82 ALAN ARTHUR BOYDEX. 



liminary titrations are first made with the homologous antigen 

 and then with various heterologous antigens in the manner just 

 described. But in all cases in which sufficient antiserum was 

 available these preliminary titrations were followed by one or 

 more double titrations made in the following manner. 



One series of tubes was set up with dilutions starting from a 

 2 per cent, standard solution. A second series was then set 

 up starting with a dilution of i : 150 of this same antigen. This 

 was made by mixing \ cc. of 2 per cent, solution (i : 50) with I cc. 

 of salt solution. By using this double series of tubes the differ- 

 ence between the concentration of one tube and its nearest 

 successor was reduced to 50 per cent, instead of 100 per cent. 

 See Fig. I. 



150 300 600 1200 2400 4800 9600 19200 



OOOOO0 O 



100 200 400 800 1600 3200 6400 12800 



0000000 



FIGURE i. A diagram to illustrate the double series of dilutions employed in 



titration. 



Now if the preliminary titration had given a titer of 6,400, 

 in the final titration only the tubes from 4,800 to 19,200 (marked, 

 with an x in Fig. i) would be given antiserum. The reading 

 in the final titration falls generally within one tube of the pre- 

 liminary titration. By using this double titration a more ac- 

 curate value of the titer is obtained without the need of using 

 excessive amounts of antiserum. The tubes containing anti- 

 serum were then put in order of increasing dilution in a rack and 

 read as described. 



EXPERIMENTAL RESULTS. 



RESULTS ON TECHNIQUE. 

 i . The Range of Error in Reading the Ring Test. 



Experimental results of any kind whatever have significance 

 only insofar as it is definitely known that they pass beyond the 

 range of experimental error. The first step in ilu- application 

 of the preripitin reaction to biological problems should then 

 consist in the establishment of the range of error in reading the 

 reaction for each investigator. 



