or urine. Others are then inoculated with a loop of
urine or much diluted feces. In this way we are apt
to have some plates with just the right amount of
colonies.
Appearance of the Colonies. Capaldi thus describes
the differentiation: Typhoid: Small, gleaming, transpa-
rent, almost colorless colonies (by reflected light, blue).
Colon: Large, milky colonies (reflected light, brown).
In using the medium it was found that even in a
pure plate of typhoid the colonies vary much in size
and appearance, while different typhoids show indi-
vidual differences in growth. In general, a medium-
sized, gray-white colony, with a few refractive granules,
438 BACTERIOLOGY.
is the typhoid (Fig. 56). However, it is often trans-
parent, without the refractive granules; sometimes
with a nuclear centre, and sometimes of equal con-
sistency throughout. Streptococci simulate typhoid,
but a high-power lens will show the coccus.
FIG. 56.
Colonies of typhoid bacilli, on Capaldi medium, slightly magnified.
Colon colonies are usually much larger than the
typhoid; a decided brown color, very large, refractive
granules, and in general quite different in appearance
(Fig. 57).
The best way to work with the Capaldi medium is to
make several plates with different typhoid cultures,
observe carefully all the variations in the colonies, and
bear these in mind when working with the mixed
plates. After these precautions have been taken the
BACILLUS TYPHOSUS.
439
medium will be found very satisfactory. The colonies,
as a rule, appear characteristically in twelve to eigh-
teen hours, and thus give a quick method of diagnosis.
The two media together (Capaldi and Eisner) work
excellently, as one is an aid to the other. When
many colonies of the typhoid bacilli were present the
FIG. 57.
Colonies of colon bacilli. Capaldi medium slightly magnified.
differentiation was usually easily seen upon both media,
and the two together made diagnosis almost certain.
The bacilli from the suspected typhoid colonies can be
quickly tested sufficiently for practical purposes on the
Hiss tube medium and by the reaction between the
bacilli and the serum from an immunized horse.
As to the comparative merits of these three media,
it is probably safe to say that any one of them will, in
the hands of one accustomed to them, reveal the typhoid
bacilli, if they are present, except perhaps when they
440 -B4 CTERIOL OGY.
exist in only the most minute numbers. The Eisner
method has the objection that it is very difficult to work
with in hot weather. The Hiss plate medium has the
objection that it is a difficult medium to prepare. If
the acidity is not just right the thread outgrowths do
not appear. Indeed, the only sure way is to test a new
batch of medium with a pure culture and alter the
reaction until the culture grows correctly. A very few
varieties of the typhoid bacillus do not produce typical
thread outgrowths from the colonies.
The Capaldi medium has the objection that some of
the typhoid and some of the colon colonies frequently
look much alike. If one, however, will always pick
out the colonies which look most like the typhoid, it
will usually turn out that typhoid bacilli have been
obtained if any were present. Personally, for general
use I prefer the Capaldi medium for the plate cultures
and the Hiss tube medium for identifying the bacilli
obtained. Through these media we are now in a posi-
tion to obtain and identify typhoid bacilli from feces,
urine, etc., within forty-eight hours.
Recently numerous investigations have been carried
out to discover how frequently and at what period in
typhoid fever cases bacilli were present in the feces
and urine. In the laboratory Hiss has recently exam-
ined the feces in forty- three consecutive cases, thirty-
seven of which were in the febrile stage and six con-
valescent. In a number of instances only one stool was
examined, but even under these adverse conditions the
average of positive results in the febrile stage was 66.6
per cent.
Out of 26 cases of typhoid fever in hospitals exam-
ined ; 21 were in the febrile stage and 5 convalescent.
BACILLUS TYPHOSUS. 441
In the febrile cases in 17 the presence of typhoid bacilli,
often in great numbers, was demonstrated. Thus in
these carefully followed cases the statistics show over
80 per cent, of the febrile cases positive. The bacilli
were isolated from these cases as early as the sixth day,
and as late as the thirtieth day, and in a case of relapse
on the forty-seventh day of the disease. The conval-
escent cases gave uniformly negative results, the earliest
examination having been made on the third day after
the disappearance of the fever. The bacilli seemed to
be more numerous in the stools from about the tenth or
twelfth day on. These observations, with regard to the
appearance of the bacilli in the stools during the febrile
stage and their usually quick disappearance after the
defervescence, have been confirmed by others. In
several cases in which no Widal reaction was demon-
strated the bacilli were isolated. From private sources
between the seventh and twenty-first day of the disease,
experience thus far obtained seems to indicate that the
bacilli may be obtained from about 25 per cent, of all
cases on the first examination and from about 75 per
cent, after repeated examinations. In some samples of
feces typhoid bacilli die out within twenty -four hours;
in others they remain alive for days or even weeks.
