584 BIOLOGY OF PNEUMOCOCCUS 



theria antitoxin, and since the technique is given in the Appendix, 

 its details need not be described here. Smith used a small number of 

 samples in developing this method as applied to antipneumococcic 

 serum, but Barnes, Clarke, and Wight 82 corroborated his findings 

 in titrations on a larger series of unconcentrated Type I and II 

 serums. 



In 1935, Felton and Stahl 432 described a method of titration by 

 determining the combining equivalents of antibody and SSS. The 

 method is based on the principle of optimal proportions, and the 

 point of equivalence is measured by testing for excess antigen in a 

 series of mixtures of antigen and antibody, rather than by deter- 

 mining the point at which particulation first occurs as in Smith's 

 method. The method reported by Felton and Stahl appears to be 

 applicable to concentrated antibody preparations as well as to na- 

 tive serum and this fact may favor its adoption in the event that in 

 vitro methods of standardization are eventually substituted for 

 mouse protection tests. 



A different method of estimating the antibody content of serums 

 was introduced by Heidelberger, Sia, and Kendall (1930), 63 ° who 

 calculated the amount of immune nitrogen in SSS-antibody pre- 

 cipitates and converted this figure into terms of immune protein 

 present in the serum. Additional details of the method and results 

 of titrations have been reported in subsequent publications. 619 ' 624 

 The essentials of the technique may be found in the Appendix, 

 pages 643 and 644. 



CORRELATION BETWEEN IN VITRO AND IN VIVO TESTS 



Much of the evidence obtained thus far concerning the relation- 

 ship between laboratory tests and therapeutic efficacy of antipneu- 

 mococcic serum has been based on mouse protection tests, and the 

 correspondence has proved to be satisfactory for general pur- 

 poses. If the mouse test is to be supplanted by a chemical or sero- 

 logical method of standardization, the correlation between the 

 results obtained by mouse protection tests and various in vitro 

 methods of assay must be determined. The best approach to this 



