoxypentose nucleate in which the viscosity was varied by adding 

 glycerol, are, however, probably of greater significance. The birefringence 

 may arise either from the orientation of a rigid particle in the stream lines 

 or from the deformation of a random coil to give increased asymmetry 

 followed by orientation; the contribution of these two effects may be 

 analyzed by varying the viscosity of the solvent (see Cerf'^^ for theoretical 

 details). On increasing the viscosity of the solvent, tan a (where a is the 

 initial slope of the curve relating extinction angle and velocity gradient) 

 should be a linear function of the solvent viscosity if the particles are rigid. 

 Schwander and Cerf'^^ and Sadron^-^ first interpreted these results (Fig. 20) 

 as indicating that the nucleate ion was a rigid particle, the discrepancy at 

 higher values of the viscosity being attributed to heating effects. On the 

 basis of a more detailed theoretical treatment of the problem, Cerf"^ con- 

 siders that the change of slope at higher viscosities (Fig. 20) is due to a de- 



i« R. Cerf, /. Polymer Sci., 12, 15 (1954). 



