PRODUCTION OF ANTIPNEUMOCOCCIC SERUM 583 



upon the determination of the nitrogen content of antigen-anti- 

 body aggregates. 



The first-named procedure, which may be termed a routine pre- 

 cipitin test, has been subjected to various modifications and inter- 

 pretations. With this method, Zozaya, Boyer and Clark (1930), 1589 

 Felton (1931), 407 Falk, McGuire, Valentine and Whitney 

 (1931 ), 387 and Barnes, Clarke and Wight 82 obtained fairly con- 

 sistent results in titrating whole antipneumococcic serum. Inaccu- 

 racies are introduced by variations in spacing of dilutions of se- 

 rums and in establishing end points, in addition to variations in 

 technical manipulations. It is customary to take as the end point 

 the highest dilution of serum affording visible precipitation with a 

 fixed amount of antigen after incubation for two hours in a water- 

 bath at 37.5° and storage overnight in the cold, although Zozaya, 

 Boyer, and Clark preferred to make readings after the preliminary 

 two-hour incubation. If the spacing of serum dilutions is too wide, 

 the titrations will not be satisfactory for comparative purposes. 

 On the other hand, difficulty in reading end points arises when the 

 increment of antibody is too small. Barnes, Clarke, and Wight 

 used the same serum dilutions employed by Sobotka and Fried- 

 lander (1928) 1299 which proved satisfactory in routine tests. The 

 details of the method used by Barnes and his associates are given 

 in the Appendix, page 644. 



The importance of optimal proportions of serum and antigen in 

 precipitin reactions was pointed out by Dean (1911), 310 and re- 

 ported upon in detail by Dean and Webb (1926). 311 Morgan 

 (1923) 912 studied the principle in relation to antipneumococcic se- 

 rum, and further observations were reported by Smith (1932), 1298 

 who stated that titration by the method would fulfill the require- 

 ments placed upon an in vitro method of testing native serum. Un- 

 fortunately, the results with concentrated preparations were not 

 wholly satisfactory. The method is similar in principle to the floc- 

 culation test introduced by Ramon (1923) 1120 for assaying diph- 



