+ *•'%

I *

I

THE

COMMON BACTERIAL INFECTIONS
OF THE DIGESTIVE TEACT •

AND THE INTOXICATIONS
ARISING FROM THEM

BY

C. A. HERTER, M.D.

PROFESSOR OF PHARMACOLOGY AND THERAPEUTICS IN

COLUMBIA UNIVERSITY, CONSULTING PHYSICIAN

TO THE CITY HOSPITAL, NEW YORK

gorfe
THE MACMILLAN COMPANY

LONDON: MACMILLAN & CO., LTD.
1907

All rights reserved

COPYRIGHT, 190T,
BT THE MACMILLAN COMPANY.

Set up and electrotyped. Published April, 1907.

J. S. Cashing & Co. — Berwick & Smith Co.
Norwood, Mass., U.S.A.

PREFACE

THIS little volume embodies views recently presented
at the New York Academy of Medicine, in a lecture
before the Harvey Society for the Diffusion of Medical
Knowledge. Although the data on which this lecture
was based have been summarized in another publication,
I wish to present them more fully in the present volume
in order that some important details, not suited for pub-
lication in a medical journal, may be brought to the
notice of practitioners and investigators. The book
does not aim at a systematic discussion of the extended
and somewhat confused field of gastro-enteric infection,
either from a clinical or a bacteriological standpoint.
Neither does it make any claim to present fully the lit-
erature of this subject.

I have laid considerable stress on methods developed
in my laboratory with a view to obtaining a better insight
into the bacterial conditions of the digestive tract than
has been hitherto possible. This I have done in the belief
that when these methods are utilized in practice they will
prove of real service in gaining a truer conception of the
nature of the bacterial processes that are operative in
disease. I am confident that the painstaking application
of these methods will furnish practitioners with new and

535528

vi PREFACE

reliable indications as to the progress of many cases of
infection of the digestive tract.

Without the cooperation of my associates I could not
have obtained the data given in this volume. I am under
especial obligations to Dr. Helen Baldwin, Miss M. L.
Foster, Dr. A. J. Wakeman, Mr. H. C. Ward, and Dr.
William R. Williams for the efficient help which they
have in various directions given me.

DECEMBER 1, 1906.

CONTENTS

PAGE

GENERAL CONSIDERATIONS RELATIVE TO THE BACTERIAL
FLORA OF THE HUMAN DIGESTIVE TRACT IN

HEALTH 1

Defensive Action of the Digestive Juices ... 6
On the Antagonism between B. coli communis and Other

Microorganisms 7

Criticism of Experiments of Conradi and Kurpjuweit 17
Influence of Reaction on the Growth and Products of

Intestinal Anaerobes ...... 21

Aerobic and Anaerobic Conditions in the Digestive Tract 23
The Bacteria of the Human Digestive Tract at Different

Ages in Apparently Healthy Individuals . • . 35

THE BACTERIAL PROCESSES IN THE DIGESTIVE TRACT OF

NORMAL NURSLINGS AND BOTTLE-FED INFANTS 37
Nursling Infants ........ 37

Distribution of the Bacterial Flora in the Digestive

Tract of the Nursling 48

The Infection of the Nursling's Digestive Tract and
the Relation of the Microorganisms to the Perma-
nent Bacterial Flora 53

Bacterial Flora of Bottle-fed Children . . . .59
Products of Decomposition in the Intestinal Tract of

Bottle-fed Children 64

Action of the Mixed Faecal Flora upon Various Media 66

THE BACTERIAL CONDITIONS IN THE DIGESTIVE TRACT
DURING CHILDHOOD, ADOLESCENCE, ADULT LIFE,

AND SENESCENCE 69

Period of Childhood and Adolescence . . . .69

Period of Adult Life 72

Period of Senescence . .75

viii CONTENTS

PAGB

CHARACTERS OF THE BACTERIAL FLORA OF CARNIVOROUS

AND OF HERBIVOROUS ANIMALS .... 80

Influence of Food on Human Bacterial Flora of the Digest-
ive Tract 86

The Reducing Action of Meat 89

The Influence of the Epithelial Cells Lining the Digestive

Tract 91

The Permeability of the Mucous Membrane of the Intes-
tinal Tract for Bacteria 93

Phylogenetic Significance of the Large Intestine . . 97
The Importance of Prompt Resorption from the Small

Intestine 99

The Phenomenon of Substitution 100

The Presence of Pathogenic Bacteria in the Digestive

Tract in Health 102

CRITERIA EMPLOYED IN THE CLASSIFICATION OF BAC-
TERIA OF THE GASTRO-ENTERIC TRACT . . 105

Methods of Investigation Ill

Character of the Microscopical Fields .... Ill
The Dimethylamidobenzaldehyde Reaction of the

Faeces 145

The Dimethylamidobenzaldehyde Reaction of the

Urine 147

COMMON BACTERIAL INFECTIONS OF THE DIGESTIVE TRACT,
CONSIDERED FROM THE STANDPOINT OF THE

MICROORGANISMS 150

The Colon-typhoid-dysentery Group 150

Colon Bacilli 150

Typhoid Bacilli 157

Paratyphoid and Allied Infections .... 167

Dysentery Bacilli . 172

Liquefying Bacteria 181

Streptococcal and Staphylococcal Infections . . . 185

Bacillus Bifidus 190

Infections through Anaerobic Bacteria . . . .191
Bacillus Putrificus . . . . . . .192

CONTENTS ix

PAGE

Bacillus A erogenes Capsulatus 196

Bacillus Botulinus 210

THE FERMENTATIVE AND PUTREFACTIVE PROCESSES PROM

THE STANDPOINT OF THEIR PRODUCTS . . 214

Oxalic Acid and Oxaluria 216

Acetone . 218

Basic Substances 221

Putrescin and Cadaverin 224

Sulphur Compounds 226

Mercaptan 226

Hydrogen Sulphide 227

Hydrogen Sulphide and its Relation to Enterogenic

Cyanosis 234

Aromatic Products of Putrefactive Decomposition . . 237

Phenol and Cresol 237

Skatol 239

Indol 241

Indicanuria ......... 257

The Possibility of the Occurrence of Indolsemia and

Indoluria 269

Indigouria 272

Individual Susceptibilities to Different Enterogenous
Poisons as Possible Factors in Determining Clini-
cal Types 274

TYPES OF CHRONIC EXCESSIVE INTESTINAL PUTREFACTION 278

I. The Indolic Type of Chronic Excessive Intestinal

Putrefaction 280

II. The Saccharo-butyric Type of Chronic Excessive

Intestinal Putrefaction 291

III. The Combined Indolic and Saccharo-butyric Type

of Chronic Excessive Intestinal Putrefaction . 306

METHODS RELATING TO THE MODIFICATION AND CONTROL
OF BACTERIAL PROCESSES CONCERNED IN CHRONIC

EXCESSIVE INTESTINAL PUTREFACTION . . 314
The Avoidance of Putrefactive Contamination of the

Food . . .317

x CONTENTS

PAGE

The Promotion of Prompt Digestion and Absorption

in the Small Intestine 322

Methods designed to reduce the Numbers of Putre-
factive Anaerobes 329

SOCIOLOGICAL CONSIDERATIONS 347

INDEX ,353

THE COMMON BACTERIAL INFECTIONS
OF THE DIGESTIVE TRACT

THE COMMON BACTEEIAL INFECTIONS
OF THE DIGESTIVE TEACT

GENERAL CONSIDERATIONS RELATIVE TO THE
BACTERIAL FLORA OF THE HUMAN DIGEST-
IVE TRACT IN HEALTH

IF one examines with the microscope the contents of
any portion of the large intestine of a human being or
of any mammal, the richness of the material in micro-
organisms is strikingly apparent, especially in stained
preparations. Their number has been estimated at
one hundred and twenty-six billions for the daily human
excreta. It is true that if the material is selected from
the lowest portion of the gut, many of the microorgan-
isms of a cultivable nature can be shown by suitable
cultural methods to be no longer living, but rather to
be undergoing a process of disintegration, partly owing
to a solution in their own juices — a process of autoly-
sis. But even the dead and dying bacterial inhabitants
of the lower intestinal tract point to the multiplicity
of bacterial life at higher levels.1 And not only are these

1 Strasburger, J., " Untersuchungen tiber die Bakterienmenge
in menschlichen Faces/' Zeitschr. f. klin. Med., xlvi, p. 413, 1902.

It has been estimated that the proportion of dead bacteria in
normal human fseces is often as high as ninety-nine per cent. It

B 1

£" INFECTIONS OF THE DIGESTIVE TRACT

bacterial inhabitants numerous but they represent
many species and varieties.

The knowledge that the digestive tract is so rich in
bacterial forms of life has led many physiologists to
inquire into the biological meaning of this remarkable
fact. Pasteur expressed a belief that these bacterial
inhabitants are essential in some way to the life of the
individual which harbors them. Nut tall and Thier-
f elder,1 in their well-known experiments, attempted to
rear guinea-pigs delivered by Csesarian section and fed
on quite sterile food. As the animals lived and increased
in weight, the experimenters concluded that the intestinal
bacteria are not essential to normal nutrition. This
view, as will be presently seen, gets support from the
observations of Levin2, that some animals of the Arctic
region, as polar bears, have no bacteria in the digestive

appears, however, that this is an under-estimate of the living bac-
teria. It may possibly hold true of the aerobic bacteria which
will grow on ordinary media. But we know that there are often
many aerobes and anaerobes in the intestine which do not grow on
ordinary media. In some instances there are very many strict
anaerobes (e.g. B. aerogenes capsulatus) which appear only on
specially prepared media on anaerobic plates. These, of course,
do not come into any count made in ordinary ways. Even the
numbers of living colon bacilli are subject to great variations. In
the same individual there may be at one tune a considerable pro-
portion of cultivable (living) bacilli. In a state of constipation the
number of cultivable colon bacilli may become very small. This
death of bacteria in the lower bowel during constipation doubtless
depends mainly on a failure of food supply and absence of moisture.

1 " Thierisches Leben ohne Bakterien im Verdauungskanal,"
Zeitschr. f. physiol Chem., xxi, p. 109, 1895; xxii, p. 62, 1896;
xxiii, p. 231, 1897.

2 " Bakteriologische Darmuntersuchungen," Skandinavisches
Archiv /. Physiol., xvi, p. 249, 1904.

INFECTIONS OF THE DIGESTIVE TRACT 3

tract. Even in temperate regions there are animals
whose alimentary tracts are comparatively free from
bacterial life. This is said to be the case with the parrot.
Other observers have, however, reached a different con-
clusion from that of Nuttall and Thierfelder. Schot-
telius * found that chickens fed on sterile food were re-
tarded in development and showed normal growth only
when given food containing bacteria. Similar results
were obtained by Madam Metchnikoff 2 in experiments
on tadpoles. Very carefully conducted experiments by
Moro 3 on the larvse of the turtle (Pelobates fureus,
Wagler) lead to the same conclusion; namely, that
intestinal bacteria are necessary to normal nutrition.

It must be admitted, I think, that none of these ex-
perimental studies are really conclusive as to the neces-
sity of bacterial action in the digestive tract for the
maintenance of health in adult mammals of the highest
type — man and various domestic animals. Experi-
ments on tadpoles and chickens cannot with confidence
be applied to the case of man. The experiments on
guinea-pigs can more justly perhaps be taken as typical
for mammals, but as the experiments of Nuttall and
Thierfelder were extended over only a short period of
time, they can hardly be held to prove that bacteria are

1 "Bedeutung der Darmbakterien fur die Ernahrung," Archiv
/. Hyg., xlii, p. 48, 1902.

2 "Note sur I'influence des microbes dans le development des
tetards," Ann. de I'Inst. Past., xv, p. 631, 1901.

3 " Morphologische und biologische Untersuchungen iiber die
Darmbakterien des Sauglings," IV. "Der Schottelius Versuch
am Kaltbltiter," Jahrb. /. Kinderheilk., xii, p. 467, 1905.

4 INFECTIONS OF THE DIGESTIVE TRACT

either essential or non-essential to the maintenance of
prolonged health during the period of adult existence.
The evidence given by the sterile intestinal contents
of certain Arctic animals is apparently conclusive for
the conditions in which this experiment of nature has
been carried out. Here we have animals born and living
in surroundings where bacteria are very few in number,
and it is probably on account of the great rarity of micro-
organisms in the air and the small number in the water l
that the intestinal contents contain so few bacteria.
These animals are able to live indefinitely in a state of
robust health. Levin examined the intestinal contents
of Arctic animals in Spitzenberg. The digestive tract
was found to be in most instances entirely sterile in
white bears, seals, reindeer, eider ducks, penguins, etc.,
although very small numbers of organisms resembling
the colon bacillus were found in one white bear and in
two seals which were examined.

Clearly then, in this case, the intestinal bacteria are
not required to carry on the ordinary digestive processes
and normal nutrition. It has been supposed that the
intestinal bacteria aid in the digestion of cellulose, which
they are undoubtedly able to decompose fermentatively.
The argument in favor of the importance of this function
of the intestinal bacteria loses much of its force if it be
true, as lately maintained by Bergman,2 that most of the

1 It was estimated that there was one organism in 11 c.c. of
water, whereas in the river Seine it was estimated that there are
about 2,000,000 in the same volume of water.

2"Studien tiber die Digestion der Pflanzenfresser," Skandir
navisches Archiv /. Physiol., xviii, p. 119, 1906.

INFECTIONS OF THE DIGESTIVE TRACT 5

cellulose eaten by herbivora (in which the digestive
function of the bacteria chiefly comes into question) is
provided with intracellular enzymes capable of decom-
posing cellulose.

The real significance of the normal intestinal flora
probably lies not in any immediate relation to processes
of digestion, but in a wholly different direction.

It is impossible to avoid the entrance of bacteria into
the digestive tract. As will be seen when we come to
consider the normal flora of the alimentary tract, the
obligate bacteria (e.g. B. lactis aerogenes, B. coli, B.
bifidus) have adapted themselves to the secretions of
this part of the body and ordinarily hold their own
against new-comers. By virtue of their adaptation
they are not ordinarily harmful to their host, but on
the contrary, they are under some circumstances cap-
able of doing a service by giving rise to conditions that
discourage the growth of many harmless and harmful
species which the human animal cannot readily exclude
from his digestive tract. I believe the chief significance
of the obligate intestinal bacteria lies in their potential
capacity for thus checking the development of other
types of organisms capable of doing injury.

Under ordinary conditions of life, in temperate cli-
mates, and still more so in hot ones, a human being is
liable to take into the alimentary tract, with food and
drink, microorganisms which are capable of doing injury
if they find opportunity to multiply in the digestive
tract. Water, milk, cheese, oysters, game, preserved
and fresh meats, etc., are liable to contain injurious

6 INFECTIONS OF THE DIGESTIVE TRACT

bacteria. Among such bacteria are pyogenic strepto-
cocci and staphylococci, the paratyphoid bacilli, typhoid
bacilli, the dysentery bacilli, proteus vulgaris, and the
spore-bearing anaerobes — B. putrificus (Bienstock),
B. aerogenes capsulatus, B. botulinus, etc.

Defensive Action of the Digestive Juices. — The normal
human organism is provided with more or less efficient
(though by no means fully understood) methods of de-
fense against these bacterial invaders. The secretion
of the gastric juice in normal abundance, after a meal,
provides a degree of acidity which acts as an effective
check upon the growth'of many non-sporula ting bacteria,
and is actually destructive to most varieties at least in
a measure. Probably the proteolytic action of the pep-
tic ferment and the tryptic enzymes leads to a very
quick destruction of any bacteria whose vitality has
been lowered by contact with the acid of the gastric
juice. If, however, bacteria are administered in very
large numbers, there is a chance that a certain proportion
of them will run the gauntlet of these defenses and find
their way into the lower part of the small intestine and
into the colon. This seems especially liable to happen
in those cases where the microbes are taken into the
empty, non-secreting stomach, or into a stomach with
defective motility which secretes little gastric juice with
a low content of hydrochloric acid — and there are many
such stomachs among persons over forty years of age
and in fair health. Thus the bacteria (with any spores
that may have developed from them or have been in-
gested as such) find their way to the region of the colon

INFECTIONS OF THE DIGESTIVE TRACT 7

and here are confronted with immense numbers of the
chief obligate race of bacteria of the digestive tract —
the representatives of the B. coli type. Another group
of obligate organisms closely allied to B. coli, B. lactis
aerogenes, is present in the upper part of the small
intestine and becomes gradually less abundant with its
descent into the colon and finally appears in relatively
small numbers in the fa3ces, if at all. As the bacteria
of the B.lactis aerogenes type grow less numerous, the
representatives of the B. coli group grow more abundant
and beyond the ileocaecal valve largely dominate the
intestinal flora.

Exactly what happens when the accidental, sapro-
phytic forms of bacterial life — the "wild races/' as
the French call them — come to close quarters with
the " obligate/' well-adapted parasitic forms of the
intestine, we do not at present know. There are, how-
ever, numerous facts which point to well-defined bio-
logical antagonisms between the "wild" forms and the
representatives of the B. coli group.

ON THE ANTAGONISM BETWEEN B. Coli COmmunis AND
OTHER MICROORGANISMS

The members of the B. coli group are organisms of
varying morphology, characterized by a certain hardiness
in growth on ordinary media, by the free production of
gas and acid on various sugars, by the coagulation of
milk, and usually by the formation of indol, and by a
sluggish motility in some fully grown forms and active

8 INFECTIONS OF THE DIGESTIVE TRACT

mo till ty in members of young surface colonies.1 They
do not retain the Gram stain. The members of the
B. lactis aerogenes group are distinguished from those
of the B. coli group by unimportant morphological
differences. More significant differences exist in respect
to biochemical characters. Among these differences
are a somewhat greater ability to form gas on sugar
media, a more rapid coagulative action on milk (often
with capsule formation), the ability to make gas from
potato starch, the more frequent failure to make indol,
and the greater luxuriance of growth on gelatin. In
general, then, the fermentative activities of B. lactis
aerogenes are somewhat greater than those of the B.
coli group, while the putrefactive powers are distinctly
less. Harden 2 has lately shown that there is a constant
difference in the behavior of B. lactis aerogenes and B.
coli when grown anaerobically on sugar bouillon, in
regard to the ratio of alcohol and acetic acid produced.
This difference appears to strengthen the right of B.
lactis aerogenes to be regarded as a distinct organism from
B. colij although the relationship between the two is
close. The behavior of these two groups toward other
organisms is probably very similar, but as the antago-

1 For a detailed discussion of B. coli communis see Escherich
(Th. Escherich u. M. Pfaundler, "Bacterium coli commune,"
Kolle & Wassermann's " Handbuch der pathogenen Mikroorganis-
men," p. 334, 1902) ; also Theobald Smith ("Note on Bacillus Coli
Communis and Related Forms"), Amer. Journ. of the Med. Sci.,
September, 1895, who gives important data as to fermentative
characters.

2 "The Chemical Action on Glucose of the Lactose-fermenting
Organisms of Faeces," Journ. of Hyg., v, p. 488, 1905.

INFECTIONS OF THE DIGESTIVE TRACT 9

nistic action of the B. coli group has been more carefully
observed I shall speak of this only, especially as the
physiological behavior of its members is the more im-
portant on account of their wider distribution in the
digestive tract.

The most far-reaching contention relating to the defen-
sive action of the B. coli group is based on the recent
observation of Conradi and Kurpjuweit l that the
members of this class make thermostabile and ther-
molabile substances which have a powerful antibacterial
action, being still active hi a dilution of 1 to 10,000
parts, and hence comparable to the antibacterial
action of carbolic acid. The inhibitory action of these
substances is stated to be not confined to alien bacteria,
but relates also to the B. coli group. The inhibition
in growth observed in old cultures was attributed to this
substance. Moreover, it was claimed that it is owing
to such bactericidal substances that the members of the
B. coli group tend to die out as they pass toward the
lower end of the bowel. It is easy to satisfy oneself
that the faeces of a healthy person contain more living
colon bacilli if the intestinal contents have somewhat
rapidly passed through the colon than if the usual so-
journ of material in the lower bowel has occurred. It
is also noteworthy that relatively few living representa-
tives of the B. coli class are present in the movements of
healthy persons with obstinate constipation. The cause

iaUeber die Bedeutung der bakteriellen Hemmungsstoffe fur
die Physiologic und Pathologic des Darms," Munch, med. Wochen-
schr., Hi, pp. 2164, 2228, 1905.

10 INFECTIONS OF THE DIGESTIVE TRACT

of this high mortality among the colon bacilli is ascribed
by Conradi and Kurpjuweit to increased opportunities
to make bactericidal substances.

In their experiments these writers found that the
growth of the typhoid bacillus and of the paratyphoid
bacillus is definitely inhibited by the activities of the
colon bacillus, and this restraint is referred by them to
the specific inhibitory products of the latter microbe.
It will be shown in subsequent pages that the claims of
Conradi and Kurpjuweit have not been successfully
sustained. The observations related by them are of
great interest, but the interpretation placed on them
appears to be erroneous.

The inhibitory action of B. coli upon the putrefactive
anaerobes is shared by a closely related group of bacteria
first described as B. bifidus by Tissier.1 This group of
bacteria was originally included by Escherich with the
colon bacilli, but for reasons which will be mentioned
later, has now been separated from them. Tissier
found that B. bifidus, under certain conditions, inhibits

1 "La Flore intestinale normale et pathologique du nourrisson,"
These de Paris, 1900.

There are certain bacteria that have the ability to check the
growth of B. coli. Streptococci derived from human faeces I have
repeatedly observed to repress the growth of B. coli from the same
individual, in the anaerobic limb of the fermentation tube. B.
lactis aerogenes may be repressed in the same way. Heinemann
(" The Significance of Streptococci in Milk," Journ. Infect. Dis.,iii,
p. 173, 1906) has found that streptococci from milk interfere with
the development of B. lactis aerogenes. Gabricewski and Mal-
jatai (" Ueber die bakterienfeindlichen Eigenschaften des Cholera-
bacillus," Centralbl. f. Bakt., xiii, p. 780, 1893) observed that B.
coli is inhibited by the growth of cholera vibrios.

INFECTIONS OF THE DIGESTIVE TRACT 11

the growth of the anaerobic microbe which he describes
as B. perfringens and which appears to be no other than
the organism known in the United States as B. aerogenes
capsulatus (Welch) and in Germany as the gas-phlegmon
bacillus (Fraenkel).

The B. bifidus of Tissier is closely related to the
organism described by Moro as B. acidophilus and shares
with the latter the ability to grow in a more strongly
acid medium than can be withstood by meat bacteria.
Recent studies of these bacteria by Bjeloussow1 have
led him to identify these acidophile microorganisms
with certain acidophile bacteria studied by him, and to
which he, as well as Mereschkowsky,2 attaches con-
siderable importance as protective inhabitants of the
digestive tract. The capacity to grow in a medium
containing one-half or even one per cent, of acetic acid
is an indication that the acidophile bacteria might as-
sume a dominant position in the digestive tract if pro-
vided with suitable food materials for the production
of acid. It is stated that if large numbers of these bac-
teria be administered to a dog by mouth, other flora may
to a large extent be temporarily suppressed. But as
in the case of feeding other kinds of microorganisms the
leading part can only be maintained by continuously
feeding large numbers of the acidophiles. Whether

*"Zur Biologie und Methodik der Ausscheidung der soge-
nannten acidophilen Bakterien," Diss. No. 76, St. Petersburg (Rus-
sian) .

2"Zur Frage iiber die Rolle der Mikroorganismen im Darm-
kanal," Centralbl. /. Bakt. Orig., xxxix, pp. 380, 581, 696, 1905,
andxl, p. 118, 1906.

12 INFECTIONS OF THE DIGESTIVE TRACT

such feeding would be practicable in man on a scale suffi-
ciently great to secure definite results in the suppression
of harmful types, such as the putrefactive anaerobes,
can only be determined by experiment. At the present
time it is known that the acidophiles are represented in
all parts of the tract, especially in the large intestine,
but we cannot yet say with confidence what is their
physiological role or what would be their influence on
the organism if their numbers should be greatly in-
creased through feeding.

The various facts that have now been advanced render
it certain that microorganisms of the B. coli type are
able under some conditions to check the growth of patho-
genic microorganisms which are often found among the
intestinal flora. The conditions that obtain in the colon
differ, however, with respect especially to the nutrient
pabulum and the digestive secretions, from any attain-
able experimental conditions, however cleverly these may
be designed to imitate what occurs in nature. We have
therefore to exercise some caution in transferring these
results, without due consideration, to the human digestive
tract. Nevertheless the evidence now available suggests
that in health the colon bacillus, both in man and the
higher mammals, exerts an important function in com-
bating the development of the injurious saprophytes
with which even in ordinary health the human intesti-
nal tract almost necessarily abounds. A long, largely
anaerobic intestinal tract, permitting gradual resorption
of the contents, is a physiological necessity in order that
a loss of water and its detrimental consequences may

INFECTIONS OF THE DIGESTIVE TRACT 13

be spared the organism. The presence in the colon of
immense numbers of obligate microorganisms of the
B. coli type may be an important defense of the organism
in the sense that they hinder the development of that
putrefactive decomposition which, if prolonged, is so
injurious to the organism as a whole.

We have in this adaptation the most rational explana-
tion of the meaning of the myriads of colon bacilli that
inhabit the large intestine. These bacilli are essential
to the life of the individual mammal as a defense against
bacterial foes which it is impracticable to wholly exclude
from the digestive tract, and not as agents in directly
facilitating the processes of digestion in the narrow
sense. This view is not inconsistent with the conception
that under some conditions the colon bacilli multiply
to such an extent as to prove harmful, through the part
they take in promoting fermentation and putrefaction.
It seems to me not unlikely that the reaction of the fluids
of the digestive tract may influence the character of the
activities of the colon bacilli, an alkaline reaction favor-
ing their putrefactive functions if peptones be present.

The following facts, first noted by Moro and Murath,
point to the existence of bacterial inhibitory powers
on the part of the fsecal flora of nurslings and bottle-fed
children, and deserve mention. If one allows faecal
matter from a normal nursling to stand in a thermostat
in a closed test-tube, the original not disagreeable odor
lasts for days or even weeks, although the material as a
rule contains small numbers of true putrefactive bacteria.
Indeed, it is possible to inoculate such normal material

14 INFECTIONS OF THE DIGESTIVE TRACT

with Bienstock's B. putrijicus quite richly without any
evidence of putrefaction, although the organisms are
kept under anaerobic conditions. Although this observa-
tion points to the absence of putrefactive processes,
it must be remembered that the growth of anaerobes is
relatively difficult where the quantity of moisture is
limited, as in the case of the foregoing experiments, and
also that the quantity of nutrient material available
for the growth of putrefactive anaerobes is small in the
movements of normal milk-fed children. A similar
result is obtainable after inoculating concentrated
fseces-agar with various bacteria, such as pyocyaneus,
prodigiosus, the bacillus of Friedlander, the vibrio of
Metchnikoff, and various strains of the typhoid bacillus.
That is to say, such f seces-agar richly inoculated with the
foregoing organisms fails to show any growth, whether
the acid reaction of the medium be retained or whether
neutralization has taken place. On the other hand, it
has been noted that B. coli, B. fluorescens, Staphylococcus
intestinaliSj streptococcus, proteus, and the paracolon
organisms, as well as putrificus are apt to grow some-
what better. From this observation one may perhaps
reach the conclusion that the bacteria of the intestinal
tract possess an elective antagonistic action against
foreign types of microorganisms.

An effort has been made to obtain an adequate ex-
planation for this behavior, and especially to determine
whether the inhibition for bacteria here noted depends
on the production of definite bactericidal substances.
The methods employed by Moro and Murath were

INFECTIONS OF THE DIGESTIVE TRACT 15

similar to those of Conradi and Kurpjuweit.1 In the
series of observations made by them the growth of the
following organisms was observed on the specially pre-
pared plates : B. coli communis, B. lactis aerogenes,
B. typhi, B. Shiga-Kruse, B. Flexneri, Staphylococcus
pyogenes aureus, Diplococcus intestinalis, Soor, B.
prodigiosus, and B. pyocyaneus. The results were by
no means uniform, but showed that the normal nursling's
stool exerts an intense inhibitory action against various
bacteria. For example, B. typhi and B. Kruse failed to
grow even in dilutions of 1 : 400. On the other hand, the
representatives of the obligate bacteria of the intestine,

1 One volume of the fresh faeces is diluted with nine volumes
of bouillon. In order to measure the fasces one may use a glass
capsule of 1 c.c. capacity, which is then emptied by means of a
glass rod of suitable size, into the bouillon. Various quantities
of the yellowish emulsion are now permitted to flow into test
tubes containing agar cooled to 42° C., so that the fluids so obtained
represent dilutions of TV, £s, -fa, T£<T» von, and T&T of the original
fsBcal material. After thorough mixing, the agar is plated in
Petri dishes. The surface of these plates is then promptly inocu-
lated with approximately the same numbers of different kinds
of bacteria.

The material to be inoculated is prepared by inoculating one
normal platinum loop of a twenty-four-hour agar culture into
10 c.c. of bouillon, and emulsifying it. From this emulsion of
bacteria a smaller loopful is smeared upon the surface of the agar
plates. It is convenient to divide up the agar into half a dozen
regions by means of a platinum needle. The surface of each one
of these islands may be smeared with a different culture of bac-
teria so that the growth of several different types of organisms
may be observed on one plate. The bacteria inoculated on one
of these islands does not extend to others. The growth is observed
at the end of twenty to twenty-four hours, and controls are made
with material of the same origin, diluted to one- tenth, boiled, and
then incubated for twenty-four hours.

16 INFECTIONS OF THE DIGESTIVE TRACT

B. coli communis and B. lactis aerogenes, usually grew
fairly well and suffered little inhibition. This was a
regular occurrence throughout the experiments. Indeed,
the phenomenon occurred so regularly that it was em-
ployed as the basis for distinguishing between typhoid
bacilli and colon bacilli. In the case of Staphylococcus
pyogenes, which is found in every nursling's stool, there
is observed very little inhibitory action. In the case
of B. pyocyaneus it was found that even in such strong
concentrations as 1 : 10 there was good growth with the
production of pigment.. It appears, therefore, that in
this case the medium favored the growth of pyocyaneus
to a certain extent. The dysentery bacillus of Flexner
in some cases behaved like the typhoid bacillus, but in
other cases showed growth in a dilution of 1 : 100. The
action of this medium was therefore more marked upon
the whole in its action against the typhoid bacillus and
the Kruse bacillus than against the Flexner bacillus.

The Diplococcus intestinalis, a constant inhabitant of
the nursling's tract, behaved like the Staphylococcus
pyogenes in developing only moderately. The growth
of the thrush organism was more strongly inhibited than
B. prodigiosus. The development of color was checked
on all the plates in which the faecal concentration was
fairly strong.

In general, the inhibitory substances to which the
restraining effect is attributed appeared to be less active
in the fseces of the nursling than in those of grown
persons.

A difference between the inhibitory action of the fseces

INFECTIONS OF THE DIGESTIVE TRACT 17

of nurslings and of bottle-fed children could not be
observed. In the case of a child suffering from a chronic
catarrhal condition of the gastro-enteric tract and
marasmus, a great diminution in the inhibitory action
was observed. It seems possible that this might be inter-
preted on the basis of a diminished growth of colon
bacilli rather than by the production of inhibitory
substances.

Experiments made with meconium showed that this
material held no bactericidal substances. On the
contrary, when mixed with agar it became an excellent
nutrient medium for bacteria, although by itself it is
only a poor medium for most varieties.

It is interesting to observe that B. bifidus, which forms
the chief representative of the nursling's flora, appeared
to be unable to make these inhibitory substances. The
question arises in the mind of the critic in connection
with this experiment, whether in the observations re-
ported there was in reality a good growth of B. bifidus,
which is an organism not easily cultivated on ordinary
media.

Moro and Murath attribute the inhibitory action of the
nursling's fseces entirely to the presence of organisms
of the B. coli type. A comparison of typhoid bacillus
with the colon bacillus with respect to the ability to
inhibit the growth of various microorganisms showed
that the typhoid bacillus possesses little or no capacity
to effect this inhibition.

Criticism of Experiments of Conradi and Kurpjuweit. —
The interpretation to be placed on the phenomenon of

18 INFECTIONS OF THE DIGESTIVE TRACT

inhibition described by Conradi and Kurpjuweit and by
Moro and Murath is not yet entirely clear. That the
observed inhibition is due to complex cellular products —
the " autotoxines " of Conradi — is questionable in view
of the studies of Manteufel,1 of Passini,2 and of Oebius.3
Manteufel showed that in many cases where the surface
of a plate indicated no growth on inspection, and where
consequently it would be assumed that an inhibitory
effect had been exerted, in reality many small micro-
scopical colonies were present. In other words, the
character of the growth may be of such a nature as to
mislead one into the belief that inhibition has occurred
where in reality such inhibition is absent or slighter than
has been supposed. In these cases the size of the in-
dividual surface colonies was smaller but their number
was increased and the surface could certainly not be
regarded as sterile. Further observations made it very
probable that the exhaustion of the culture medium
incidental to the growth of the faecal flora is in part the
cause of the failure of the surface colonies to grow.
The assumption that there are produced in the bouillon
culture thermolabile, unfilterable, and dialyzable sub-
stances which check the growth, appears unwarranted in
the light of certain experiments with reference to this

1 "Untersuchungen iiber die 'Autotoxine' (Conradi) und ihre
Bedeutung als Ursache der Wachstumshemmung in Bakterien-
kulturen," Berl. klin. Wochenschr., xliii, p. 313, 1906.

2 "Die bakteriellen Hemmungsstoffe Conradis und ihr Einfluss
auf das Wachstum der Anaerobier des Darms," Wien. klin.
Wochenschr., xix, p. 627, 1906.

3 " Ueber spontane Wachstumshemmung der Bakterien auf
kiinstlichen Nahrboden," Med. Klinik, ii, p. 598, 1906.

INFECTIONS OF THE DIGESTIVE TRACT 19

point. It also appears doubtful whether there are heat-
resisting substances formed by B. coli which are of value
in checking surface growths made under the conditions
already mentioned. It was found that there was no
difference in the surface growth of B. coli if the agar me-
dium containing twenty-four-hour cultures of B. coli was
heated for thirty minutes at 90° C. or 100° C. Of course
in this experiment it was necessary to exclude the factor
of an exhausted medium previous to making the surface
cultures, since in the presence of such an exhausted
medium organisms would naturally not grow upon the
surface. These various experiments speak against
any specific inhibitory substances. It would, indeed,
be a remarkable and apparently contradictory phenom-
enon if organisms should form substances in the course
of their growth which are more injurious to themselves
and their own species than to foreign ones. Yet this
view is apparently assumed by Conradi to be correct.
In order to explain the dominance of the colon bacilli
in the intestine he has had to make the assumption that
these organisms have given rise to substances against
which they have become immunized — a complex
assumption. It is also singular that the autotoxines
of Conradi should act in an inhibitory manner without
causing actual death, if it be true, as claimed, that these
autotoxines are inhibitory in an even higher degree than
carbonic acid. Conradi and Kurpjuweit, in spite of this
fact, attribute the bactericidal action of fresh faeces to
these inhibitory substances. It is difficult to under-
stand how, if these supposed restraining substances are

20 INFECTIONS OF THE DIGESTIVE TRACT

really only inhibitory and not bactericidal, leaving the
bacteria still capable of growth when brought into better
nutrient media, so many of the bacteria in the fseces
have been killed. The proof remains to be furnished
that the death of these bacteria in the lower part of the
intestinal tract and the inhibition at higher levels are
due to the same cause.

