3

here, once and for all, to mention the technique employed and to state the pre-
cautions taken with a view to control these observations. All agglutination
experiments were made macroscopically in the capillary tubes devised by Wright,¹²
by means of which also the serum dilutions were effected. Equal quantities of
the various dilutions of serum, made with normal salt solution, and of sterile
emulsions of the micrococcus were employed in all instances. The preparations
were allowed to stand at laboratory temperature (about 18° C.) for 24 hours,
when the results were recorded. The emulsions were all made in exactly the
same way, namely, an uniform growth on agar of from 4 to 7 days was emulsified
with sterile normal salt solution, about 2.5 cc. of the latter being used for each
agar culture; the bacteria were then killed by heating at 60° C. for 15 minutes
and finally 0.5 per cent. carbolic acid was added.

      During the period that the observations to be related anon were in progress
controls with the bloods of healthy individuals and of persons suffering from other
diseases were made almost daily. In this way an ample number of control ex-
periments were obtained both with the stock organism and with the several races
isolated from our cases. Thus, we tested as controls the following bloods :—

      (1) The bloods of 51 healthy individuals, all natives of India.—Of these,
11 were servants at the Pasteur Institute and 40 were persons who were under-
going inoculation against rabies. In the case of one of the latter individuals a
complete reaction in dilution of 1 in 160 was obtained both with the stock culture
and with one of the races isolated by us. An enquiry into the history of this
person elicited the information that one and a half months previous to our
examination of his blood he had suffered from remittent fever for at least a
fortnight. During this illness he was at home in his village and did not come
under medical observation. But he states that there were no objective symp-
toms except fever and profuse sweating.

      In the remaining 50 cases no reaction was obtained in dilution of 1 in 10
with the stock M. Melitensis or with any of the strains isolated in India.

      (2) All bloods sent to the Pasteur Institute of India for serum diagnosis of
typhoid fever.—During the period over which our present investigation lasted
more than 150 such bloods were examined with the emulsions of M. Melitensis
which were in daily use. Not one of these bloods gave a trace of agglutination
reaction in dilution of 1 in 20, the only dilution in which the preparations were
put up.

      These two series of control experiments, amounting in all to over 200 ob-
servations with different races of M. Melitensis, support our contention that this
bacteriological test is a delicate one and that it affords an easy and absolutely
trustworthy method of diagnosis between this disease and other fevers which
clinically simulate it.