490 THE TUBERCLE BACILLUS GROUP 
estimate the nature and extent of tuberculous processes, to determine 
the vahie of various vaccines, extracts and other products derived 
from tubercle bacilli as curative agents, and to estimate the prognosis 
in individual cases. 
The results on the whole have been somewhat confusing: some 
observers report positive reactions in a large proportion of all cases 
examined, 1 others have found only a small proportion of known tuber- 
culous patients out of a large number give a positive reaction. Some 
of the variance in results depends upon the antigen used, and doubtless 
also the details of the procedure are of importance in evaluating the 
final results. 
Upon the whole, the consensus of opinion at the present time is 
summarized by Petroff ,- who has made the most extensive studies upon 
this subject, as follows: 
"We consider the complement-fixation test in tuberculosis more 
specific than the Wassermann test, basing our conclusions on experi- 
mental data. Complement-fixation is only one of the many links in the 
tuberculosis diagnostic chain. In order to obtain reliable results, the 
technique must be standardized and the test performed only by well 
trained workers." 
As a routine procedure in institutions where tuberculous patients are 
cared for, the test performed several successive times undoubtedly 
gives results that will be of value. As a single, unrepeated test, it is as 
yet of undetermined significance. Generally speaking, in the earlier 
cases where the tuberculous process is demonstrable by clinical methods, 
a high percentage of positive results may be confidently expected. In 
the later stages of the disease, a negative reaction usually results.^ 
C. Bacteriological Diagnosis. — 1. Principle Involved. — (a) Micro- 
scopic E.vomination.—Fhikh, tissues or exudates suspected to contain 
tubercle bacilli are stained preferably by the Ziehl-Neelsen method 
for the demonstration of acid-fast bacilli having the morphology of 
the tubercle bacillus. Other acid-fast bacilli may be confused with 
the tubercle bacillus and their presence must be borne in mind when 
microscopic examinations are made. These will be discussed under 
their appropriate headings. 
It is essential to use absolutely new slides for examination of the 
tubercle bacillus; old slides which have been used for this purpose 
not infrequently retain tubercle bacilli. It is also advisable not to 
make a positive diagnosis unless ten typical tubercle bacilli can be 
demonstrated in the preparation. Among thousands of smears which 
have been examined by Arms, formerly of the Boston State Board of 
Health Laboratory, only one has failed to show ten tubercle bacilli, 
1 For detailed account of various antigens and procedures see Stimson: Hygienic 
Lab. Bull. No. 101, August, 1915. 
2 Am. Rev. Tuberc, 1920, 3, 683. 
5 For a detailed description and discussion, see Brown and Petroff (Am. Rev. Tuberc, 
1918, 2, 525). Petroff: Ibid., p. 523; 1920, 3, 683. 
