278 IMMUNOLOGICAL METHODS OF DIAGNOSIS 



mended the use of an antihuman instead of an antisheep 

 hemolytic system. I have found, however, that this is not only 

 inconvenient, but also unnecessary, as the natural antisheep ambo- 

 ceptor can be readily removed by merely diluting the inactivated 

 serum of the patient with five times its volume of the corpuscle 

 emulsion, and incubating the mixture for thirty minutes in the 

 water-bath at 37 C.; the corpuscles with the anchored "natural" 

 antisheep amboceptor are then thrown down with the centrifuge, 

 when the supernatant, now already diluted, serum is ready for 

 use. This step is now carried out as a matter of routine in my 

 own laboratories, and assures perfectly satisfactory results; with the 

 use of the Noguchi antigen, and this modification the objectionable 

 "Nachlosung" (continuing hemolysis at the end of the experiment) 

 is no longer a source of error and hence of worry. 



The test-tubes which we use measure 4 inches in length by I of an 

 inch inside diameter. 



METHOD. When everything is in readiness the complement and 

 amboceptor are adjusted to one another, using dilutions of 1 to 1000, 

 1 to 2000, 1 to 3000 up to 1 to 6000 of the amboceptor; 0.5 c.c. is 

 our unit of measure, and we accordingly combine 0.5 of the various 

 amboceptor dilutions with 0.5 c.c. of the complement (1 in 10) and 

 0.5 c.c. of the corpuscle emulsion (5 per cent.). The tubes are placed 

 in the water-bath at 37 C., and are frequently shaken. At the 

 expiration of thirty minutes the highest dilution is noted at which 

 complete hemolysis occurs. The amboceptor dilution to be used in 

 the actual experiment is then made 2i to 3 times as strong. Thus, 

 if complete hemolysis occurred at 1 to 6000, we would use a 1 to 

 2500 or a 1 to 2000 dilution. 



The antigen has been previously tested, as described. With 

 human heart antigen, one can usually use a dilution of 1 in 10. 



The titers of the various reagents having thus been ascertained, 

 the experiment proper can now be carried out (tubes marked E), 

 using 0.5 c.c. of the patient's serum (1 in 5) combined with 0.5 c.c. 

 of complement (1 in 10) and 0.5 c.c. of antigen (1 in 10). At the 

 same time controls (tubes marked C) are prepared, in which the 

 antigen is left out, so that 0.5 of each serum is combined with 0.5 c.c. 

 of complement and 0.5 c.c. of saline (in place of the antigen). The 

 E and C tubes properly marked with the patient's numbers are 

 placed in the water-bath for thirty minutes and then receive, 



