TEST FOR BACTERIUM PULLORUM IlirFECTION IK FOWLS. 13 
On June 26 all the 47 birds and 9 controls were reinjected and ex- 
amined 5 hours later. At this time every bird, including controls, 
displayed a swelling varying in the different individuals from a 
trace tq five times the normal thickness. The observation merely em- 
phasizes the fact of the occurrence of a nonsignificant swelling fol- 
lowing, injection with the diagnostic agent. 
At 24 hours 39 birds displayed swelling of the wattle varying from 
a trace to enlargement to five times the normal thickness. Autopsy 
revealed undoubted, lesions in 30 of these, questionable lesions in 
7, and no lesions in 2. Total agreement between the results of this 
reading, the agglutination test, and autopsy findings occurred in 70 
per cent of the birds tested. In two cases, or 4 per cent of the 
birds, the positive readings by the agglutination test were not sup- 
ported by the negative results of the intradermal test and the 
autopsy. In 1 case, or 2 per cent, negative results of the intradermal 
test were contradicted by the positive results of agglutination test 
and autopsy. Thus, the results yielded by the first and second 24- 
hour readings of the test on supposedly infected birds vary but 
little. 
The results yielded by the test on the control birds were perfect, 
as confirmed by the autopsy. The only two birds that displayed 
traces of swelling proved on autopsy to be infected. 
The fact that the results of the agglutination test, intradermal 
test, and autopsy are in complete agreement in 70 per cent of the 
cases, coupled with the fact that the absolute diagreements are very 
small, indicates that the two tests are equally accurate. 
The results obtained at the autopsy of the birds emphasize the 
difficulty of determining a standard for comparison of the accuracy 
of the two tests under trial. Thirty-one cases, or 64 per cent, were 
found to possess unquestionable lesions consisting of the angular ova 
characteristic of the infection. All of the cases had given positive 
reactions to one or both tests. In nine cases, or 10 per cent, the 
autopsy was inconclusive in that there were present only very .small 
dark ova or cysts. Of these 9 questionable cases 3 had given ques- 
tionable agglutination readings but positive intradermal reactions. 
In two cases the agglutination and intradermal tests disagreed. In 
fcur cases both tests had given positive results. 
SIGNIFICANCE OF SWELLING AS AN INDICATION OF A REACTION. 
In determining the significance in diagnosis of an edematous 
sAvelling of a wattle one is confronted with the fact that in all birds 
such swelling occurs shortly after injection. The problem is to 
determine the point of time after injection to read the test when 
this preliminary swelling has disappeared, yet not too late to escape 