This seems to depend on the bacteria present in the
feces and upon its chemical formation. Probably the
presence of typhoid bacilli in some stools and their ab-
sence in others must be explained largely upon the
characteristics of the intestinal contents. The short
life of the typhoid bacillus in many specimens of feces
suggests that stools be examined as quickly as possible.
In fact, unless the physician wishes to take the
trouble to have the sample of feces sent immediately
442 BACTERIOLOGY.
to the laboratory, it is hardly worth while for the bac-
teriologist to take the trouble to make the test.
Typhoid Bacilli in the Urine. Of even more interest
than the presence of the bacilli in feces is their frequent
occurrence in great numbers in the urine. The results
of the examinations of others as well as those of our
own indicate that the typhoid bacilli are not apt to be
found in the urine until the beginning of the third week
of the fever, and may not appear until much later.
From this on to convalescence they appear in about 25
per cent, of the cases, and usually in pure culture and in
enormous numbers. Of nine positive cases examined
by Richardson 1 two died and seven were discharged.
At the time of their discharge their urine was loaded
with typhoid bacilli. We have noted similar cases. In
one the bacilli persisted for five weeks. Undoubtedly
in some cases they persist for months. When we think
of the chances such cases have to spread infection as
they pass from place to place, we begin to realize
how epidemics can start without apparent cause. The
more we investigate the persistence of bacteria in con-
valescent cases of disease the more difficult the preven-
tion of their dissemination is seen to be. The disin-
fection of the urine should always be looked after in
typhoid fever, and convalescents should not be allowed
to go to places where contamination of the water-supply
is possible without at least warning them of the neces-
sity of great care in disinfecting their urine and feces
for some weeks. Richardson made the interesting dis-
covery that after washing out the bladder with a very
weak solution of bichloride of mercury the typhoid
bacilli no longer appeared in the urine.
1 Journal of Experimental Medicine, May, 1898.
BACILLUS TYPHOSUS. 443
The Detection of Typhoid Bacilli in Water. This sub-
ject is considered on pages 247 and 248. There is
absolutely no doubt that the contamination of streams
and reservoirs is the frequent cause of the outbreak of
epidemics of typhoid fever, but the actual finding and
isolation of the bacilli is a very rare occurrence. This
is owing to the contamination often having occurred
and passed away before the bacteriological examination
is undertaken, and also because of the great difficulties
met with in detecting a few typhoid bacilli when they
are associated with large numbers of other bacteria.
The greater the amount of contamination which is
thrown into the water, and the shorter the time which
elapses between the infection and the drinking of the
water, the greater is the danger.
The typhoid bacillus and the colon bacillus of
Escherich resemble each other in many respects. The
characteristics of each, which allow us to differentiate
the one from the other, will be considered at the end of
the description of the colon bacillus.
CHAPTER XXIV.
BACILLUS COLI COMMTJNIS (OR COLON BACILLUS OF
ESCHERICIl).
THIS organism was first described by Emmerich
(1885), who obtained it from the blood, various organs,
and intestinal discharges of cholera patients at Naples.
It was afterward found by Escherich (1886) in the
feces of healthy milk-fed infants and by Weisser in
the alvine discharges of healthy men. It has since
been demonstrated to be a normal inhabitant of the
intestines of man and of many of the lower animals.
Morphology. The size and shape of the bacillus coli
varies considerably in its morphology according to the
sources and the culture media from which it is obtained.
The typical form is that of short rods with rounded
ends, from 0.4/z to 0.7/* in diameter by I/JL to 3// in
length; but sometimes the rods are so short as to be
almost spherical, resembling micrococci in appearance,
and, again, they are somewhat oval in form or are seen
as threads of 6^ or more in length. The various
forms may often be associated in the same culture (Fig.