It is only fair to say that Conradi and Kurpjuweit
have not succeeded in establishing their contention that
the inhibitory action of colon bacilli is due to specific
toxines which they form. The inhibitory action de-
scribed by them is certainly due in part to the exhaustion
of the media by the freely growing colon bacilli and in
part to the production of acid, indol, phenol, or other pu-
trefactive or fermentative substances. It would be going
too far, however, to deny that substances at present
unknown to us may possibly have a part in bringing
about the phenomena of inhibition.

It is clear that the colon bacilli can act protectively
to the digestive tract only when they are able to grow
freely. If the conditions are such in the tract that they
cannot multiply abundantly, they can neither exhaust
the nutritive materials on which other races grow nor
make products endowed with inhibitory powers. There
are many examples of disease in which the colon bacilli
disappear not merely from the faeces, but from the
diarrhceal stools. Other races take their place. This
condition is quite unintelligible unless we assume that
the colon bacilli no longer find a suitable nutrient medium
in the mixture of digestive juices and food. Without

INFECTIONS OF THE DIGESTIVE TRACT 21

radical changes in the character of the food the colon
bacilli have in many instances been observed to resume
their prominent position in the intestinal contents. It
is reasonable to believe that the return of more normal
secretory conditions in the digestive tract is a very
influential factor in determining the reinstatement of
the legitimate flora. If this be true, the therapeutic
value of introducing colon bacilli into the intestine
(with food or by high colon injection) with a view to
utilizing their restraining action upon injurious races
must be less than would at first sight appear to be the
case. There is need for much careful research in regard
to this question.

Influence of Reaction on the Growth and Products of
Intestinal Anaerobes. — If mixed faecal flora be grown
in sugar bouillon and in sugar bouillon containing
calcium carbonate for one week at 37° C., it will be found
that a considerable amount of volatile fatty acids has
been formed in each case. The quantity will usually
be found to be greater, often considerably greater, in the
flask containing calcium carbonate than in the flask with-
out it. Similar results are obtained if one substitutes
magnesium carbonate for the calcium salt. The observed
difference in the quantity of acids is therefore probably
not due to a specific effect of the calcium salt, but to its
action in maintaining a neutral reaction.

In a series of observations of the sort just mentioned,
which were carried out by Dr. A. J. Wakeman, two things
are worthy of note. First, it was found that in nearly
all instances the molecular weights of the volatile fatty

22 INFECTIONS OF THE DIGESTIVE TRACT

acids were at least somewhat higher in the case of the
cultures in the sugar-bouillon-carbonate media than
in the corresponding growths in sugar bouillon. This
fact suggests a relatively greater activity of putrefactive
bacteria in the neutral media, for it is known that such
bacteria tend to form the higher rather than the lower
volatile fatty acids. This view was confirmed by the
study of the bacteria dominant in the two series of flasks.
Secondly, it was found by Dr. Wakeman that in almost
every instance the proportion of non-volatile acids
(mainly lactic) was greater in the sugar-bouillon cultures
than in the neutralized cultures. The most reasonable
interpretation of these results is that a neutral medium
favors the development of putrefactive bacteria (especi-
ally of the anaerobes), whereas an acid medium (such as
is maintained in the sugar-bouillon flasks) restrains the
growth of these bacteria and favors the growth of the
lactic acid makers. It is probable that in the case of
the neutral flasks some of the lactic acid formed is con-
verted into butyric acid.

These observations and similar ones with pure cultures
indicate that in the digestive tract the growth of putre-
factive anaerobes must be favored by a neutral reaction
and restrained by the presence of acid. The favorable
influence of milk food, containing lactic-acid formers,
in controlling putrefactive decompositions in the digest-
ive tract finds its explanation partly in the inhibitory
action exerted by such bacteria and in part in the
presence of preformed lactic acid in the food at the
time it is ingested.

INFECTIONS OF THE DIGESTIVE TRACT 23

AEROBIC AND ANAEROBIC CONDITIONS IN THE
DIGESTIVE TRACT

There are many conditions which influence the char-
acter and extent of bacterial decompositions in the ali-
mentary tract. The chemical character of the food
(quite aside from the bacteria it may contain) , the
solubility of the food in the digestive juices, the volume
and composition of these digestive juices, all affect in
important ways the ultimate fate of the food in its
relation to microorganic decomposition. Some phases
of these influences will be discussed in these pages in
connection with special problems, although it must
be owned that our knowledge of them is far from being
full, owing partly to the great difficulties of experi-
mental investigation. Intimately intermingled with
these factors of food and secretory activity is the influ-
ence of aerobic and anaerobic conditions in the digestive
tract upon the nature of the bacterial activities that
occur there.

Pasteur, with his sharp insight into biological phe-
nomena, was first in recognizing the ability of certain
microorganisms (including a butyric acid producer,
yeast plants, aspergillus, and some mucors) to live in the
absence of oxygen and thereby first to distinguish be-
tween aerobic and anaerobic life. He embodied his views
in a famous epigram, the validity of which has been much
discussed — " Fermentation is life without oxygen."
This statement, although too sweeping and exclusive,
has in it a large element of truth and has a special

24 INFECTIONS OF THE DIGESTIVE TRACT

significance in its application to the bacterial processes
in the digestive tract. This significance comes from two
fundamentally important facts : first, that the initiation
of putrefactive decompositions in the digestive tract
(as elsewhere) depends very largely, though probably
not exclusively, on the activities of obligate anaerobes ;
secondly, that an important portion of the digestive
tract is most of the time under anaerobic conditions.
The importance of the role taken by obligate anaerobes
in bringing about putrefactive decomposition in native
proteids is only beginning to be appreciated. Bien-
stock * made an important contribution to biology when
on definite experimental evidence he insisted on the
inability of the ordinary facultative anaerobes to attack
native proteids, and showed that obligate and strict
anaerobes like B. putrificus are able, quite unaided, to
break down proteids. Dr. Rettger,2 in his numerous
experiments on putrefactive decomposition, confirmed
and extended the work of Bienstock. He showed that
an egg-meat mixture is not attacked by B. lactis aerogenes,
B. coll. Streptococcus pyogenes, B. alcaligenes fcetidus,
proteus vulgaris, etc., whereas it is regularly decomposed
by strict anaerobes such as B. putrificus, B. maligni
cedematis, and B. anthracis symptomatici (bacillus of
quarter-evil). This is a severe test, and there are putre-
factive anaerobes which cannot attack the egg-meat

1 " Untersuchungen iiber die Aetiologie der Eiweissfaulnis,"
Archiv /. Hyg., xxxvi, p. 335, 1899; ibid., xxix, p. 390, 1900.

2 "Studies on Putrefaction," Journ. of Biol. Chem., ii, p. 71,
1906.

INFECTIONS OF THE DIGESTIVE TRACT 25

mixture but which are nevertheless intensely putre-
factive when grown in a medium containing a low per-
centage of sugars as well as native proteids. This is
true of B. aerogenes capsulatus. I can confirm the obser-
vation that B. coli cannot decompose media consisting
of native proteids even on long standing. On the other
hand, there is no more striking type of putrefaction than
that induced in proteids by B. putrificus upon egg-meat
mixture.

At the present time I know of no evidence to show
that native proteids can be actively decomposed by any
of the ordinary facultative anaerobes, and in respect to
the digestive tract the facts all point to the correctness
of the view that we largely owe the initiation of proteid
cleavage there to the agency of the strict anaerobes.
But it must be remarked in passing that it does not
follow from what has just been said that intestinal
putrefaction is carried on through the sole activity of
strict anaerobes. The fact that these organisms are the
great destroyers of native proteids does not tell the
story of the entire putrefactive process. The intestine
abounds with microorganisms which are able to attack
albumoses and peptones and to effect a further degra-
dation of the proteid molecule, thus entering into a
symbiotic action with the strict anaerobes.

It appears desirable to state at this point just what is
meant by the expression " anaerobic life'7 as applied to
the strict anaerobes to which reference has been made.
When we speak of anaerobic life, we are liable to vaguely
picture a state of existence in which the microorganisms

26 INFECTIONS OF THE DIGESTIVE TRACT

concerned are subsisting without access to any oxygen
whatsoever, and the term "strict anaerobe" tends to
give fixity to such a conception. Such an idea, however,
would be far from expressing the actual truth, since there
is no known exception to the rule that living microorgan-
isms cannot maintain themselves without producing
carbon dioxide, a process which involves the oxidation
of organic materials within the cell. But the quantity
of oxygen required to maintain life may be extremely
small, and it is in this sense only — which is expressed in
the term "micro-aerophile," suggested by Beijerinck1 —
that we are justified in speaking of anaerobic micro-
organisms. It has been calculated by Matzushita2
that anaerobes may live although the surrounding air
contains not more than 0.0031 per cent, of oxygen.
This fact is liable to mislead us if we accept it as meaning
more than that the presence of very little oxygen in
the air is consistent with the maintenance of some
forms of microorganic life. For, on the one hand, it
seems clear that microorganisms may obtain their small
portion of essential oxygen not from the air, but from
a medium containing a decomposable substance which,
like glucose, contains oxygen which becomes available
for the support of life. In other conditions the body of
the microorganism may contain material which is capable
of yielding oxygen for a time, in minute amounts.

1 Arch. Norland, ii, 1899.

2"Zur Physiologic der Sporenbildung der Bacillen, nebst
Bemerkungen zum Wachstum einiger Anaeroben," Archiv /. Hyg.,
xliii, p. 327, 1902.

INFECTIONS OF THE DIGESTIVE TRACT 27

It is conceivable that under such conditions as those
just mentioned, it becomes a matter of indifference to
the anaerobes how little oxygen is contained in the sur-
rounding air. On the other hand, it is well known that,
paradoxical as it seems, the strictest anaerobes known
are able to grow freely in apparently close contact with
a fairly abundant supply of oxygen, provided aerobic
bacteria are also present. Pasteur, to whom this note-
worthy symbiotic phenomenon was known, ascribed it
to the circumstance that the aerobes remove the oxygen
present in so thorough a way that the anaerobes are
unhindered in their growth. This doctrine, which
found wide acceptance, has lately been questioned by
Kedrowski * (and others), who attribute the ability of
anaerobes to grow symbiotically with aerobes to the
formation of a "ferment" by the latter, which renders
the anaerobes immune to the action of oxygen. It
cannot be said that this hypothesis has received sub-
stantial support, and we are compelled to return to the
teaching of Pasteur or to some modification of it.2 There
is a limit to the amount of oxygen which the aerobes

1 " Ueber die Bedingungen unter welchen anaerobe Bakterien
auch bei Gegenwort von Sauerstoff existiren konnen," Zeitschr.
f. Hyg., xx, p. 358, 1896.

2 See Walter von Oettingen, "Anaerobic iind Symbiose,"
Zeitschr. /. Hyg., xliii, p. 475, 1903. This writer contends that
the primary action in aerobic and anaerobic symbiosis is the split-
ting action of the anaerobes in sugar with their associated inability
to appropriate the oxygen contained in carbohydrates, that oxy-
gen being eagerly seized by the aerobes for their use. It is hard to
reconcile this view with the retarded growth of the anaerobes (as
compared with the aerobes) in some symbiotic experiments. In
this paper will be found the literature on the subject.

28 INFECTIONS OF THE DIGESTIVE TRACT

can divert from the anaerobes, and it is easy to check
the growth of anaerobes which are not facultative by
passing oxygen with some freedom through the fluid
medium, no matter how many aerobes are present.

A study lately made by Willimsky * on the adaptation
of aerobic bacteria to anaerobic conditions shows that
aerobic bacteria like the spirillum of cholera, B. alkali-
genes, B. fluorescens non liquefaciens, are able to adapt
themselves to an atmosphere in which there remains
only a minimal trace of oxygen and that this adapta-
tion is the more efficient the more slowly the oxygen is
withdrawn, sudden withdrawal of oxygen greatly slowing
the multiplication of the organisms. The multiplication
of organisms was found, however, to be much less rapid
where oxygen had been thus withdrawn than where it
is present. When an absolutely anaerobic condition was
brought about, the organisms died, and their destruction
was the more rapid, the more abrupt the withdrawal of
oxygen.

The symbiosis of aerobes and anaerobes is a biological
phenomenon of much consequence in determining the
distribution of anaerobic bacterial processes in the diges-
tive tract. Without such symbiotic action the develop-
ment of strict anaerobes would be confined to those parts
of the digestive tract into which oxygen rarely passes,
and then only in small amounts. The large intestine
is seldom visited by free oxygen, but it is probably
usual in man for the small intestine to contain a little

1 "Ueber des Verhalten der aeroben Keime gegeniiber der ab-
soluten Sauerstoffentziehung," Archiv f. Hyg., liv, p. 375, 1905.

INFECTIONS OF THE DIGESTIVE TRACT 29

air. Boycott * in his observations on the gases of the
intestine found that the small intestine of the cat gener-
ally contains a little oxygen, while the large intestine
holds either no oxygen or a smaller percentage than is
present in the small intestine. Sometimes the proportion
of oxygen in the gas from the large intestine was a little
above one part per hundred. Boycott noticed that oxy-
gen passes very readily out of the intestine, partly
by diffusion, but mainly by being appropriated by the
mucous membrane. There is apparently very little
direct exchange with the blood. In the stomach and
cavity of the mouth there is always a considerable
quantity of oxygen — unless, indeed, the oxygen in the
stomach be displaced by the development of an abundance
of carbon dioxide, as sometimes happens. It is probably
safe to assume that in the mouth the free presence
of oxygen constantly acts as a deterrent to anaerobic
growth. In spite of this, however, anaerobic life is
possible, and the indications are accumulating which
point to the important part played by anaerobes in setting
up pathological states in the mouth. Caries of the teeth,
which was formerly referred to aerobic bacterial action,
seems clearly the result of the invasive action of anae-
robes on the tooth pulp. In certain derangements of
digestion, I have noticed that butyric acid fermentation
or putrefaction takes place with great rapidity in the
food particles that have lodged between the teeth, espe-
cially if carbohydrate food has been taken. It must be

1 " Observations on the Gaseous Metabolism of the Small In-
testine of the Rabbit," Journ. of Physiol., xxxii, p. 343, 1904-05.

30 INFECTIONS OF THE DIGESTIVE TRACT

admitted that proof is wan ting that butyric acid formation
in the mouth is always due to anaerobes. The growth of
anaerobes in the mouth is favored by want of cleanliness
there, since the accumulation of food-masses encour-
ages anaerobic conditions beneath the surface. Probably
the intelligent and free use of the tooth-brush is a most
potent means of discouraging anaerobic growth in the
mouth. In removing decomposing food-masses one
not merely admits air to the anaerobes, but also removes
many aerobes which, through the symbiotic action
already mentioned, facilitate the multiplication of the
former.

In the normal human stomach the conditions for the
development of anaerobic bacteria are usually poor,
partly on account of the presence of air, but mainly
owing to the action of the gastric juice. Nevertheless
the interior of a food bolus must often afford good
opportunities for the growth of anaerobes that happen
to be taken with the food. In a stomach which secretes
little or no hydrochloric acid, and which is sluggish in
emptying its contents, the chances for anaerobic develop-
ment are good, and hence we frequently find, under these
circumstances, that there are evidences of putrefactive
decomposition of food that has been unduly retained
in the stomach (e.g. production of sulphuretted hydro-
gen, mercaptan, butyric acid, etc.). Sometimes, too,
we have definite infections of the stomach wall with bac-
teria which are facultative anaerobes (e.g. streptococci),
and which are capable of inducing acute lesions of the
stomach. On the whole, however, I think one may say

INFECTIONS OF THE DIGESTIVE TRACT 31

that in the course of chronic gastric affections, the
number of anaerobic microorganisms in the stomach is
seldom great.

We are able to form an opinion as to the bacterial con-
ditions that prevail in the stomach because this organ is
readily accessible through the aid of the stomach-tube.
Of the conditions of bacterial life in the small intestine
we know really very little, because of the inaccessibility
of the contents of this portion of the digestive tract
during life. It is, of course, well established that the
duodenum and jejunum contain few bacteria and that
these are usually not of the kind that produce pathogenic
effects or induce putrefaction. For example, Gushing
and Livingston * found that in gunshot wounds of the
upper part of the small intestine, the bacteria which
escaped into the peritoneal cavity did little or no harm
-a result in sharp contrast with what happens after
injuries at lower levels of the digestive tract. Again, if
we take the intestinal contents from the jejunal or duo-
denal level of the gut of a person recently dead of an
acute disease not especially implicating the digestive
tract (e.g. pneumonia), it will be found that the bacteria,
while acting readily on sugars, owing to the presence of
B. lactis aerogenes, do not induce putrefactive decompo-
sition on proteid media. There are exceptions to this

1 " Experimental and Surgical Notes upon the Bacteriology of
the Upper Portion of the Alimentary Canal, with Observations
on the Establishment there of an Amicrobic State as a Preliminary
to the Operative Procedures on the Stomach and Small Intestine,"
"Contributions to the Science of Medicine by the pupils of
William Welch," p. 543, 1900.

32 INFECTIONS OF THE DIGESTIVE TRACT

rule, but they do not alter the general applicability of the
statement that putrefactive microorganisms are com-
monly few in the upper two thirds of the small intestine.
The chief factors in bringing about this state of relative
freedom from anaerobic organisms are apparently the
defensive action of the stomach juices and the rapid
passage of the chyme through the upper intestine, which
offers little chance for the multiplication of anaerobes,
or indeed of bacteria of any kind. To what extent the
moderate amount of oxygen present inhibits bacterial
growth is uncertain.

In the human ileum, the character of the bacterial
decompositions in healthy persons appears to differ
rather widely according to a variety of conditions, some
of which are still obscure. Perhaps on account of the
approach to the ileocsecal valve, which offers some
degree of mechanical difficulty to the passage of the con-
tents of the small intestine, there is an accumulation of
bacteria in the ileum. This mechanical obstacle, in the
human intestine and in that of many lower animals
(e.g. dog), consists partly in the narrowing of the lumen
of the gut, but partly also in the presence of inspissated
intestinal contents on the further side of the valve, in
the colon itself. In man there is within a foot or two
of the colon a marked increase in the number of bacteria
in the ileum, and this increase relates not merely to the
total number of bacteria present, but also to the varieties.
The mere accumulation of material affords opportunity
for anaerobic conditions of life, and we find in fact that
the numbers of strict anaerobes in the lower part of the

INFECTIONS OF THE DIGESTIVE TRACT 33

human intestine and in that of dogs is much increased
in the ileum as compared with higher levels. Hence
we find that the mixed faecal bacteria taken from this
level of the lower ileum are capable of inducing putre-
factive changes in native proteids and in more simple
nitrogen-holding media. This is true of the human ileum
in health and at all periods of life, but the accumulation
of putrefactive bacteria is even more pronounced in
certain cases of disease — in other words, anaerobic con-
ditions of bacterial life are exaggerated in pathological
states. We may indeed look on the ileum as the debat-
able land of the digestive territory. In ideal health
this region contains relatively few anaerobes and patho-
genic bacteria; in acute and chronic intestinal patho-
logical processes the ileum is probably inhabited by
flora capable of producing substances which may
damage the organism either locally or generally, and
among these the anaerobes are apt to be prominent.

In the large intestine we find the most dense accumu-
lation of bacteria and the best conditions for anaerobic
growth. The transition from small to large intestine
is in this respect very striking, more regularly so, per-
haps, in dogs than in human beings. In dogs the large
intestine is usually closely crowded with bacteria beyond
the ileocaBcal valve, whatever may be the conditions in
the ileum. The anaerobic conditions are well main-
tained throughout the large intestine, and it is here that
we find the greatest numbers of anaerobes and the most
pronounced evidence of putrefaction. There is, however,
a gradual fall hi the number of living bacteria beyond the

34 INFECTIONS OF THE DIGESTIVE TRACT

ileocaecal valve, so that in the rectum the numbers of culti-
vable bacteria are very much less than in the ascending
colon. But although in the ascending colon we have usu-
ally the greatest numbers of bacteria, it should be noted
that their variety is often not so great as in the ileum.

With a view to learning something about the anaerobic
conditions in the intestine, and their relation to the re-
ducing power of bacteria, a number of experiments were
made in which dogs were fed on meat containing methyl-
ene blue.1 The animals were killed and the contents of
the intestine examined under boiled water in order to
determine at what point in the intestine reduction of
methylene blue had occurred. The following typical
result was obtained in one instance. A dog weighing
fourteen and one-half pounds had received methylene
blue in water, about two hundred milligrams in all.
On the third day he received one hundred milligrams
of methylene blue, placed inside a piece of cooked meat.
He ate heartily of this meat and four hours after the meal
the animal was killed and the intestine examined.
The following diagram of the large and small intestines
indicates the points at which the contents were examined.

SMALL INTESTINE

A

B

C

D

E

F

The stomach was found to be about one-half full, and
intensely blue. In the duodenum, at A, the intestinal

1 1 am indebted to Dr. A. J. Wakeman for the series of proto-
cols on which these statements are based.

INFECTIONS OF THE DIGESTIVE TRACT 35

contents were intensely blue, the color diminishing in
intensity toward B until at C the color had almost
disappeared and had entirely disappeared between C
and D. The contents of the intestine obtained from D
blued only when exposed to air. When boiled with
hydrochloric acid, water, and hydrogen peroxide, they
became intensely blue. The contents of the colon
(between E and F) were not blue under water, but be-
came blue very quickly when exposed to air. The
mucus adhering to the mucous membrane of the intes-
tinal tract behaved in general like the intestinal contents
as regards the color. This experiment indicates that
beyond the middle of the small intestine the conditions
become rapidly anaerobic or very nearly so, since the
reduction of methylene blue in this concentration could
not occur in the presence of the air. The main factors
in the reduction process are, of course, the bacteria, and
it is noteworthy that the reduction of the methylene
blue begins at just that level of the intestine in which
the bacteria begin to be numerous. There is probably
in these cases not merely a reduction of the methylene
blue, but decomposition of this salt, with the liberation
of hydrochloric acid and the production of the methylene
blue leucobase. This is indicated by the behavior of
the leuco-body with hydrochloric acid.

THE BACTERIA OF THE HUMAN DIGESTIVE TRACT AT DIF-
FERENT AGES IN APPARENTLY HEALTHY INDIVIDUALS

While it is true that at all periods of life the human
digestive tract is the seat of the life activities of myriads

36 INFECTIONS OF THE DIGESTIVE TRACT

of microorganisms, it is also true that the biological
characters of these microorganisms are not the same at
all times of life. If we compare the bacterial flora of
apparently healthy persons of about the same age; we
may find considerable differences in those parts of
the digestive tract that are easily accessible to study,
and yet, upon the whole, the resemblances in the bio-
logical characters of the bacteria so compared will
ordinarily be more striking than the differences, pro-
vided the comparison be made between the flora of in-
dividuals living in a temperate climate and on the kind
of food that is usual in countries commonly regarded as
civilized. But if it were practicable to compare the
bacterial flora of the same individual at different periods
of life, it would be found that there are wide and
perhaps characteristic variations — variations dependent
in part on differences in food, but in part on other
influences. Such comparisons have probably seldom
been made in the same individual over a long period
of time, but comparisons between different persons of
unequal ages, but in general similarly conditioned, have
been repeatedly made and support the statement that
the normal bacterial flora characteristic of different
ages, present different biological characters and are
responsible for different types of decomposition in the
digestive tract.

THE BACTERIAL PROCESSES IN THE DIGEST-
IVE TRACT OF NORMAL NURSLINGS AND
BOTTLE-FED INFANTS

Nursling Infants. — In babies fed on mother's milk
the alimentary tract is the seat of conditions of bacterial
activity that possess a high degree of interest for one who
wishes to obtain an insight into the physiology of diges-
tion. For in nurslings one finds a relatively simple bac-
terial flora which gives a clew to the more complex and
puzzling bacterial conditions that characterize normal
adult life and many states of disease. The healthy
nursling's bacterial flora is not merely relatively simple
as regards the varieties of microorganisms; it is also
remarkably definite in the individual child during the
nursling period as well as in different children who are
nourished on mother's milk.

By means of the Gram stain, carefully employed, it
is possible to divide all the bacteria present at a given
level into two groups, those which retain the stain and
thus appear dark blue in the microscopical field (Gram-
positive bacteria) and those which do not retain the
stain (Gram-negative bacteria), and are hence suscep-
tible to coloration by a counter-stain (a red dye such
as fuchsin or saffranin being usually chosen). As these
different ways of behaving to the stain correspond to
biochemical peculiarities of the bacteria, we have at our

37

38 INFECTIONS OF THE DIGESTIVE TRACT

disposal a rough means of differentiation which is often
of great service in the microscopical identification (in
so far as this is possible) of bacterial types. If this
method of study be supplemented by making aerobic and
anaerobic cultures on gelatin plates, sugar-agar plates,
and blood-agar plates (with or without sugar), one ob-
tains a fair idea of the really representative bacteria
present in the alimentary tract. Some of the bacteria
present may, however, fail to grow on any of these media.
Gram-stained microscopical fields prepared from the
normal slightly acid, orange-yellow, rather soft fa3ces of
a nursling child from any portion of the large intestine,
present essentially the same characteristic appearances.
In the first place, the bacteria are seen to be extremely
abundant — more so than in the f seces of normal adults
and very much more so than in the meconium. A very
large proportion of them are Gram-positive. The
great majority of these Gram-positive bacteria present
the same appearances and may give to a field the appear-
ance of consisting of a nearly pure culture of one bac-
terial type. These bacteria are slender, often slightly
curved bacilli of moderate size. They constitute the
characteristic microorganisms of the human nursling's
digestive tract and have been named (for reasons which
will later appear) B. bifidus by Tissier, who first described
them and took them out of the group of colon bacilli,
with which they had been grouped by Escherich.
Mixed in with B.bifidus is another Gram-positive bacillus,
which is closely related, but grows readily on a strongly
acid medium and has hence been designated B. acidophi-

INFECTIONS OF THE DIGESTIVE TRACT 39

lus (Moro). These bacilli are not surely distinguishable
from B. bifidus in the Gram-stained field. Although
B. bifidus thus dominates the typical field, other Gram-
positive bacteria can always be found. Of these the
most important (though very few in number) are large,
plump bacilli, occurring usually singly, but sometimes
in pairs. This is the B. aerogenes capsulatus (B. Welchii
or gas-phlegmon bacillus of Fraenkel). As will be seen
later, it may play an important part in the pathology
of the digestive tract. Another Gram-positive organism
which is regularly present in small numbers in the field
is a diplococcus which frequently grows in chains. This
coccus is commonly associated with a similar organism
which is Gram-negative. The chief remaining Gram-
negative bacteria are small coccal or coccoid forms and
shorter and longer bacilli which by cultural methods
can be shown to have the biological characters of the
B. lactis aerogenes group and the B. coli communis
group.

Thus by means of the Gram method of staining one
finds that the typical field from the nursling's stool is
predominantly Gram-positive and consists very largely
of slender bacilli which but for their retention of the dye
might be regarded as colon bacilli, and were indeed so
regarded by Escherich, who in his original studies em-
ployed gentian violet as the sole staining agent. The
fact that the bacterial fields from the nursling are Gram-
positive is now familiar, but it came to those who first
observed it as a distinct surprise. It was known that
the B. coli group of bacteria are Gram-negative, and as

40 INFECTIONS OF THE DIGESTIVE TRACT

it was naturally assumed that the main bacteria of the
large intestine of the nursling must be organisms of this
group (as in the case of bottle-fed children), the failure
of the nursling's fields to conform to this assumption at
once threw doubt on the view that the nursling's domi-
nant fsecal bacteria are really colon bacilli. The view
was put forward that the Gram-positive character of
the fields is due to the influence of the large fat content
of the mother's milk,1 but this had soon to be abandoned.
It was only through the use of suitable cultural methods
that the mystery was solved.

If one inoculates agar, gelatin, or bouillon media with
the mixed flora got from a healthy nursling and cultivates
the microorganism in the presence of air, bacteria of the
B. coli type show themselves regularly to be the dominant
forms, although they may be associated with diplococci,
and sufficiently refined methods give evidence of the
presence of the closely allied B. lactis aerogenes. If,
however, one grows the mixed flora on a suitably pre-
pared acid medium,2 it becomes easy to restrict or even
quite check the growth of the members of the B. coli
group and to demonstrate the presence of a type of
microorganisms which grows readily in this acid medium,
has a morphology like that of the slender bacillary colon
forms, coagulates cow's milk but does not act on human
milk, and is definitely Gram-positive. This micro-
organism, or rather group of microorganisms, accounts

1 Alex. Schmidt, "Zur Kenntniss der Bakterien der Sauglings-
faeces," Wien. klin. Wochenschr., v, p. 643, 1892.

2 Acid beerwort or acetic-acid bouillon, as suggested by Moro.

INFECTIONS OF THE DIGESTIVE TRACT 41

for a small number of the Gram-positive bacilli seen
in the microscopical field. Moro, who first described
these bacteria under the title of B. acidophilus, supposed
that their presence explained fully the normal Gram-
positive bacillary field, but this proved to be an error.

On subjecting the mixed bacterial flora to strict
anaerobic conditions, Tissier found he could regularly
cultivate from the nursling's faeces certain slender, Gram-
positive bacilli of moderate length and thickness which
not infrequently presented a bifurcated extremity and
to which he accordingly gave the name B. bifidus.
There can be no doubt that a large majority of the
bacteria seen in the microscopical fields may be classed
as B. bifidus. The bacteria belonging to the B. bifidus
group possess a varied morphology which it seems
desirable to notice here in view of the physiological
importance of these bacteria during infancy. The ap-
pearance of the organisms is somewhat different accord-
ing as we meet with them in the faeces or in culture
media. In the microscopical fields they are recognizable
under three chief forms: (1) a plain bacillary form,1

1 The individuals of this class are rods of varying dimensions
(length 3-5 /A ; breadth, 0.2-0.4;*). They are usually straight
or slightly bent with somewhat pointed or rounded ends. Oc-
casionally, owing to attenuation of one end and excessive develop-
ment of the other, they assume a comma form. There is nothing
characteristic about the arrangement of the bacilli in smears.
Although distinctly Gram-positive, they vary much in the degree
in which they retain the dye, both in material from the same in-
fant at different times, and in different individuals. The entire
bacterium is frequently not stained with even intensity, some parts
of the organism retaining the color well, while other portions give
it up. The result is the appearance which Escherich expressed

42 INFECTIONS OF THE DIGESTIVE TRACT

(2) a bifid form, and (3) a cephalated (knobbed) or
headlet form. The simple form is by far the commonest,

in the course of his studies in the name "punctate " bacilli. These
unevenly colored bacteria are commonly stained in their middle
portion, while their ends are colorless, and this condition arises
presumably on account of a central concentration of the proto-
plasm, which leaves relatively empty adjoining spaces. There
are, however, other variations in the distribution of the colored
protoplasm within the bacillus, which it is unnecessary to describe
here. Frequently it will be noted that several bacilli are joined
end to end, forming one long, unevenly stained structure. As
the decolorized portion of the simple form of B. bifidus is still
able to take up the counter-stain (e.g. carbol f uchsin) with avidity,
we get apparently Gram-negative bacilli in those cases where the
protoplasm has been rarefied throughout.

In addition to the simple forms of bifidus, a careful examination
of the smear will sometimes reveal the presence of the bifid form
from which the group takes its name. Sometimes the bifid form
is very scanty or absent; at times it is moderately abundant;
but it is never in health the dominant form. As pointed out by
Moro, the bifurcation is peculiar in these bacteria in being a true
division of the end or ends of the organism rather than a lateral
branching, such as one sees in certain actinomyces. It is not
necessary to describe here the numerous variations in morphology
exhibited by these bifid bacteria. (For fuller details see Moro,
" Morphologische und biologische Untersunchungen tiber die
Darmbakterien des Sauglings," Jahrb. /. Kinderheilk., Ixi, p. 687,
and p. 870, 1905; also Tissier, loc. cit. and Ann. de VInst. Pasteur,
xix, p. 109, 1905; and Rodella, " Repartition des Microbes dans
Tlntestin du Nourrison," Ann. de VInst. Pasteur, xix, p. 404, 1905.)

A third form of bacterium, which apparently belongs in the
same group with the simple and bifid forms just described, is
characterized by a small enlargement at one or both ends — the
headlet or "Kopfchen" form. These headlet bacteria are seldom
numerous in the faeces of normal nurslings, and it is usually
necessary to look carefully over the field to find them, although
they are seldom quite absent. The bodies of these bacteria are so
like the bodies of the simple form in their morphology that one is
tempted to regard them as simply slightly modified examples of
this simple form — a view which is borne out by the results of

INFECTIONS OF THE DIGESTIVE TRACT 43

and in many preparations is the exclusive representative
of the bifidus type.

cultural studies. Indeed, the microscopical picture itself shows
transitional forms — bacilli with slight indications of the accumu-
lation of protoplasm at one end and headlet forms which are
bifid — which look like the simple and the headlet forms. Dumb-
bell forms are at times seen in which a headlet has developed at
either end of a bacillus. In preparations which have been fixed
by heat and treated with Lugol's solution it is possible to demon-
strate the presence of granules in the bodies of the bacteria of the
bifidus group, located at the peripheral ends of the organism and
generally limited to this position. In hanging drops made from
fresh material one finds that many of the rod-shaped bacteria
possess a sluggish motility which is soon lost.

Certain relationships between the polymorphous Gram-positive
bacteria which dominate the microscopical fields (prepared from
the nursling's faecal bacteria) can be made out by means of anaero-
bic cultures upon suitable media, such as agar or sugar-agar,
containing acid. Pure cultures of the bifidus bacteria are not,
however, always quite easy to obtain, as the B. coli group, and
some diplococci, which are facultative anaerobes and grow well on
the media just mentioned, are apt to restrict the growth of the
less hardy bifidus. The development of colonies on sugar-agar is
slow, for the tiny white lens-shaped smooth-edged growths first
appear in the deeper parts of the sugar-agar after about three
days. (Colonies like these in all respects except in having a toothed
border are sometimes seen, and consist of mixtures of B. bifidus
with B. acidophilus. On further inoculation from such colonies
there grows, according to Kahn, only B. acidophilus.)

