TEST FOE BACTERIUM PULLORUM INFECTION IN FOWLS. 9 
reactions rated as either slight or positive. Thus, the 24:-hour in- 
terval yielded the largest percentage of reactions. Practically all 
birds, both those inoculated and controls, exhibited a swelling shortly 
after injection and therefore no diagnostic value has been attributed 
to swellings observed before the lapse of 24 hours. 
Three birds gave negative readings at both 24 and 48 hours. 
Autopsy of two of these revealed unquestionable lesions of pullorum 
infection, from which the organisms were obtained, while in the 
third one the lesions were questionable and no culture was obtained. 
Thus, the test failed to detect 6 per cent of the birds in which lesions 
were found. 
In all but two cases the same birds were retested after an interval 
of 7 or 8 weeks. Of the 30 birds retested 22, or 73 per cent, gave a 
reaction rated as either a trace, slight, or positive at 24 hours on the 
second test. At 48 hours on the second test only 8, or 26 per cent, 
displayed reactions rated as a trace, slight, or positive. Further, 8 
birds, or 26 per cent, showed no reaction at either 24 or 48 hours. 
It is evident that a retest after an interval of about 8 weeks is far 
less reliable than a first test. 
Of the 32 birds tested, autopsy revealed unquestionable lesions in 
18, or 56 per cent. In 8, or 25 per cent, the lesions were regarded as 
questionable. In 6 birds, or 16 per cent, no lesions were found, 
although all were positive to the test at 24 hours. 
TAventy-six controls were tested for the first time. These had 
been gathered from various sources and there was no assurance that 
they were free from infection. Of these, 5 at 24 hours after injec- 
tion displayed swellings rated as slight or positive and 4 displaj^ed 
the same condition at 48 hours. At autopsy 2 were found to be in- 
fected, and 1 through accident was not examined. No lesions were 
found at autopsy of 2 ; however, 1 of these came from the same flock 
as one of the unquestionabl}^ infected controls, and had been in the 
same cage as the infected bird. 
While agglutination tests were made on serum drawn from inocu- 
lated birds, after injection with the diagnostic agent, and the results 
appear in the various tables, it is realized that agglutination would 
naturally be expected as a result of the various injections. We have 
observed that as a result of the artificial infection with cultures of 
Bacterium pullorum^ the agglutinating value of the serum of these 
birds varied within a wide range. Some birds gave an agglutina- 
tion at a dilution of 1 : 1,000, while others that had been repeatedh^ 
injected with the test fluid gave no agglutination, owing to the strong 
bacteriolytic properties of their sera, presumably resulting from the 
various injections. Negative control birds after one injection with 
the test fluid gave an agglutination titer of 1 : 50. 