58). The bacilli may occur as single cells or as
pairs joined end-to-end, rarely as short chains. In
unfavorable culture media in stained preparations they
may present unstained spaces (vacuoles) and more in-
tensely stained portions at the extremities, closely
resembling spores, but these are due, according to
BACILLUS COLI COMMUNIS. 445
Escherich, to degenerative changes in the protoplasm.
The colon bacillus does not form spores. There is
nothing in the morphology of this bacillus which is
characteristic or which may aid in its identification,
for in this respect it simulates many other organisms.
FIG. 58.
Colon bacilli. Twenty-four-hour agar culture. X 1100 diameters.
The colon bacillus stains readily with the ordinary
aniline colors; it is quickly decolorized by Gram's
method.
Biology. It is an aerobic, facultative anaerobic, non-
liquefying bacillus. It is motile, but its movements
are so sluggish that a positive opinion is often difficult,
being exhibited often by one or two individuals, in fresh
cultures, and at a high temperature only. These move-
ments are produced by flagella, which may be demon-
strated by Loffler's method of staining, though not
usually in such numbers as are seen to occur on the
bacillus typhosus.
446 BACTERIOLOGY.
Growth on Gelatin. In gelatin plates colonies are
developed in from twenty-four to forty-eight hours,
which vary considerably in their appearance according
to their age, and in different cultures in the same
medium. They resemble greatly the colonies of the
typhoid bacillus, except that they are somewhat larger
for the same period of growth. When located in the
depths of the gelatin and examined by a low-power
lens they are at first seen to be finely granular, almost
homogeneous, in structure round, and of a pale yellow-
ish to brownish color; later they become larger, denser,
darker, and more coarsely granular. In shape they
may be round, oval, or " whetstone-like. " The super-
ficial colonies appear as small, dry, irregular, flat,
blue- white points that are commonly somewhat dentated
at the margin.
In stab cultures on gelatin the growth usually takes
the form of a nail with a flattened head, the surface
extension generally reaching out rapidly to the sides
of the tube.
On Nutrient Agar and Blood-serum. On nutrient
agar and blood-serum an abundant, soft, white layer is
quickly developed in the incubator, but the growth
is not characteristic.
In Bouillon. In bouillon the bacillus coli produces
diffuse clouding with sedimentation; in some cultures
a tendency to pellicle formation on the surface is seen
occasionally. In old cultures, in the absence of sugar,
the reaction becomes alkaline, and a decided fecal odor
may be noticed.
The colon bacillus produces indol in bouillon and in
peptone solutions, this reaction being most pronounced
after a week's development in the incubator. It pos-
BACILLUS COLI COMMUNIS. 447
sesses also a considerable reducing power, converting
nitrates into nitrites, as may be demonstrated by the
addition of sulphuric acid in the proper proportion to
a bouillon or peptone culture, when a pink coloration
results.
On Potato. On potato the growth is rapid and
abundant, appearing after twenty-four to thirty-six
hours in the incubator as a yellowish-brown to dark
cream-colored deposit covering the greater part of the
surface. But there are considerable variations from
the typical growth on potato; there may be no growth
at all, or it may be scanty and of a white color. These
variations are due at times to the bacillus, but more
often to variations in the potato.
Gas-production. The bacillus coli grows rapidly
in media containing glycerin and sugar, particularly
glucose, causing active fermentation with liberation of
carbonic acid and hydrogen gas. Cultivated in solid
media, to which glucose has been added, the gas-pro-
duction is recognized by the appearance of numerous
bubbles along and about the points of growth ; in fluid
media it may be demonstrated in the fermentation- tube.
Grown on lactose-litmus-agar, the colonies are pink
and the color of the surrounding medium is changed
from blue to red, showing the production of acid.
Milk is coagulated by the growth of the bacillus
coli after twenty-four to seventy-three hours in the in-
cubator, with the production of gas and acid; very
rarely acid may be produced and no coagulation occur.
The coagulation of the milk is hastened by warming.
The thermal death-point of the colon bacillus from
feces was found by Weisser to be 60 C., the time of
exposure being ten minutes. When the bacilli from
448 BACTERIOLOGY.
a bouillon culture were dried upon thin glass covers
they failed to grow after twenty-four hours (Weisser).
Waliczek found that when dried upon pieces of sterile
filter-paper they failed to grow at the end of eighteen
hours. These results give confirmation to the view
that the colon bacillus does not form spores.