On microscopical examination of these colonies, it is apparent
that we have to deal with differently shaped bacteria in different
colonies and that even in the same colonies there is a tendency
to polymorphism. Sometimes the bacilli (always Gram-positive)
correspond closely to the simple form already described, both in
their form and in the variations in their staining peculiarities.
Some colonies are made up of bacteria very much larger than
those of the simple form, and frequently knobbed or bifurcated or
both. There seems to be little doubt that we are dealing here
with various forms of B. bifidus, and the frequent development of
bifid and headlet bacteria of varied morphology on the agar medium

44 INFECTIONS OF THE DIGESTIVE TRACT

The B. putrificus of Bienstock can sometimes be ob-
tained from the nursling's intestine by employing a

which has been inoculated from the simple form of bifidus con-
stitutes the best evidence of close relationship of these varieties.
In lactose-bouillon fermentation tubes which have been inoculated
from the faeces of nurslings one finds in the sediment large numbers
of the bifid form of B. bifidus, after incubation for twenty-four hours.

In anaerobic sugar-agar plates made from the faeces of a
human nursling, one finds colonies which develop before those
made up of the bifidus type, especially colonies of Staphylo coccus
pyogenes albus (which sometimes grow luxuriantly on the surface),
and at times one finds varieties of streptococci. An opaque,
white, rumpled layer may be seen on aerobic plates made from
the first dilutions of the faecal material, which shows itself, under
the microscope, to be made up of bacilli possessing motility which
on further study prove to be the common potato bacilli (B. mesen-
tericus vulgatus, Fliigge).

In the study of the bacterial flora of the intestinal tract (es-
pecially in its lowest portions, colon, sigmoid, and rectum) the
routine employment of pasteurization at 80° C. for fifteen or twenty
minutes previous to making aerobic and anaerobic plates, is of
much value. Most vegetative forms of bacteria present in the
human intestine are killed by this exposure to heat, and only
those races survive which can form spores and have done so.
The application of this procedure is important both in the study
of physiological conditions and for the detection of pathological
peculiarities of the intestinal flora. On employing this method
in the study of the intestinal flora of nurslings, it is found that by
the use of suitable culture media, several forms of spore-bearing
bacteria are present as regular or almost regular inhabitants of
the lower intestine.

Among the microorganisms which are brought to notice by
these methods is a variety which makes on agar small, cottony
colonies, of somewhat indefinite outline, giving rise to small gas-
production which may cause slight rents in the medium. These
colonies consist of Gram-positive bacilli, often colored only in places
by the dyestuff, which sometimes carry a knob at one end, which
can, in certain instances, be shown to be spores. These spore-bear-
ing bacilli are to be regarded as identical with the headlet bacteria
already mentioned, and are thus the representatives of the resistant

INFECTIONS OF THE DIGESTIVE TRACT 45

medium of sterilized egg-white, but its presence in more
than small numbers is exceptional and it may be entirely
absent.

A more important anaerobic spore-bearing micro-
organism than B. putrificus is B. aerogenes capsulatus*
which can be demonstrated by its growth on sterilized
milk to which an abundance of faecal bacteria has
been added or by its ready growth on blood-agar with
or without the addition of sugar. So long ago as 1894
Fliigge found that the faeces of children nourished wholly
on mother's milk contained Botkin's Bacillus butyricus,
and we now know from the critique of Grassberger and
Schottenfroh that Botkin's bacillus in reality represents
two distinct microorganisms, each of which is capable

or spore-forming variety of B. bifidus. The headlet bacteria have
a morphological resemblance to another spore-forming anaerobe,
B. putrificus, which by appropriate methods of culture can some-
times be isolated from the nursling's stool. The resemblance is
so close that in faecal smears submitted to microscopical examina-
tion it is hardly possible to make a distinction with confidence.

1 This organism is known by different names in different coun-
tries. The name B. aerogenes capsulatus was given to it by Pro-
fessor Welch and is the name by which it is best known in the
United States. The organism is identical with the gas-phlegmon
bacillus of Fraenkel and probably with the granulo-bacillus im-
mobilis liquefadens of Grassberger and Schottenfroh. In France
the organism is known as B. perfringens (Veillon and Zuber).
In England Klein described an organism which he called B. en-
teritidis sporogenes,b\it it is doubtful if he was dealing with pure
cultures, although it appears probable that he had B. aerogenes
capsulatus in his cultures. A trinomial name is considered objec-
tionable by botanists, and hence Migula proposed the name B.
Welchii for B. aerogenes capsulatus, but the latter name is now so
firmly established in the United States that it is doubtful if it
will be displaced. Welch's bacillus is often spoken of as "the
gas-bacillus."

46 INFECTIONS OF THE DIGESTIVE TRACT

of inducing decomposition with the production of
butyric acid. Of these the granulo-baciUus immobilis
liquefaciens is probably identical with B. aerogenes
capsulatus (but not pathogenic), and was described by
Passini in 1902 as occurring in the faeces of normal nurs-
lings. This is undoubtedly true, but I wish to empha-
size the fact that the organism normally occurs under
these conditions in only very small numbers as com-
pared with bacteria of the bifidus type. One may look
through many Gram-stained fa3cal fields without find-
ing any evidence of its presence. The physiological
significance of this fact will later be pointed out. The
other organism which appears to have been included
in Botkin's description of his B. butyricus, is one closely
related to B. aerogenes capsulatus in morphology. It is,
however, motile, does not liquefy gelatin, and readily
forms spores — features which distinguish it sufficiently
for the present enumeration * of the anaerobic flora.
This organism has been found in the faeces of nurslings,
but is of much less importance than B. aerogenes capsu-
latus. Occasionally, too, the nursling's fa3ces have been
found to contain the bacillus of malignant oedema
(Passini) and a very closely related microorganism which
differs from the bacillus of malignant cedema in being
Gram-positive and in making skatol.

One may summarize the results of cultural studies of
the bacterial flora of normal human nurslings as follows.
The stools and the contents of the large intestine from

1 It is known as the granulo-bacillus mobilis nons-liquefadens
of Grassberger and Schottenfroh.

INFECTIONS OF THE DIGESTIVE TRACT 47

these infants number among their inhabitants the
following: B. bifidus communis and the headlet bac-
teria (which are probably varieties of B. bifidus), B. coli
communis, and B. lactis aerogenes, B. acidophilus, B.
aerogenes capsulatus, and certain intestinal streptococci.1
It should be noted that of these bacteria the dominant
form, B. bifidus (and in the upper part of the large intes-
tine, B. coli communis and B. lactis aerogenes), is one
which forms lactic acid on sugar media, is antiputrefac-
tive, and does not attack proteids. It should also be
observed that B. aerogenes capsulatus, the most impor-
tant representative of the putrefactive type of in-
testinal bacteria, is an extremely scarce organism, and
finally, that spore-holding bacteria and free spores are
of rare occurrence (relatively) in the faeces and contents
of the large intestine.

What has been already said regarding the results of
microscopical and cultural investigation of the nursling's
faeces makes it clear that the picture which we get from
the former method is quite different from that furnished
by the latter. Whereas one would conclude from the
microscopical field that he was dealing chiefly with one
race of bacteria (B. bifidus) supplemented by very few
other forms, cultural methods which include anaerobic

1 Less constant inhabitants are the following, Staphylococcus
pyogenes albus, intestinal diplococci, B. putrificus, the motile
butyric acid bacillus (Amylobacteria of Briiber or granulo-badllus
mobilis non-liquefadens of Grassberger and Schottenf roh) ; B.
mesentericus vulgatus (Fliigge) ; certain actinomyces, especially
Actinomyces chromogenes; sarcinse ; Oidium albicans, and various
yeasts.

48 INFECTIONS OF THE DIGESTIVE TRACT

studies on suitable media show that the flora is in
reality more varied and that forms physiologically im-
portant are present of which one obtains no suggestion
from the fields alone. Such a picture, taken by itself,
is misleading in certain respects, but representative in
others. The microscopic fields of the faeces showing
mainly one form represent with a fair degree of accuracy
the bacterial conditions in the faeces, for here the other
types are in reality very feebly represented. The cul-
tural methods which it is almost impossible to employ
in such a way as to make them give an accurate insight
into the actual quantitative relations of the bacteria
present, might easily mislead one into supposing that
the subsidiary varieties thus brought to light are more
abundant in the faeces than is actually the case. But
if we look at these cultural results as indications, not of
the quantitative relations of the faecal flora, but rather
as indices of the conditions prevailing at higher levels
of the large intestine, where one type is no longer so
dominant, these results become instructive instead of
confusing.

Distribution of the Bacterial Flora in the Digestive
Tract of the Nursling. — Although the distribution of the
bacteria of the digestive tract in normal or approximately
normal nurslings has not yet received much attention,
it is a subject which cannot be ignored, for it is capable
of giving us information in regard to the nature of the
decompositions that are possible at different levels of
the tract in health, these decompositions being often of
a different nature in disease from those which occur in

INFECTIONS OF THE DIGESTIVE TRACT 49

health. Escherich, one of the few students of this sub-
ject, found that the number of intestinal bacteria in-
creases gradually from above downward in the small
intestine, undergoing a sudden and extreme increase at
the level of the caecum. Even as high as the duodenal
level moderate numbers of microorganisms could be
found corresponding to the types of the B. coli and B.
lactis aerogenes groups. Schmidt, making use of Weigert's
method for staining fibrin, noticed that a preponderance
of Gram-positive over Gram-negative forms does not
begin until the level of the colon is reached. From the
middle of the colon downward, the characteristic Gram-
positive fa3cal flora begins to dominate the picture and
appears exclusively present in the rectum — a striking
evidence of the fundamental difference between the flora
of the upper and lower portions of the intestine.

I have had an opportunity to make microscopical
examinations (using Gram's method of staining) of the
contents of the digestive tract of nursling infants dying
within the first six months from conditions not closely
connected with the digestive tract (e.g. acute broncho-
pneumonia). On comparing the results of these exami-
nations with those obtained by Moro by the use of the
Escherich-Weigert fibrin stain, they were found to be so
nearly alike in their leading features as to make it
probable that the following description, based on these
two sets of observations, is generally applicable to the
case of healthy nursling children.

In normal nurslings the mouth contains few bacteria,
and such as it contains are for the most part organisms

50 INFECTIONS OF THE DIGESTIVE TRACT

derived from the skin and nipple — Staphylococcus
pyogenes aureus, bacilli of the B. coli group and B. lactis
aerogenes. In the stomach the number of bacteria is
also small, and the bacterioscopic picture shows usually
a few positive or negative diplococci or streptococci or
negative cocco-bacilli or positive or negative bacilli,
suggesting the B. coli and B. lactis aerogenes group. The
duodenum and jejunum generally contain remarkably
few microorganisms, owing mainly to the rapidity with
which the food passes through this region. When food
in the jejunum is present in larger amount than is usual,
the numbers of bacteria are correspondingly increased,
but it is the rule that the greater part of the small intes-
tine is almost free from bacteria and that they do not
become really numerous until the level of the lower ileum
is reached. Here there is almost regularly an abrupt
increase in the bacterial flora. In the csecum (and like-
wise in the vermiform appendix) the bacterial increase
is enormous. The normal bacteria of the greater por-
tion of the small intestine are Gram-negative, short bacilli
of the coli and lactis aerogenes variety, mixed sometimes
with a few positive and negative coccal forms. In the
lower ileum the Gram-negative forms become mixed with
those of the bifidus type, especially the punctate bacilli.
The transition from lower ileum to caecum marks, as a
rule, a striking change in the proportions of coli and
bifidus types, since the former lose their dominant nu-
merical position. This ascendency of the bifidus type is
so great in the transverse and descending colon that all
other bacteria in these regions are numerically incon-

INFECTIONS OF THE DIGESTIVE TRACT 51

spicuous and, as in the rectum, this type has the ap-
pearance of being present in pure culture.

The bacterial flora of the intestinal tract of the nurs-
ling is thus only moderately numerous as regards variety.
The bacteria are concentrated in the regions that lie
between the lower ileum and the anus, the ileocsecal
region presenting more kinds of bacteria than the lower
colon and more organisms capable of being cultivated.
The comparatively small numbers of bacteria found in
the small intestine has its explanation partly in the
small amount of food that lodges there, partly perhaps
in the bacteriolytic action of the succus entericus.
Wherever particles of transformed casein are found, there
also will be found bacteria in abundance, but with the
exception of the lower ileum the small intestine does not
harbor food masses to any considerable extent. The
bacteriolytic action of the succus entericus is moderate
but appreciable. The epithelial cells are said to contain
an antitryptic ferment, and this passes to some extent
into the succus entericus, where it is perhaps capable
of exerting a restraining influence on that peptonization
of proteids which is the first essential step toward putre-
factive decomposition.

Various attempts have been made to form a numerical
estimate of the numbers of living and dead bacteria
present at different levels of the colon and to deter-
mine what proportion of bacteria have been rendered
lifeless during their descent through the colon. Eberle,
who compared the numbers of bacteria seen in micro-
scopical preparations with the results of plate cultures,

52 INFECTIONS OF THE DIGESTIVE TRACT

concluded that from four and five-tenths to ten per cent,
of the faecal bacteria are capable of growing on the
common culture media. Klein got only about one per
cent, to grow; Schmidt and Strasberger, basing their
estimates on a different method of procedure, which in-
volved weighing masses of bacteria in order to avoid the
errors incidental to counting, came to the conclusion that
only about 0.07 per cent, of the bacteria are viable. It
is, however, easy to see that none of these methods can
claim to give even approximately correct results, because
we have in the nursling's faeces many anaerobic bacteria
which grow only under special conditions, and acidophile
bacteria which require special media for their cultiva-
tion, and these special conditions have not yet been
furnished by those who have made numerical estimates.
Moreover, it is extremely probable that on their
way through the large intestine many bacteria undergo
autolysis and wholly disappear. Hence the problem
of determining the fate of the bacteria in their descent
through the nursling's colon is one beset with diffi-
culties so considerable as to make numerical com-
parisons an unpromising field for study. There can,
however, be no doubt that, as already stated, the living
bacteria of the intestine usually undergo a marked
diminution in numbers on their passage through the colon
and that a large proportion of them seen under the
microscope in preparations from the faeces are not viable,
even on anaerobic plates. The histological signs of dis-
integration are often visible. Among the causes of the
death of intestinal bacteria, under normal conditions,

INFECTIONS OF THE DIGESTIVE TRACT 53

the diminution of food and the lack of moisture prob-
ably are among the most effective.

A satisfactory study of the products of the mixed fsecal
flora from normal nurslings has not yet been made. One
fact nevertheless stands out, namely, that on sugar
bouillon containing blood the volatile acid or acids pro-
duced give a molecular weight corresponding closely to
that for acetic acid. The presence of only insignificant
amounts of higher volatile fatty acids points to the
absence of considerable numbers of anaerobic putrefac-
tive bacteria, since these form the higher acids, pro-
vided anaerobic conditions be secured.

In harmony with this is our observation that ten or
twenty cubic centimeters of a ten per cent, suspension
of fsecal bacteria from a normal nursling may be in-
jected into a rabbit, which is then incubated, without
inducing a gas-liver from putrefactive anaerobes (B.
aerogenes capsulatus).

The mixed fsecal flora when grown on plain bouillon
makes indol, doubtless owing to the multiplication of
colon bacilli.

The Infection of the Nursling's Digestive Tract and the
Relation of the Microorganisms to the Permanent Bac-
terial Flora. — One naturally asks the question : What is
the origin of the normal intestinal flora of the nursling
child? How does it happen that the meconium, which
is a sterile medium derived from the intestinal juices
and epithelial cells, becomes the seat of bacterial life?
As the mother's milk is normally free from bacteria
when it issues from the breast, one cannot attribute the

54 INFECTIONS OF THE DIGESTIVE TRACT

first infection of the digestive tract to necessary bacterial
inhabitants of the milk. But since a portion of the
milk may come in contact with the nipple, and the child's
lips come into contact with the skin of the breast as well
as the surface of the nipple, it is clear that any bacteria
living on or near the nipple may find their way into the
digestive tract of the child. It might be supposed that
the gastric juice or the enterokinase of the small intes-
tine would form an obstacle to the passage of living bac-
teria into the intestine, but it has been well demonstrated
that these defenses even in the fully developed form in
which they exist in adults, are only partial, and that many
living bacteria readily pass from the stomach into the
small intestine in newly born children. In the case of
very young children, the secretion of a bactericidal gastric
juice is probably but a feeble protection against the pas-
sage of living bacteria into the intestine, and even were
it in part effective, there might frequently be oppor-
tunity for bacteria, located in the interior of curds, to
pass from the stomach into the intestine. Moreover,
if the bacteria were present in large numbers in the stom-
ach, some of them would almost certainly pass into the
intestinal tract, although, as has already been seen in
discussing the distribution of the bacteria of the intes-
tinal tract, it is usual to observe a considerable falling
off in the numbers of microorganisms as one passes from
the stomach to the duodenum and jejunum. It seems
probable, then, on the grounds already stated, that an
infection of the intestine by bacteria occurs through the
mouth. This probability was rendered a certainty by

INFECTIONS OF THE DIGESTIVE TRACT 55

the observations of Escherich on a case of atresia ani
in which he was able to find in the intestine the obligate
milk stool bacteria, B. coli and B. lactis. Moro, fur-
thermore, found B. acidophilus in human milk and on
the areola in nursing women and drew the inference that
this organism reaches the intestine by way of the mouth.
The passage of this particular organism which is so
resistant in acid media is, of course, not surprising.
Other observations were made by Moro upon human
milk in the hope of finding there B. bifidus communis and
of thus explaining the presence of this microorganism or
group of bacteria in the intestine of nursling children;
but although he made numerous anaerobic sugar-agar
cultures from various samples of human milk (where no
precautions had been taken to cleanse the nipple), he
was unable to obtain anything except negative results,
although relatively large quantities of milk were regu-
larly used in making the cultures. Moro was indeed
unable to cultivate any strictly anaerobic bacteria from
human milk. On the other hand, he regularly found
Staphlococcus albus, which is well known to be present
on or near the epidermis; sometimes colon bacilli and
pseudodiphtheria bacilli ; and in exceptional cases strep-
tococci, sarcinse, and Micrococcus candicans; and once
a fluorescent bacillus, but never anaerobes.

A further observation bearing on this question was
made by Moro. A newly born child received only ster-
ilized milk from the second until the sixth day. The
excretion of meconium took place in the usual way and
the characteristic bacterial types were found in it.

56 INFECTIONS OF THE DIGESTIVE TRACT

The first milk stool appeared on the fifth day and con-
tained the physiological flora. It is evident, therefore,
that the organisms present in this case must have had
another source. The possibility of the presence of
B. bifidus in the mouth of the child was then excluded by
the observation that anaerobic organisms do not grow
from the mouth secretion of newly bora children. As
it thus appeared clear that the infection of the intestinal
tract by bifidus does not occur by way of the mouth,
there remained only one possibility ; namely, that the
infection occurred through the anus. A study of the
organisms of the meconium has made it appear very
probable that this is the mode of infection with B.
bifidus.

The bacteria of the meconium were first carefully stud-
ied by Escherich, who pointed out that the originally
sterile meconium very easily acquires bacteria charac-
terized by their relative sparcity and considerable variety
and the presence of certain spore-bearing organisms.

Among the constant flora are coccal forms in relatively
large numbers; some short bacilli, including B. coli
communis, and some varieties of spore-bearers. One
form of spore-bearer is the so-called headlet bacterium
already mentioned, and a thicker form which was iden-
tified by Kohn as B. subtilis. On gelatin plates the most
abundant organisms are usually B. coli communis, which
is sometimes associated with liquefying streptococci. In
fact, the presence of liquefying bacteria in the meconium
is rather characteristic of the meconial period, and this
peculiarity, under normal conditions, ceases abruptly

INFECTIONS OF THE DIGESTIVE TRACT 57

with the onset of the milk-stool period. B. lactis aero-
genes was not found in the meconial stools. It appears
from Moro's observations, which I can corroborate, that
bacteria of the B. coli type may appear in agar plates
as early as ten hours after birth. Somewhat later,
various coccal forms may appear ; in the course of forty-
eight hours the bacterioscopic fields may show bacilli
which are Gram-positive and which in some instances
show typical bifurcations and through anaerobic culture
on sugar-agar may prove to be B. bifidus communis.
Still a little later, B. aerogenes capsulatus may make its
appearance, and about this time, too, spore-bearing cylin-
drical bacilli may appear. The very early occurrence
of B. bifidus in the meconium is a striking fact, since it
appears hardly possible to explain its presence there
so soon after birth on the supposition that bifidus has
entered by the portal of the mouth. It appears, there-
fore, upon the whole, extremely probable that the infec-
tion of the gut in newly born children with B. bifidus
communis is by way of the anus and that the air is the
source of the infection. It is interesting to note that
even before the appearance of B. bifidus in the meconial
stools the colon bacillus is represented there. The pres-
ence of the colon bacillus, like the presence of B. bifidus,
cannot be explained at this early period by an invasion
through the mouth, although, as already stated, the colon
bacillus may enter the intestine by way of the mouth at
a later period. Both these organisms, B. bifidus and B.
coli, possess motility, and it is probably by virtue of this
motility that they find their way into the lower intestinal

58 INFECTIONS OF THE DIGESTIVE TRACT

tract, and into the meconium on which they are able to
live, although it is only a fair culture medium. The
meconium thus contains forms which are the forerunners
of the permanent intestinal bacteria. In the case of
B. aerogenes capsulatus, it appears to me that its presence
in the meconium can hardly be explained on the ground
of its possessing motility, since it is questionable whether
this organism is ever at all motile, and certain that as a
rule it is non-motile. It is possible that it may be found
in the meconium in extremely small numbers as a result
of contamination at the time of the passage of the meco-
nium, since B. aerogenes capsulatus is so widely distributed
and so common an inhabitant of the air. Possibly, also,
the same explanation would apply to the presence of B.
putrificus, a spore-bearing organism which is common in
dust.

We see thus that while non-motile forms of organisms
like B. lactis aerogenes, B. acidophilus, staphylococci, and
streptococci can find their way into the intestinal tract
only by way of the mouth, through the milk (being un-
endowed with motility), the members of the colon
bacillus group may invade the digestive tract either
through the mouth or the anus, and, finally, that B.
bifidus, the chief obligate bacterium of the milk-stools
of nurslings and a regular and permanent inhabitant of
the meconium, reaches the intestinal tract only by way
of the anus. It is noticeable that the meconium contains
a considerable proportion of spore-forming bacteria;
that is, of bacteria capable of resisting unfavorable in-
fluences. Whether they find their way into the intes-

INFECTIONS OF THE DIGESTIVE TRACT 59

tine as spores or in their vegetative form is not clear
and is probably not important. They do, however,
sporulate in the meconium, which may perhaps be inter-
preted as indicating that the meconium is a poor rather
than a good nutrient medium for these organisms, al-
though I do not feel sure that this interpretation is
absolutely safe. It appears that the meconium is in
general a poor medium for bacteria, and this fact ex-
plains satisfactorily its relative poverty in bacteria.
The poor nutrient character of the meconium may be
due in part to its inspissation and in part to its relatively
high content of bile acids, which would tend to prevent
the rapid multiplication of invading organisms. It is
doubtless true that the meconium, in consequence of
these properties, exercises a selective action on the in-
vaders of the lower intestine, discouraging all those
vegetative forms that are non-resistant either through
the inability to adapt themselves to the peculiar con-
ditions of the medium or through the inability to form
spores. The meconium is therefore to be regarded as
possessing physiological functions instead of being an
indifferent substance. The main physiological function
of this meconial plug is apparently the exclusion of
numerous foreign types of bacteria and the encourage-
ment of those organisms which are destined to become
the permanent inhabitants of the nursling's digestive
tract.

Bacterial Flora of Bottle-fed Children. — If one makes
a comparison of the bacteria of the digestive tract of
children fed on cow's milk with the flora which has

60 INFECTIONS OF THE DIGESTIVE TRACT

already been described as characteristic of the digestive
tract in breast-fed children, one finds many points of
resemblance but also some typical and important differ-
ences. In general it may be said that the number of
bacterial forms present in the digestive tract of children
nourished with cow's milk is considerably greater than
in the case of breast-fed children. This is true even
where children are fed^on cow's milk that has been
sterilized by boiling or by pasteurization. And the
difference in number is even greater in those cases where
sterilization has not been practiced, since this procedure
destroys a large proportion of the bacteria present in
the milk and especially the lactic acid producers. Where
sterilized milk is employed the increase in the number of
bacteria in the digestive tract as compared with the
number of varieties present in the intestines of breast-
fed children is dependent at least in part upon the
presence of anaerobic bacteria or facultative varieties
capable of forming spores.

That the number of bacteria should ordinarily be
greater in children fed on cow's milk than in children
fed on mother's milk can cause no surprise, when one
remembers how abundant are the microorganisms present
in most specimens of cow's milk. The number ordi-
narily ranges from one or two thousand bacteria per cubic
centimeter * in cow's milk that has been collected with
the utmost precautions in regard to cleanliness"of pres-
ervation, up to several million per cubic centimeter.

1 Under very special conditions milk can be obtained which has
a far smaller number of bacteria.

INFECTIONS OF THE DIGESTIVE TRACT 61

Although the character of the bacteria is much more
important than their number, those specimens of milk
which contain the largest numbers are also most likely
to contain the forms which, aside from the inciters of
typhoid fever and dysentery, are most injurious; that
is, especially, the streptococci and the putrefactive anae-
robes. Ordinarily good milk frequently contains from
thirty to one hundred thousand bacteria per cubic cen-
timeter. Thus we see that there is good reason why the
mouth and stomach are liable to contain much larger
numbers of bacteria in the case of children fed on cow's
milk than in the case of breast-fed children, and it is
further clear that the actual number will vary somewhat
with the character of the milk that is employed. As in
the case of breast-fed children, there is a marked dimi-
nution in the number of bacteria found in the upper part
of the small intestine as compared with the bacteria in
the stomach. This diminution, which probably de-
pends in part on the bactericidal action of the gastric
juice and intestinal juices, tends to bring about in the
small intestine a greater equality in the number of bac-
teria in breast-fed and bottle-fed children than is ob-
served at higher levels. Nevertheless, even in the lower
part of the small intestine and throughout the large
intestine, the number of bacterial species appears to be
as a rule greater in the case of children fed on cow's milk
than in the case of nurslings.

Many of the bacterial forms which have already been
described as regular inhabitants of the nursling's intes-
tinal tract are also inhabitants of the digestive tract of

62 INFECTIONS OF THE DIGESTIVE TRACT

bottle-fed children. We find especially that this resem-
blance holds true with regard to the B. lactis aerogenes
group, the B. coli group, the B. acidophilus group, and
the group of organisms which we have come to know as
B. bifidus. As regards B. bifidus, however, we find that
there is a striking difference between the nursling and the
bottle-fed child. In nurslings the B. bifidus is the domi-
nant type throughout the greater part of the large intes-
tine; in the bottle-fed child the group of B. bifidus is
fairly well represented, as a rule, throughout the large
intestine and also in the lower part of the small intestine,
but it is by no means the dominant organism. The place
of preponderance is, in this case, occupied by organisms
of the B. coli type, and we thus find that the microscopi-
cal picture in the case of the bottle-fed child shows us a
field in which a majority of the microorganisms present
are Gram-negative instead of Gram-positive, as is the
case in nurslings. In addition to these leading types
of organisms in the case of bottle-fed children we have to
include certain positive and negative diplococci, some
of which are seen in chains while others appear as di-
plococci only. An organism described by Kruse,1 and
known as the Streptococcus lacticus, is almost always
present and appears to be identical with the streptococ-
cus of Hirsh-Libbmann. It possesses the power of
coagulating milk. Among the less constant aerobic
bacteria are the white staphylococci. Sometimes a

1 " Das Verhaltnis der Milchsaurebakterien zum Streptococcus
lanceolatus (Pneumoniecoccus, Enterococcus,u.s.w.)," Centralbl. f.
Bakt., 1st Abt., Orig., xxxiv, p. 737, 1903.

INFECTIONS OF THE DIGESTIVE TRACT 63

prominent inhabitant is B. cloacce, which is a more active
gas former than B. coli. In fermentation tubes contain-
ing milk it may induce an active gas production and
disruption of the coagulum which suggest the action of
B. aerogenes capsulatus. Sarcinse are often found, both
in the small and large intestine, and white yeasts are
occasionally present. As was first pointed out by
Rodella,1 there are several forms of bacteria present in
the large intestine which are capable of dissolving and
peptonizing casein. These forms, which are sometimes
facultative anaerobes, sometimes obligate anaerobes, are
usually distinctly more abundant in bottle-fed children
than in nurslings. Staphylococcus pyogenes albus is one
of the peptonizing forms bacteria often present and
possesses the ability to liquefy both casein and gelatin.
Probably in the case of bottle-fed children this organism
is less important as a liquefier than some of the faculta-
tive and strict anaerobic bacilli.2

1 " Ueber die Bedeutung der im Sauglingsstuhle vorkommenden
Mikroorganismen mit besonderer Beriicksichtigung der anaeroben
Bakterien," Zeitschr. f, Hyg., xli, p. 470, 1902.

2 The facultative and strict anaerobes are much more abundant
in the large intestines of bottle-fed children than in the intestines
of nurslings. I attach especial importance to the presence of B.
aerogenes capsulatus which, as already mentioned, is present in
only small numbers in the case of breast-fed children. The number
of these organisms is not necessarily larger in the intestines of
children fed on cow's milk, but as a rule even in children con-
sidered quite normal, their number, though still small, is greater
than in the case of nurslings. As compared with the number pres-
ent in the faeces of adults these bacteria are few. Another an-
aerobic organism which is sometimes found is the motile butyric
acid bacillus of Schottenfroh and Grassberger. Furthermore, B.
putrificus can be found in some instances in the stools of bottle-

64 INFECTIONS OF THE DIGESTIVE TRACT

Products of Decomposition in the Intestinal Tract of
Bottle-fed Children. — It has already been pointed out
that the products of intestinal decomposition in normal
nurslings are remarkably small in amount when we
consider the large numbers of bacteria that inhabit the
lower part of the intestinal tract. Almost the same thing
holds true of the intestinal tract of bottle-fed children.
If we make extracts of the contents of any portion of the
large intestine from a normal, bottle-fed child, we find
by the most delicate methods merely a trace of indol,
or even no trace at all. Traces of skatol are much less
frequent than traces of indol. Corresponding to the
fact that indol is in health always slight in amount and
frequently absent, is the observation that aqueous
extracts from any part of the contents of the krge in^
testine give no reaction with an acid solution of dimeth-
ylamidobenzaldehyde. It will be shown when dealing

fed children, but their number is usually small and their presence
inconstant. The bacillus of malignant O3dema has also been found
in the stools of children fed on cow's milk, but it is likewise incon-
stant, and I believe ordinarily of little physiological importance.
The anaerobes just mentioned — the motile butyric acid bacillus,
B. aerogenes capsulatus , B. putrificus, and the bacillus of malignant
03dema, are all spore-producers, although they do not produce
their spores with equal readiness. B. aerogenes capsulatus indeed
sporulates only under special conditions and probably seldom in
the intestinal tract of a normal bottle-fed child. The number of
spores present in the faeces of bottle-fed children who may be re-
garded as normal is ordinarily very small as compared with the
numbers found in some conditions of disease. The character of
the spores formed is at present uncertain; that is to say, it is
difficult to decide in the case of free spores to what organisms they
belong. Where sporulating bacilli are present, they are probably
not members of the group of B. aerogenes capsulatus.

INFECTIONS OF THE DIGESTIVE TRACT 65

with the contents of the digestive tract in adults that a
strong red coloration with this aldehyde is commonly
obtained in persons in fair health and that a moderate
reaction is usual in persons who are well. The reaction
depends in part on the presence of indol and perhaps in
part also on the presence of urobilinogen. In the case
of a bottle-fed child, there is usually only a faint pink
tinge when the conditions of the reaction are such as will
give a distinct red in the case of material from adults.
On standing, however, for several days, a blue-gray
coloration may set in, the nature of which is not at
present clear.

If we acidify a watery suspension made from any por-
tion of the contents of the intestinal tract, we find that a
moderate amount of volatile acid is obtained in the dis-
tillate. The molecular weight of such acids lies usually
between sixty and sixty-five; that is to say, it corre-
sponds closely to the molecular weight for acetic acid.
There is no doubt that acetic acid forms by far the larger
amount of the acids normally present in the distillate
prepared as just mentioned. The presence of acetic acid
rather than of higher fatty acids indicates that such
bacterial processes of decomposition as occur within the
intestinal tract are of a fermentative rather than of a
putrefactive nature. It is a fact of interest that if we
take the intestinal contents of a normal horse and make
a distillation for the determination of the volatile acids
present, this volatile acid also corresponds in molecular
weight to the figures corresponding to acetic acid. The
same thing is true of young pigs. These facts, and others

66 INFECTIONS OF THE DIGESTIVE TRACT

which might be brought forward, plainly point to the
presence of acetic acid fermentation as a normal process
in many healthy animals. If the material employed be
distilled after the addition of alkali instead of acid, one
obtains the volatile alkali; namely, ammonia. The
quantity of ammonia found is moderate and about suf-
fices to neutralize the acetic acid which is present.

Action of the Mixed Fcecal Flora upon Various Media.
- As already mentioned in describing the action of the
mixed faecal flora from the nursling, the products of
bacterial action are determined in important ways by
the nature of the medium, and both their quality and
quantity are influenced in important ways by the readi-
ness with which bacteria grow upon the various media.
If we grow the mixed flora from a normal bottle-fed child
upon sugar bouillon, there is always a production of gas
hi the fermentation tube, and this production of gas is
usually greatest in lactose bouillon. A less production of
gas is seen in dextrose and levulose and the slightest is
observed in saccharose. By making strongly alkaline
with sodium hydroxide one may measure the amount of
carbon dioxide that has been formed, and such observa-
tions have shown that from one-third to two-thirds of
the gas produced under these circumstances is carbon
dioxide. The composition of the remaining gas is not
yet definitely known, but it consists mainly of hydrogen
and probably in part of methane. It will be found
on using litmus paper that there has been a production
of acid at the same time that gas was produced.
Acetone is not usually detectable in sugar-bouillon

INFECTIONS OF THE DIGESTIVE TRACT 67

cultures. If the sediments in fermentation tubes
prepared in the manner just mentioned be examined
microscopically, it will be found in every case that there
has been a fairly abundant growth of microorganisms
of the B. coli type. The staphylococci and Gram-
positive diplococci will also be found to have multiplied
in the anaerobic limb of the tube. As a rule there is
little or no evidence of the development of strict
anaerobes.

The gas production and the acid formation which have
been noted are probably to be ascribed almost exclusively
to the action of organisms of the B. coli group. The
presence of streptococcal forms in abundance appears
to inhibit the formation of gas, probably by checking the
development of the colon bacilli.

