52 THE AMERICAN MONTHLY [Feb 
idence must not be separated from the clinical. The bac- 
teriologist may be able to identify a germ as soon as seen 
under the microscope, but to be absolutely certain, it is 
necessary to make cultures and animal tests. The meth- 
ods are too cumbersome and take too much time for most 
diagnostic purposes. Ifthe bacteriologist knows the clin- 
ical symptoms of the patient at the time the culture was 
taken, it might remove any uncertainty in the identifica- 
tion of the germ. For instance, a Health Department box 
was inoculated and sent to a pathological laboratory for 
examination. 
The bacteriologist reported the finding of the diphthe- 
ria bacilli. Had the culture been accompanied with the 
information that it had been inoculated from a healthy 
vagina, the germ would have been recognized as the ba- 
cillus vaginalis. The method of taking the culture is also 
important. The condition of material sent to the labor- 
atories often shows very careless methods or ignorance 
of bacteriology. The surface of the medium is scarcely 
touched with the inoculating swab; cotton swabs come 
wrapped in newspaper, envelopes, or dirty bottles, Some 
of the fluids of the body are destructive togerms. When 
taking blood for examination, it should be at once diluted 
by large quantities of broth to prevent.it from destroy- 
ing the germs. It is the verdict of the bacteriologist that 
swabs from suspected anginas should be used at once to 
inoculate culture media. If this is not done at once, the 
action of the saliva may destroy the bacilli and thus pre- 
vent the detection of their presence. 
For this reason, swabs alone cannot be sent by mail. It 
must be remembered that culture media furnishes a dif- 
ferent environment than the human tissues and modifi- 
cations in the morphology and virulence of germs may oc- 
cur. Most students of bacteriology know the variations 
which occur in the Leeffler bacillus when grown on agar 
agar and blood serum; but it is not so well known that 
