964 THE ANIMAL VIRUSES 



Svedberg's hands it has been used with conspicuous success in estimating the mole- 

 cular weight of proteins. 



Bauer and PickeLs (1936, 1937) have devised a high-spe«d centrifuge combining some 

 of the principles of the Henriot and Huguenard and some of the Svedberg model. Elford 

 (1936) has shown that, by introducing a capillary tube into the fluid to be centrifuged 

 and coUecting the virus particles on blotting paper at the bottom, much of the difficulty 

 caused by " mixing " can be overcome. Schlesinger (1936) has adapted the Sharpies 

 centrifuge for the concentration of virus by using a thin layer of suspension in a hollow 

 cyhnder rotating vertically, so that the total distance the virus particles have to travel 

 is a fraction of a millimetre. Mcintosh and Selbie (1940) have modified the Sharpies 

 centrifuge so as to permit the coUection of the virus particles on a sheet of cellophane ; 

 this method has the advantage that a continuous stream of the virus to be purified can 

 be fed to the machine. The preparation of pure virus suspensions is sometimes facilitated 

 by prehminary tryptic digestion of the matrix m which they are embedded. 



Not only is it possible, now, with some of the high-speed centrifuges, to throw 

 down completely the larger viruses, but their approximate size can be calculated 

 from measuring their rate of sedimentation. Bechhold and Schlesinger (1931) have 

 worked out a formula from which the size of evenly dispersed spherical particles 

 submitted to a constant centrifugal force may be determined. Further, by measur- 

 ing the rate of concentration, it can be ascertained whether the particles are of 

 uniform size. It can be shown . for instance, that the logarithm of the concentration 

 of particles in the supernatant fluid is proportional to the length of time of centri- 

 fugation. If they are of unequal size, the larger particles will be thrown down 

 rapidly, and the curve formed by plotting the logarithms of the concentrations 

 against time will not be a straight line. 



The larger viruses, of 0-1-0-2 f.i in diameter, can be thrown down completely 

 under suitable conditions in about half an hour by a centrifuge revolving at 10,000 

 r.p.m. (see Amies 1933), while particles of about 60 m^, such as the staphylococcal 

 bacteriophage, require a speed of 40,000 r.p.m. maintained for 1 to 1| hours 

 (Mcintosh 1935). Where centrifuges of only 3,000 r.p.m. are available, and it is 

 desired to concentrate the suspension, the virus may sometimes be adsorbed on to 

 kaolin, animal charcoal or blood corpuscles, and the deposit subsequently sus- 

 pended in a protein-free medium (see Levaditi and Nicolau 1923, Gins and Krause 

 1923, Tang 1932, Francis and Salk 1942). Viruses vary, however, in their reaction 

 to different adsorbing agents (Lewis and Andervont 1927), and this method is 

 therefore not always successful. The purity of centrifuged virus suspensions is 

 dependent on the number of particles of foreign matter present resembling in 

 sedimentation rate that of the virus which is being concentrated. Methods for 

 determining the degree of purity have been suggested by Smadel, Rivers and 

 Pickels (1939) and Luria 1940). 



Morphology. — Information on the shape of filtrable virus particles has been 

 furnished by Barnard, who has been successful in photographing some of the 

 larger viruses in ultra-violet light. One of the most carefully studied is that of 

 ectromelia, a virus with a diameter of about 120 m// (Barnard and Elford 1931). 

 This organism is coccoid, and frequently occurs in pairs. Isolated organisms 

 are spherical and highly refractile, the refractivity apparently decreasing with 

 shortening of the wave-length used for illumination. Reproduction is by binary 

 fission, and elongation is evident before division. The final separation of the 

 two organisms takes place quickly, but a very fine connecting filament may be 