If suspensions of the mixed flora be inoculated into the
fermentation tubes containing bouillon without sugar,
the results of the bacterial decomposition are different.
Here we find either very small gas production or none
whatever, and such acids as may be produced are, more-
over, bound by alkali, so that the medium, originally
neutral, either remains so or becomes faintly alkaline.
There is a moderate production of ammonia. This
ammonia is united mainly to acetic acid. There is also
a production of hydrogen sulphide under these circum-
stances. Not rarely there is a small quantity of mercap-
tan, but as a rule the amount of mercaptan produced is
little more than a trace, and very frequently — perhaps
usually — it is wholly absent. The formation of indol
is a constant phenomenon on the bouillon medium, and

68 INFECTIONS OF THE DIGESTIVE TRACT

in the course of twenty-four or forty-eight hours a moder-
ate quantity is usually formed. Skatol has not been
observed, but traces of phenol were regularly found.

Although the putrefactive decompositions in the
intestinal tract are normally very limited in both nurs-
lings and bottle-fed children, they are often somewhat
higher, as the foregoing pages indicate, in the case of
breast-fed children. The reasons for this are not en-
tirely clear.

The urine of healthy bottle-fed children does not
differ from that of nurslings in any essential particulars.
The ethereal sulphates perhaps tend to be excreted in
rather larger amounts in proportion to the total nitro-
gen and preformed sulphates. Indican is usually absent
or present only in traces.

THE BACTERIAL CONDITIONS IN THE DIGEST-
IVE TRACT DURING CHILDHOOD, ADOLES-
CENCE, ADULT LIFE, AND SENESCENCE

THE bacterial conditions in the digestive tract between
the period of childhood and that of old age differ in
health so considerably from the conditions that exist
during the period of milk feeding that they call for sepa-
rate consideration. The difference depends mainly on
the character of the diet, which grows more varied at
the end of the milk period and stays so throughout the
remainder of life. With the more varied diet comes
increased opportunities for the entry of many sorts of
bacteria into the digestive tract. It is impossible to
briefly picture the bacterial conditions in the digestive
tract in such a way as to take account of the many
individual peculiarities which are met, but fairly typical
descriptions are possible. It must be distinctly under-
stood that these descriptions are based on well-cared-for
individuals and not on the study of neglected persons
or persons following peculiar occupations which subject
them constantly to irregular conditions of life.

Period of Childhood and Adolescence. — The bacterial
conditions in childhood * and during adolescence differ
somewhat from those of later life, and it is essential to
recognize these differences in any discussion of such
conditions. During normal childhood and adolescence

70 INFECTIONS OF THE DIGESTIVE TRACT

all the bacteria described as present in the tract during
the period of bottle-feeding are likely to be still present,
for the use of cow's milk is generally continued into
the epoch of a mixed diet. Microorganisms of the
B. bifidus type are usually much less numerous in the
intestine in childhood and adolescence than in the cow's-
milk period, but other types of bacteria are commonly
more numerous. This is true of the putrefactive anae-
robes, certainly of B. aerogenes capsulatus and probably
of B. putrificus.1 Still the number of these anaerobes
is small2 and putrefactive processes in the intestine

1 An organism corresponding to Nencki's B. liquefaciens ilei
is usually abundant in the lower ileum, ascending colon, and per-
haps the caecum. I have not yet had an opportunity to study the
biochemical characters of this organism, but suspect that it may
be a vegetative form of B. putrificus.

2 The following notes relate to typical microscopical fields pre-
pared from the digestive tract of a boy aged fourteen years : —

CASE OF SUDDEN DEATH, MALE, MT. 14

Autopsy showed heart dilated, aorta 1.5 cm. in diameter;
stomach intensely congested; intestines normal. The skull was
not opened. There was no history of infectious disease to ex-
plain the cardiac dilatation, and no endocarditis. No cause for
this sudden death was revealed by the autopsy findings. The
slides were made two hours after death.

Microscopical Findings

I. Stomach. Very few bacteria are seen. There are small
numbers of positive diplococci, sometimes in chains, all medium-
sized. A few short, negative bacilli and a few small, negative
diplococci are also seen.

II. Duodenum. There are a few positive diplococci; other-
wise no bacteria. Considerable masses of epithelial cells and
very few small negative diplococci are found.

III. Jejunum. A few medium-sized, positive diplococci and

INFECTIONS OF THE DIGESTIVE TRACT 71

are not active. This is shown by the character of the
urine and by the nature of the decomposition products

negative cocci, ill-defined as regards staining, are seen. Epithe-
lial cells are numerous. The number of bacteria here is extremely
small.

IV. Ileum. Bacteria are more numerous here, but the number
of varieties is small. There are a few forms resembling B. coli,
short, stout, negative bacilli; also bacilli of about the same
width and three to four or five times the length of typical coli;
also a medium-sized or rather large, positive diplococcus and a
few small, positive bacilli about the size of colon bacilli. The
long, slender, Gram-negative bacillus is rather prominent. Some-
times it takes the form of a fairly long thread. Considerable
epithelium is present in this section.

V. Caecum. This field presents an entirely different picture
from the preceding. The bacteria are numerous. The field is
mixed about evenly as regards the staining. Negative bacteria
consist of three forms — small diplococci which are moderately
numerous; short bacilli, suggesting colon bacilli in the typical
form; slender, negative bacilli, more slender than colon and two
or three times as long, sometimes in threads. The positive forms
consist of oval diplococci medium-sized, occasionally large, single,
oval cocco-bacilli, a few forms suggesting colon bacilli (but these
are not numerous) and a very few organisms the size and form of
B. aerogenes capsulatus, though tending to be a little small for this
organism ; and a few positive bacilli of this type containing spores
located between the middle and end. No free spores are observed.
The rather long, stout negative bacilli observed at higher levels
are also seen occasionally. There are a few slender, negative
bacilli showing irregularities in staining; that is, "punctate"
bacilli.

VI. Sigmoid flexure. Bacteria are numerous. The negative
forms are more prominent than in the last level. Organisms of
the colon type are numerous, as are also slender, negative bacteria.
Positive diplococci are considerably less numerous than in higher
levels. There are moderate numbers of organisms suggesting
capsulatus, but their number is, however, very small in comparison
with the total number of bacteria. Spore-holders are not seen,
nor are free spores.

VII. Rectum. The fields are mixed. Colon bacilli are numerous,

72 INFECTIONS OF THE DIGESTIVE TRACT

in the faeces. Indol and phenol are found in the faeces
in only very small quantities — a few milligrams in
one hundred grams of the moist faeces. Frequently,
indeed, only a trace of indol is detectable in the intestinal
contents. The character of the urine as regards putre-
factive products corresponds to the faeces. The ethereal
sulphates are low in proportion to the preformed, and
there is little or no indican in the urine. There is also
little phenol. The reaction with dime thy lamidobenz-
aldehyde is slight or moderate. Frequently it is so
slight that its existence is questionable. During a
period of temporary digestive derangement the ethereal
sulphates may show a considerable increase, and indican
may be present, but as the derangement subsides the
urinary indications of putrefactive decomposition in
the intestine likewise grow less and gradually fall back
to the previously existing normal level.

Period of Adult Life. — During adult life factors
usually enter into the lives of human beings which are
not without influence upon the bacterial conditions
that prevail in the stomach and intestine. Great
differences exist in the habits of different persons at
this time of life, and these differences are in a degree
reflected hi the nature of the bacterial processes that

as are also slender, negative bacilli. There are moderate numbers
of positive diplococci and diplobacilli. Organisms of the capsulatus
type are a little more numerous than at the last level. There are
a very few free, oval spores, and a few spore-holding bacilli, very
much enlarged, suggesting clostridia. There are also some few
slender Gram-positive bacilli about the length of B. aerogenes
capsidatas or longer, but about one-half the diameter of capsulatus.

INFECTIONS OF THE DIGESTIVE TRACT 73

hold sway in the digestive tract. In adult life the
individual experiences new responsibilities, new ambi-
tions, new dangers, an enhanced emotional life, and in
very many instances a marked change in the direction
of more sedentary habits incidental to a larger proportion
of indoor life. The dietary is apt to undergo an altera-
tion in the direction of increased and frequently injudi-
cious liberty and the use of tea and coffee, also the use of
tobacco and alcoholic drinks is either increased or begun.
For a time these influences may not make themselves
definitely felt, but sooner or later they lead to slight
derangements of digestion which manifest themselves
clinically. There are of course the greatest individual
differences in respect to these manifestations. One
sometimes meets with persons more than fifty years
of age in whom the bacterial conditions in the intes-
tine do not differ materially from those observed dur-
ing the period of normal adolescence. These persons
are usually well nourished, muscularly strong, and capa-
ble of sustained mental and muscular work. They have
a high percentage of hemoglobin and their red blood
cells approximate the upper physiological limit in regard
to number. The volume of blood is large. These
persons are remarkably free from nervous disturbances,
even under trying conditions of work. They generally
have at their disposal a surprising amount of reserve
energy.

These are, however, exceptional instances of robust
health. A large portion of the population in the United
States presents different physical conditions by the time

74 INFECTIONS OF THE DIGESTIVE TRACT

they reach the fiftieth year of life, although they are
in no sense in a state of invalidism. They are able to do
large amounts of work without excessive fatigue and
most of the time they feel well. While in such persons
the faecal flora shows nothing striking, it is usually not
difficult to demonstrate that the number of putrefactive
anaerobes in the intestine is larger than is the case with
healthy individuals in the period of adolescence. The
presence of B. aerogenes capsulatus can be determined
by anaerobic plating in blood-agar, and the number of
colonies will be found ordinarily to exceed the number
obtainable from the faeces in the earlier period of life.
B. putrificus may also be moderately abundant. The
colon bacilli are well represented. The urine shows the
presence of a slightly larger amount of ethereal sulphates
than is ordinarily found in earlier life. The ratio of
ethereal to preformed sulphates varies from 1 : 8 to
1 : 12 most of the time. A little indican is of frequent
occurrence, and not rarely the reaction is strong with
Obermeyer's reagent — not regularly strong, perhaps,
but frequently so for a few days at a time. The phenol
excretion may also be a little greater than that noted
during earlier life.

In short, then, we find in middle life a large number of
persons whose health is good or fair, but in whom the
putrefactive processes in the intestine are distinctly
more active than is the case with most younger persons
who are representative of normal health. These persons,
though hi good health, are not robust. A period of sus-
tained hard work is followed by considerable and per-

INFECTIONS OF THE DIGESTIVE TRACT 75

haps annoying mental and physical fatigue. Moreover,
these persons have found by experience that they must
be more careful than formerly in respect to food and
drink, emotional and sexual excitement, etc. Dining
out and the use of alcoholic drinks are indulgences
quickly followed by unpleasant consequences. Physical
exercise out of doors becomes more and more a necessity
to this group of individuals. They are conscious that
it requires careful living to keep them in a condition
compatible with the performance of their duties.

Period of Senescence. — The age of an individual
must be measured rather by the physiological potential
of his cells than by the number of his years. There are
men who at seventy have cells with functional capacities
superior to those of other men who are little beyond
forty, and who show their superiority in the ability to
work without fatigue, to digest without any conscious-
ness of the digestive processes, and to make large out-
puts of mental and muscular energy without ill effects.
These persons retain soft arteries, are well nourished, and
exhibit little atrophy of the subcutaneous areolar tissues,
and hence show little wrinkling of the skin. They are,
in short, candidates for an advanced age.

If we examine the intestinal bacteria and the urine of
such people, we find conditions wholly in harmony with
the unusual preservation of general functional powers
and with the freedom from signs of disordered digestion.
The faeces contain an abundance of viable bacilli of the
B. coli group and the putrefactive anaerobes are few
in number. Analyses show the presence of mere traces

76 INFECTIONS OF THE DIGESTIVE TRACT

of indol in the intestinal contents (perhaps one or two
milligrams in one hundred grams faeces) a few milligrams
of phenol, and no skatol. The urine is free from indican
almost habitually, and only moderate quantities of
phenol are excreted (twenty to forty milligrams in
twenty-four hours). The ethereal sulphates are corre-
spondingly low. The dimethylamidobenzaldehyde re-
action is feeble or moderate in intensity.

Persons of this type are apt to die at an advanced age
of some condition apparently quite distinct from disease
of the digestive tract, especially thrombosis of cerebral
vessels (from atheroma) or chronic myocarditis.

I have not had a wide experience in the careful study
of the bacterial conditions in persons above sixty-five
years of age, but believe it safe to say that a large majority
of such persons (perhaps seventy per cent, of the popula-
tion in the United 'States above this age) give evidence
of distinct putrefactive processes in the digestive tract.
These processes are in many instances characterized by
their mildness. But if we compare these mild processes
of putrefaction with those existing in the majority of
persons under twenty years of age, we find a distinct
difference between the two. I think it quite clear that
the conditions in youth are much more close to the ideal
physiological state of infancy and childhood than are
those of senility. The difference lies mainly in the
direction of the greater abundance of putrefactive
bacteria in old age, but with this there is apt to be some
diminution in the number of typical colon bacilli, if
we may form a judgment on this point from the appear-

INFECTIONS OF THE DIGESTIVE TRACT 77

ance of the microscopical fields and the gas production
in the fermentation tubes. These conditions of bacterial
activity in the period of senility are like those which
were just described a& being common in people about the
age of fifty. The main difference between the putrefac-
tive conditions found at fifty and at seventy is that at
the latter period they are a little more marked in their
intensity and affect a much larger proportion of the pop-
ulation. The subjects in question at this later period
of life are not ill, but in order to keep fairly well have to
be very careful as to their habits of living. They are
moderately anaemic and easily develop slight disorders
of digestion. They weigh less than formerly, and though
they may still be well nourished in appearance, are con-
scious of losing strength from year to year. They are
undergoing what is usually regarded as normal involution.
Thus it is clear that there exists a distinct difference
between normal childhood and adolescence on the one
hand and normal old age on the other in respect to the
intensity of the putrefactive processes that go on in the
digestive tract. The origin and precise significance of
this difference are at present not clear. It appears
likely that the tendency to an increasing degree of putre-
factive decomposition in the intestine is connected with
the repeated but not necessarily severe derangements
of intestinal function that are experienced from time to
time by most individuals. In the course of these acute
derangements, there is frequently an increased opportu-
nity for the development of putrefactive anaerobes, and
it is possible that owing to slight but persistent damage

r& INFECTIONS OF THE DIGESTIVE TRACT

to the epithelial and other structures of the intestinal
mucous membrane, these anaerobes are gradually afforded
more and more favorable opportunities for their growth.

As to the relation between the relatively active anaerobic
life in the intestine during senility and the development
of the involutional alterations in the tissues, it is difficult
to formulate an opinion. Many factors may doubtless
enter into the production of these cellular alterations, and
it is difficult to assign to each of these its just position.
One statement may, however, be confidently made:
the onset of senility may be distinctly accelerated through
the development of intestinal infections in which the
putrefactive anaerobes are prominently represented.
I have observed this in cases where it has appeared to
me a certainty that other toxic causes of premature
senility could be excluded.

The view that intestinal infections stand in a causative
relation to old age is by no means new. It has been
advanced by Metchnikoff on several occasions. It is
probable that the "wild races" of bacteria of which
he speaks as responsible for senile changes consist largely
of putrefactive microorganisms. I do not know that
Metchnikoff has singled out any definite group of micro-
organisms as especially concerned. I am inclined to give
prominence to B. aerogenes capsulatus as the most impor-
tant factor in the production of the putrefactive decom-
positions of advanced age. It is extremely probable
that persons with apparently equal numbers of gas-
bacilli in the intestinal tract do not necessarily suffer an
equal degree of harmful influence from their putrefactive

INFECTIONS OF THE DIGESTIVE TRACT 79

products. Some races of the gas-bacillus are much less
pathogenic for animals than others, and these differences
in pathogenicity may be of considerable significance to
human beings harboring them.

CHARACTERS OF THE BACTERIAL FLORA OF
CARNIVOROUS AND OF HERBIVOROUS ANI-
MALS

IN the course of the study of anaerobes of the human
intestine it appeared desirable to learn something about
the characters of the bacterial flora inhabiting the large
intestine of various domestic and wild animals. It
was noticed that in the dog, which is frequently exclu-
sively carnivorous, the intestinal contents often showed
the presence of large numbers of spores, spore-bearing
bacilli, and vegetative forms of anaerobes. The numbers
present in the faeces were noted to be especially large in
some animals which had been exclusively fed on meat.
A study of a grown cat fed upon raw meat showed the
presence of Gram-positive vegetative anaerobes from
one end of the digestive tract to the other. Flora derived
from the stomach, small intestine, and large intestine
were inoculated and grown in bouillon flasks and showed
an abundant production of methyl mercaptan as well
as of hydrogen sulphide. The numbers of colon bacilli
present in this case were relatively small as compared
with the anaerobes. The study of the colonies obtained
on anaerobic plates showed that a large portion of the
organisms present in the intestinal tract were B. aero-
genes capsulatus. Intravenous infusion of these organisms

80

INFECTIONS OF THE DIGESTIVE TRACT 81

into a rabbit which was afterwards killed and incubated
showed in a high degree the typical gas formation.1

Observations on other cats showed the presence of
considerable numbers of spore-holding bacilli and free
spores, sometimes in chains, in addition to vegetative
forms of anaerobes. The position of these spores and
spore-holding bacilli has not been established in a bacte-
riological sense. Observations were also made upon the
intestinal contents of the wolf, tiger, and lion. Several
different tigers were studied, and the observations were
not confined to the examination of one lion and one wolf.
The material from the lion showed the presence of many
free spores. It also showed the presence of considerable
numbers of Gram-positive bacilli, suggesting B. aero-
genes capsulatus. Gram-stained preparations from wolves
showed pictures similar to those observed in the lion
except that the spore-holding bacilli were more numerous.
The findings in the case of supposedly healthy tigers
were not essentially different from those in the case of
the wolf and lion. In the case of one tiger, suffering
from osteomalacia, greatly impaired nutrition, and loss
of strength, the microscopical fields derived from several
different samples of faeces revealed the presence of im-
mense numbers of free spores and smaller numbers of
immature Gram-negative spore-holding bacilli. These
spores developed into organisms which possessed all
the generally known cultural and biochemical characters
of B. aerogenes capsulatus, including the ability to develop

JThe method by which these incubation experiments were
carried out is explained on p. 84.

82 INFECTIONS OF THE DIGESTIVE TRACT

a high grade of gas formation in rabbits injected and
incubated.

It was found that bouillon cultures of the mixed faecal
flora from the lion, tiger, wolf, and cat all developed
quickly a sufficient quantity of methyl mercaptan to give
a prompt, strong reaction with isatin in sulphuric acid.

Experiments were made with the mixed fsecal flora
from these carnivorous animals to determine their patho-
genicity when injected into the subcutaneous connective
tissue. It would have been better to have worked
with pure cultures of the anaerobes in question, but
opportunity has not yet arisen to isolate them. The
result of the inoculations into guinea-pigs was the same
in each instance. The animals died within twenty-
four hours and usually in fifteen to eighteen hours.
At autopsy the subcutaneous connective tissues were
haemorrhagic, oedematous, and showed necrotic changes
which extended in some instances to the muscles. Gas
formation was not usually noted as a prominent feature.
These pathological alterations were not confined to the
site of inoculation, but had extended to the subcutaneous
connective tissues throughout the body and were espe-
cially pronounced in the axillae and in the groin. It is
unnecessary to enter here into the details as to the
character of the organisms recovered from these lesions.

We may contrast with these findings the observations
made upon herbivorous animals, including the buffalo,
goat, horse, elephant, and camel. In the case of the
camel, elephant, and horse the preponderant bacteria
in the Gram-stained fields were small, Gram-negative

INFECTIONS OF THE DIGESTIVE TRACT 83

organisms which were regarded as special forms of B.
coli. In the case of the goat the fields contained some
Gram-positive bacteria, and of the Gram-negative ones
a considerable number were of greater length than the
dominant small forms which were regarded as belonging
in the class of colon bacilli. In the case of the buffalo,
mixed fields were found as regards the Gram-staining
and many of the positive organisms were found to be
small diplococci and small bacilli. In none of these
animals were seen any organisms suggesting B. aerogenes
capsulatus, excepting in the case of the buffalo, where
the number of bacilli of this type was very small. Spore-
holding organisms were not observed, but moderate
numbers of free spores were noticed in all the fields except
those from the elephant. In the fields showing the
largest number of spores their occurrence was far less
frequent than in the lion, tiger, wolf, or cat.

The mixed flora of these different herbivorous animals,
grown upon peptone bouillon, failed to show the produc-
tion of methyl mercaptan excepting in the case of the
horse, where a moderate reaction was obtained.

Observations were also made upon the effect of sus-
pensions of the mixed flora from herbivorous animals
when injected subcutaneously. The quantities of
suspension employed were usually about twice as great
as in the case of the suspensions from the carnivorous
animals. With the exception of the suspensions obtained
from the horse, the pathogenicity of these suspensions
was found to be slight, the guinea-pigs frequently
living two or three days or entirely recovering. In the

84 INFECTIONS OF THE DIGESTIVE TRACT

animals injected with faeces from the horse were found
hsemorrhagic and cedematous lesions with necrosis,
similar to those found in the carnivorous animals.
These lesions were, however, less pronounced than hi the
case of the suspensions from the carnivorous animals.
In the case of the elephant a considerable quantity
of fibrinous exudate was found about the point of in-
oculation. No oedema or necrotic change was observed
in the subcutaneous tissues.

A further confirmation of the radical differences
existing in the intestinal tracts of carnivora and her-
bivora is furnished by a series of observations with the
Welch-Nut tall incubation test. Suspensions were made
from the fseces of all the types of animals mentioned,
and equal quantities of these suspensions were infused
intravenously into a series of living rabbits. The rabbits
were then quickly killed and incubated. On examination
after twenty-four hours it was found that all the rabbits
infused with suspensions from carnivora showed in an
extreme degree the characteristic putrefactive changes
hi the liver, cellular tissues, etc., induced by pure cul-
tures of B. aerogenes capsulatus or of the bacillus of
symptomatic anthrax. The rabbits infused with sus-
pensions made from the fseces of the herbivora showed
similar but very much slighter changes in each case.
The results for each group of animals separated the her-
bivora sharply from the carnivora. Examination of the
livers showed the number of bacteria hi the carnivorous
series to be many times greater than in the herbivorous
series. The microorganisms were regarded as being

INFECTIONS OF THE DIGESTIVE TRACT 85

almost certainly B. aerogenes capsulatus on account of
their morphology and failure to sporulate. The bacilli
of symptomatic anthrax readily sporulate in the incu-
bated rabbits. The gas-bacillus (B. aerogenes capsulatus)
does not sporulate under these circumstances.

These differences in the appearance and behavior
of the bacteria derived from typical carnivora and
herbivora suggest that the habit of living upon a diet
consisting exclusively of raw meat entails differences
in the types of bacteria that characterize the contents
of the large intestine. The occurrence of considerable
numbers of spore-bearing organisms in" the carnivora
points to the presence of anaerobic putrefactive forms
in great numbers. The results of subcutaneous inocula-
tions into guinea-pigs bear out this view and indicate
that the numbers of organisms capable of producing a
hsemorrhagic oedema with tissue necrosis, with or without
gas production, are very considerable. Unfortunately
the data pertaining to the biological properties of these
pathogenic anaerobes are at present insufficient to permit
us to classify them or to say more of their nature than
that they are organisms representative of a definite
group of putrefactive anaerobes which make butyric
acid and hydrogen and exert a peptonizing action upon
living tissues. Nevertheless the observations here re-
corded are of much interest in relation to the bacterial
processes and nutrition of herbivorous * as distinguished
from carnivorous animals, and are significant furthermore

1 Many of the herbivora yielded mixed flora incapable of making
gas on dextrose bouillon.

86 INFECTIONS OF THE DIGESTIVE TRACT

for the interpretation of bacterial conditions found in
man. The question arises whether the abundant use of
meat over a long period of time may not favor the devel-
opment of much larger numbers of spore-bearing putre-
factive anaerobes in the intestinal tract than would be the
case were a different type of proteid substituted for meat.
Inquiries made of Dr. Blair, the pathologist in the New
York Zoological Gardens, elicited the fact that while,
upon the whole, the carnivorous animals are apt to live
somewhat longer than the herbivorous animals of about
equal size, the carnivora are much more likely to develop
conditions of advanced anaemia in the later years of their
lives than is the case with the herbivora. Dr. Blair
states that it is usual in the later years of life for the car-
nivora to show a much diminished volume of blood and
at least a moderate fall in the haemoglobin. Instances
are stated to be not uncommon in which a pernicious
type of anaemia has developed in the carnivora. On
the contrary, among the herbivora it is said that pro-
nounced anaemias are vejy occasional. The examples
of severe anaemia encountered among the herbivora
were said by Dr. Blair to be in nearly all instances
referable to gross animal parasites.

INFLUENCE OF FOOD ON HUMAN BACTERIAL FLORA OF
THE DIGESTIVE TRACT

Surprisingly little is known about the influence of
different classes of food upon the nature of the micro-
organisms in the digestive tract. It is evident that
this subject is one that deserves the most careful kind

INFECTIONS OF THE DIGESTIVE TRACT 87

of study by means of modern methods. That a knowl-
edge of the influence of different foods upon the flora
in health and disease would not only be of great bio-
logical interest but would also give many indications
of a practical sort for the use of different types of
foods in pathological conditions, requires no argument.
Some observations have been recorded upon the influ-
ence of food which have led to the conclusion that there
is little difference between a, diet consisting of vegetable
food on the one hand and a mixed diet with abundance
of meat on the other. The methods employed for de-
termining the presence of differences in the character
of the flora were, however, so crude that the work in
question cannot be regarded as in any manner conclusive.
I have observed that the number of Gram-positive
organisms in the faecal fields was much increased when
an adult subject who had previously been on a mixed
diet began to live on a dietary consisting almost exclu-
sively of meat. The microscopical appearances indicated
that an increase in the number of putrefactive anaerobes
was largely responsible for the change from a mixed
faecal field to one which was dominantly Gram-positive.
In this connection I think the observation noteworthy
that the intestinal contents of animals living on a diet
of raw meat tend to give mixed or dominantly Gram-
positive fields, whereas similar material from herbivo-
rous animals tends to give fields in which Gram-negative
organisms are predominant. As mentioned in connection
with the discussion of the faecal flora in herbivorous
animals as compared with carnivorous animals, there is

88 INFECTIONS OF THE DIGESTIVE TRACT

a great difference in the numbers of anaerobes in the two
groups, the number of these organisms being greater
in the case of the carnivora.1 Some observations of
interest on the flora of dogs we owe to Lembke,2 who
found distinct differences between the flora after a diet
of bread and after one of meat. Of especial interest is
the fact that a bread diet rendered the faeces much
richer in anaerobes than did a diet of meat. The bacteria
observed upon the diet containing a great abundance of
fat resembled closely those noted on a bread diet. The
changes which are induced by alteration in the diet are
said by Lembke to be generally of a temporary sort.
On this subject he speaks hi the following way: "If
the diet is changed, there appear on the fsecal plates
new colonies of the most varied sort. In the course of a
few days these are materially reduced and the colonies
of B. coli again gain the upper hand." It appears that
with almost any change in diet new varieties of bacteria
are introduced in large numbers and somewhat obscure
the predominance of the colon bacilli in the fsecal
fields. Very soon, however, the colon bacilli regain their
original predominance, doubtless owing to the fact that
they are better able to adapt themselves to the nutrient
conditions than are the facultative forms which have
been introduced. Lembke states that if one limits the
entry of bacteria by sterilization of the diet, the foreign

1 Escherich found in a young dog whose milk-fseces flora was
very similar to that of normal nurslings, that after a pure meat
diet there was only an extremely small proportion of colon bacilli,
whereas there appeared large numbers of liquefying colonies.

2 "Beitrag zur Bacterienflora des Darms," Archiv f. Hyg., xxvi,
p. 325, 1896.

INFECTIONS OF THE DIGESTIVE TRACT 89

types of microorganisms after a time disappear and the
plates show exclusively colonies of the colon bacillus
and B. lactis aerogenes in almost pure culture. On the
other hand, there is good evidence that the Gram-nega-
tive bacilli of the f seces may be temporarily replaced by
feeding Gram-positive acidophile bacteria.

It appears probable that in considering the influence
of foods upon the flora of the intestinal tract one should
take into account .the factor of rapid digestion and
absorption in the upper part of the digestive tract.
For example, in cases where a patient takes daily a large
quantity of meat which is imperfectly masticated, there
is much more opportunity for the development of pu-
trefactive anaerobes in the lower part of the intestine
than if the same quantity of meat is thoroughly sub-
divided by mastication. I believe also that the influence
of diet must be largely modified by the character of the
dominant organisms in the intestinal tract and that this
influence may come to the front in a telling way in cases
of chronic infections of the large intestine.

THE REDUCING ACTION OF MEAT

In the study of the anaerobic conditions that prevail
in the digestive tract it is important to recognize every
factor that may enter into the production of anaerobic
conditions there. It seems extremely probable that
different articles of food have a different significance in
respect to their influence on the presence or absence of
oxygen in the digestive tract. It is known that fresh
tissues of animals exert a considerable degree of reduc-

90 INFECTIONS OF THE DIGESTIVE TRACT

ing power and that this reducing power is much more
active in general than the reducing action of vegetable
cells. If one places in a cylinder or flask a quantity of
finely divided fresh muscle of liver from a mammal and
immerses this material in a dilute solution of methylene
blue, under such conditions that oxygen from the air is
largely excluded from the region occupied by the divided
tissues, it is soon noticeable that the blue color begins to
disappear, owing to the conversion of methylene blue
into leucomethylene blue. The reducing action of fresh
liver has been successfully employed by Professor Theo-
bald Smith in rendering the closed arm of the fermen-
tation tube more strictly anaerobic and thus facilitating
the growth of* certain strictly anaerobic bacteria. With
these facts in mind one naturally asks whether the use
of large quantities of raw muscle may not considerably
aid in the production of anaerobic conditions in the di-
gestive tract. In the case of carnivorous animals living
on raw meat there seems little doubt that anaerobic
conditions may exist throughout the digestive tract, and
I think it probable that the reducing action of the meat
in the upper part of the tract may materially contribute
to diminish the quantity of oxygen carried into the di-
gestive tract. Even in the case of man this factor is one
which cannot be entirely ignored. Although human
beings in civilized countries for the most part eat their
meat after it has been subjected to cooking, there are
many instances in which large quantities of raw or nearly
raw meat are eaten. Moreover, meat that has been
slightly cooked still retains considerable reducing power.

INFECTIONS OF THE DIGESTIVE TRACT 91

How far the excessive use of meat that is raw or
slightly cooked may influence the production of anaerobic
states in the digestive tract it is at present impossible
to say, but it seems not unlikely that there are cases
of excessive intestinal putrefaction dependent on the
excessive activity of anaerobes hi which the conditions
of anaerobiosis are distinctly favored by excessive meat
eating.

THE INFLUENCE OF THE EPITHELIAL CELLS LINING
THE DIGESTIVE TRACT

We know at present little of the influence of changes
in the epithelia lining the digestive tract upon the
physiology and pathology of the human organism.
There are, however, certain general considerations that
require mention here. It may be safely assumed that
every cell possesses a certain life potentiality ; that is,
has an inherent capacity if undisturbed by injurious
agencies to live a certain period of time. Every
epithelial cell of the digestive tract doubtless pos-
sesses a high capacity for reproduction. This power
must, nevertheless, be a limited one, and if the cells of
the digestive tract be injured by too many demands
upon them, they may fail after a time to reproduce
normally. Superficial cells which have under these
circumstances undergone desquamation are thus not
so easily replaced. It may happen that in some portions
of the digestive tract the epithelial layer is much thinner
than normal, owing to excessive desquamation as com-
pared with their power of reproduction. For example,

92 INFECTIONS OF THE DIGESTIVE TRACT

we see in the case of Miiller's superficial glossitis (geo-
graphical tongue) a condition in which varying areas
of the tongue suffer a temporary or permanent loss of
the superficial epithelial layers. In many cases of chronic
disorder of digestion of very long standing, the papillae of
the entire dorsal surface of the tongue show excessive
desquamation of the epithelium. In how far similar
conditions occur in other parts of the digestive tract
we do not know, but there is reason for believing that in
some chronic derangements of digestion there is in the
stomach and in the intestinal tract a similar thinning
of the epithelial layer. How far conditions of the tongue
may be regarded as an index to the conditions in other
parts of the digestive tract is uncertain.

Of the pathological effects of excessive desquamation
of epithelium we have at present little definite knowledge.
It appears reasonable to suppose that where desquama-
tion is excessive without corresponding reproduction of
epithelial elements there must be diminished secretion,
provided the cells in question take any part whatever
in providing a secretion containing digestive enzymes.
It seems likely also that the process involved in transu-
dation may be abnormal under conditions in which
the epithelial layer is greatly thinned. There are some
patients in whom the tongue indicates an excessive
desquamation of the epithelium and in whom cathartics
such as salines no longer act promptly or efficiently.
It seems to me possible that the not uncommon appear-
ance of a loss of response to cathartics may be connected
with the partial failure in function of the epithelium of the

INFECTIONS OF THE DIGESTIVE TRACT 93

intestinal tract. The correctness or incorrectness of this
view can only be established by experimental methods.
Still another influence of diminished function in con-
sequence of a pathological thinning of the epithelial
layers may be found in a diminished capacity on the part
of the intestinal epithelium to act upon products of
decomposition in the intestine. In experiments made
many years ago with indol, it was found that the epithe-
lium of the digestive tract possesses in a high degree
the capacity to bind indol in such a way that this sub-
stance cannot be recovered by distillation. I think it
likely that in cases where there is excessive production
and absorption of indol in the intestinal tract the epithe-
lium acts as a protective agency to the organism as a
whole. This action of the epithelial cells is certainly
not confined to indol. It seems not unreasonable to
suppose that where the epithelium has been very much
thinned, the organism suffers from the enfeeblement of
this function incidental to the loss of epithelium. Here
again experimental methods should throw more light
upon this protective action of the intestinal epithelium.

THE PERMEABILITY OF THE MUCOUS MEMBRANE OF
THE INTESTINAL TRACT FOR BACTERIA

Another function of the epithelial cells of the digestive
tract has to do with the protection of the body from
the invasion by bacteria within the tract. There exists
some experimental evidence indicating that an intact,
fully developed layer of epithelium is an important
barrier to the entry of at least some kinds of bacteria

94 INFECTIONS OF THE DIGESTIVE TRACT

into the mucous membrane. How far the presence
of mucus on the surface of the epithelial layer is a
factor in rendering the mucous membrane impermeable
is still a question. It appears from the work of
Hilgermann 1 and others that the normal mucous
membrane is much more permeable to some bacteria
during the period of infancy than in later life. It was
found by Hilgermann that when young rabbits and
guinea-pigs are fed with tubercle bacilli, these organisms
pass through the wall of the stomach and penetrate
the digestive tract throughout its entire extent. It
was noticed that the number of bacteria passing through
the small intestine, especially in its upper third, was
considerable, whereas the number which penetrated the
wall of the large intestine was much smaller. In the
vermiform process the conditions for the passage of
tubercle bacilli appear to have been the same as in the
case of the small intestine.