Pathogenesis. The colon bacillus is pathogenic in
varying degrees for test animals, though the results of
the inoculations, as with the typhoid bacillus, cannot
always be predicted with certainty. Intraperitoneal
injections of from 0.1 to 1 c.c. of fresh, virulent cul-
tures usually produce death in mice at the end of from
one to eight days, but death does not invariably follow.
The more rapidly death ensues the greater the number
of bacilli found in the body; they are always more
abundant in the abdominal cavity than in the blood;
in other words, the result is to be attributed to the
toxic rather than to the infective properties of the
culture used. But the fact that the bacilli are found
in the blood and internal organs when death rapidly
follows inoculation proves that they do multiply to
some extent in the body. When less virulent cultures,
however, are injected and death results, this is due to
the poisonous products formed by the bacilli and given
up at their death. The lesions produced are those of
enteritis : the duodenum and jejunum are found to con-
tain fluid, the spleen is somewhat enlarged, and there
is marked hypersemia and ecchymosis of the small in-
testines, together with swelling of Peyer's patches.
Intraperitoneal and intravenous inoculation of guinea-
pigs and rabbits may also produce death, which, when
it follows, usually takes place within the first forty-eight
hours, accompanied by a decided fall of temperature,
BACILLUS COLI COMMUNIS. 449
the symptoms of enteritis, diarrhoea, etc., and finally
fibro-purulent peritonitis.
When subcutaneous inoculations of mice and guinea-
pigs are made it requires the introduction of much
larger quantities of the culture to produce infection;
in rabbits this is followed only by abscess formation at
the point of inoculation. Dogs and cats are similarly
affected.
Bazy and Guyon have succeeded in producing infec-
tion of the bladder in animals by injection of pure
cultures into the blood with simultaneous tying of the
ureters; Albaran and Halle have caused cystitis and
pyelonephritis by direct injections into the bladder and
ureters, the urine being artificially suppressed; Chassin
and Roger produced angiocholitis and abscess of the
liver in the same way. Loruelle, Fraenkel, and Bar-
hacci, by injuring or tying the intestines and intro-
ducing dirt into the abdominal cavity, with or without
the simultaneous injection of cultures of the colon
bacillus, succeeded in causing diffuse peritonitis in
animals. Akermann produced osteomyelitis in young
rabbits by intravenous injections of cultures. So far
all attempts to produce experimental infection of the
intestines by the ingestion of cultures of the colon
bacillus have failed to give positive results (Emmerich
and Korkunoff).
Certain peculiar effects have been observed by Black-
stein and by Gilbert and Lion as the result of intra-
venous inoculation of rabbits with pure cultures of the
bacillus coli, which are worthy of note. The former
of these authors found, from eight to thirty-eight days
after injection, that the liver frequently contained
opaque, whitish or yellowish-white spots, and streaks
29
450 BACTERIOLOGY.
of irregular size and shape, giving a peculiar mottling
to the organ when present in large numbers. By micro-
scopical examination these were found to represent
places where the liver cells had undergone necrosis,
accompanied by emigration of leucocytes, and the cells
about them were in a condition of fatty degeneration.
In sections of the liver, masses of the bacilli were dis-
covered in and about the necrotic foci. The bacilli
were not found generally distributed through the body,
but only in the bile, liver, and occasionally in the
spleen. Gilbert and Lion found in addition that hemi-
plegia and paraplegia were often produced in conse-
quence of atrophy of the cells of the cord. These
observations have been confirmed by Thoinot and
Massilin, but in their experiments the nerve-lesions
were not commonly present.
From experiments on animals it would, therefore,
appear that the true explanation of the palhogensis of
the colon bacillus is undoubtedly to be found in the
toxic effects of the chemical products of the organism
rather than in its mechanical presence in the tissues.
