There are certain factors that should be taken into consideration in interpreting these serological results. The fact that the sera of all the suspicious horses and controls were heated at 56°C. for one-half hour would tend to cut down the agglutinating titer of those sera. However, as the same technic was carried out in both test and control animals the results have a true relative significance. This fact may account for the absence of agglutinins in such a large percentage of the normal control horses.
The results definitely indicate that in horses giving a suspicious intra-palpebral mallein reaction of this type the complement fixation test is of more value than the agglutination reaction. In the former 75.4 per cent were positive while in the latter only 57.4 per cent were positive, even when all the borderline cases were included.
These observations agree fairly well with published statistics. In Povitzky's recent article (this Journal, 1918, 3, 463) she records work done in the laboratories of the New York City Department of Health on 123 horses which were proven at autopsy to have glanders. The complement fixation reaction was positive in 75.6 per cent of these cases, while the agglutination reaction was positive in 64.2 per cent. In her work she used horse sera which had not been heated to 56°C. for one-half hour, but again this would be partly neutralized by her using a titer of 1:1000 as necessary for a positive result. The consensus of opinion is that the agglutination reaction gives higher positive results in acute cases, while in chronic or subacute cases the complement fixation is the more reliable.