The influence of halide concentration on the metabolism of Penicillium sclerotiorum Van Beyma.

albumin in these conditions whereas the others cannot. Whether or not the presence of salt is necessary to start the combination of two proteins, it certainly greatly increases it. As in causing flocculation, the efficiency of different salts in influencing complex formation depends on the valency of the ion with the charge opposite to that of the protein particles. The two processes, however, are not identical. If the protein is denatured by heating in the absence ofsalts, the subsequent addition ofsalt causes flocculation. By contrast, salts influence complex formation only if they are present in protein mixtures during heating, for the addition of salt to a protein mixture heated salt-free does not lead to formation of any measurable amount of serologically nonprecipitating complexes. This difference could be explained by assuming that individual proteins differ in their rates of aggregation, and that these rates are much greater than rates of denaturation. Then, if salt is present during heating, particles of different proteins would be able to combine with one another as soon as they are denatured, but if salt is added to a mixture of proteins already denatured by heating in salt-free solution, each protein would tend to aggregate separately. The results with antibodies almost exactly parallel those with antigens. The two antisera used were to bushy-stunt virus and to human-serum albumin. When heated in the presence of salt, these behave similarly, the antibodies forming non-precipitating complexes which inhibit precipitation. In the absence ofsalts the two antisera behaved differently. The virus antiserum again ceased to precipitate and instead inhibited precipitation, though less strongly thanwhen heated with salt, whereas the antiserum to human albumin still precipitated. This difference at first sight suggests differences in the heat stabilities ofthe two antisera. A similar difference between the behaviour of antisera to 'O' and 'H '-type antigens is well known and was for long interpreted as evidence that these antigens produced antibodies with different properties. It has now been shown, however, that antibodies to antigens of both types undergo similar changes when heated, and that the difference in their behaviour occurs because the precipitability of the antigens is affected differently by the presence of antibodies changed into non-precipitating complexes [Kleczkowski, 1941 b; Bawden & Kleczkowski, 1941, 1942]. Similarly, the difference between the behaviour of dialysed antisera to human albumin and to bushy-stunt virus respectively, may be because the two antigens are affected differently by the presence of non-precipitating antibody complexes, which are probably formed to a small extent when both antisera are heated in the absence of salts.