2

of 17 men, many of whom had been treated with tuberculin. He records having
in some instances observed sedimentation in a dilution of 1 in 500; in dilution of
1 in 10 and 1 in 20 nearly all gave complete sedimentation. It was also noted
that all the strains used did not give the same results, but that no difference was
observed between patients previously treated with tuberculin and those not so
treated. On this point Konrich's results are in direct opposition to those obtained
by Birt and Lamb,⁸ who in a series of over 150 controls found no sedimentation
in dilution above 1 in 10, and also to those of M. C. Nicolle,⁹ who, after an
examination of the sera of 35 persons suffering from other diseases, concluded
that a complete reaction in dilution of 1 in 10 was strong evidence of the case
being one of Mediterranean fever.

         Cornwall ¹⁰ examined the blood of nine healthy men and of nine cases of long-
continued, but ill-defined, fever which were clinically not characteristic of
Mediterranean fever, as regards their agglutination reaction with M. Melitensis.
He relates having got in some of these cases a complete or partial reaction in
dilution varying from 1 in 20 to 1 in 400, and concludes by saying "that even in
dilutions of 1—80 a complete agglutination of M. Melitensis is not pathognomonic
of Malta fever and that the blood of persons suffering from various febrile
conditions may have a higher agglutination power on the micrococcus than that
of persons in good health."

         In spite, however, of the observations of Konrich and Cornwall we still hold,
that the serum agglutination method for the diagnosis of Mediterranean fever,
when properly and rigidly controlled, is one of the most delicate bacteriological
tests with which we are acquainted and ought to take its place as an important
addition to the ordinary clinical methods for the diagnosis of this disease. This
opinion is evidently shared by the Commission ¹¹ for the investigation of Mediterra-
nean fever which is at present at work under the supervision of an Advisory
Committee of the Royal Society. A perusal of the first two parts of their report,
the only parts which have as yet appeared, will show the great use which these
workers make of this test both in the diagnosis of cases and as a means of
identifying the specific coccus. This opinion is strengthened by the result of
the investigations which we have now to bring forward.

         These investigations are concerned with observations made on several cases of
continued fever in India, in which the diagnosis of Mediterranean fever was arrived
at by means of the serum sedimentation test. Further, in a considerable number
of these cases this diagnosis was proved to be correct by the isolation from the
spleen during life of a coccus bacteriologically indistinguishable from M. Melitensis.
As the preliminary step in all our cases consisted in testing the serum as regards
its agglutinin content, and as much reliance was placed on the serum reactions
of the specific organism when isolated from the spleen, it will be convenient