2IO METHODS OF SERO-DIAGNOSIS. 



stant and the quantity of antigen varying according to the dilu- 

 tions in which it is added. Another control tube is necessary in 

 addition to those used in the previous tests, in order to show 

 that the antigen itself has no hemolytic action. The tubes are 

 now incubated as before and, on examination, the first tube of 

 the series showing complete haemolysis of the contained blood 

 indicates the largest quantity of antigen which does not interfere 

 with haemolysis, and we use, as above-mentioned, half of this 

 quantity in the final test. 



In making the test proper the following procedure is 

 adopted :■ — 



The serum from the horse to be examined is obtained by 

 withdrawing a small quantity of blood, the serum being collected 

 after coagulation. This serum is now inactivated by heating for 

 about half an hour from 58 to 62 C. Heating at the higher 

 temperature seems to have an effect in reducing the quantity of 

 certain non-specific bodies in the horse serum, which can fix a 

 certain amount of complement by themselves and may, if not 

 destroyed, interfere with the test. 



Now, we take four tubes and in these we place the comple- 

 ment in the determined amount, and to the first two we add .1 oc 

 and to the second pair .2 cc of the serum to be tested. Then to 

 one tube of each pair we add the previously determined suitable 

 dose of the antigen. The volumes of the fluid in the tubes are then 

 made up to a certain fixed quantity, and the tubes are placed in 

 the incubator at 37 ° C for one hour. If the specific glanders 

 antibody is present, the complement should now, through its 

 medium, combine with the antigen, and we see if this has occurred 

 by placing in each of the tubes, at the end of the hour, the already 

 e.-timated suitable amount of inactivated haemolytic serum and, 

 along with it, 1 cc of the 5 % suspension of sheep's red corpuscles 

 in physiological salt solution. The tubes are now shaken and 

 again placed in the incubator, at 37 C, for two hours. The 

 necessary controls for the complement, inactivated haemolytic 

 serum, antigen, and physiological salt solution are, of course, 

 also employed. Now the tubes are removed from the incubator 

 and put aside for some time in a cool place, in order that their 

 contents may settle, and then the reading of the test is made. 



If no haemolysis has occurred in the tubes to which the 

 antigen was added, it indicates that the complement has entered 

 into .union with the antigen, and that none is left free to combine 

 with the mixture of inactivated haemolytic serum and sheep's 

 -corpuscles subsequently added; but, as it can only do this when 

 the specific glanders anti- or immune-body is present, we con- 

 clude that the horse is affected with glanders. In the tubes to 

 which no antigen was added haemolysis should occur, and they 

 are merely controls to show that the serum of the subject by itself 

 does not prevent or interfere with the occurrence of haemolysis. 



This method of diagnosis by complement fixation has proved 

 of great value in the case of glanders, in the diagnosis of which 

 it is sometimes applied exclusively, but perhaps most often in 

 combination with the agglutination test, and the precipitin test is 