Hilgermann attempted to determine what factors
are concerned in the passage of tubercle bacilli through
the digestive tract in animals. He was at first inclined
to regard the passage as accidental and due to small
lesions, but this view was abandoned because it was found
that the penetration did not occur at single points, but
throughout the length of the gastro-enteric tract. It
was found, moreover, that there is no evidence of pene-
tration occurring in corjsequence of an irritation by a
considerable number of bacteria acting locally. If this

x"Die Bakteriendurchlassigkeit der normalen Magendarm-
pchleimhaut im Saulingsalter," Archiv f. Hyg., liv, p. 335, 1905.

INFECTIONS OF THE DIGESTIVE TRACT 95

had really been the case, the bacteria must have passed
much more abundantly than they did, instead of in
the typical manner and in the definite stages observed
by Hilgermann, who is inclined to believe with Behring
that the mucous membrane in early life is lacking in
natural protective substances capable of hindering the
penetration of bacteria.

Of considerable interest in connection with the study
of the penetration of the digestive tract by bacteria
are the observations of Ficker * on the influence of
exhaustion and of hunger upon the passage of bacteria.
He calls attention to the fact that states of inanition
favor the occurrence of infections, especially those arising
from the intestinal tract, but that on the other hand
the best nourished and strongest organisms may show
a high degree of sensitiveness toward such infections.
The experiments of Ficker indicate that excessive
exertion is a factor which at times leads to an exhaustion
which is a predisposing condition to infection, especially
in the case of typhoid fever. This influence of exhaustion
on the development of infection has been to some ex-
tent studied by Charrin and Roget on white rats which
were fatigued in a rotating drum. These animals were
shown to be more sensitive to anthrax than the control
animals. Ficker selected dogs for his experiments, in-
ducing fatigue in them through work in a treadmill. He
found that a combination of fasting and fatigue facilitated
in an extraordinary way the passage of germs through

1 " Ueber den Einfluss des Hungers auf die Bakteriendurch-
lassigkeit des Intestinaltraktus," Archiv /. Hyg., liv, p. 354, 1905.

96 INFECTIONS OF THE DIGESTIVE TRACT

the intestinal tract. In a dog that had hungered eleven
days, a period of three hours of exercise sufficed to per-
mit the free passage of microorganisms. It appears that
neither of these factors is capable of giving rise to this
degree of permeability when operating separately. Indi-
vidual peculiarities were, however, observed in respect
to the permeability of the tract, and these must have
their explanation in special conditions in the intestine.

Ficker does not maintain that a full understanding
of the conditions favoring the permeability of the
intestinal tract has been reached through his experi-
ments. He holds very definitely that in fully grown
dogs, in which the intestinal tract is ordinarily little
permeable, it is possible through inanition or fatigue
or a combination of the two, to facilitate a penetration
of the intestinal mucous membrane which is analogous
to that observed in the infantile tract and further re-
sembles the conditions present in the dying organism.
Ficker suggests that under the influence of powerful
bodily action the volume of the gastric juice and of
other digestive juices is greatly reduced and that peri-
stalsis is diminished, while on the other hand the lymph
and blood flow are actively accelerated, all these being
factors which might favor the penetration of bacteria.
He suggests that perhaps the increased leucocytosis
induced by active muscular exertion is responsible for
the increased extravascular bactericidal power of the
serum which he observed, and thus that the leucocytes
can be regarded as carriers of bacteria from the intestinal
tract. Such a transportation of bacteria by leucocytes

INFECTIONS OF THE DIGESTIVE TRACT 97

would presumably be facilitated in a high degree by the
accelerated movement of the blood serum during active
exercise. Ficker thinks it probable, also, that at the
time of greatly increased muscular exercise the body
cells generally, including those of the intestinal tract,
suffer a temporary check in their metabolism and hi
their ability to liberate energy. This temporary cell
infirmity is supposed in some way to influence the intes-
tinal epithelium so as to favor the passage of bacteria.
Ficker further suggests that through his experiments
are explained the well-known observations that the
flesh of animals that have been slaughtered after having
been driven long distances very quickly decomposes
after the killing, while, on the other hand, the flesh is
much better preserved in animals that are allowed to
rest for several days before being killed. He also raises
the question whether the penetration of the intestinal
tract by bacteria may not explain some of the phenomena
that have been noted after great fatigue, for example,
the so-called "fever of exhaustion," and also the state
designated by the older physicians as "autotyphization."

PHYLOGENETIC SIGNIFICANCE OF THE LARGE INTESTINE

In the study of the physiology and pathology of
the digestive tract it is necessary to realize the impor-
tance of certain functions in the development of the race
which may be too readily overlooked because they no
longer possess the significance for the maintenance of
life that was formerly the case. There can be little
doubt that one important function of the large intestine

98 INFECTIONS OF THE DIGESTIVE TRACT

is connected with the inspissation of the contents of this
part of the tract. In some lower animals and in some
human individuals the desiccation of intestinal contents
reaches so high a grade that the fseces voided contain
only a very low per cent, of water. By means of this
absorption of moisture the organism is protected against
a loss of water which under certain conditions might
prove a detrimental condition in the struggle for existence.
It is easily conceivable that under primitive conditions
of life, where water is not always readily obtainable,
those animals which were best able to conserve their
water and salts would have a material advantage over
animals less well able to prevent this waste. In many
mammals urine of very low specific gravity is excreted,
although in the glomeruli of the kidneys in these same
animals the concentration of the urine is extremely
high. If it were not for the powerful resorptive action
of the epithelia of the convoluted and other tubules of
the kidney, the organism would be subjected to an
enormous loss of water. Under conditions where
water is freely obtainable such a loss would, perhaps,
have little significance, bu,t where the water supply is
uncertain such wastefulness on the part of the organism
would soon lead to death. Thus in the large intestine as
well as in the tubules of the kidney we have to recognize
mechanisms protective to the life of the individual
through provision against the waste of water. In the case
of man this function is of relatively little significance
at the present time since water may usually be obtained
in abundance. It seems probable that in addition to the

INFECTIONS OF THE DIGESTIVE TRACT 99

function just mentioned the large intestine has often
served as a reservoir for food, but this function certainly
is superfluous in man. These considerations are of
importance ; for if it can be shown that the large intestine
is not needed either for purposes of inspissation or as a
reservoir or as a place of importance in carrying certain
digestive processes to an end, there is no reason why the
intestine should not, under certain conditions of disease,
be shortened by the elimination of a large part of the
colon. There is some difference of opinion as to whether
animals can live without the large intestine, but the bal-
ance of evidence is in favor of their being able to main-
tain life even after the large intestine has been excluded.
There are also instances in human beings that point to
the same conclusion. Dr. Bryant of New York tells
me that in one instance he practiced the insertion of the
ileum into the rectum. The patient was a woman who
lived a considerable period of time in this condition
without suffering in nutrition.

THE IMPORTANCE OF PROMPT RESORPTION FROM THE
SMALL INTESTINE

It is almost self-evident that the prompt resorption of
food from the small intestine is one of the most impor-
tant factors in preventing the occurrence of excessive
putrefactive conditions in the digestive tract. The pas-
sage of large quantities of partially digested proteid
material into the region of the intestine where anaerobic
conditions prevail must necessarily greatly facilitate the
bacterial decomposition of proteids in the digestive tract.

100 INFECTIONS OF THE DIGESTIVE TRACT

As will be pointed out in dealing with the methods of
diminishing chronic excessive intestinal putrefaction,
those measures which are designed to secure prompt
digestion and prompt absorption from the small intestine
are of the greatest significance in limiting bacterial de-
compositions*

The administration of a cathartic which acts imper-
fectly (that is, which carries partly digested food from
the small intestine into the colon without, however,
securing an evacuation) leads to conditions similar to
those following an excessive meal. In some persons
this miscarriage of a cathartic is followed by headache
and flatulence, in others by more serious signs of intoxi-
cation such as weakness in the muscles. In normal
persons these effects are relatively slight; in persons
whose digestive tracts are the seat of infection with
putrefactive anaerobes the evidence of excessive fer-
mentation and putrefaction is much more pronounced.

THE PHENOMENON OF SUBSTITUTION

I am convinced that what may be termed the phenom-
enon of substitution of flora in the digestive tract is a
common and significant occurrence in the course of many
derangements of the digestive tract. By substitution is
meant the temporary replacement of one type of micro-
organism normally abundantly present in the digestive
tract by an allied form which, though perhaps normally
present in small numbers, never assumes a dominant place
during health. The substitution may be complete or in-
complete and may be a temporary or a prolonged phe-

INFECTIONS OF THE DIGESTIVE TRACfr 101

nomenon. After a time the substituting microorganism
loses its prominent place and in its stead the obligate
bacteria of the tract again become dominant.

The phenomenon of substitution is one that most
frequently involves temporary replacements of the colon
bacilli in the lower part of the large intestine and also in
higher portions. An example of what is meant by such
substitution is afforded by the folio whig case. A woman
of twenty-five years of age developed measles and soon
after convalescence became ill with what was apparently
a condition of mucous colitis, lasting about one month
and attended by slight fever. Studies of the flora made
toward the end of the period of mucous colitis showed
that the faecal fields contained an excessive number of
diplococci, and the sediments of the fermentation tubes
exhibited streptococcal growths in great abundance.
On litmus gelatin plates made from faecal suspensions
it was found that the dominant organism resembled
colon bacilli in appearance, but differed from them in
forming acid very slowly. Many colonies were fished
from these plates and studied, but none were found which
could be called colon bacilli. The organisms obtained
differed from colon bacilli in making no indol and in
possessing only feeble powers of gas production and of
acid formation on glucose. They irregularly fermented
levulose and lactose. They coagulated milk only after
a lapse of ten days, and then the coagulation was in-
complete. They were actively motile Gram-negative
bacteria. Two weeks later, during the progress of con-
valescence from the condition of mucous colitis, it was

102 INFECTIONS OF THE DIGESTIVE TRACT

found that the fsecal suspensions gave on the gelatin
plates large numbers of bacteria of the type just men-
tioned and in addition contained considerable numbers
of true colon bacilli. One month after the first exami-
nation the fsecal suspensions showed the presence of
large numbers of colon bacilli, which had now become the
dominant type. From the litmus gelatin plates a very
small proportion of bacteria were isolated which were
incapable of coagulating milk, made no indol, and had
little effect^jifLQn the sugars. It was evident that the
substituted bacteria were now in a small minority.
The substituted microorganisms in this case resembled
the bacilli of typhoid fever except in the fact that they
were able to ferment glucose and possessed the ability
to ferment lactose and levulose in an irregular manner.
I have in several instances met with cases of substi-
tution in which the colon bacilli have disappeared from
the stools and have been replaced temporarily or per-
manently by other types of bacteria. Dysentry bacilli
may temporarily replace obligate colon bacilli.

THE PRESENCE OF PATHOGENIC BACTERIA IN THE
DIGESTIVE TRACT IN HEALTH

Evidence is gradually accumulating which goes to
show that pathogenic microorganisms may be present
in moderate or even considerable numbers in the intes-
tinal tract under some conditions without giving rise
to clinical manifestations of deranged function. It is
now well known that moderate numbers of typhoid
bacilli may in some instances be found in the stools

INFECTIONS OF THE DIGESTIVE TRACT 103

of patients who have lately had typhoid fever.
Here the explanation of the absence of symptoms is
doubtless to be sought in local or local and general con-
ditions of immunity. There appear, however, to be
instances in which there is no evidence that typhoid
fever has occurred but in which nevertheless moderate
numbers of organisms may be obtained from the fseces.
The same facts hold true for the bacilli of dysentery.
Duval has shown that dysentery bacilli are present
occasionally in small numbers in the stools of normal
individuals, and this observation has been confirmed by
others.1 We have noted the presence of B. pyocyaneus
occasionally in the digestive tract of persons apparently
in the best of health. It is likely, however, that in all
these cases the pathogenic organisms in question are
held in check by the bacteria present in the digestive
tract or by the bacteria and the intestinal secretions so
that they are unable to multiply in a significant manner
or to gain entry into the cells of the mucous membrane.
It seems not unreasonable to suppose that errors in diet
or depressed general conditions favor the multiplication
and penetration of pathogenic bacteria that have for
some time been present in a slumbering state. Irritant
foods may possibly so alter the secretions of the digestive
tract as to favor definite infection by the semi-parasitic
bacteria that are present.
The considerations just mentioned as applying to the

1 In some of the apparently normal children from whom Dr.
Wollstein obtained dysentery bacilli there were subsequently
developed the clinical indications of dysentery.

104 INFECTIONS OF THE DIGESTIVE TRACT

bacilli of typhoid fever, of dysentery, and of B. pyocyaneus
probably hold equally true of the more saprophytic forms,
such as the microorganisms concerned in chronic exces-
sive intestinal putrefaction. It is certain that the intes-
tine may harbor considerable numbers of B. putrificus
and B. aerogenes capsulatus or both of these together
without the development of clinical manifestations. A
variety of conditions may be presumed to so favor the
development of these anaerobes that their products,
instead of being formed in such small amounts as to
be harmless, begin to exert a detrimental effect upon
the organism. Especially important in this connection
are influences which alter the character of the secretions
in the large intestine or bring into the large intestine
unusually large quantities of partly digested proteid food.
There is an important practical aspect to the fact
that pathogenic organisms inhabit the intestinal tract
without giving obvious clinical signs of their presence.
A good quality of milk or meat, free from pathogenic
bacteria, may be blamed for bacterial decompositions of
a harmful kind which are in reality due to abnormal bac-
terial conditions prevailing in the digestive tract before
the use of the food under suspicion. In certain con-
ditions of the digestive tract an excessive or even a
moderate meal of proteid food will precipitate an intoxi-
cation or a seizure of vomiting or diarrhoea. There
are cases classed as "ptomaine poisoning" in which the
digestive tract, rather than the food, is responsible for
the observed disorders.

CRITERIA EMPLOYED IN THE CLASSIFICATION
OF BACTERIA OF THE GASTRO-ENTERIC
TRACT

IT appears not out of place to refer briefly to the cri-
teria employed in classifying the bacteria of the gastro-
enteric tract although these standards of judgment do
not differ from those of bacteria generally. The cul-
tural characters of the growths upon ordinary media
must be regarded as relatively less important in form-
ing a judgment as to the character of given organisms
than was formerly the case, since in the early days of
bacteriology such cultural characters constituted our
chief reliance in identification. While the cultural char-
acters of a microorganism derived from the digestive
tract must in every instance be taken into account,
these do not usually suffice to give more than a suspicion
of the identity of the bacterium in question. In general
it may be said that the cultural characters do not pos-
sess a high grade of permanence and are, moreover, often
not specific within certain groups. For example, within
the group of colon bacilli one finds many organisms which
must be regarded as distinct but which nevertheless give
practically undistinguishable growths on the ordinary
culture media. Indeed, a strain of colon bacillus may in
time come to vary quite widely from its original appear-
ance upon a given medium. The cultural characters of
microorganisms are often much influenced by physical

105

106 INFECTIONS OF THE DIGESTIVE TRACT

conditions pertaining to a culture medium. For example,
Dunham obtained widely different appearances in growths
of the same typhoid bacillus when grown in gelatin
plates in which the gelatin possessed varying degrees of
concentration and slightly different chemical characters.
The morphological characters of organisms are cer-
tainly important to consider and must be known as
a matter of course in any study of bacteria, but within
groups of closely related organisms morphology itself
gives us little aid. The question of motility is one of a
good deal of importance and should always be considered.
In general within a group of colon-like organisms a high
grade of motility suggests pathogenicity, but the absence
of motility or a low grade of it does not point in the
opposite direction, since pathogenic types of bacilli in
the dysentery group presumably not distantly related
to the colon bacilli are non-motile. The determination
of the staining characters of the organism with the Gram
method is most helpful provided a uniform method of
procedure be employed. In the absence of a uniform
technique the Gram method is less good than a simple
stain, since it serves only to confuse. Of great impor-
tance for the classification of the bacteria of the gastro-
enteric tract, as of bacteria generally, are the biochemical
properties. These biochemical characters include the
cleavage abilities of the organism with special reference
to their action on the soluble carbohydrates1 with the
production of acid and gas, their ability to produce indol

1 The great importance of this method of studying bacteria
was first pointed out by Professor Theobald Smith in connection
with the differentiation of typhoid and colon bacilli.

INFECTIONS OF THE DIGESTIVE TRACT 107

and skatol, and to liberate sulphur products such as
hydrogen sulphide and methyl mercaptan. The high
grade of permanence of the fermentative characters of
bacteria give this test much value. This permanence
is probably more marked within certain groups such as
the colon-typhoid group than in certain others, as, for
example, the group of streptococci. Strains of colon
bacilli may be carried through a long series of sugar-
bouillon fermentation tubes without appreciably modify-
ing in a quantitative way their ability to form gas and
acid. In a series of experiments in this direction which
was carried out by Mr. H. C. Ward, it was found that
strains of colon bacilli which had grown for a long time
upon a succession of sugar-bouillon tubes grew less well
upon peptone bouillon than had previously been the case.
On being cultivated in peptone bouillon after a long
period of growth on sugar bouillon it was found that they
produced only about one-half as much indol as had pre-
viously been the case . This result might easily have been
interpreted as an evidence that the organisms in question
were losing their ability to make indol upon peptone
bouillon. Careful examinations of the growths, however,
showed that the diminished production of indol was
to be referred rather to an impaired ability of the colon
bacilli to grow upon peptone bouillon, than to a real
inability to make indol. The formation of indol is a
character of considerable importance in the classification
of bacteria, but there is some evidence that it is not so
permanent and invariable a characteristic as the ability
of the bacteria to ferment certain sugars. Much more

108 INFECTIONS OF THE DIGESTIVE TRACT

work is required to establish the exact position of indol
production as a criterion for the classification of bacteria.
Unfortunately, nearly all that we know at the present
time about the production of indol is based upon the
nitroso-indol test, which possesses two grave disadvan-
tages. First, this test is not highly sensitive in the detec-
tion of indol, and secondly, the test as usually practiced
does not exclude the possibility of the presence of other
substances which react with indol. I consider it desir-
able to replace the old nitroso-indol test by the dimeth-
ylamidobenzaldehyde reaction first described by Pro-
fessor Ehrlich and which is much more delicate than
the nitroso-indol test. For quantitative determinations
of indol the yS-naphthaquinone-sodium-monosulphonate
method which I have described is unquestionably the
best. There is little doubt that in the classification of
the intestinal bacteria more and more stress will be laid
on the biochemical characters of the organisms in ques-
tion. The pathogenicity of an organism is, of course, a
highly important character, but is far more variable in
general than the biochemical characters just mentioned.
Virulent bacteria may easily lose their virulence or have
it greatly increased, according to the circumstances under
which they grow. The character of the conditions under
which an organism has been cultivated must, therefore,
be carefully borne in mind in drawing conclusions as to
this point. An organism which has long been grown in
the laboratory may easily lose its virulence on ordinary
media. On the other hand, some microorganisms,
such as diphtheria bacilli, may be grown for years on

INFECTIONS OF THE DIGESTIVE TRACT 109

artificial culture media of the same composition and still
remain strikingly constant with respect to the activity of
the toxins produced. The aerobic or anaerobic charac-
ters of microorganisms are of great importance in the
classification of intestinal bacteria, since the putrefactive
processes in the digestive tract are carried on largely
through the agency of strict anaerobes. No study of the
bacteria of the gastro-enteric tract can be considered
thorough which does not take the strict anaerobes into
account. This, of course, involves the use of anaerobic
technique. Some highly aerobic bacteria, like the micro-
organisms of cholera and the bacilli of tuberculosis,
multiply very poorly under strictly anaerobic conditions.
The majority of pathogenic microorganisms in man are,
however, facultative anaerobes capable of growing under
anaerobic as well as aerobic conditions, although the anae-
robic growth may be much less active than the develop-
ment in the presence of air. The ability of organisms to
make spores is another feature which has to be taken into
consideration in any classification of bacteria. Most of
the strict anaerobes of the intestinal tract — possibly
all of them — are capable of sporulating under certain
conditions, and this is a feature of great significance for
their persistence in the gastro-enteric tract, since in the
absence of the ability to sporulate in the tract the vege-
tative forms might easily lose their hold and die out. Of
late years much attention has been given to the aggluti-
native properties of bacteria growing in the intestinal
tract, with results of great importance for the classifica-
tion of these bacteria. It was at first supposed that the

110 INFECTIONS OF THE DIGESTIVE TRACT

agglutinative characters of certain intestinal bacteria
such as typhoid and dysentery organisms with human
or experimentally induced sera possess a high degree of
specificity. There was also at first an inclination to
believe that two bacteria of different origin possessing
the same agglutinative characters with one serum were
brought very close together by virtue of this fact. It is
now clear that the failure to distinguish between group
or common agglutinins and specific agglutinins may lead
one to think bacteria are identical when this is really not
so. The group agglutinins may to a considerable extent
be removed by the procedure of absorption which has
lately come into use. By means of absorption methods
it becomes possible to largely remove the common
agglutinins from a serum which contains them by bring-
ing this serum into contact with bouillon cultures of the
bacteria in question and after a certain period of contact
filtering off the bacteria, which have absorbed from the
serum a large part of the common agglutinins. The
specific agglutinative properties of the serum now come
very clearly into play and may be a great aid in deter-
mining the relationship of two microorganisms suspected
of standing in close' relation to each other. The value
of this absorption test is somewhat in question in some
laboratories, but it has given a high degree of satisfaction
in the laboratory of the Board of Health in New York
City, where it has been employed extensively and sys-
tematically by Dr. Park and Dr. Collins.1

1 The absorption method has also been successfully employed
by Dr. E. K. Dunham in his careful study of the agglutinins of the
meningococcus.

INFECTIONS OF THE DIGESTIVE TRACT 111

The agglutinative characters of many pathogenic bac-
teria are remarkably fixed properties. Dr. Theobald
Smith attempted to induce variations in the agglutina-
tive characters of hog-cholera bacilli and colon bacilli
by treating a series of sensitive experimental animals
with the bacteria to be tested. He met with no success.
Dr. Smith tells me, however, that he thinks the chances
for success in modifying the agglutinative characters of
the bacteria would have been greater had he tried more
resistent animals than were used, since this would have
afforded more opportunity to excite antagonism on the
part of the microorganisms, and with this, some altera-
tion in agglutinative properties. It is stated that the
agglutinative characters of some streptococci are more
subject to variation than those of the B. coli group.

METHODS OF INVESTIGATION

It is desirable to speak briefly here of the methods
that have been employed in some of the investigations
with which this volume deals. Some of these methods
are well known, others are new. They relate in part
to the study of the morphological and cultural charac-
ters of the bacterial organisms found in the digestive
tract under different conditions, but they have to do
mainly with the products of the life activities of these
bacteria upon different nutrient media.

Character of the Microscopical Fields. — By far the
most helpful method of studying the microscopical
fields is with the aid of the Gram stain. The routine

112 INFECTIONS OF THE DIGESTIVE TRACT

examination of the intestinal contents by means of this
stain is a real aid in forming a judgment as regards the
presence or absence of certain types of bacteria in the
digestive tract.1 There are many conditions in which,
as must be obvious, it can be of little service. There
are instances in which Gram-stained microscopical fields
appear entirely normal, but after cultural and chemical
studies it is evident that abnormal decompositions are
at work and that pathological microorganisms are pres-
ent. On the other hand, one may often form an opin-
ion from the mere study of the fields as to whether the
bacteria present are capable of initiating putrefactive
processes or not. In order to make the staining of any
value it is necessary that the method should be carried
out in a uniform manner. The suspension of the flora
used should not be too concentrated, as the pictures may
be confused through an excessive number of bacteria.
Suspensions of fsecal material in the proportion of one

1 In studying the bacteriology of the intestinal tract it is greatly
to be desired that very fresh specimens of the intestinal contents
be employed, since it may happen that certain delicate forms of
organisms may otherwise quickly die. In the case of colon bacilli,
of many coccal forms, and of most of the anaerobes a period of delay
appears to make little difference. In some of our investigations
we have worked with strictly fresh material, but it has necessarily
often happened that there was a delay of a few hours and some-
times of even greater time before cultures were made. The danger
in these cases is simply the danger of losing some types which
might otherwise be represented in the plates and fermentation
tubes. It is stated by some writers that B. bifidus is sensitive to
a depression of temperature, but we have not always found this
to be the case with the temperature of the ice box, since we have
been able to obtain growths of that organism in sugar bouillon
even after several days' residence of the material upon ice.

INFECTIONS OF THE DIGESTIVE TRACT 113

part by weight to ten parts of normal salt solution answer
very well as a material for making the smears. The
following technique serves satisfactorily for routine work.
One stains for three minutes with the Gram solution
of gentian violet. The solution is then allowed to run
off the slide, but the latter is not washed. Lugol's solu-
tion of iodine is then permitted to act for two minutes.
The slide is now rinsed with water and decolorization
is practiced for one-half minute by means of absolute
alcohol. The smear is then restained with a fuchsin
solution. Unless one employs approximately the same
technique, varying and confusing results are obtained
which make it almost impossible to say with confidence
which organisms are Gram-negative and which are Gram-
positive.

From the use of the Gram method one obtains some
idea as to the numbers of microorganisms resembling
the colon bacilli in morphology; one may form a judg-
ment as to the state of their preservation, as to the pres-
ence or absence of slender, long, Gram-negative organ-
isms of the type of B. liquefaciens ilei; a judgment may
be formed as to the numbers of Gram-positive diplococci
and other coccal forms ; an opinion may be formed as to
the probability of the presence of B. bifidus, especially
in its unbranched form; and finally it is possible, after
one has had experience, to make an estimate as to whether
free spores and spore-holding organisms and vegetative
anaerobic forms are present in excessive numbers. None
of these microscopical appearances can be regarded as
positive evidence of the identity of the dominant bacteria

114 INFECTIONS OF THE DIGESTIVE TRACT

present until this has been determined by cultural
methods. It may, however, aid one in forming conclu-
sions that are valuable on account of their high degree
of probability. In the case of a given individual, iso-
lation by careful plating and identification by suitable
methods of any dominant organism becomes a great aid
to subsequent interpretation of the Gram-stained field,
even though one may not have the opportunity to make
further isolations and identification in the same patient.
There are many applications of the Gram stain that
may advantageously be made in the study of intestinal
disorders, quite aside from the direct study of the faeces
and intestinal material obtained through the use of
cathartics. A study of the types of bacteria that are
dominant at different levels of the intestinal tract in
persons dying of various diseases is greatly facilitated by
the use of Gram staining. This method of study of the
bacteria at different levels has not yet received the at-
tention it deserves. I consider it especially important
to determine the distribution of the anaerobes in the
small intestine in numbers sufficiently great to enable
them to put their stamp on the character of the intestinal
decomposition. The Gram stain further has proved
highly useful in the study of the sediments in all the
fermentation tubes inoculated with the mixed faBcal
flora. Moreover, the same method has given consider-
able information in the study of the sediments obtained
in various media on which certain mixed flora have been
grown from seven to ten days. It is sometimes desira-
ble to make use of a staining method for spores in order

INFECTIONS OF THE DIGESTIVE TRACT 115

to obtain an idea of the numbers of free spores present.
Without this procedure one is apt to underestimate their
number. The faecal flora may also be stained directly
for capsules by Welch's acetic acid method.

I shall not attempt here to discuss fully methods of
isolation and identification of individual bacteria. The
methods generally in use for making plate cultures are
those which have been employed. In the case of the
bacilli of typhoid fever, dysentery, and allied organisms
this subject has already been fully developed by highly
trained workers. The study of the diplococci and other
coccal forms is one that has been relatively neglected
(except in the case of acute infections) and is deserving
of much fuller development. On the subject of the anae-
robes of the large intestine almost nothing has been done,
partly, perhaps, on account of the difficulties of anae-
robic technique. Some acute infections with anaerobes
have indeed been studied, notably by Tissier and by
Klein, but until recently the chronic processes dependent
on excessive development of putrefactive anaerobes have
escaped attention. The necessity of furnishing anaerobic
conditions for the growth of some of these bacteria con-
siderably increases the difficulties of identification, but
they are by no means insuperable. Where a high degree
of anaerobiosis has been desired, we have found it help-
ful to make use of a stream of compressed hydrogen
instead of using hydrogen generated in the laboratory.
This is followed by the employment of the pyrogallic
method for removing the last trace of oxygen. Where
a somewhat less high degree of anaerobiosis suffices it

116 INFECTIONS OF THE DIGESTIVE TRACT

is very convenient to make use of the simple method,
recently described by Hans Zinsser 1 from the Depart-
ment of Pathology in Columbia University. Here two
crystallizing dishes are made to do service for obtaining
anaerobic growths . The smaller dish containing the inoc-
ulated agar is inverted inside the larger dish, the oxygen
being exhausted by the pyrogallic method. Water, on
the surface of which oil is placed, serves as a seal. This
method should serve to remove some of the obstacles to
the study of anaerobes.

The identification of the anaerobes calls for a good deal
of patience and care. The anaerobic life of the large
intestine (especially in disease) having been so little
studied, it is not unlikely that new varieties will be found
which have heretofore escaped notice. A growth of
B. bifidus and other acidophiles occurs under anaerobic
conditions in both glucose and plain agar, if one takes
the precaution to make up the media so that they con-
tain 0.5 per cent, of acetic acid.2 In the study of B.
aerogenes capsulatus it is important to use blood agar in
order to obtain the conditions proper for the growth
of these bacilli. The use of one cubic centimeter of
defibrinated rabbit's blood to eight cubic centimeters of
sugar agar answers well.

In studying the characters of the organisms for the
purpose of identification the fermentation tubes are ex-
tremely helpful, as first pointed out by Professor Theo-

1 " A Simple Method for the Plating of Anaerobic Organisms,"
Journ. of Exper. Med., viii, p. 542, 1906.

2 A beer-wort medium may be conveniently used.

INFECTIONS OF THE DIGESTIVE TRACT 117

bald Smith. Anaerobic organisms in general grow much
better in the closed limb of the tube on sugar bouillon
in the presence of bits of sterile tissue (such as liver from
a guinea-pig or rabbit), and in some cases, as in that of
B. aerogenes capsulatus, they do not grow in the ab-
sence of this aid. The identification of the anaerobes
involves the study of their cultural characteristics, of
their ability to form gas on sugar media, and the de-
termination of their gas formula. The estimation of
the ratio between the hydrogen and the carbon dioxide
formed is an important point in identification first
suggested by Theobald Smith. It is easy to determine
this point by adding caustic potash to the contents of
the fermentation tube, this being followed by absorp-
tion of the carbon dioxide.

The influence of the growth of anaerobes upon milk,
especially litmus milk, is easily studied in the fermenta-
tion tubes and gives considerable aid in identification.
For example, B. aerogenes capsulatus quickly makes acid
on milk and sets up a stormy fermentation due to the very
rapid production of gas from milk-sugar. It also acts
proteoclastically on the casein, and the coagulated milk
is partly digested and broken into small masses. This
behavior is in marked contrast to that of the bacillus
of malignant oedema. Very rapid and abundant gas
production from milk occurs through the agency of B.
cloacce, an organism not rarely found in the human
intestine.

The pathogenicity of the isolated anaerobes is also
a point of importance in the establishment of their

118 INFECTIONS OF THE DIGESTIVE TRACT

identity. There is a considerable group of strict anaer-
obes whose members are able to induce in guinea-pigs
(when subcutaneously injected) certain characteristic
lesions, including hsemorrhagic oedema, necrosis of cells
due to proteoclastic enzymes, and in certain cases evolu-
tion of gas. Finally, the introduction of cultures of
putrefactive anaerobes into the circulation of a living
rabbit which is killed in the course of a few minutes and
then incubated is sometimes a method very helpful for
the establishment of the identity of an anaerobic micro-
organism. This method, which was first employed by
Welch and Nuttall,1 must be regarded as an extremely
important means of revealing not only the gas-forming
properties of a pure culture of anaerobes, but also of
showing those of a suspension of the mixed fa3cal flora.
Here the blood and tissues of the rabbit act as a pecul-
iarly favorable culture medium for the growth of the gas-
bacillus (B. aerogenes capsulatus) , the bacteria having
been thoroughly spread by the blood through the body
and the conditions of growth being highly anaerobic.
Welch and Nuttall made use of their procedure to isolate
the gas-bacillus and to demonstrate its ability to make gas
on a proteid medium containing little sugar. The bacil-
lus of rauschbrand (symptomatic anthrax) also induces
the conditions obtained by the gas-bacillus, but unlike
the gas-bacillus, sporulates under these conditions.
The bacilli of botulism and of malignant oedema do not

1 "A Gas-producing Bacillus (B. aerogenes capsulatus, nov. spec.)
Capable of Rapid Development in the Blood Vessels after Death,"
Butt, of the Johns Hopkins Hosp., iii, p. 81, 1892.

INFECTIONS OF THE DIGESTIVE TRACT 119

form gas in the incubated rabbit, but induce other evi-
dences of putrefaction.

It seems singular that almost no use has been made by
subsequent investigators of this ingenious and extremely
valuable method of studying the gas-bacillus. Acting
on the suggestions carried by the paper of Welch and
Nuttall, we have used the method not merely as an aid
in the identification of B. aerogenes capsulatus, but also
(with somewhat unexpected success) as a means of deter-
mining whether the gas-bacillus is present in considerable
numbers in the faeces.

The conditions found at autopsy after twenty-four
hours' incubation of a rabbit previously injected intra-
venously with a pure culture of B. aerogenes capsu-
latus have been so fully described by Professor Welch
that it is unnecessary to add anything to his description
aside from emphasizing the fact that the almost intoler-
able odor of putrefaction which is developed during the
incubation experiment is dependent in part on the pro-
duction of butyric or a closely allied acid.1 It should
also be added that an odor very similar to that charac-
teristic of the incubation experiment can be observed in
the faeces of some persons with chronic disturbances of
digestion, and frequently in persons with advanced anae-
mias associated with irregular diarrhoeal conditions.

The incubation method of Welch and Nuttall has
apparently never been employed in connection with the

1 1 do not know of another equally impressive example of the
ability of microorganisms to induce rapid putrefactive decompo-
sition.

120 INFECTIONS OF THE DIGESTIVE TRACT

study of the human faeces, but I believe it has here an
important clinical application. Although it is true that
B. aerogenes capsulatus can be isolated from the fseces
of a majority of adult individuals, including very many
who are in excellent health, it is also true that there are
wide differences in the number of capsulati habitually
present in the case of different individuals. There are
some young persons between the ages of five and twenty
years from whom it is either very difficult to obtain
B. aerogenes capsulatus by plating or whose movements
give no evidence whatever of its presence. If we pre-
pare a suspension of the fseces from such individuals by
grinding one gram of the fresh material with nine cubic
centimeters of 0.85 per cent, salt solution and filtering
through absorbent cotton, we can inject intravenously
one or two cubic centimeters of this suspension into a
rabbit and then incubate the quickly killed rabbit for
five hours at 37° C. without obtaining evidence of the
abundant presence of the gas-bacillus. On opening a
rabbit which has been thus incubated, one finds none of
the signs of the activity of the gas-bacillus — no accu-
mulation of gas in the liver or vessels or in the con-
nective tissues or serous cavities. Moreover, smears made
from the liver blood or the auricular blood either do not
show the presence of capsulatus at all, or these organisms
are present only in small numbers. If, however, the
foregoing experiment be made with the fsecal material
derived from a patient with pernicious anaemia or from a
person suffering from a capsulatus diarrhoea, one gener-
ally gets an entirely different result. At the end of five

INFECTIONS OF THE DIGESTIVE TRACT 121

hours the liver is soft and friable, crepitates between
the fingers, and on section shows the presence of many
bubbles of gas. There may also be a small accumulation
of gas in the peritoneal cavity. Smears from the he-
patic blood and from the auricular blood swarm with
organisms of the capsulatus type. The spleen also con-
tains such organisms in great numbers.

