ANTIBODIES. 105 



toneal cavity of a guinea-pig. This test requires from one- 

 half to one hour for completion. By reversing the test an 

 unknown bacterium or an unknown serum may be identified. 



In making the " in vitro " test-tube reactions small defi- 

 nite quantities of bacteria are mixed with the serum to be 

 tested. This serum must be inactivated by heating to 56 C. 

 A small quantity of normal serum to furnish the complement 

 is added and the whole placed in the incubator at 37.5 C. 

 After a few hours plates are poured and the colonies counted. 

 The extent of bacterial destruction can be determined by the 

 growth or non-growth. Control sets must, of course, be 

 made. 



For the demonstration of hsemolysins washed red blood- 

 corpuscles of the specific animal against which the reaction 

 is to be made must be prepared. They are prepared by re- 

 ceiving blood in a flask or large tube, according to the quan- 

 tity it is desired to obtain, containing some glass beads or a 

 looped wire, so that by shaking fibrin may be removed and 

 clotting prevented. Other methods are immediate centri- 

 fuging and washing with salt solution, or receiving the blood 

 directly into a solution of sodium citrate 1 per cent, and 

 sodium chloride f per cent, and centrifuging. For the actual 

 tests a 5 per cent, suspension of the corpuscular sediment in 

 0.85 per cent, salt solution is made use of. A small portion 

 of this suspension is mixed with a small portion of the serum 

 to be tested, and the whole brought up to a definite amount 

 with 0.85 per cent, salt solution. If it is desired to have a 

 fixed amount of complement the serum must be first heated 

 ,to 56 C. to inactivate, and a definite quantity of fresh com- 

 plement added (guinea-pig serum). Control tubes with nor- 

 mal serum are also made. The contents of the tubes are 

 thoroughly mixed and incubated for one hour at 37.5 C. 

 If haemolysis takes place, a deep Burgundy red color will per- 

 vade the tube, which is positive. If the serum is not active 

 for the corpuscles, no haemolysis will take place and the cor- 

 puscles will settle down without any change of color. A 

 wide variety of results between these two extremes is occa- 

 sionally observed. Bordet and Gengou, while following their 



