H 
Journal of Agricultural Research 
Vol. VIII, No. i 
In order to establish whether a tuberculin injection would interfere 
with the results of the tests, two vigorous young animals were sub¬ 
cutaneously injected with 2 c. c. of Bureau of Animal Industry tuberculin. 
Blood samples were obtained from these animals prior to the injection 
and subjected to the complement-fixation test with negative results. 
Subsequent to the tuberculinization, blood was drawn twice weekly from 
them. On the fourth day following the injection a positive reaction was 
obtained; after 10 days the fixation was complete, persisting for four 
weeks; it was then followed by partial fixation, and completely subsided 
after six weeks. It is possible that in some animals the fixation might 
persist for a still longer period; therefore it is essential to determine 
whether an animal has been injected with tuberculin, and if so, how long 
ago, before judgment can be passed on the results of the complement- 
fixation test. 
In connection with these experiments a peculiar phenomenon was 
observed in several instances. In using an antigen which contained a 
higher amount of precipitate than the standard established by titration, 
a reaction which might be designated as a reverse reaction followed—that 
is, instead of the expected positive reaction it proved negative, and vice 
versa. 
While it is not intended to offer a theoretical explanation for this 
peculiarity, it serves to emphasize the necessity for a careful standardiza¬ 
tion of the antigen each time the test is undertaken. The apparent deli¬ 
cacy of the test in the diagnosis makes it essential to employ more than 
one known positive and known negative serum in standardizing the anti¬ 
gen. In the experiments cited the antigens were always standardized to 
a series of known positive and negative serums. 
COMPARATIVE TESTS WITH VARIOUS ANTIGENS 
The advantages of an antigen prepared by our method have already 
been described. In order, however, to establish its comparative value 
various other antigens were used and the following brief mention is. made 
of the methods of their preparation and t*he results obtained. 
For the antigen prepared from tubercle bacillary emulsion, dilutions were 
made from a stock emulsion in which 30 mgm. were suspended in each cubic 
centimeter of the emulsion. The stock suspension was never heated and 
was prepared by milling the bacilli in a porcelain mill for about 40 days. 
The dilution used as antigen consisted of 3 mgm. per cubic centimeter in 
either physiological salt solution or in hypotonic solution. It was 
observed that the emulsion made up with the normal physiological salt 
solution gave an excessive number of positive reactions, whereas the one 
prepared with the hypotonic solution gave much closer results. Never¬ 
theless the comparative results proved this antigen to be unsatisfactory. 
Its greatest disadvantage proved to be that the difference between the 
