IMMUNITY AND ANAPHYLAXIS 207 



of precipitates, due to precipitins in the rabbit's serum for 

 sheep's serum (from insufficient washing of the injected 

 corpuscles), acting on any sheep's serum present, from 

 insufficient washing of the corpuscles used in the test. The 

 serum is inactivated by heating to 56 C. The quantity 

 of haemolytic serum to be used must be accurately deter- 

 mined,' in relation to a definite amount of complement. 

 This is the more necessary since it has been shown that, 

 (contrary to Ehrlich's commonly accepted conception 

 that immune -body and complement react in definite 

 combining proportions), the haemolytic immune-body and 

 a complement react in inverse proportions ; that is, the 

 more immune-body, the less complement is required. The 

 bearing of this on the test is that a very small quantity of 

 complement left over by the combination of syphilitic- 

 immune-body and antigen might suffice, in presence of 

 large excess of h hemolytic immune-body, to cause haemo- 

 lysis ; thus giving a negative result in a positive case. 

 The quantity is determined by putting up a number of 

 mixtures, each containing o-i c.c. of fresh guinea-pig 

 serum to give complement, and 1 c.c. of the 5 per cent 

 emulsion of sheep's corpuscles. To each mixture 

 inactivated haemolytic serum is added, in smaller and 

 smaller quantities. The smallest quantity which gives 

 complete haemolysis is taken as the unit amount to be used 

 in the test. The haemolytic unit may therefore be defined 

 as " the smallest quantity of inactivated haemolytic 

 serum which, in the presence of a stated amount of com- 

 plement, is able to cause complete haemolysis in 1 c.c. of 

 a 5 per cent emulsion of washed red blood cells." 



4. The Complement. This is obtained by drawing blood 

 from a guinea-pig, allowing it to clot, and pipetting off the 

 serum, or separating by the centrifuge. Such serum, 

 kept at a low temperature, preserves its complement in 

 a fairly constant amount for three days. 



5. The Serum to be tested is got from the patient under 

 aseptic precautions, from the median basilic vein, from 

 the finger, or from the ear. Three to four c.c. are with- 

 drawn, as 1 c.c. of clear serum is required to go over 

 the tests and controls. It is inactivated by heating to 

 56 C. in a water-bath for 20 minutes. Noguchi advises 



