246 H. K. SCHACHMAN AND R. C. WILLIAMS 



viruses are obsolete. It is fortunate that the newer techniques are neither 

 more difl&cult nor more time-consuming than the older methods. Versatile 

 equipment, although expensive and complicated, is now available com- 

 mercially from several sources. No longer is a diffusion measurement rele- 

 gated to a few select laboratories throughout the world where different 

 investigators pioneered in the design and use of this highly specialized 

 technique. 



After many years of experimentation leading to various ingenious designs 

 for diffusion cells, there now seems to be widespread agreement that the 

 Tiselius electrophoresis cell, with only shght modifications, is the ideal cell. 

 This conclusion results in large part from the important proposal of Kahn 

 and Poison (1947) that poorly formed bomidaries between the solvent and 

 solution can be made almost ideal by a capillary siphoning procedure. This 

 technique, called boundary sharpening, has permitted the formation of 

 almost infinitely sharp boundaries between the solvent and solution. More- 

 over, the bomidary is located in the cell at a level which permits observation 

 and photography even during the formation of the boundary. Thus the large 

 and somewhat uncertain corrections required because of imperfections in 

 the initial boundary are no longer necessary. 



As already indicated, the major innovations were concerned with the 

 rediscovery and adaptation of neglected optical principles and practices. 

 For many years the spreading of boundaries due to diffusion of macro- 

 molecules was followed by means of schlieren optical procedures of one type 

 or another. These techniques were convenient, and procedures were de- 

 veloped whereby the photographic patterns were translated into diffusion 

 coefl&cients by a variety of calculation methods. In effect, the patterns were 

 plots of the concentration gradient versus distance. All of the procedures 

 involve the assumption, which was subject to test, that the plots were 

 Gaussian in shape obeying the following equation: 



where c^ is the initial concentration of solute and the other symbols have the 

 meanings already assigned. This equation is the residt of integration of 

 Fick's second law (Equation (11). Most popular of the various methods of 

 treating the data is the so-called height-area method by means of which the 

 diffusion coefficient is calculated from any of the individual photographs. 

 With time, the height of the curve decreases in a prescribed manner while 

 the width increases so as to maintain a constant area under the curve. The 

 shape of the bomidary must have a definite form, according to Equation (13), 

 and deviations of the experimental curves from the theoretical shape are an 