The following protocols are instructive in this con-
nection : —

EXPERIMENT 1. One cubic centimeter of a filtered faecal suspen-
sion, from a normal person sixteen years of age, almost free from
putrefactive products in the urine, was injected into a rabbit which
was immediately killed by a blow on the neck. After five hours'
incubation, the animal was examined. Abdomen distended slightly
from distension of large intestine; liver firm, slightly friable, and
free from gas. A stained liver smear shows a few short bacilli
(not capsulatus} . Heart's blood shows short bacilli in abundance,
rarely a bacillus of capsulatus type.

EXPERIMENT 2. Two cubic centimeters of filtered faecal sus-
pension from a healthy man (aet. 42) recently recovered from uni-
versal eczema, were injected intravenously into a rabbit which
was then promptly killed by a blow on the neck. Animal incu-
bated for five hours at 37° C. On exhibition, no odor of butyric
decomposition; liver firm and without gas bubbles. Smear from
heart's blood showed a few bacteria of capsulatus type.

EXPERIMENT 3. Two cubic centimeters of filtered faecal sus-
pension from a healthy breast-fed baby were infused intravenously
into a rabbit which was then promptly killed. Examination after
five hours' incubation at 37° C. reveals no odor of butyric putre-
faction, and liver is firm and free from gas. Smears from the
heart's blood show it to be free from bacteria of any kind.

These experiments may be contrasted with the follow-
ing : —

EXPERIMENT 4. Two cubic centimeters of filtered faecal sus-
pension from an anaemic baby with irregular diarrhoea were in-

122 INFECTIONS OF THE DIGESTIVE TRACT

fused intravenously into a rabbit which was then promptly killed
and incubated for five hours at 37° C. On examination the
tissues gave a strong butyric acid odor. Liver soft, friable, and
filled with bubbles of gas. Smears from heart's blood show a
great abundance of bacteria of the capsulatus type.

EXPERIMENT 5. One cubic centimeter of filtered faecal suspen-
sion from a patient with pernicious anaemia was infused intra-
venously into a rabbit which was then incubated for five hours
at 37° C. On examination the abdomen was slightly distended.
Characteristic butyric odor. Liver crepitant, contains a few ob-
vious gas bubbles. Bacilli of capsulatus type abundant in heart's
blood, in almost pure culture. Capsulati also abundant in liver.

EXPERIMENT 6. Two cubic centimeters of filtered faecal suspen-
sion from a patient with pernicious anaemia whose faeces contained
a great abundance of free (capsulatus ?) spores were injected intra-
venously into a rabbit which was incubated at 37° C. for twenty-
four hours. At autopsy the animal was greatly distended with gas,
and bloody fluid was oozing freely from nose, mouth, etc. Gas
escaping from abdominal cavity burns with blue flame. Extremely
offensive odor of butyric decomposition. Tissues in advanced
state of putrefactive liquefaction. Blood from heart shows bacilli
of capsulatus type to be extremely abundant. Most of these were
Gram-positive and occurred characteristically in diplobacillus
form, but there were also many long threads which were doubtless
capsulati. Gram-negative forms also occur and of these one
variety was especially prominent. This was a long organism
bearing a large spore near either end. It is probably to be re-
garded as a young form of capsulatus about to undergo division
midway between the spores. This organism was Gram-negative.

The foregoing experiments are typical of a large group
and show plainly enough the difference in capsulatus
activity in the case of material derived from normal
and pathological faeces. It appears to be true that a
ten per cent, suspension of faeces does not excite an active
formation of gas in the liver, if the material injected has
been derived from persons free from intestinal derange-
ments, and with scanty evidences of putrefactive pro-

INFECTIONS OF THE DIGESTIVE TRACT 123

ducts in the urine. On the other hand, material from
persons whose stools contain an abundance of capsulatus
(including a large proportion of persons suffering from
"primary" pernicious ansemia) induces with great
regularity the peculiar alterations in incubated rabbits
which have been already described. The short period
of time (four to six hours) which is required to bring
about the liberation of gas in the liver is a point worthy
of notice.

The distribution of B. aerogenes capsulatus in the
bodies of rabbits that have been subjected to the incu-
bation test is a matter of some interest. Where normal
rabbits are employed for this test one finds large num-
bers of the gas-bacillus in the liver and often considerable
numbers in the heart's blood. One finds also moderate
or considerable numbers of the bacilli in the spleen,
where they have doubtless been arrested owing to me-
chanical conditions. In the brain, in the suprarenals,
in the kidneys, and in other organs, one finds very few
gas-bacilli. One may experimentally change somewhat
the distribution of these organisms if previous to the
injection and killing of the rabbit the animal be fed oh
large quantities of dextrose — say twenty grams daily for
three days. Under these conditions, if the inoculation
test be carried out, it will be found that not only do
the liver and blood contain much larger numbers of gas-
bacilli than is ordinarily the case, but the spleen also con-
tains great numbers of the same organism. Moreover,
the spleen which in normal rabbits gives no gas, will be
found to contain numerous bubbles of gas and therefore

124 INFECTIONS OF THE DIGESTIVE TRACT

to float on the surface of water. This greatly enhanced
growth of the gas-bacillus in animals fed upon sugar is
doubtless connected with the increased blood content
of the organism in glucose. It is paralleled by the
conditions which we find at autopsy in persons dying of
diabetes. I have tried to rid the body of its carbo-
hydrates by means of phlorhizin poisoning to such a
point that the growth of the gas-bacillus is impaired in
the incubation test, but have not succeeded in distinctly
altering the habitual bacteria.

The selective action of the dead organism for B. aero-
genes capsulatus is certainly a striking feature. The in-
fusion of any faecal suspension means the introduction
of many varieties of living bacteria. Yet after a few
hours of incubation in the dead rabbit the number of
microorganisms in the blood has been narrowed either
to capsulatus alone or to capsulatus and one of two com-
panions— often positive diplococci, sometimes spore-
bearing bacilli resembling capsulatus, but not positively
identified. A longer period of incubation usually elimi-
nates from the blood all organisms except those of the
capsulatus type. The initial bactericidal power of the
blood and cell juices may suffice to kill many of the bac-
teria of the faeces, while the strict anaerobic conditions,
so necessary to the multiplication of capsulatus, in itself
cuts out many varieties.

It is believed that the more refined application of this
method to the study of the faeces will prove of clinical
value in several directions. It may also prove of utility
in the study of milk. If, as seems probable, B. aerogenes

INFECTIONS OF THE DIGESTIVE TRACT 125

capsulatus is really identical with the granulo-badllus
immobilis liquefaciens of Grassberger and Schattenfroh
(as is claimed by some observers and denied by others,
including Fraenkel), this method may prove helpful in
connection with the study of some diarrhoeal diseases
which have been attributed to the use of infected milk.

There is reason to think that the faecal suspensions
from most persons in good health would induce gas pro-
duction in incubated rabbits, since the faeces of most
apparently normal persons contain moderate numbers
of B. aerogenes capsulatus. The difference, as regards
the outcome of the incubation experiment, between the
flora of these persons and the flora of persons with cap-
sulatus infections is probably one of degree and not of
kind. Some idea of the relative numbers of capsulati
present might perhaps be gained by determining the small-
est volume of a given fsecal suspension that will just suf-
fice to induce the distinctive gas production in incubated
rabbits within a given time.

As it was obviously impracticable to isolate in every
instance the dominant microorganisms of the faeces, it
was resolved to study the action of the mixed faecal flora.
This was done in the hope of obtaining information which
should be of service in determining what bacteria are
really dominant in the intestine. The effort has not
been wholly successful, but has, nevertheless, proved
useful in several ways. A suspension of the mixed flora
was prepared in normal salt solution in the proportion
of about one part of material by weight to ten parts
of the solution. From this suspension a series of

126 INFECTIONS OF THE DIGESTIVE TRACT

fermentation tubes was inoculated in a uniform manner.
The fermentation tubes contained dextrose bouillon,
levulose bouillon (made with Schering's diabetin), lactose
bouillon, saccharose bouillon, peptone bouillon and plain
bouillon, litmus milk, dextrose bouillon containing methyl-
ene blue, dextrose bouillon containing neutral red, and
dextrose bouillon containing methyl violet. The tubes
containing the dyes were employed to determine the
reducing activity of the bacteria. As the information
yielded by these color tubes did not appear especially
helpful, the method was temporarily abandoned. Far
more useful was the study of the gas production in the
sugar-bouillon tubes. These tubes contained concen-
trations of the sugars already mentioned equal to two
per cent, in each case. It was found that the least
gas was usually obtained on the saccharose medium,
and the most on the dextrose or lactose. The quantity
of gas produced in conditions of health by the mixed
flora is somewhat variable, but may be roughly stated
as varying ordinarily from fifteen to thirty per cent,
of the height of the anaerobic limb. This estimate is
based not on any one sugar tube, but on the average of
the four tubes, the gas production, as already stated,
having been unequal in these different tubes. In normal
children on a milk diet the gas production is often some-
what less than in adults and may be not more than ten
to fifteen per cent. In conditions of disease the gas
production was found to be usually considerably less
than the average production in health. This is true
both of adults and of children. In the case of well-

INFECTIONS OF THE DIGESTIVE TRACT 127

marked examples of saccharo-butyric putrefaction the
quantity of gas produced may be one-half, one-third,
one-quarter, or even one-fifth of the normal gas produc-
tion. I am disposed to attribute this mainly or wholly
to an elimination of the colon bacilli. The gas produc-
tion in the sugar-bouillon tubes has in several instances
been observed to correspond fairly closely to the volume
of gas produced by the colon bacilli from the same case.
It is known that different strains of colon bacilli liberate
different quantities of gas on the same sugar-bouillon
medium, and I am therefore disposed to attribute
the fluctuations observed in health mainly to the different
gas-producing character of organisms of the colon bacillus
group. It is possible that B. lactis aerogenes plays a
part, but I believe this organism to have a subordinate
role. There are cases in which the intestine contains
yeast organisms, and under these circumstances there
is a very much larger gas production than is ordinarily
the case. A very abundant gas production in the sugar-
bouillon tubes inoculated from the faecal flora should
excite suspicion of the presence of yeasts. Sometimes,
however, the excessive gas production is dependent on
the growth of B. lactis aerogenes. As I have already
stated, this organism in pure culture does not ordinarily
grow in the closed limb of the fermentation tube, and
in order to enable it to grow it has to be aided by the
presence of blood or bits of sterile tissue. We have
observed instances, however, in which B. aerogenes cap-
sulatus has grown in the closed limb of the fermentation
tube without the addition of tissue, and I attribute this

128 INFECTIONS OF THE DIGESTIVE TRACT

to the transference to the tube of some nutritive sub-
stance contained in the intestine.

By far the greatest value of the gas production in the
fermentation tubes is in those cases in which the gas
production is much below normal. This diminution in
gas formation is not a variable occurrence, but in the
case of the same individual is persistent so long as the
conditions of diet and habits of life remain the same.
It is, in fact, an individual peculiarity. As already
stated, it appears to depend on the elimination of typical
colon bacilli. This view has been supported repeatedly
by the results of plating upon litmus gelatin. The dis-
appearance of the colon bacilli is occasionally met with
in persons apparently in good health, but I believe this
to be a very exceptional occurrence. Ordinarily the
inability to produce gas in normal abundance is a sign
of a temporary or persistent alteration in the character
of the intestinal flora. There are instances, indeed, in
which we have been unable to obtain any gas whatever
in some or all of the sugar-bouillon tubes. These have
been usually instances of an extremely marked form of
saccharo-butyric putrefaction and have not been very
uncommon among cases of pernicious anaemia associated
with an infection with B. aerogenes capsulatus and the
disappearance of the colon bacilli. This disappearance
has been noted also in some cases of mucous colitis
in which a foreign race of colon bacilli (or organisms
intermediate between colon bacilli and typhoid or para-
typhoid bacilli) have been found to be the dominant
representatives of the colon bacillus group.

INFECTIONS OF THE DIGESTIVE TRACT 129

The influence of diet must not be overlooked in con-
nection with the question of gas production in the fer-
mentation tubes by the mixed faecal flora. It seems
probable that a diet containing an abundance of carbo-
hydrates leads to greater gas production than a diet in
which carbohydrates are much restricted. A patient
on a diet consisting chiefly of meat will harbor organisms
making less gas than if he were on a mixed diet. The fall
in gas on a meat diet may amount to forty or fifty per
cent, of the total gas production on a mixed diet, perhaps
in some cases to more than this. This factor is thus one
which must be taken into consideration in interpreting
the results of the gas production by the mixed faecal
flora.

From the peptone-bouillon tube it is possible by
means of the Ehrlich dimethylamidobenzaldehyde re-
action to form a rough estimate of the abundance of the
indol formed. Some idea of the quantity of ammonia
and other volatile bases produced may be obtained by
the use of Nessler's reagent. By diluting the bouillon
and using definite quantities of Nessler's reagent in
making the tests, one may gain some idea as to quantita-
tive differences in regard to the production of bases.
The addition of strong hydrochloric acid to the peptone-
bouillon tube causes in some cases a liberation of sul-
phureted hydrogen, which collects in the upper portion
of the closed limb and may give an indication of an un-
usual formation of this gas. That it is mainly or wholly
hydrogen sulphide that collects under these circumstances
is shown by the fact that the addition of a solution of

130 INFECTIONS OF THE DIGESTIVE TRACT

an iron or cadmium salt leads to the reabsorption of
the sulphur compound, giving rise to sulphides of the
metals.

In some cases fermentation tubes containing bouillon
(or sugar bouillon) have had added to them calcium or
magnesium carbonate. By maintaining the neutrality
of the culture medium the carbonates have influenced
the character of the dominant organisms in the fermen-
tation tubes and have thus modified the proportions of
the products formed and even to some extent their
character.

It has been found useful to examine regularly the
sediments of the fermentation tubes which have been
inoculated with the mixed fa?cal flora. This has been
done as a matter of routine in a large number of in-
stances, the examination having been made especially
in the case of the four sugar-bouillon tubes and the
peptone-bouillon tube. The appearance of the Gram-
stained flora gives as a rule, but not always, an indication
of the dominant flora in the lower part of the intestine.
In tubes inoculated from normal persons, Gram-negative
organisms corresponding in size and form to bacteria of
the B. coli type grow abundantly in all the tubes. As
a rule they constitute the dominant flora in all the sedi-
ments. Generally one sees mixed with them moderate
numbers of Gram-positive and Gram-negative diplo-
cocci. Moderate numbers of organisms morphologically
like B. aerogenes capsulatus are seen, but these forms
may be wholly absent. There is an aerobic, Gram-
positive bacillus, resembling closely in its morphology

INFECTIONS OF THE DIGESTIVE TRACT 131

the anaerobe just mentioned, and the two organisms
can be positively distinguished only by making blood-
agar cultures under aerobic and anaerobic conditions.
The growth of the organisms is usually slighter in a
sugar-free tube than in the ones containing sugar. The
diplococcal forms, especially, are apt to grow poorly
in the peptone bouillon and may quite fail to multiply.
In material containing large numbers of the plain form
of B. bifidus one finds in the sugar tubes large numbers
of this organism in its bifid form. Very beautiful
branching forms, presenting considerable variety in
their morphology, may be observed. One has this ex-
perience especially with the lactose and dextrose bou-
illon, and I think the growths of Hfidus are apt to be
especially luxuriant in the lactose-bouillon tube. The
so-called punctate form already described is very com-
mon in the fermentation tubes.

Under pathological conditions one may meet with
different bacterial elements in the sediments. In cases
in which the intestine contains large numbers of positive
diplococci or streptococci, there is an abundant growth
of positive diplococcus forms and streptococcus forms in
the sugar tubes. One may find streptococcus forms very
abundant even hi cases where the tubes have been inocu-
lated from flora not containing a strikingly increased
number of such bacteria. As a rule a .very great pre-
ponderance of streptococcal and diplococcal forms in the
fermentation tubes points to an excess of such organisms
in the digestive tract. By the use of this method atten-
tion has sometimes been drawn to this -type of bacterial

132 INFECTIONS OF THE DIGESTIVE TRACT

activity in the digestive tract where it would otherwise
have been overlooked. In some instances, also, one
sees a great development of anaerobes in the fermentation
tubes. Sometimes one sees sedimentary fields which
consist wholly of streptococcal forms and of forms which
are probably B. aerogenes capsulatus. This has been
repeatedly noticed in some cases of advanced chronic
saccharo-butyric putrefaction with an extreme degree
of ansemia. The tendency for the organisms of the B.
coli type to be poorly represented in cases of chronic
saccharo-butyric putrefaction is evident in the sediment
of the fermentation tubes. In some cases one finds that
the fermentation tubes develop a greenish color in the
open aerobic arm of the tube and also a greenish pellicle.
Culture shows this to be due to the B. pyocyaneus.
This organism has been found in only a small number of
cases and usually, but not always, in persons presenting
signs of digestive disorder.

It cannot be claimed that we now know the full value
or the limitations of the information to be derived from
the careful study of the sediments of the fermentation
tubes, but it may safely be said that this method of
investigation has often been helpful in gaining a clew to
the nature of the dominant organisms in the intestinal
tract. One cannot rely upon it alone, but in connection
with data derived from other methods it helps us to
form a conception of the bacterial types present and
active in the lower part of the digestive tract. One
finds in the same case almost regularly the same kinds of
sediments, and this gives a degree of confidence hi the

INFECTIONS OF THE DIGESTIVE TRACT 133

value of the sediments, as an indication of the well-de-
fined character of the bacterial conditions that give rise
to these individual peculiarities. A highly interesting
observation which has been repeatedly made is that the
character of the sediments may be greatly altered as a
patient grows better or develops a more intense degree
of a condition from which he is suffering. The gradual
reappearance of the colon bacilli in the sedimentary
fields, with the corresponding disappearance of an exces-
sive number of coccal and streptococcal forms and of
anaerobes, has repeatedly been noticed as a concomitant
of a pronounced improvement in clinical conditions. In
addition to the study of the mixed faBcal flora (as grown
upon fermentation tubes) many observations have now
been made upon larger volumes of media in which the
mixed flora has grown for relatively long periods of time,
usually a week. As a routine procedure in the investiga-
tion of many normal persons and numerous pathological
conditions, four flasks containing about five hundred
cubic centimeters each of medium have been inoculated
with the suspensions of the mixed flora. The media em-
ployed have been peptone bouillon, peptone bouillon with
calcium carbonate, sugar bouillon and sugar bouillon
with calcium carbonate. The organisms have been per-
mitted to grow for a period of seven days. Under the
conditions prevailing in these flasks a large part of the
growth has been anaerobic and a high degree of anaero-
biosis has been maintained, in part owing to the forma-
tion of reducing products such as hydrogen, which are
incidental to the development of fermentative and

134 INFECTIONS OF THE DIGESTIVE TRACT

putrefactive cleavages. Two objects have been served
through the employment of these flask cultures: first,
the study of the bacterial sediments after considerable
periods of growth; and, secondly, the study of the pro-
ducts formed on a representative medium. As regards
the characters of the organisms found in the sediments
(at the end of seven days) it is not possible at present to
state definitely to what extent the biochemistry of the
digestive tract can be furthered by the mere study of the
Gram-stained fields, without resort to cultural methods
permitting positive identification of the dominant forms.
The appearances in the flasks usually vary considerably
with the different media employed and also according
to the origin of the suspensions employed for inoculation.
It has been found, in general, that the anaerobes grow
more abundantly in the flasks in which a neutral re-
action is maintained, owing to the presence of calcium
carbonate. Whether the presence of calcium ions as
such influences the growth of the organisms in a specific
way is not known. It is certain that hi some cases of
excessive intestinal putrefaction attention has been
called to the prolific growth of certain anaerobes which
have been much less prominent in the intestinal con-
tents themselves and in the sediments of the fermenta-
tion tubes (in which the suspended bacteria have been
permitted to grow for a considerably shorter time than
in the flasks). In several instances microorganisms of
the morphological type of B. putrificus have been ex-
tremely abundant in the sediments of the calcium-car-
bonate flasks. Later it has been noticed in some of

INFECTIONS OF THE DIGESTIVE TRACT 135

these cases that the growth of what we may assume to
have been B. putrificus is no longer evident, and this
change in the character of the growths in the flasks has
corresponded to a clinical improvement in the condition of
the patients concerned. As regards B. aerogenes capsula-
tus, it cannot be said that we have learned more from the
seven-day flasks than from the fermentation tube sedi-
ments already mentioned. In a few instances it has
been observed that the peptone-bouillon flask has con-
tained enormous numbers of spores when the corre-
sponding calcium-carbonate flask has shown large num-
bers of vegetative forms of putrefactive microorganisms
and relatively few spores. Further studies are neces-
sary to determine the value of the seven-day-flask
method and to show how the bacterial fields are to be
interpreted.

The chemical examination of the seven-day flasks
has included two different series of procedures. In the
case of the peptone-bouillon flask and the peptone-
bouillon flask containing calcium carbonate, the con-
tents have been examined with respect to hydrogen
sulphide, methyl mercaptan, volatile fatty acids, am-
monia, indol, skatol, and phenol.1 Quantitative deter-

1 A number of observations have been made upon the bouillon
and bouillon-carbonate flasks inoculated from the mixed flora
with a view to determining whether indol acetic acid is ever formed.
The method consisted in applying the three tests of Salkowski
to the material which remains behind after the distillation of
indol and skatol. In one case only has this residual material given
reactions with the three tests of Salkowski. This was in the case
of flasks prepared from a patient with peripheral neuritis and
pronounced psychosis resembling that of alcoholic intoxication.

136 INFECTIONS OF THE DIGESTIVE TRACT

minations have regularly been made in the case of the
volatile fatty acids, ammonia, indol, skatol, and phenol.
Alcohol and acetone have also been tested for in these
flasks. In the case of the sugar-bouillon flasks the
contents have been examined for alcohol, acetone, the
volatile fatty acids, and the non-volatile organic acids.
The molecular weight of the barium salts of the volatile
fatty acids has regularly been determined in order to
obtain a conception of the chemical nature and quantity
of the chief fatty acids produced. The interesting ob-
servation has been made that in the flasks containing
calcium carbonate the molecular weight obtained for
the volatile fatty acids has nearly always been some-
what higher than in the case of the molecular weight
obtained from the volatile fatty acids of the sugar-
bouillon flasks. This fact corresponds to the appearances
in the smears (already noted) that putrefactive anaerobes
are found in greater abundance in the neutral flasks
than in the sugar-bouillon flasks to which calcium car-
bonate has not been added.

The fact that acetone has been found in many instances
in the flasks containing peptone bouillon and peptone
bouillon with the addition of calcium carbonate will be
referred to in reviewing the facts regarding the nature
of the products of fermentative putrefactive decomposi-
tion.

The method employed to determine the presence of

2

In this case, however, alcoholic intoxication could be almost
certainly excluded. There were various other indications that
the neuritis and psychosis were due to intestinal intoxications.

INFECTIONS OF THE DIGESTIVE TRACT 137

methyl mercaptan is as follows: One hundred cubic
centimeters of the culture to be examined are trans-
ferred to a flask communicating through a calcium
chloride tube with an Erlenmeyer flask containing
isatin dissolved in concentrated sulphuric acid. A
current of air is then drawn through both flasks so
that any mercaptan given off from the culture flask
will enter the isatin-sulphuric-acid flask. The pres-
ence of mercaptan is indicated under these conditions
by a gradual change of the isatin solution from red to
olive-green or grass-green.1 In order to save isatin
the reaction may be carried out substituting a test-
tube containing about ten cubic centimeters of the
isatin-sulphuric-acid mixture for the Erlenmeyer flask.
The method is not adapted for quantitative deter-
minations, but some idea can be gained through it
of the quantity of mercaptan present in the culture, and
it further serves to indicate differences in the amount
formed in different cultures. Twenty-five milligrams of
a one per cent, solution of methyl mercaptan suffice
gradually to alter the isatin red solution (about fifty cubic
centimeters) to a deep green in the course of ten minutes.
Reactions as strong as this are occasionally obtained
from one hundred cubic centimeters of a bacterial cul-
ture in the course of five or ten minutes. Frequently

1 This method has been used by Niemann, ("Ueber die Abspal-
tung von Kohlensaure, Mercaptan und Schwefelwasserstoff beim
Kochen einiger animalischen und vegetabilischen Nahrungsmittel,"
Archiv f. Hyg., xix, p. 126, 1893); also by Bauer, ("Ueber die
Einwirkung gespannter Wasserdampfe auf Keratin," Zeitschr. f.
physiol. Chem., xxxv, p. 346, 1902).

138 INFECTIONS OF THE DIGESTIVE TRACT

one obtains after the lapse of fifteen minutes only a
browning of the isatin-sulphuric-acid solution with no
definite appearance of a green tint. This appearance is
probably to be interpreted as dependent on a mere trace
of methyl mercaptan.

The method used for the determination of indol
obtained from distillation of putrefactive mixtures is
that based on the formation of a color compound due to
the combination of two molecules of indol with one
molecule of /3-napthaquinone-sodium-monosulphonate.
This method has proved in a high degree serviceable
and may be recommended on account of its accuracy
and quickness. The main features of this method
are as follows: —

The method is dependent upon the almost quantitative
precipitation of indol in alkaline solution by a dilute
solution of /3-naphthaquinone-sodium-monosulphonate.
A dilute solution of indol (1 : 500,000 parts of water)
made slightly alkaline with potassium hydroxide, gives
with one drop of a two per cent, solution of /3-naph-
thaquinone-sodium-monosulphonate a blue or green-blue
color. A more concentrated solution of indol will give
a precipitate which upon examination will be found to
consist of well-defined acicular needles, bluish in color
and closely felted together. The compound, evident
either as precipitate or as coloration, is completely
soluble in chloroform, with the production of a red color.
From this extract the indol may be determined quantita-
tively either by evaporating the chloroform or by match-
ing the color to a standard in the Duboscq colorimeter.

INFECTIONS OF THE DIGESTIVE TRACT 139

A slight blue precipitate forms in solutions containing
one part of indol to 256,000 parts of water; in greater
dilution the coloration is green, and fails entirely when
the dilution is 1 : 1,024,000 parts. Even in this extreme
dilution chloroform will indicate the presence of the di-
indyl - di - hy dronaphthalme - keto - sodium - monosulpho-
nate by its faintly pink color. The reaction is essentially
a time reaction, and even in dilute solution it has been
found desirable to allow a period of not less than ten
minutes to elapse before shaking out the condensation
product with chloroform. The naphthaquinone solution
should be added in sufficient excess to tinge the filtrate
slightly yellow.

As indol is sometimes associated in the course of putre-
faction with skatol (and this is not uncommonly so in
the case of the contents of the large intestine), it becomes
important to have a method for the separation of indol
and skatol. These putrefactive decomposition products
may be separated by means of their picrates, but the
method involves so much time as to make it unfit for
clinical or ordinary chemical investigations. It is
believed that the method described by Herter and
Foster l will prove useful in effecting the quick and nearly
complete separation of indol from skatol. Moreover,
this method serves also for the determination of the
quantity of skatol present.

The procedure is based on the fact that by means of
the naphthaquinone method for indol which has just

1 " On the Separation of Indol from Skatol and their Quanti-
tative Determination," Journ. Biol. Chem., II, p. 267, 1906.

140 INFECTIONS OF THE DIGESTIVE TRACT

been described, it is possible to remove almost completely
the indol from a solution containing both indol and skatol
and that the skatol remaining after the removal of
the indol can be distilled and recognized by means of
the dimethylamidobenzaldehyde reaction described by
Ehrlich. If one takes a putrefactive mixture containing
both indol and skatol, these bases should first be dis-
tilled either in acid or alkaline solution — preferably
with the aid of steam if the quantities are considerable.
In the distillate the skatol passes over earlier than the
indol, as can easily be shown by means of the blue color
which it gives on boiling with Ehrlich's aldehyde. To
the distillate containing the indol enough sodium
or potassium hydroxide is added to render it slightly
alkaline. An excess of the /3-naphthaquinone-sodium-
monosulphonate is now added to this solution. As
already described, this substance in the course of a few
minutes reacts almost completely with the indol present,
but not with the skatol. The blue or purplish-blue
precipitate of the newly formed indol-naphthaquinone
compound is now removed by filtration. In cases where
the concentration of indol is too small to give rise to
a precipitate when treated with the naphthaquinone
compound the solution simply develops a green or
greenish-blue color. The solution is now acidified and
subjected to distillation, with or without the use of steam.
The skatol passes over into the distillate, whereas the in-
dol is held back in the form of the indol-naphthaquinone
compound, with the exception of a very small uncom-
bined portion, which passes over with the skatol. The

INFECTIONS OF THE DIGESTIVE TRACT 141

amount of indol, however, which passes over after treat-
ment with the /3-naphthaqumone-sodium-monosulpho-
nate is so small that it is practically negligible, although
its presence is detectable through the red color which it
gives when acted upon by dimethylamidobenzaldehyde.
The distillate containing skatol is boiled with a solution
of Ehrlich's aldehyde in sulphuric acid.1 A slight amount
of dilute hydrochloric acid is now added and has the
effect of intensifying the blue color produced by boiling
the skatol with the Ehrlich aldehyde solution. A little
experience is required to find the amount of hydro-
chloric acid which gives the maximal intensification of
the reaction. An excess of hydrochloric acid causes the
blue color to fade. It is important to use an excess of the
Ehrlich aldehyde solution in order to develop fully the
color reaction with skatol. The color obtained through
the action of Ehrlich's aldehyde upon skatol is purple-
blue rather than blue so long as the solution is hot.
On cooling it under the tap, the blue color asserts itself
more strongly, and the solution may become some-
what opalescent from the separation of uncombined
dimethylamidobenzaldehyde. Chloroform is now added
to the solution containing the blue product. On agita-
tion with the solution this carries out the blue color,
and the chloroform assumes a pure blue tint. By means
of a good colorimeter the quantity of skatol present
in the original solution may be approximated by the
intensity of the color reaction.
On evaporating the chloroform containing the blue
1 Five per cent, aldehyde in ten per cent, sulphuric acid.

142 INFECTIONS OF THE DIGESTIVE TRACT

color resulting from the action of the dimethylamido-
benzaldehyde on skatol, one obtains an amorphous blue
material which can be partially purified from the ad-
mixture with Ehrlich's aldehyde by the use of petroleum
ether. The nature of this compound is not at present
known. The melting point of our preparation lay be-
tween 65° C. and 66° C.

The method of separating indol and skatol here de-
scribed has been used in a routine way during the past
year in connection with the study of the faeces and has
given satisfaction. To twenty-five grams of the mate-
rial have been added twenty cubic centimeters of water
and one to two cubic centimeters of a ten per cent,
sodium hydroxide solution. The suspension is then
subjected to distillation with the aid of steam, until the
distillate no longer gives a color reaction when boiled
with the dimethylamidobenzaldehyde solution. Ordina-
rily the indol and skatol present go over completely within
an hour, but where the material develops frothing a more
prolonged distillation may be necessary. When the
distillate has been obtained, it is treated in the manner
above outlined for the separation of indol and skatol.
In some cases one obtains only skatol from the faeces,
but as a rule indol is also present. In the presence of
indol the chloroform extract, instead of being a pure
blue, may have a slightly purplish tinge, owing to una-
voidable admixture with a slight amount of indol.

It is important in making the colorimetric estimations
of the quantity of skatol present to employ a standard
color solution for comparison with the color obtained

INFECTIONS OF THE DIGESTIVE TRACT 143

from the distillate containing the skatol. Various dyes
have been tried with a view to obtaining a standard
solution which will retain its color unchanged. Ex-
perience has shown that the best standard color solution
is one obtained from a solution of skatol. Although
such a solution fades after a few days, especially when
exposed to the light, and may assume a greenish tint,
in the dark it may last several weeks without undergoing
appreciable change. Moreover, as the skatol standard
solution is readily prepared, there is little disadvantage
from being compelled to renew the color solution from
time to time. This solution may be conveniently pre-
pared by dissolving five milligrams of skatol in water and
acting upon it with an excess of dimethylamidobenzalde-
hyde. It commonly requires from one hundred to one
hundred and fifty cubic centimeters of chloroform to
extract completely the blue coloring matter which has
already been described. The quantity of coloring mat-
ter present is sufficient to impart a deep blue color to
this volume of chloroform. Ten cubic centimeters of
this solution are placed in the receptacle of the Duboscq
colorimeter and used as a standard for comparison with
the chloroform color solution obtained from the distil-
late to be tested. The matching of the colors can
usually be made very closely. In cases where the
quantity of skatol is so small that the trace of indol
present influences the color of the chloroform solution,
changing it to violet or even purple, it is more difficult
to obtain a satisfactory matching of colors. In this
case is may be necessary to add a small quantity of indol

144 INFECTIONS OF THE DIGESTIVE TRACT

to the skatol employed in making the standard color
solution. This then imparts to the standard color solu-
tion a violet tint like that obtained from the distillate
to be matched. It seems unnecessary to give further
details. After some experience with the method of match-
ing colors it is possible to employ the method so that it
will give satisfactory quantitative results.

It is not desirable here to enter into a description of
the chemical methods of studying the f seces and the urine.
The methods ordinarily employed for this purpose are
fully described in text-books relating to these subjects.
It may be stated, however, that numerous observations
have been made upon the acids and bases present in
human fseces in health and disease and the intestinal
contents have been studied for hydrogen sulphide and for
the presence of mercaptan. The reaction of the fseces
with concentrated solution of mercuric chloride, accord-
ing to the manner first described by Schmidt,1 is of much
value in determining the presence of a substance sup-
posed to be hydrobilirubin. The marked red color
with yellowish fluorescence which is a characteristic of
Schmidt's reaction is supposed to depend on the oc-
currence of a combination between the mercuric salt
and hydrobilirubin. An analogous combination exists
in Jaffe's zinc-chloride-urobilin compound. Both sub-
stances give the same spectroscopic picture; namely,
a band between the lines b and E. An intensification
of the red color obtained in the Schmidt reaction may
often be secured by transferring a portion of the tested

1 " Verhandl. d. Congresses f . inn. Medicin," xiii, p. 320, 1895.

INFECTIONS OF THE DIGESTIVE TRACT 145

material to a beaker of distilled water. Under these
circumstances the color compound diffuses out into the
water and imparts to it a distinct and often strong red
color which is frequently more pronounced than one
would have predicted from the appearance of the mate-
rial which has developed the red color under the influ-
ence of the Schmidt procedure as ordinarily carried out.

