THE PHYSICAL PROPERTIES OF INFECTIVE PARTICLES 245 



free of some of the objections of the older cells is shown in Fig. 1. The solution 

 is placed in the lower compartment and the solvent in the upper to effect 

 gravitational stability, and stirring is accomplished mechanically with 

 rotating magnets. After the ceU is assembled, the diffusion is allowed to 

 proceed with periodic removal of the upper liquid and immediate replace- 

 ment with fresh solvent. This is repeated several times until an assay of the 

 removed liquid reveals that the rate of transport of solute into the upper 

 compartment is constant. This indicates that a steady state condition has 

 been achieved as a result of the formation of the appropriate concentration 

 gradient within the channels of the sintered glass disk. When this condition 

 has been attained (several days may be required for particles as large as 

 viruses), the two solutions are removed, their concentrations determined 

 and the diffusion coefficient is calculated from the following: 



In this equation, the superscripts o and t refer to zero time and to a subse- 

 quent time, t; ^ is the ceU constant {A[L){llvu + l/vj), where ■?;„ and Vi and 

 c„ and Ci are the volumes and concentrations of the upper and lower liquids 

 and A and L are the effective area and thiclaiess of the porous disk. Evalua- 

 tion of the ceU constant, ^, is performed through the use of a substance of 

 known diffusion coefficient. Potassium cliloride is the favored substance for 

 this determination, but the use of a virus or protein of known diffusion 

 coefficient is preferable to reduce the difference in size between the cali- 

 brating substance and the unknown agent. Equation (12) wiU be recognized 

 as the product of integration of Equation (10). Thus this technique consists 

 simply of the direct determination of the rate of transport of material, 

 dmjdt, across a selected plane. 



It is obvious that adsorption of any virus particles to the surface of the 

 porous diaphragm wiU lead to erroneous results, as will the use of a sintered 

 glass disk with pores so small as to be impermeable to the vims particles. In 

 view of the simpHcity of the technique, it is advisable to conduct experiments 

 with different cells so that irregularities or deficiencies in an individual cell 

 may be detected. 



ii. Free Diffusion. In the past ten years major modifications have occurred 

 in the techniques employed for the study of free diffusion, with the result 

 that today such diffusion experiments can be performed routinely and with 

 an accuracy far greater than that expected by the most optimistic research 

 workers of the 1930's and early 1940's. This tremendous advance in experi- 

 mental precision was due largely to the rediscovery of interferometric optical 

 methods and their adaptation to diffusion studies. As a consequence of these 

 developments, the older, more conunon procedures used for proteins and 