Variation in Virulence. The virulence of the colon
bacillus varies considerably as derived from different
sources. An attempt has been made to establish
certain rules for this. Thus, Lesage and Macaigne
express the opinion that when obtained from a healthy
body it is only slightly virulent, while that isolated
from a diseased person is much more virulent. The
infective power is thought to bear a definite relation
to the severity of the disease with which the organism
is associated; for instance, to be greatest in cultures
taken from cholera patients and least in those obtained
from pus. Dreyfus also confirms this view. He found
BACILLUS CO LI COMMUNIS. 451
by experiment that 1 c.c. of a fresh bouillon culture
of the B. coli from normal feces was required to kill
guinea-pigs by intraperitoneal and rabbits by intraven-
ous injection, whereas less than one-fifth as much of a
culture from a fatal case of cholera nostras was suf-
ficient to kill the same animal; but this rule has prob-
ably many exceptions, even if it be true in some cases.
All observers, however, agree that the virulence
of the B. coli is diminished by continued cultivation
through successive generations, and that it is increased
by passage through animals.
Immunization. Immunization against colon infec-
tion is comparatively easy to produce in the usual way
by the inoculation of gradually increasing doses of
cultures of the living bacilli or dead bacilli.
Occurrence in Man and Animals. The bacillus coli
communis is a common inhabitant of the intestinal
canal in man and in many animals. According to
Fremlin, it is found normally in dogs, mice, and rab-
bits, but not in rats, pigeons, or guinea-pigs. Accord-
ing to Dyas and Keith, it occurs in goats, rabbits, dogs,
cats, swine, and cows, but not in horses. Grimbert
claims to have found it in the intestines of almost all
domestic animals, and in the mouth as well as the in-
testines of man. It is also frequently found in water
and food (milk, etc.), so that it is one of the most wide-
spread saprophytic bacteria known. Formerly it was
thought that the presence of the B. coli in water was
sufficient proof of its contamination by feces; but the
recent investigations of Weichselbaum, Kruse, Beck-
mann, and Refith would seem to show that there are
no grounds for this assumption, as the colon bacillus
may reach the water from many different sources.
452 BACTERIOLOGY.
From its common seat in the intestines it may,
under favoring conditions, penetrate other organs after
death which fact may account for its being found
so often at autopsy in the interior of the body; but it
may also be absorbed during life, more especially if
there is obstruction of the intestines or if the mucosa
has been deprived of its epithelium. For this reason,
no doubt, the B. coli is so frequently found in cholera,
typhoid fever, and dysentery, producing often a second
infection. The absorption of the colon bacillus from
the intestinal canal plays an important part, probably,
in the production of many diseases, such as cystitis
and other inflammatory affections. It has been con-
sidered to be the cause of epidemic infectious enteritis
and cholera nostras, this assumption being based upon
the facts that the colon bacillus in these diseases is
found in greater abundance than usual in the alvine
discharges and often in pure culture ; that it then pos-
sesses an increased virulence, and that it often pene-
trates the interior organs, as has been shown by autop-
sies. But the conclusion drawn from these facts as to
the etiology of the diseases above mentioned is not
positive, though it cannot be denied that under certain
conditions the colon bacillus may be productive of dis-
ease. This is brought about, according to the com-
monly accepted view, either by an increase of virulence
of the B. coli normally present in the intestines or by
the introduction of especially virulent bacilli in the
food. The colon bacillus has also been assumed to be
the cause of cholera infautum; but the investigations
of Booker, Baginsky, Escherich, and Fliigge would
seem, to indicate that this disease is of a much more
complicated origin. The B. coli, moreover, is associ-
BACILLUS COLI COMMUNIS. 453
ated with dysentery, probably as a secondary affection,
as in amoebic or tropical dysentery. It is also found
frequently in cases of diffuse and circumscribed peri-
tonitis, appendicitis, etc., either alone or together with
other bacteria which play a part in the etiology of these
diseases along with certain chemical ferments and toxins
and foreign bodies in the intestines. The origin of in-
fections of the gall-ducts (with at times the production
of gallstones) and multiple abscess of the liver is also
explained in this way by Dmochowski and Jauowski,
though, according to Letienne, the mere presence of
the B. coli in the bile, in which it has been found
under normal conditions, is not sufficient to account
for these affections. Puerperal fever is not infre-
quently caused by the colon bacillus by infection of
the vagina or uterus. Other diseases to which the
colon bacillus seems to stand in a certain relation,
though rarely, are : Endocarditis, meningitis, tropical
abscess of the liver, bronchopneumoriia and an irregu-
lar type of lobar pneumonia, fetid bronchitis, chronic
amygdalitis, and abscess of the lachrymal sac. The B.
coli has been found in a case of urethritis (pseudo-