836 COMPLEMENT FIXATION 



result of twenty years' experience with the test, it is still crude. The Wassermann 

 technique, with its allowance for so-called "anticomplementary" and "non-specific" 

 activities of the reagents, is only an approximate measure of the specific activity of 

 the sera. Yet this is the form of test which has been used in the majority of the studies 

 of the bacterial antigens, whereas accurate titration of the activity of the specific im- 

 mune serum is essential. This fact, apparently, has been appreciated by Calmette 

 and Besredka in their application of complement fixation to the diagnosis of tubercu- 

 losis. A similar conclusion was forced upon us in our work with the tubercle antigens, 

 and the test was therefore remodeled to provide an accurate titration of the activity 

 of all the sera tested. Thus, only as a result of long experience with the test performed 

 in this manner is any standardization of the antigen possible. The ranges of activity 

 possessed by normal healthy individuals, as compared with those of infected indi- 

 viduals, must be accurately determined with a large number of sera. In fact, it is 

 largely owing to the difficulty of determining these ranges of activity that the practical 

 application of the test to the diagnosis of disease has been so unsatisfactory. Long 

 experience with the Wassermann test was necessary to determine these ranges of 

 activity, and we have only very limited data recorded with the diff'erent bacterial 

 antigens, the non-specific fixation of which has so often overlapped the specific 

 activity. 



In general, the results of the practical application of the complement fixation 

 test with the bacterial antigens in the diagnosis of infection has been beset with pit- 

 falls which have been avoided only in a very limited number of the researches, and 

 one finds in these reports conclusions invariably so qualified as to destroy the prac- 

 tical value of the test. Thus, positive reactions are obtained in normal healthy indi- 

 viduals or those undergoing other infectious processes, and the infected individual oc- 

 casionally faUs to react. This, however, is also true of the Wassermann test in the 

 diagnosis of syphilis, but experience has so refined the test in its procedure and con- 

 trol as to eliminate such a large proportion of these instances that only a negligibly 

 small percentage of error remains. In the diagnosis of tuberculosis, experience with 

 the better methods is now accumulating so rapidly that results comparable with 

 the Wassermann test in syphilis are assured. It requires a careful, precise technique 

 and, as knowledge increases, the test may acquire prognostic significance. 



To sum up, the physico-chemical balance essential to the antigen and the elements 

 concerned in it are still so far from standardized that the specific, refined, purified ex- 

 tracts, although in general superior to the cruder antigens which have been empirically 

 "established, are not yet entirely reliable and our knowledge is not yet sufficiently precise 

 to place their standardization upon a scientific basis. Similarly, some of the crude bac- 

 terial antigens empirically established may prove even more serviceable but unless 

 carefully controlled and, so far as possible, standardized, there is great danger of ob- 

 taining misleading results. The significance of the results really depends upon the 

 precision and extent of the comparisons and controls. Although essential, fundamental 

 knowledge regarding the nature and preparation of the antigens is still lacking, mate- 

 rial progress has been made in the study of the different extracts and jireparations. 

 Research, however, has not very generally recognized the importance of adapting the 