In the course of a systematic examination of the faeces
from normal individuals and from many pathological
series it was noted that the mercuric chloride reaction
was strongest in persons suffering from intestinal dis-
orders, especially in those with intestinal putrefaction.
The weakest reactions were found in the case of alcoholic
stools and in those of children and young adults present-
ing only slight indications of intestinal putrefactive
decomposition (low ethereal sulphates, absence of indican,
and low phenol).

The Dimethylamidobenzaldehyde Reaction of the Fceces.
- If one extracts human faeces with an aqueous sodium
chloride solution (two grams of fseces to twenty grams
of 0.85 per cent, sodium chloride solution), the extract
will usually yield a color reaction with a suitably
prepared acid solution of Ehrlich's aldehyde.1 In the
case of healthy children or adolescents on a mixed diet
the color obtained is usually a light rose, and may be very
faint. Between this light tint and a very deep cherry-
red, all transitions are met.2

1 Water, 270 c.c.; concentrated H2S04, 30 c.c.; Ehrlich's alde-
hyde, 15 grams.

2 We have employed a graded color scale in order to record our
results with some degree of accuracy.

146 INFECTIONS OF THE DIGESTIVE TRACT

The explanation of the chemical basis of the Ehrlich
aldehyde reaction of the faeces is not yet wholly sat-
isfactory. Baumstark l thought it could be ascribed
to the indol of the faeces and based a quantitative
method for indol on the reaction. Bauer,2 however,
showed that the faeces contain another substance which
reacts with the aldehyde, and claims that this substance
is urobilinogen. I reached a similar conclusion inde-
pendently, after noticing that the faeces in one instance
gave an intense red reaction after the indol had been
distilled off. In other cases a strong reaction was ob-
tained in spite of the fact that the faeces were free from
indol from the outset. That the reaction from this non-
volatile part of the faeces depends wholly on urobilino-
gen does not appear to me to have been convincingly
shown. It is true, however, that one may reduce
urobilin (Schuchardt's) with alkali and zinc dust and
thus obtain a substance which gives a stronger and
more characteristic Erhlich aldehyde reaction than
the urobilin itself. Probably both urobilinogen and
a skatol derivative are implicated in the Ehrlich

1 " Bestimmungen der Faulniss Produkte im Urin und in dem
Faces mit Beniitzung der Ehrlichschen Aldehydreaktion,"
Munch, med. Wochenschr., 1, p. 722, 1903; also Arch. f. Ver-
dauungskrankh., ix, p. 201, 1903.

2 This distillation should be conducted in an atmosphere of
carbon dioxide in order to prevent oxidation of the reacting sub-
stance, which is sensitive both to air and to the action of sunlight.

Bauer fell into error in claiming that the reaction from the
faeces is not due to indol and that the method does not serve
for the detection of this substance. I am confirmed in this
view by Ury ("Die Ehrlich'sche Reaction im Stuhl," Zentralbl.
f. innere Medizin, xxvii, p. 41, 1906).

INFECTIONS OF THE DIGESTIVE TRACT 147

aldehyde reaction of the urine, but it is possible that
other substances are also concerned.

The relation of the hydrobilirubin reaction to the
urobilinogen reaction of the faeces with Ehrlich's
aldehyde is a point of interest on which one cannot
at present express an opinion. There are instances in
which the faeces, after distillation of all the indol pres-
ent (as shown by the aldehyde reaction, which is very
delicate), still give a reaction in the cold with a di-
me thylamidobenzaldehyde solution. This reaction is
believed to depend on the presence of urobilinogen. It
has been fairly well marked in some of our anaemia
cases, but there is as yet no evidence that the reaction
bears any definite relation to the hydrobilirubin
reaction, although the two substances urobilinogen and
urobilin (hydrobilirubin ?) are closely related chemically.

The Dimethylamidobenzaldehyde Reaction of the Urine.
— It is well known that on the addition of acid
solution of dimethylamidobenzaldehyde to certain
urines a red color is obtained, sometimes in the cold,
more often only on the addition of heat. The explana-
tion of the reaction has been the occasion of considerable
discussion. Ehrlich * was inclined to attribute it to
glycosamin; Neubauer2 and Bauer3 referred it to

1"Ueber die Dimethylamidobenzaldehydreaction," Med.
Woche, p. 151, 1901.

2 " Ueber die neue Ehrlich'sche Reaction mit Dimethylamido-
benzaldehyd," Sitzungsber. d. Gesellsch. /. Morphol. u. Physiol., ii,
p. 32, 1903.

3 "Die Ehrlich'sche Aldehydreaktion im Harn und Stuhl,"
Zentralbl. f. inn. Med., xxvi, p. 833, 1905.

148 INFECTIONS OF THE DIGESTIVE TRACT

urobilogen, which has passed from the intestine into the
urine. I have found that the administration of skatol
to men and to monkeys is followed by some intensifica-
tion of the Ehrlich aldehyde reaction, and believe that
where skatol is produced in the intestine it may con-
tribute to intensify the reaction.1 Dr. Adolf Meyer,
Director of the Pathological Institute of the State
Hospitals for the Insane, tells me that in some observa-
tions made under his direction the administration of
skatol to patients has been followed by a similar though
slight intensification of the dimethylamidobenzaldehyde
reaction of the urine.

This reaction is seldom found in its most intense form
except in the urines from persons suffering from patho-
logical conditions. It was noticed by Ehrlich that the
intensity of the reaction is apt to be great in the case
of patients with phthisis, typhoid fever, and chronic
enteritis, and Clemens observed that persons with digest-
ive disorders are among those whose urines are likely
to give a strong reaction. It has appeared to me that
the urines of persons with the urinary evidences of exces-
sive chronic intestinal putrefaction (excessive indican and
phenol and high ethereal sulphates) are especially liable
to exhibit the deepest cherry-color tints on being treated
with the reagent. The urine of normal children, in
whom the processes of intestinal putrefaction are apt
to be mild, usually give negative or almost negative

1 " On a Relation between Skatol and he Dimethylamidobenz-
aldehyde Reaction of the Urine," Journ. Biol. Chem., i, p. 251,
1905.

INFECTIONS OF THE DIGESTIVE TRACT 149

results. Although I have come to associate an intense
Ehrlich aldehyde reaction of the urine with the occurrence
of a high degree of intestinal putrefaction, it must be
admitted that not every case of excessive intestinal
putrefaction shows this reaction in its very marked
form.1

1 A good discussion of Ehrlich's dimethylamidobenzaldehyde
reaction in the urine by C. E. Simon may be found in the American
Journal of the Medical Sciences, cxxvi, p. 471, 1903. Bauer (loc.
tit.) brings forward evidence to support the view that the Ehrlich
reaction depends on urobilinogen. This view appears to me not
irreconcilable with the fact that the administration of skatol
intensifies or even gives rise to this reaction, since more than one
factor may be operative in certain instances. I am not certain
of the identity of the reactions obtained by Bauer and those
dependent on the administration of skatol, as I have failed to
make use of spectroscopic methods of study.

COMMON BACTERIAL INFECTIONS OF THE DI-
GESTIVE TRACT, CONSIDERED FROM THE
STANDPOINT OF THE MICROORGANISMS

THE COLON-TYPHOID-DYSENTERY GROUP

A SURVEY of the bacteria which we may include in
the colon-typhoid-dysentery group is beset with many
difficulties with respect to satisfactory classification,
arising from an incomplete knowledge of the biological
characters of these bacteria. We know enough, however,
to see that we are dealing with microorganisms not sep-
arated from one another by unbridged gaps, but rather
linked in series by most intimate affinities.

Colon Bacilli. — If we ask the question, what patho-
logical consequences do the colon bacilli of the human
intestine bring to the human species, the answer is
by no means simple. Many conditions formerly at-
tributed to the colon bacilli were given their position
without adequate reason. To-day we have the right to
insist on knowing the fermentative and agglutinative
as well as the cultural characters of the colon organisms
which are stated to be the causes of disease, and too
often this information is not given.

At the present time the view is generally held by
pathologists that colon bacilli are capable of doing
damage in two ways, as infective agents outside the

150

INFECTIONS OF THE DIGESTIVE TRACT 151

digestive tract and as infective agents inside the tract.
We are not concerned here with those examples of colon-
bacillus infection in which the organisms find their way
into the urinary tract from the exterior and set up a
cystitis or pyelitis. Nor shall I do more than refer to
those infections in which colon bacilli have passed during
life through some local spot of injury in the digestive
tract into the blood or the peritoneal cavity, setting up
in the former case a septicaemia or pyaemia, and in the
latter a peritonitis. With less serious consequences
than these, the colon organisms probably often pass
temporarily from the digestive tract by way of the blood,
to the kidney, with a resulting pyelonephritis or pyelitis,
or at least a bacteriuria. It is noteworthy that the
localization of colon bacilli under such pathological
conditions as these is often attended by a heightened
virulence for small animals as compared with the vir-
ulence of the original bacteria of the intestine.

The question which does especially interest us here is
whether the colon bacilli of an individual ever do injury
to that individual while still confined to the digestive
tract. In attempting to answer it, I must divide the
question into two — one relating to acute effects of the
colon bacilli, the other relating to prolonged pathological
activities.

If we go back to the time when Escherich first de-
scribed B. coli commune as an obligate inhabitant of the
human digestive tract, we find that there was a dis-
position to regard it as the cause of various diarrhreal
affections. A close examination of facts relating to the

152 INFECTIONS OF THE DIGESTIVE TRACT

identity of the organisms in many instances makes it
clear that affections attributed to colon bacilli were thus
classed without sufficient justification. For example,
many instances of supposedly colon-bacillus diarrhoea
must almost certainly have been due to the dysentery
bacillus. Remembering the cultural similarities of certain
colon bacilli and dysentery bacilli, it is easy to under-
stand how confusion might have occurred when the
fermentative characters were left out of account, as
was usually the case. Even an organism so different
from B. coli as the bacillus of typhoid fever was con-
founded with it in the early days of bacteriology, when
limited cultural criteria alone were employed. Thus
progress in methods of identification has step by step
circumscribed the conception of the colon bacillus and
given it sharper definition. There are indications that
this process of circumscription has not reached its final
stage.

It seems established that organisms having the cultural
fermentative and agglutinative characters of B. coli at
times are endowed with a degree of virulence for small
animals which does not pertain to colon bacilli derived
from healthy human individuals. Such pathogenic
bacilli have been found in great abundance in some
diarrhceal conditions and appear to have been the cause
of these acute derangements. It is not clear whether
in such cases the dominant normal organisms have
acquired increased irritant properties or whether there
has occurred a process of substitution by which organisms
present normally in small numbers or introduced from

INFECTIONS OF THE DIGESTIVE TRACT 153

outside temporarily gain the numerical ascendency in
consequence of alterations in the nutrient media which
favor this race while repressing the ordinary dominant
one. In addition to a colon-bacillus diarrhoea associated
with colon bacilli of more than ordinary virulence, it is
probable that we must recognize a similar derangement
in which there is no increase of virulence on the part of
the bacilli, but simply an enormous physiological increase
which verges on the pathological. Almost any healthy
child who has eaten a largely excessive quantity of
ripe fruit, rich in sugar, will develop loose movements in
which colon bacilli are present in very large numbers. I
do not know whether these organisms are any more viru-
lent than the type dominant previous to the excessive
indulgence, but it seems extremely unlikely that they
suffer any appreciable change in the course of the short
interval of excessive fermentative activity. Acetic,
lactic, and other organic acids are formed in the course
of the fermentative process rendered possible by the
sugar-holding fruit, and it is to the irritant action of
these acids, and perhaps also to an excessive liberation
of gas, that the symptoms must be ascribed. Very
likely, too, in a case like this, bacilli of the group of B.
lactis aerogenes aid in the fermentation. In this instance
we are dealing clearly with a process on the borderland
between the physiological and the pathological.

Of the influence exerted by colon bacilli in the course of
chronic processes within the digestive tract we have as
yet little knowledge, partly because etiological relations
are very difficult to establish in the case of obligate

154 INFECTIONS OF THE DIGESTIVE TRACT

bacteria. Attention has lately been called by Dr.
Turck * of Chicago to the fact that he has experimentally
induced in dogs the formation of ulcers of the stomach
in many respects comparable to those occurring in man.
The fact is in itself one of considerable interest to pathol-
ogists, although its bearing on the etiology of human
gastric ulcers is obscure. The reported experimental re-
sults were obtained by feeding dogs, during considerable
periods, bouillon cultures of bacilli derived from patients
with gastric ulcer. The bacilli are stated to have been
colon bacilli, but it is regretable that the cultural and
biochemical characters of the bacteria are not mentioned.
Moreover, while it seems likely that human colon bacilli
may in disease migrate to the gastric region and multi-
ply freely there, this is not now known to be the case in
ulcer.2

In considering the part played by the colon bacilli in
chronic disorders one fact stands out distinctly and gives
us a clew to the nature of the activity of these bacilli.
Human colon bacilli have practically no power to dis-
solve and peptonize native proteids, such as casein or
egg albumen. I conclude from experiments which I

J" Ulcer of the Stomach; Pathogenesis, Pathology," Journ.
Amer, Med. Assoc., xlvi, p. 1753, 1906.

2 The association of colon-bacillus infection of the stomach with
gastric ulcer would not of course prove an etiological relation, as
the infection might be a secondary phenomenon. Moreover, it
must be remembered that the production of ulcers in dogs by means
of human colon bacilli involves the action of foreign organisms.
The presence of the obligate and homologous colon bacilli of man
in the human stomach is thus not strictly comparable with the
conditions induced in Turck's experiments.

INFECTIONS OF THE DIGESTIVE TRACT 155

have made on milk that the peptonizing action of these
bacilli is extremely slight, though not absolutely lacking.
On the other hand, the colon bacilli are able to cleave
energetically peptones prepared by other microorgan-
isms. Then they give rise to familiar products of pu-
trefaction such as ammonia, volatile fatty acids, phenol
and indol, and hydrogen sulphide. These facts have an
important bearing on the significance of these bacteria
within the digestive tract, which may be summarized
as follows. If there is good absorption of proteids above
the lower ileum, that is, above the level where there is
a predominance of colon bacilli, little proteid material
finds its way into the colon. Hence whatever may be the
nature of the bacterial inhabitants of the large intestine,
but little putrefactive decomposition occurs there. But
if, owing to a superabundance of proteid food consid-
erable native proteid finds its way into the colon, its
fate depends on the character of the flora there. In a
normal intestine, containing few peptonizing bacteria,
there will be little putrefaction, because the dominant
races, members of the coli group, cannot initiate active
putrefaction of native proteids. But if there be present
also putrefactive anaerobes capable of peptonizing pro-
teids, the colon bacilli take an active part in breaking
down the hydrolyzed proteids with which they are thus
supplied. In this way the colon bacilli frequently be-
come active participants in excessive intestinal putre-
faction. I shall recur to this subject in the discussion
of the putrefactive anaerobes.
The disappearance of typical colon bacilli from the

156 INFECTIONS OF THE DIGESTIVE TRACT

fasces of persons in whom such bacteria have previously
been present is a phenomenon of much interest. It is
observed in a variety of pathological conditions : in cases
of excessive saccharo-butyric putrefaction due to infec-
tion with B. aerogenes capsulatus, in some cases of
mucous colitis, in certain instances of dysentery due to
the Shiga bacillus. Probably it occurs also in cholera
and typhoid fever.

The disappearance of B. coli from the contents of the
lower bowel is in itself no proof that these organisms
are not present higher up, but I think it presumptive
evidence that the growth of these obligate bacteria is at
least more restricted at higher levels than is normally
the case. Frequently the failure to find living colon
bacilli in the faeces is representative of the entire gastro-
enteric tract, since in these cases diarrhceal movements
(spontaneous or due to purgatives) also fail to contain
them.

One is justified in suspecting a complete or partial
suppression of B. coli if the mixed faecal flora fails to
form gas in sugar-bouillon fermentation tubes or makes
gas in smaller quantities than is usual for normal flora.
This suspicion, if correct, is confirmed by the failure of
colon bacilli to appear on litmus gelatin plates or by
their appearance in abnormally small numbers.

The possibility that some cases of mucous colitis are
due to colon bacillus infection has been suggested by
Sir A. E. Wright, who tells me that he has cured certain
obstinate and extreme instances of this affection by
inoculations made with killed colon bacilli. The bacilli

INFECTIONS OF THE DIGESTIVE TRACT 157

used for inoculation were obtained in one instance from
the cystitic urine of the patient. The study of mucous
colitis from this standpoint appears promising. It is
very desirable that the biochemical characters of the
bacilli concerned in this disease should be well established.
The possibility has not been excluded, I think, that in
the cases mentioned by Dr. Wright the organisms con-
cerned were not typical colon bacilli, but closely related
pathological varieties.

Typhoid Bacilli. — It has just been pointed out that
organisms of the strict colon bacillus group are not at
present known to be the cause of severe acute intestinal
diseases and that their chief role in chronic disorders
of the intestine depends on their participating in putre-
factive decompositions. We may briefly consider a
microorganism which, while closely affiliated to the colon
bacillus, is morphologically and in cultural characters
different from it in important biochemical properties
and is immeasurably more pathogenic for the human race.
This is the bacillus of typhoid fever discovered by Eberth
and Gaffky independently in 1884. This organism as
compared with the colon bacillus is far less hardy in its
growth on ordinary culture media and requires much
more special conditions in order to thrive — conditions
which it often finds in the human body. We do not know
in what relation it stands to the colon bacillus in the long
process of natural evolution. The claim that colon
bacilli can be experimentally transformed into typhoid
bacilli has never been made good, and still we cannot
deny that what bacteriologists cannot do in the laboratory

158 INFECTIONS OF THE DIGESTIVE TRACT

may possibly have been done by nature and may
perhaps be still in progress. The existence of bacteria
intermediate between colon bacilli and typhoid bacilli -
the paratyphoid and paracolon bacteria — is now es-
tablished, and the fact that there are these transitional
forms suggests that they may have been descended
from a common ancestor through successive slow modi-
fications.1

1 The studies of Doebert ("Die verwandtschaftlichen Bezie-
hungen zwischen den Bacillus faecalis alkaligenes und dem Ty-
phusbazillus," Archiv f. Hyg., lii, p. 70, 1905) indicate that
there exists a close resemblance between the bacillus of typhoid
fever and Petruschky's B. fcBcalis alcaligenes. A culture of the
latter organism furnished by Petruschky rendered litmus milk
blue in twenty-four hours, forming a pellicle on the surface of the
milk, and on potato gave a yellowish brown growth in one day.
With typhoid serum the organism agglutinated in 1 : 800. It is
stated that it was easy to modify the organism of Petruschky by
passing it through guinea-pigs so that it acquired the cultural
characters of the true typhoid bacilli. The organism modified by
Doebert experimentally by passage through guinea-pigs was not
agglutinated by the serum from guinea-pigs in 1 : 100, but was
agglutinated by typhoid serum in a dilution of 1 : 15,000. An-
other strain of B. fcecalis alcaligenes studied by Doebert could not
be converted in this manner into an organism so closely resembling
the typhoid bacillus by means of passage through animals.

There appears to be little doubt that B. fcecalis alcaligenes is
sometimes responsible for the occurrence of a febrile disease at
present indistinguishable from mild forms of typhoid fever on
purely clinical grounds. Though closely related to the bacillus
of typhoid fever, it appears proper to regard the bacillus of Pe-
truschky as a distinct bacterium. (For details as to the distinction
of B. fcecalis alcaligenes and B, typhi see Piorkowski, "Zur Differ-
enzierung der Typhusbacillus und Bacillus fsecalis alkaligenes,"
CentralbL /. Bakt., I Abt., Orig., xl, p. 437, 1905-06.) It seems
most reasonable to look upon B. fcecalis alcaligenes as a variant of
the typhoid bacillus which has lest the capacity to act fermenta-
tively on dextrose. Something might be said in favor of regarding

INFECTIONS OF THE DIGESTIVE TRACT 159

While colon bacilli possess in a marked degree certain
commonplace properties such as the ability to coagulate
milk, to reduce neutral red, to make indol, and to ferment
sugars, with the liberation of gas and acids, these prop-
erties have been lost in the case of the typhoid bacilli,
excepting only the capacity to form acid from sugars.
In respect to these characters, the paracolon and para-
typhoid bacilli occupy an intermediate position. The
loss of these relatively banal properties has been asso-
ciated with the development of pathogenic properties
for man — properties which are doubtless dependent
on substances chemically well defined but at present
quite unknown. It has, for several years, been recog-
nized by students of immunity that in respect to the
formation of these toxic substances there is a striking
inconsistency in the behavior of the typhoid bacilli
in the body and in artificial culture media; for in the
former case, while they make poisons oftefi sufficiently
active to kill, in the latter only feeble poisons are obtain-
able. In partial explanation of this the view was ad-
vanced that, unlike the bacilli of diphtheria, which make

it as a hog-cholera variant or even as an extreme modification of
B. coli. Its active motility, however, places it nearest to the
typhoid or paratyphoid groups. Brion and Kayser ("Die nosolo-
gische Stellung des Symptomkomplexes 'Abdominaltyphus,'"
Deutsches Archiv /. klin. Med., Ixxxv, p. 552, 1905-06) have de-
scribed an organism differing distinctly from typhoid bacillus,
but which was obtained from a patient showing a fever clinically
resembling a mild typhoid. From the blood of this patient an
organism was cultivated which agglutinated with the patient's
serum in a dilution of 1 : 500. It had the property of liquefying
gelatin and formed yellow colonies. They gave to it the name
of Bacterium flavosepticum.

160 INFECTIONS OF THE DIGESTIVE TRACT

soluble poisons diffusing readily from the bodies of the
bacteria, the typhoid bacilli hold their poisons tena-
ciously in their bodies, and that these endotoxins may
be liberated in the body of the host, when the bacterial
bodies are dissolved, but are not so liberated in ordinary
culture media. A more adequate hypothesis was formu-
lated by Professor Welch. According to this, the typhoid
bacilli are stimulated in the body by means of its defen-
sive juices to form and liberate their poisons. Such a
stimulus being wholly lacking in artificial culture media,
where their existence is not actively threatened, the
bacilli are not forced to make their poisons. It may be
that the difference in the poisons formed in the body
and in culture media is mainly one of quantity rather
than of kind. There is at least some ground for think-
ing that the character of the immunity is the same,
whether the poison comes from the bodies of bacteria
in culture or from similar bacteria under the stimulus
of the protective juices of an animal.

The word "aggressins" has been introduced by Kruse
and Oscar Bail 1 to designate substances of unknown
constitution which are formed by most pathogenic
bacteria when in antagonism with the animal organism,
as for example when introduced into the peritoneal
cavity of a guinea-pig — substances which break down in
some way the natural defenses of the body, so that the
bacterial invaders, previously held in check, become
reproductively and invasively active. The "aggressins"

1 " Untersuchungen liber Typhus und Cholera Immunitat,"
Archiv /. Hyg., lii, p. 272, 1905.

INFECTIONS OF THE DIGESTIVE TRACT 161

were obtained from the peritoneal exudate and carefully
separated from cellular elements by the centrifuge.
It was found that numbers of typhoid or cholera bacilli
below the fatal dose become fatal when injected simul-
taneously with "aggressins," that it is possible by means
of "aggressins" to inhibit the protective action of a
bacterial immune serum in the peritoneal cavity of the
guinea-pig, and, finally, that one may by suitable treat-
ment with "aggressins" induce immunity to fatal
doses of typhoid or cholera bacilli. Bail regards this
"aggressin" immunity as something distinct from a
bactericidal immunity because a bacterial immune serum
does not protect against "aggressins," but very recent
studies by Wassermann and Citron 1 indicate that
"aggressins" are really bacterial extracts, and that their
influence in promoting infection depends on their ability
to bind defensive substances largely derived from leu-
cocytes and designated complements in Ehrlich's ter-
minology. It is not yet clear whether the "aggressins"
really act through binding complements in vivo or not,
and this is a crucial point.2 Hence we cannot feel certain
how to interpret Sir A. E. Wright's observations on the
immunization of English soldiers against typhoid fever,
in which he met with at least a measure of success by

1 Wassermann and Citron, " Ueber die Bildungsstatten der
Typhusimmunkorper," Zeitschr. f. Hyg., 1, p. 331, 1905; Citron,
"Ueber nattirliche und kiinstliche Aggressine," Centralbl. f.
Bakt., I Abt., Orig., xli, p. 230, 1906; "Die Immunisierung gegen
die Bakterien der Hogcholera (Schweinpest) mit Hilfe von Bak-
terienextrakten," Zeitschr. f. Hyg., liii, p. 515, 1906.

2 Wolff-Eisner, "Die Aggressinlehre," Centralbl. f. Bakt., I Abt.,
Ref ., xxxviii, p. 641, 1906.

K

162 INFECTIONS OF THE DIGESTIVE TRACT

vaccinating with typhoid bacilli killed by heating to
60° to 65° C. — a procedure which he dared to practice
after learning from Pfeiffer that he had in man obtained
the specific agglutination reaction of typhoid fever after
the subcutaneous inoculation of a heated culture of
typhoid bacilli. Chantemesse in 1904 obtained ap-
parently favorable results in the treatment of typhoid
fever by the use of a serum that developed in horses
treated with a filtrate from typhoid bacilli which had
grown on a medium containing splenic pulp and human
defibrinated blood. It is possible that this serum acts
by binding the "aggressins" made by the typhoid bacilli
as they grow, thus removing in some degree a condition
favorable for the multiplication of the typhoid bacilli.

We may conveniently classify bacteria in a rough
way according to the degree of parasitism they exhibit.
Thus the plague bacillus in man and the anthrax bacillus
in guinea-pigs are highly parasitic, for they find in their
hosts the conditions that permit their rapid multiplica-
tion and invasion, although the original infection may
have been established by very few bacteria, perhaps even
by a single bacterium. Directly opposed to the parasites
stand the saprophytic bacteria, which have the utmost
difficulty in multiplying in the tissues of their host
even when introduced in large numbers, but may prove
injurious if inoculated in enormous quantities. Many
of the bacteria of the intestinal tract are of this character.
Midway between these groups stand the hemiparasites,
microorganisms which are invasive only when inoculated
in considerable numbers and are pathogenic only when

INFECTIONS OF THE DIGESTIVE TRACT 163

the way has been prepared by poisons which neutralize
the defensive substances of the host; for example, by
"aggressins." The bacilli of typhoid fever and of cholera
may be classed as hemiparasites, and their virulence
appears to stand in a close relation to the efficiency of
their "aggressins."

The ordinarily accepted view that the portal of infec-
tion for the inciters of typhoid fever is the digestive
tract is doubtless true in general and still it may not
represent the whole truth. There is some evidence to
support the idea that at times the bacilli enter by way
of the throat, for the organisms have been found in the
tonsil, and a patient may be in the early period of the
disease without showing the presence of the bacilli in
the stools even when modern methods of isolation are
employed. It is conceivable that typhoid bacteria in
the intestine do not necessarily point to an original
intestinal invasion, as they may at times come mainly
from ulcerated Peyer's patches in which the bacilli
have accumulated as a result of the invasion of these
lymphatic structures in the course of a general bacillary
infection which strongly tends to implicate the lym-
phatics. It would be of special interest to know whether
the axillary and other lymph nodes at a distance from
the intestine are early invaded in typhoid fever. Even
where the invasion is by the intestinal path the typhoid
bacilli, which tend to be irregularly distributed in the
intestinal contents, may perhaps come from ulcerated
Peyer's patches which have become infiltrated and ulcer-
ated, secondarily to the typhoidal bacilla3mia. We should

164 INFECTIONS OF THE DIGESTIVE TRACT

then be dealing with an eliminative process for bacteria,
such as pharmacologists have long known to exist for
poisons like carbolic acid, corrosive sublimate, and urea,
and such as Dr. Flexner has shown to exist for ricin
poisoning and for intoxications arising from the auto-
lyzed dysentery poison.

That the blood contains typhoid bacilli early in the
course of the fever is now well established for a very
large proportion of all cases appropriately examined, and
is probably true of all cases of typhoidal type associated
with infection by typhoid bacilli. This bacillsemia
explains many of the complications and sequelae of the
disease, the passage of bacilli into the spleen with en-
largement of the organ, their passage into the liver and
hence into the gall-bladder (frequently with cholecystitis,
with or without gall-stone formation) and the entry
of the bacilli into the urine, where they may be found
for some weeks after the subsidence of the fever. That
these living bacilli should find their way abundantly
into the urine after convalescence has begun shows that
the microorganisms must still have a foothold in the
host, and points to unnumbered possibilities in respect to
relapses and the spread of the bacilli. It could hardly
have been predicted that convalescence is possible
while the living inciters of the disease still persist in the
body of the host. The explanation of this apparent
paradox must be sought in the development of protective
powers on the part of the host, powers which accrue to
the host in consequence of the specific excitation and
reaction of many different types of cells through sub-

INFECTIONS OF THE DIGESTIVE TRACT 165

stances made by the bacilli. It is quite aside from the
present design to discuss problems of immunity, but it
is perhaps desirable to note certain facts relative to
the protective action aroused by typhoid bacilli. The
experimental inoculation of human beings with typhoid
cultures exposed to heat has been shown by Wright to
result in the following changes in the blood; (1) an in-
crease m the bactericidal power of the blood; (2) an
increase in bacterioly tic power of the blood ; (3) a devel-
opment of antitoxic properties as indicated by the miti-
gation of the severe constitutional signs of intoxication
that ordinarily follow a first inoculation with typhoid
vaccine; (4) an increase in the ability of the leucocytes
to take up and digest typhoid bacilli ; and (5) the devel-
opment of specific agglutinating powers of the blood for
typhoid bacilli. With the exception of the agglutinating
action of the blood, which bears no definite relation to the
immunity of the patient, conditions similar to these
apparently constitute at least a part of the mechanism
of defense hi the case of typhoid fever. How far the
bactericidal property of the blood as measured in vitro
is an evidence of effective bactericidal activity within
the body is an open question. Bail emphasizes the fact
that an experimental immune serum which is active
outside the body, does not cause an effective solution
of typhoid bacteria when injected with such bacteria
into the blood stream. The antitoxic power noted by
Wright is very likely dependent on "anti-aggressins"
that have arisen in response to "aggressins." Effective
"anti-aggressin" action is probably extremely important

166 INFECTIONS OF THE DIGESTIVE TRACT

in the struggle betwee^ the organism and the bacteria ;
for if the latter cannot restrain the natural protective
forces of the animal they are attacking, they cannot
efficiently multiply, and they ultimately fall a victim
to the action of the leucocytes. During convalescence,
after the subsidence of fever, the typhoid bacilli found in
the urine can hardly represent bacteria that are multi-
plying in the blood and tissues, but rather are micro-
organisms that have come from the intestine from cer-
tain foci in the tissues and have escaped the action of the
leucocytes and bactericidal serum. A state of balance
between opposing forces then exists, and this may be
readily disturbed to the detriment of the host by factors
favoring multiplication of the bacilli in the intestine
(such as improper food) or checking the formation of
"anti-aggressins." It is a noteworthy peculiarity that
a patient may harbor the inciters of typhoid or cholera
in the intestine and still not develop the specific disease.
The blood in these persons need not contain an increase
in bactericidal substances, and this suggests that the
individual in question has acquired a local immunity
implicating the intestinal structures, which prevents the
invasion of typhoid bacilli. Wassermann and Citron
have shown that immunity substances are made in
connective tissues as well as in the lymphatic system,
spleen, and bone marrow, and furnish evidence pointing
to the possibility of inducing a local intestinal immunity.1

1 For a review of the recent literature relating to typhoid fever,
see Kolle and Wassermann's " Handbuch d. path. Mikroorganis-
men," Supplementary Volume I, 1906; Kutscher, "Abdominal-
typhus," pp. 251, 255, 269, etc.

INFECTIONS OF THE DIGESTIVE TRACT 167

Paratyphoid and Allied Infections. — By means of
careful biochemical studies, including especially the study
of the fermentative action on sugars (first systematically
employed by Professor Theobald Smith *) and the agglu-
tination reactions and pathogenic properties, it has been
found that there is an important group of pathological
bacteria which according to their characters range them-
selves between the group of colon bacilli and the typhoid
bacilli. The group is large and somewhat heterogeneous
as regards certain properties, such- as agglutinative be-
havior and pathogenic qualities. The group includes
the inciters of hog-cholera,2 of the spermophile disease,
of mouse typhoid, of guinea-pig disease, of certain forms
of pseudo-tuberculosis, of calves' diarrhoea, of psittacosis
or parrot plague (which may cause an atypical pneu-
monia in man), and it includes the paracolon and para-
typhoid bacilli, B. enteritidis and Sanarelli's bacillus of
yellow fever.3 A glance at this somewhat formidable
list shows that many of the bacteria in question are the

Gahrungskolbchen in der Bakteriologie," Centralbl.
f. Bakt., vii, p. 504, 1890; "Zur Unterscheidung zwischen Typhus-
und Kolonbacillen," ibid., xi, p. 367, 1892.

2 The agglutinative characters of some of the bacteria in this
group are very similar though probably not really identical.
Thus Smith found the agglutinative behavior of certain hog-
cholera bacilli to be indistinguishable from that of Sanarelli's
bacilli of yellow fever. It is possible that common group agglu-
tinins were so abundant in the sera used as to mask the action of
very specific agglutinins that may have been present.

3 Another group of aberrant colon-like bacilli includes the spe-
cific incitants of various infectious diseases of wild and domestic
birds. These organisms retain the ability to coagulate milk and
make indol, but differ from colon bacilli in showing weakened
fermentative powers.

168 INFECTIONS OF THE DIGESTIVE TRACT

cause of animal diseases. There is little doubt that they
have their normal habitat in the intestinal tract of
domestic animals, and this accounts for their very wide
distribution and for the occurrence of some of them in
the human intestine. It is especially those members of
the group that are pathogenic for man that interest us
here, and of these we may select for brief consideration
the bacteria grouped under the term "paratyphoid."

Only five years ago Schotmiiller first maintained that
there are cases of fever which run a clinical course
indistinguishable from typhoid fever but differing from
it in being incited by the paratyphoid bacilli. This
position is now firmly established by numerous observa-
tions ; for although some of the characteristic lesions of
typhoid fever, such as ulcerated Peyer's patches, are
relatively seldom found in paratyphoid infections, it
is not clear that there are any essential differences in
the lesions observed.1 That the poisons made by one
variety of paratyphoid bacilli are probably closely re-
lated to the poisons made by the typhoid bacteria is
suggested by the very close agglutination relationships
that have been discovered between these bacteria.

It became necessary to make a separation of the para-
typhoid bacilli into two groups owing to agglutinative
and cultural differences — paratyphoid A and B. The
bacillus of paratyphoid B is probably widely distributed.
It is the cause of many instances of disease running a

1 "Weitere Mittheilungen iiber mehrere das Bild des Typhus
bietende Krankheitsfalle, hervorgerufen durch typhusahnliche
Bacillen (Paratyphus)," Zeitschr. /. Hyg., xxxvi, p. 368, 1901.

