VI. SEROLOGICAL REACTIONS 643 



any of the others ; occasionally two tubes run parallel, indicating that 

 the true optimal antigen concentration lies midway between them. 



(3) Final Test with Narrow Range of Antigen Dilutions: Eight 

 tubes are numbered and buffer solution is added from a graduated 

 pipette; nil to tube 1, 0.1 cc. to tube 2, 0.2 cc. to tube 3, and so on. 

 With a fresh pipette, enough antigen solution is added to each tube to 

 bring the total content up to 1 cc. The dilution of SSS solution to be 

 used is decided from the result of the rough test, the object being to 

 have the optimal tube at about the middle of the series. Serum is 

 added in 1 cc. quantities as before and the tubes are incubated. 

 Closer observation is necessary than in the previous test because, with 

 this narrow range of antigen dilutions, precipitation extends rapidly 

 from the optimal tube to those on either side. A hand lens is used 

 until one of the precipitates can be seen by the naked eye, but even 

 so it is often impossible to be sure which tube first shows precipita- 

 tion. The optimal tube, however, is distinguished more by the fact 

 that in it the particles remain larger for some considerable time than 

 those in any other tube and also the fluid part of the mixture looks 

 clearer and more transparent than in any of the others. The antigen- 

 antibody ratio is then calculated by dividing the antigen dilution of 

 the optimal mixture by the serum dilution. 



(4) Confirmation of the Precipitation Result by Tests for Excess 

 of Antigen : Occasionally two, three, or even four tubes run so nearly 

 parallel that it is impossible to be sure which represents the optimal ; 

 that is to say, an optimal band is produced instead of an optimal 

 tube. In such cases it is still possible to find the optimal ratio by 

 centrifuging the mixtures after precipitation is complete in all of 

 them, and testing the supernatant fluids for excess of antigen. In 

 practice it is sufficient to show that antigen remains free only in those 

 tubes to the left of the optimal. A simple method is to layer the su- 

 pernatant fluid over a potent serum suitably diluted in normal horse 

 serum; the formation of a white ring at the junction of the fluids in- 

 dicates that polysaccharide has remained in the supernatant fluid. 



b. Nitrogen determination (Heidelberger, Sia, and Kendall 630 ). If, 

 on addition of a 1 to 10,000 saline solution of the specific polysaccha- 

 ride to a portion of the immune serum, precipitation is rapid and heavy, 

 0.5 cc. of the serum should be used for the quantitative determination. 

 This amount will suffice for all except very low grade serums, of which 

 it is better to use 1 cc. The serums should be measured in duplicate 