INFECTIONS OF THE DIGESTIVE TRACT 169

course like that of typhoid fever, but as a rule milder —
this milder course corresponding to the intermediate
position of the bacilli between colon bacilli and typhoid
bacilli. The justification for separating the paratyphoid
bacilli from those of typhoid has been questioned by
some, but the biochemical distinctions between them
are definite, and as the recognition of these distinctions
is serviceable to man, it is desirable to emphasize them.1
With the recognition of these different etiological factors
the unity of the condition we call typhoid fever is swept
away. This unity is in fact still further broken by the
fact that another bacillus of the colon group, B. facalis
alcaligenes, has lately been shown to be the inciter of
febrile states clinically inseparable from mild forms of
the fever caused by the typhoid bacilli.

It is an inevitable event in the development of bacterio-
logical science that different investigators should some-
times discover the same microorganism in its relation
to different diseases and that under these conditions
of discovery the organisms in question should come to
bear different names. Thus it has lately been insisted
that the bacillus of paratyphoid B is identical with the
bacillus of mouse typhoid and with the well-known
B. enteritidis 2 of Gartner. In this case the identity

1 So long as the possibility remains that closer study may re-
veal some differences of a characteristic sort between the lesions
and pathogenesis of the disease caused by typhoid bacilli and the
disease caused by paratyphoid bacilli it is perhaps permissible to
speak of the latter class of cases as paratyphoid fever.

2 Kutscher u. Meinicke, " Vergleichende Untersuchungen iiber
Paratyphus-, Enteritis-, und Mausetyphusbakterien und ihre im-
munisatorischen Beziehungen," Zeitschr. /. Hyg., Hi, p. 301, 1906.

170 INFECTIONS OF THE DIGESTIVE TRACT

apparently does not extend to completely reciprocal
pathogenicity, for mouse typhoid bacilli have never been
known with certainty to cause the typhoidal disease in
man, whereas the paratyphoid bacillus B is pathogenic
for mice.1 Of greater importance is the fact that the
widely distributed bacilli of paratyphoid B are cer-
tainly the cause of many outbreaks of meat poisoning,2
and that the bacilli of certain cases of meat poison-
ing are identical with some which cause typhoid fever.
A highly instructive instance is one lately reported in
detail, in which seven persons living hi one house suffered
from an infection by paratyphoid B bacillus.

All the members of a family fell acutely ill with vomit-
ing, diarrhoea, and fever. At the end of three days the
temperature had returned to normal in all the patients
except one, and in this instance the illness ran the course
of a moderately severe typhoid, the fever lasting eighteen
days. The sera from all these patients gave positive
agglutinations with the bacillus of paratyphoid B.

Why it is that some persons develop an acute gastro-
enteritis after infection with this bacillus while others
have typhoid fever, is still somewhat obscure. In cases
of meat poisioning, the infected meat contains not only
the bacilli but certainly also poisons derived from them.
The presence of these poisons accounts for the acute

1 It is also possible that the use of the absorption methods which
reveal common agglutinins would show that the agglutinations
for these three organisms (now claimed to be identical) are not in
reality absolutely identical.

2 Levy and Fornet, " Nahrungsmittelvergiftung und Paraty-
phus," Centralbl. /. Bakt., Orig., xli, p. 161, 1906.

INFECTIONS OF THE DIGESTIVE TRACT 171

onset of symptoms.1 The number of bacilli ingested
doubtless exerts an important influence on the character
of the illness. If few bacilli are associated with the
poison, they usually do not suffice to set up an infection,
and the seizure is transitory. If many bacilli are present,
they may establish themselves, freely multiply, and in-
vade the organism, causing the phenomena of typhoid
fever. Hence in a case where the poison present in
the food is associated with many paratyphoid bacilli,
the fever incidental to the intoxication may be directly
succeeded by a period of fever due to the bacillary in-
fection. Animal experiment teaches that in certain
virulent bacterial infections the period of incubation is
in a rough way inversely proportional to the number of
bacteria taking part in the infections. The ingestion of
very large numbers of paratyphoid bacilli may thus be
followed by a short period of incubation, whereas many
days may probably intervene before the onset of fever
if the number of bacilli barely suffices for them to gain
a foothold in the blood- vascular and lymphatic structures.
Although we have no definite information on this point
in the case of paratyphoid infection, it seems safe to
assume that the character of the anti-bacterial defenses
of the organism is a factor of equal importance with the
number of bacilli in determining infection. These con-

1 Dr. Holt tells me it is common for typhoid fever in young
children to begin acutely. Probably the defenses of the digestive
tract against the bacilli are much feebler during infancy than in
later life, and an insult from an error in diet may injure these local
defenses so as to permit rapid and free multiplication of the typhoid
bacilli.

172 INFECTIONS OF THE DIGESTIVE TRACT

siderations regarding the influence of the number of the
bacilli and the vigor of the defenses doubtless apply also
to the case of infections with the typhoid bacilli, but here
there is probably never any primary intoxication from
the presence of toxins in food, for the reason that the
typhoid bacilli, requiring such special conditions for
multiplication, do not often find the requisite condi-
tions for giving off their poisons except within the human
body.

Dysentery Bacilli. — Very few years have passed since
it became apparent that certain microorganisms which
had previously been regarded as colon bacilli are in
reality quite distinct from them and stand in a direct
causative relation to the widespread intestinal disorders
of temperate climates which are grouped under the
term dysentery. Takati Shiga, a gifted young Japanese
with a German training, was able to show that the in-
testine in dysentery contains large numbers of bacilli
which, while resembling colon bacilli in some respects,
yet differ from them in their fate, in their appearance
on agar plates, in their lack of motility, in their failure
to make acid from sugar media, in their failure to pro-
duce indol or make gas, in their power to set up in man
a disease clinically like dysentery, and in their ability
to confer on the human blood serum the property of
specific agglutination in high degree. This discovery
was soon followed by the researches of Flexner in Manila,
which showed that there is another type of dysentery
inciter allied to the Shiga bacillus but differing from
it in the direction of retaining more fully some of the

INFECTIONS OF THE DIGESTIVE TRACT 173

characters of typical colon bacilli, especially the capacity
to ferment sugars. Just as in the case of the typhoid
and paratyphoid bacilli, it is evident that the bacilli
farthest removed from colon bacilli (the typhoid bacilli)
in their characters incite the severest form of the disease,
so in the case of the dysentery bacilli it is true that the
severest forms of disease are dependent on those bacteria
farthest removed from the colon bacilli in their biochem-
ical characteristics; namely, the Shiga bacilli. The
Shiga bacilli are the cause of the severe epidemics of
Japan and of Germany, but in the latter country bear
the name of Kruse. The Flexner bacilli are apparently
the cause of dysentery in young children much more
frequently than the Shiga bacilli. The two types of
bacilli are further distinguished by their agglutination
reactions and somewhat less sharply by their behavior
toward immune sera.1 The Shiga and Flexner organisms,
however, taken together do not quite cover the entire
range of bacillary dysentery ; for, as Park and Hiss have

1 It has been proposed to group these types of dysentery bacilli
under the name paradysentery. As it appears to have been
conclusively shown that the lesions caused by these bacteria are
identical with those caused by the Shiga bacilli, there is no reason
why the "paradysentery" bacilli should not be regarded as dysen-
tery bacilli equally with the Shiga organisms. To use the term
"paradysentery" on the ground that the bacilli so designated stand
between the Shiga bacilli and the B. coli group in cultural, fermen-
tative, and other characters does not seem appropriate, for there
are doubtless many bacteria intermediate between the Shiga and
colon organisms which do not cause dysentery. For the present
it seems best to employ the names " Flexner type " and " Park-Hiss
type" in referring to the organisms in question. Doubtless
subdivisions within these types will prove desirable from a bac-
teriological standpoint.

174 INFECTIONS OF THE DIGESTIVE TRACT

shown, there are- pathogenic bacteria intermediate in
type between the two. This further distinction of a
third variety is justifiable on grounds of convenience.
Indeed, Shiga1 has lately adopted the classification of
Hiss, which recognizes four groups based on fermenta-
tive characters and has added to this a fifth, interme-
diate between the acid bacilli and the non-acid bacilli.

Like the typhoid bacilli, the bacilli of dysentery are
hemiparasites ; that is, their invasive and destructive
properties are developed only when the bacteria have
been introduced into the organism in considerable num-
bers or have had an opportunity to multiply owing to
feeble powers of resistance on the part of the infected
individual. Like typhoid bacilli, the bacilli of dysentery
form substances — the so-called "aggressins" -which
are capable of inhibiting or even completely paralyzing
the natural protective action of the body fluids and thus
removing the most important obstacle to the invasion
of the bacteria which up to the moment of this oppor-
tunity had led a saprophytic existence. It is not clear
whether these "aggressins" are identical with the soluble
products of the dysentery bacilli which one obtains on
filtering an old bouillon culture or are bodies wholly
or in part distinct. It is well established that guinea-
pigs may be actively immunized against the severest
intraperitoneal infection by means of two or more
injections of sterile guinea-pig exudate containing
"aggressins" derived from the peritoneal cavity of a suit-

1 "Observations on the Epidemiology of Dysentery in Japan,"
The Philippine Journ. of Sci., i, p. 485, 1906.

INFECTIONS OF THE DIGESTIVE TRACT 175

ably prepared animal. It has further been found that
as the result of long preparation with such exudates
guinea-pigs, rabbits, and sheep furnish a serum which in
amounts of one-half cubic centimeter is able to protect
animals from intraperitoneal infection. It is claimed
by Kikuchi l that the immune serum obtained through
treatment with "aggressins" does not show in vitro the
well-known characteristics of a bacteriolytic serum and it
is maintained that this " anti-aggressin " action is the
basis of a new type of immunity. The validity of the
latter claim cannot be successfully upheld without further
research in this complex problem. There are now two
well-known antidysenteric sera designed for human use :
one prepared by Shiga by immunization with the bacilli
of the originally discovered strain of dysentery; the
other made at the instigation of Dr. Flexner by means of
the variety of bacillus bearing his name. These sera
are regarded as being essentially bactericidal, but there
is room for doubts as to how far the bactericidal action
exhibited in vitro is operative after the serum has been
put into the dysenteric patient. That these sera,
though not universally applicable, are quickly effective
in some instances of severe dysentery is certain. The
effect consists mainly in a diminution in the number of
movements, alleviation of tenesmus, some lowering of
the temperature, and a calming of the nervous system
which is sometimes very striking. These effects may

1 "Weitere Erfahrungen fiber Aggressinimmunitat gegen den
Shiga-Kruseschen Dysenteriebazillen," Archiv f. Hyg., liv, p. 298,
1905.

176 INFECTIONS OF THE DIGESTIVE TRACT

depend in part on an antitoxic action, which is perhaps
mainly anti-aggressive. It seems clear that if the solu-
ble poisons formed by the dysentery bacilli in the body
could be promptly neutralized, the bacilli would soon
fall a prey to the natural bactericidal process of the
serum and leucocytes. It is reasonable to believe that
sera may yet be obtained which will better fulfill this
need than do those at present in use. It must, however,
be clearly stated that even a powerful antitoxic or anti-
aggressive serum could not be counted upon to cure all
cases of dysentery, for in this disease there is in many
subjects a period when other bacteria than the bacilli of
dysentery (especially streptococci) enter the field of
action in a significant manner. A secondary infection
of mixed character is then set up, and against this the
specific sera of dysentery must necessarily be ineffective.
Finally it may be mentioned that there is sometimes
a therapeutic difficulty of a wholly different character,
due to the fact that a serum prepared with one strain
of dysentery bacillus is apt to be effective only for this
strain and to have little action on other varieties of the
bacillus. A serum may thus be worthless in a given
case because it is not adapted to the variety which is
acting. Shiga has recently proposed the use of a
universal or polyvalent serum designed to forestall
embarrassments of this sort. Shiga's advice is as fol-
lows. Two horses should be simultaneously immunized,
one with the bacilli of Shiga's types I and II, the
other with the bacilli of types I and IV. As soon as
a high grade of immunity is attained serum should be

INFECTIONS OF THE DIGESTIVE TRACT 177

taken from both horses and mixed in equal volumes.
"By this means the best universal serum can be ob-
tained" (Shiga).

I have mentioned that the dysentery bacilli resemble
those of typhoid fever in being hemiparasites capable
of infecting a susceptible individual only if present
in such considerable numbers that his anti-aggressive
powers are neutralized, thus clearing the way for a free
multiplication of the bacilli. There are, however, some
apparently fundamental differences in the biological
characters of these two important pathogenic forms.
While the typhoid bacillus, an actively motile organism,
is highly invasive and sets up a bacillsemia with con-
sequences already noted, the dysentery bacillus, a non-
motile organism, is only slightly invasive, remains almost
limited to the intestinal mucous membrane, and leads
to neither bacillsemia nor bacilluria. The typhoid bacilli
produce soluble poisons with difficulty, at least in fluid
cultures. Some dysentery bacilli, namely those of the
original Shiga or promptly alkali-making variety, show
a considerable capacity to yield soluble poisons in fluid
media. They differ in this respect from the acid-making
varieties, which yield their active toxic product only
within the animal organism.1 The chemical nature of
the poison is at present wholly unknown, but significant
facts in respect to its mode of action in animals, including
probably the human species, have been lately brought
to light, chiefly through the experimental studies of

1 Dr. Flexner tells me he has lately succeeded in obtaining toxins
from his dysentery bacilli by means of a special procedure.

178 INFECTIONS OF THE DIGESTIVE TRACT

Dr. Flexner 1 and his associates. Of these facts the most
important have to do with the pathological alterations
observed in the intestinal mucous membrane in the
course of human dysentery and in the course of experi-
mental intoxications. By intravenously injecting into
rabbits the poison obtained by permitting the bacilli
to undergo a process of autolysis or self-digestion, Dr.
Flexner was able to reproduce intestinal lesions analogous
to those observed in human dysentery ; namely, inflam-
matory, sometimes diphtheritic processes, especially in
the large intestine and often attended by hsemorrhagic
changes. It is quite clear that these lesions are inciden-
tal to an elimination of the injected poison into the in-
testine and that the damage done to the intestinal
structures is due primarily to the contact with the
poison in the blood during the act of elimination or
secretion. It seems probable that a secondary invasion
of the damaged structure by intestinal bacteria forms
an essential part of the dysenteric process. These pri-
mary eliminative lesions are, as pointed out by Flexner,
analogous to those experimentally called forth by
corrosive sublimate and by ricin. I observed, many
years ago, hsemorrhagic and necrotic intestinal lesions in
dogs after intravenous infusions of urea, and these in-
juries also belong to the group of eliminative manifes-
tations. The special value of Flexner's studies lies, it
appears to me, in the fact that they show for the first

1 "The Pathogenesis of Experimental Colitis, and the Relation
of Colitis in Animals and Man," Journ. of Exper. Med., viii, p. 514,
1906.

INFECTIONS OF THE DIGESTIVE TRACT 179

time that the poison of human pathogenic bacteria have
the power of breaking down the resistance of the intesti-
nal structures in the eliminative process. It has always
been customary to think of the inflammatory, necrotic,
intestinal lesions of human dysentery as the result of
strictly local conditions, and latterly the dysentery bac-
teria have fallen under suspicion as the inciters of the
lesions, in conjunction with the gross mechanical insults
which might be offered by the hard contents of the lower
bowel. Objections to this hypothesis of purely local
causation are the fairly even distribution of the dysentery
bacilli through a much larger extent of the intestine
than that occupied by the severe lesions, and, secondly,
the frequent absence of any ascertainable local and aux-
iliary factors. The newly acquired experimental facts
force us to look at the local dysentery lesions in a
wholly different light; namely, as being perhaps in
part the result of the action of soluble toxins which
have found their way by the blood to the gut in a
changed (perhaps autolyzed) form and in considerable
concentration. Although this mode of origin cannot
now be positively affirmed for these local human
dysenteric lesions, it must be regarded as extremely
probable. There is reason to believe that the dysentery
toxins are excreted in the bile as well as through the
intestinal mucous membrane. These toxins so excreted
may thus take part in doing damage to the intestinal
structures.

One further phase of the action of the dysentery poison
calls for mention; namely, its effect upon the nervous

180 INFECTIONS OF THE DIGESTIVE TRACT

system. Dopter * has shown that the toxin prepared from
the Shiga bacilli may cause paralysis in rabbits and that
this paralysis is usually referable to acute lesions in the
gray axis of the spinal cord or ponto-bulbar region. As
the result of the action of a similar poison Flexner saw
small haemorrhages in the brain and softening in the gray
substance of the spinal cord with or without haemorrhage.
Rabbits are unequally susceptible to this nervous poison
or neurotoxin, and only a small proportion of them de-
velop nervous lesions after treatment with the poison,
whether this be autolyzed or not. The fact that such
lesions may be sometimes induced — and perhaps regu-
larly under suitable experimental conditions — has a pos-
sible bearing of great importance for human pathology.
It has long been clear to clinicians that young children
are susceptible to paralyses referable generally to lesions
in the gray substance of the spinal cord and that these
paralyses are apt to come on at a time of disordered intes-
tinal digestion; often indeed at the time of teething, to
which process the paralysis has often been somewhat
vaguely referred. As we have only learned in recent
years to see a specific bacillary dysentery in many cases
of slight intestinal disorder in children, it is likely that
many instances of infantile spinal paralysis have been as-
sociated with a true dysenteric infection. Of the numer-
ous cases of dysentery that occur yearly in children, only
a relatively small number are attended with paralyses,
but in view of the experimental data now in our posses-

1 "Effets exp Sriment aux de la toxine dysent6rique sur le
systeme nerveux," Ann. de I'Inst. Pasteur, xix, p. 353, 1905.

INFECTIONS OF THE DIGESTIVE TRACT 181

sion this small number assumes a new interest. It re-
mains for the future to definitely prove or disprove the
role of dysentery toxin in the causation of poliomyelitis.
I have endeavored to sketch with few words some
of our leading acquisitions and problems in the colon-
typhoid-dysentery group of bacteria while touching only
lightly the realm of hypothesis. It is certain that
research will distinguish further varieties possessing
interest for human pathology within this large group.
Nor can we doubt that the methods by which we now seek
to combat the infections due to members of this group
will gain in efficiency and precision. Our attention has,
perhaps, been too exclusively fixed on the specific excit-
ants, and the role played by associated bacteria must re-
ceive more study, for it is clear that they sometimes play
a significant part in determining the outcome of an infec-
tion. The difference that decides whether a man will
live or die must frequently be a slight one, looked at from
the standpoint of the processes of battle within the body.
To learn to recognize more clearly in what this difference
consists and how to use such knowledge to turn the tide
of warring forces to the advantage of the life that hovers
on the verge of extinction is among the problems which
intelligent, careful research will help us to solve.

LIQUEFYING BACTERIA

In the foregoing section has been considered very
briefly the highly important group of Gram-negative
microorganisms which do not liquefy gelatin and
which are facultative anaerobes. It is desired here

182 INFECTIONS OF THE DIGESTIVE TRACT

to refer briefly to certain Gram-negative facultative
anaerobes which liquefy gelatin and which are regarded
as having the ability to induce putrefaction in proteid
material. The most important member of this group
is the organism called B. vulgaris, or proteus vulgaris,
which is known to make a tryptic ferment and
to peptonize casein and attack carbohydrates. It is
said by A. E. Taylor to produce indol, skatol, and diam-
ino-acids when grown on casein. The organism is one
which is sometimes found in moderate numbers in the
fseces of healthy individuals where, as a rule, liquefying
aerobic organisms are not numerous. B. proteus vulgaris
was formerly regarded as a very important agent in the
production of putrefactive decompositions outside the
body and it has been suspected of inducing active putre-
faction within the human intestine. Recent investiga-
tions indicate clearly that other types of microorganisms
belonging to the class of strict anaerobes are much more
important factors in putrefaction generally than is the
B. proteus vulgaris, which is sometimes, however, asso-
ciated with these anaerobes. It appears that the ability
of B. proteus vulgaris to attack native proteids has been
overestimated. I have gamed the impression from the
study of the action of this organism on milk in fermenta-
tion tubes that its peptonizing action on casein is ordi-
narily not considerable during a period of four to five
days' sojourn in the incubator at body temperature.
That B. proteus vulgaris is capable of inducing acute dis-
ease of the human gastro-enteric tract appears to be well
established. There are instances on record of severe

INFECTIONS OF THE DIGESTIVE TRACT 183

meat poisoning apparently due to the contamination of
meat by this organism. A number of instances have
been recorded in which the poisoning following the eat-
ing of sausages has been attributed to the action of
proteus vulgaris.1 It has been found in such cases that
the nitrate of the bouillon culture of the proteus organ-
ism derived from the contaminated meat or from the
intestine has been fatal to experimental animals. There
are also recorded some instances of poisoning through
cheese which had become contaminated by B. proteus
vulgaris. In one instance a man ate abundantly of a
soft cheese containing this organism and soon developed
seizures of vomiting associated with great prostration
and cardiac depression followed by death after five
days. Other persons who ate of the same material more
moderately were similarly though less violently ill and
ultimately recovered.2

B. proteus vulgaris is an organism which in general
shows little tendency to invade the circulation from the
digestive tract although it may be the cause of the sever-
est gastro-enteric symptoms.

Other Gram-negative liquefying facultative anaerobes
are sometimes found in the human intestine in great
abundance. Their relation to the production of disease
is not yet clearly established. Possibly some varieties
of liquefiers are harmless if present in moderate numbers.
In a case of pernicious anaemia from which I was unable

1 A case of this kind is that reported by Schumburg, Zeitschr.
f. Hyg., xli, p. 183, 1902.

2 Berl. Molkereizeitung, xiii, p. 78.

184 INFECTIONS OF THE DIGESTIVE TRACT

to cultivate any colon bacilli, the gelatin plates indicated
the presence of very large numbers of an organism cor-
responding closely to that described by Ford * as B.
entericus. This organism liquefies gelatin with great
rapidity.

Lubenau 2 has lately described a spore-bearing Gram-
positive aerobic organism which liquefies gelatin and
which he looks upon as the cause of an outbreak of severe
diarrhosa among the inmates of an institution. This
outbreak was characterized by persistent vomiting, great
prostration, severe headache, and frequently by mental
confusion. The diarrhoeal seizures were in some in-
stances as frequent as twenty or more in twenty-four
hours and were attended by great abdominal pain. The
general condition of the patients improved rapidly, but
the vomiting and diarrhrea lasted for two or three days.
The temperature was only slightly elevated and even
this slight rise was observed only in a small number of
the cases.

The organism described by Lubenau as the cause of
this outbreak resembled the common hay bacillus in
morphology. It was an actively motile aerobe which
formed no gas on sugar-agar and attacked milk slowly,
producing alkali and peptone. It sporulated readily
and in its sporulating stage was non-pathogenic
for animals. When obtained hi large numbers in its
vegetative form, through growth upon milk to which

1 Loc. tit., p. 40.

2 " Bacillus peptonificans als Erreger einer Gastroenteritis
Epidemic," Centralbl. f. Bakt., etc., I Abt., Orig., xl, p. 433,
1905-06.

INFECTIONS OF THE DIGESTIVE TRACT 185

Witte's peptone had been added, it induced pro-
nounced diarrhoea in young dogs, sometimes associated
with blood.

STREPTOCOCCAL AND STAPHYLOCOCCAL INFECTIONS

The biological characters of the streptococcal and
staphylococcal forms of bacteria met with in the human
intestine in health and in disease have not yet received
sufficient attention. There are indications that this sub-
ject is likely within a few years to undergo a much-
needed development. In the meantime it is impossible
to deal with the subject of streptococcal and staphylo-
coccal infections in a really satisfactory way.

The normal intestine (both small and large) usually
contains Gram-positive diplococci in moderate numbers.
In culture media these may grow freely in chains, the
individual organisms undergoing division hi a plane
transverse to the line of the chain development. In
bouillon cultures streptococcal organisms are more
likely to maintain any virulence they may possess than
when grown in sugar media. Large quantities of bou-
illon cultures of the Gram-positive diplococci of the nor-
mal intestine (which may have developed into chains
on the bouillon media) may be injected into guinea-pigs
either intraperitoneally or subcutaneously without giving
rise to any symptoms or lesions. The human intestinal
tract under normal conditions is probably most of the
time free from pathogenic streptococci and in normal
adults such pathogenic bacteria introduced with milk or

186 INFECTIONS OF THE DIGESTIVE TRACT

water are ordinarily quickly destroyed in the upper part
of the tract.

During infancy the digestive tract is much less re-
sistant to streptococcus infections, and invasion of the
mucous membrane by streptococci is a frequent occur-
rence and may be associated with disturbances of al-
most any grade of severity. It has been shown by
Booker in this country and by Escherich in Germany
that some of the severest forms of infantile ileocolitis
are associated with streptococcus infections and are
probably dependent upon them. As the digestive tract
of the infant is readily damaged so as to become per-
meable by bacteria, it is not surprising that streptococcus
septicaemia is of frequent occurrence. The condition is
one which frequently but not necessarily ends fatally.
A less common result of the penetration of the intestinal
wall by streptococci is purulent peritonitis.

In children there are some instances of severe ileo-
colitis in which streptococci appear to be the primary
inciters of disease. At least one may say that other
bacteria which might fall under suspicion, such as dysen-
tery bacilli and pathological types of colon bacilli have
not been found in the cases in question. It is doubtful,
however, whether the biochemical characters and viru-
lence of certain colon-like forms have received sufficient
attention in those cases which are regarded as instances
of pure streptococcus infections. In many examples of
infantile diarrhoea it appears clear that the streptococci
which are prominent in the stools are really secondary
invaders. Here the primary infection through colon

INFECTIONS OF THE DIGESTIVE TRACT 187

bacilli or atypical colon bacilli or dysentery bacilli
appears to have prepared the way for such a secondary
infection. In the case of the dysentery bacilli it is
probable that they have the ability to make "aggres-
sins" which inhibit the local or even the general defenses
so as to permit the free multiplication of pathogenic
streptococci. A similar relationship is seen in some
instances of smallpox, scarlet fever, and measles. It
seems clear that in many of these cases, in which the role
of the streptococci is that of a secondary invader, this
infection is responsible for severe exacerbations or death.
In adults the intestinal tract is not rarely the seat of
invasion by strep tococcal microorganisms. Where these
are virulent, they are apt to be associated with diar-
rhceal disorders. There are, however, instances in which
the streptococci are only slightly pathogenic for animals,
such as guinea-pigs, but are present throughout con-
siderable periods of time in excessive numbers in the
stools. Their presence may be demonstrated in the
Gram-stained microscopical fields, where they are seen
as short chains or as diplococci. In such cases these
organisms will be found to be abundantly represented
in sugar-bouillon fermentation tubes which have been in-
oculated from the mixed faecal flora. We have observed
instances in which these Gram-positive streptococcal
forms enormously preponderate in the sedimentary
fields from the fermentation tubes. It has some-
times happened, also, that even though the Gram-posi-
tive fields fail to show the streptococcal forms in such
abundance as to attract attention, they outgrow other

188 INFECTIONS OF THE DIGESTIVE TRACT

types of bacteria in the fermentation tubes. In one
instance in which such an overgrowth of the fermenta-
tion tubes was regularly observed and in which the move-
ments regularly showed the presence of excessive num-
bers of epithelial cells, it was found that the patient had
for many months been infecting himself daily from an old
abscess next a carious tooth. Although the abscess was
treated surgically, the improvement in the clinical con-
ditions was very slow, and after a period of six months
under very favorable conditions of outdoor life evi-
dences of streptococcal infection still persisted to some
extent. In this case loss of weight and strength, mental
depression, and moderate anaemia were the prominent
clinical conditions. The movements contained in ad-
dition to streptococci somewhat excessive numbers of
B. aerogenes capsulatus.

The presence of large numbers of intestinal strep-
tococci and diplococci in the faeces is a prominent
feature in some cases of pernicious anaemia. The
significance of this is not yet clear, partly because the
streptococcal infection has been associated in these
cases with the presence of greatly excessive numbers
of putrefactive anaerobes, especially B. aerogenes cap-
sulatus. It seems probable that a streptococcus in-
fection of the intestinal tract may greatly contribute
to bring about a deterioration in the physical condition
of a patient already infected with the gas-bacillus. In
several instances of severe anaemia — some of them
of the pernicious type — a prolonged streptococcal
diarrhoea has been observed to precede the onset of

INFECTIONS OF THE DIGESTIVE TRACT 189

more grave manifestations of alterations in the blood
and loss of muscular power.

It is now well known that there are occasional instances
of phlegmonous gastritis due to streptococcal infec-
tions. I have observed instances of extreme hyper-
chlorhydria which showed in the vomited material during
a seizure of protracted vomiting large numbers of Gram-
positive streptococci and many leucocytes.

The relation of streptococcus infections to appendicitis
is still unsettled despite the considerable work that has
been done on the etiology of this affection. In many
instances the dominant flora in the diseased appendix
are streptococci, but the question arises whether they
are primarily or secondarily concerned. I have obser-
vations on two fatal cases of appendicitis in which the
intestine between the appendix and the anus contained
enormous numbers of streptococci and diplococci at all
points. It appears to me probable that the careful
microscopical and cultural study of the intestinal con-
tents will prove an aid to diagnosis in some obscure
forms of local intestinal infection by streptococci.

Of the position which should be assigned to infections
of the intestinal tract by staphylococci there is little to
be said at present. One sometimes observes in the sedi-
ments of the fermentation tubes or bouillon flasks which
have been inoculated from mixed faecal flora numerous
clumps of Gram-positive cocci. The biochemical char-
acters of these organisms have not been studied and their
classification is at present uncertain. It is noteworthy,
however, that in two cases of severe staphylococcus acne

190 INFECTIONS OF THE DIGESTIVE TRACT

in which the opsonic index was low for these organisms,
the mixed fsecal flora regularly showed staphylococcal
forms in great abundance in the sedimentary fields
from the fermentation tubes. Such organisms, however,
have been detected hi great abundance in the flora from
some persons suffering from excessive intestinal putre-
faction in which acne was not a feature. The very free
growth of organisms of this type in bouillon inoculated
with the mixed flora has been observed especially in
the case of material derived from persons suffering from
disorders of digestion, but it is uncertain what signifi-
cance should be attached to these observations.

The possibility of controlling streptococcal and staphy-
lococcal infections of the digestive tract by means of the
vaccination methods employed by Wright is worthy of
serious consideration. There are cases of advanced
anaemia in which it appears likely that streptococcus
infection is an important etiological factor, perhaps
through the agency of hsemolytic poisons and certainly
through the causation of exhausting diarrhoea. If in
such cases the streptococcal element in the infection
could be controlled, it seems clear that the outlook for
the patients would be improved.

BACILLUS BIFIDUS

It has already been mentioned that the intestinal tract
in childhood contains large numbers of the not extremely
anaerobic microorganism described under the name of B.
bifidm. Many if not all adolescents and adults harbor
B. bifidiis, although in smaller numbers than in early

INFECTIONS OF THE DIGESTIVE TRACT 191

life — a fact not widely known. I have met with excep-
tionally robust adults in whom bifidus was very promi-
nently represented, if one may trust the microscopical
fields and the primary * fermentation-tube sediments.
Whether bifidus ever assumes a pathological character in
man is uncertain. I may refer here to the observation
that a Gram-positive bacterium indistinguishable from
typical bifidus in its microscopical characters is sometimes
almost the exclusive inhabitant of the lower digestive
tract in children with chronic intestinal indigestion char-
acterized by abdominal distention and retarded growth.
In sugar bouillon these fsecal bacteria show the typical
varied morphological characters of B. bifidus of nurslings,
but whether the organisms from these pathological cases
are really identical with those obtained from the in-
testine of normal infants is uncertain because no ade-
quate biochemical observations have been made. I
think the possibility should be borne in mind that a bac-
terial form which is entirely physiological when dominant
during infancy may not be equally physiological when
it constitutes the dominant type in later years. The
mere persistence of B. bifidus as the leading large-intes-
tinal microorganism may carry with it certain physio-
logical disadvantages, such as relatively feeble powers of
defense against some harmful bacteria.

INFECTIONS THROUGH ANAEROBIC BACTERIA

The anaerobes of the human intestine have received far
less attention than the aerobic and optionally anaerobic

1 Made from direct inoculation of the faeces.

192 INFECTIONS OF THE DIGESTIVE TRACT

bacteria. This neglect may be ascribed to the technical
difficulties connected with the study of microorganisms
whose growth is checked by minute traces of oxygen,
and not to any belief that such microorganisms are
unimportant.1 There is in fact growing evidence that
the strict anaerobes play an important part in some
pathological processes which have their seat in the digest-
ive tract, and one may hazard the prediction that careful
studies of them will give us wholly new light on some
obscure forms of disease.

Bacillus Putrificus. — The organism described by
Bienstock as B. putrificus is a spore-forming strict
anaerobe, capable of actively attacking and hydrolyzing
native proteids, and like many such anaerobes able to
give rise to characteristic products of putrefaction,
including butyric acid and hydrogen. It forms hydro-
gen sulphide and mercaptan but not indol. Consid-
erable difference of opinion exists as to whether B. putri-
ficus is or is not a regular inhabitant of the normal
human intestine.2 This question is bound up with the
question of the identity of the bacterium. Bienstock,
it should be noted, has lately brought forward evidence

1Tarozzi ("Ueber ein Leicht in aerober Weisse ausfiihrbares
Kulturmittel von einigen bis jetzt fur strenge Anaeroben gehal-
tenen Keimen," Centralbl. f. Bakt., xxxviii, I Abt., Orig., p. 619,
1905) has lately described a method by which it is possible to
obtain aerobic growths of organisms which have hitherto been
regarded as strict anaerobes, including B. aerogenes capsulatus,
B. botulinus, etc.

2 Passini holds that he can generally obtain it; Bienstock states
that he cannot find it in normal persons.

B. putrificus is widespread and may readily become a contami-
nation as it is common in laboratory dust.

INFECTIONS OF THE DIGESTIVE TRACT 193

to show that two distinct bacteria are included under
the name of B. putrificus.1 One of these, the true B.
putrificus, attacks native proteids such as serum albumin ;
the other, though morphologically indistinguishable from
B. putrificus fails to attack native proteids but decom-
poses sugars. Bienstock calls this organism B. para-
putrificus. He states that he has found it not infre-
quently in the normal human intestine and believes it is
paraputrificus and not putrificus that has been found in
the contents of the lower bowel. As paraputrificus makes
acid from sugars and does not cleave proteids, Bienstock
inclines to the belief that the organism is antagonistic
to the true putrefactive microorganisms of the intestine
and thus supports the colon bacilli in this function. But
it is not clear that acid-forming bacteria necessarily act
as a check to the development of putrefactive bacteria
except under special conditions, and it appears ques-
tionable whether B. paraputrificus is really a factor in
checking the multiplication of harmful anaerobes.

The role of the putrefactive type of B. putrificus in
pathological processes has not yet been determined.
That it plays a part in certain putrefactive disorders
appears to me probable on the following grounds. An
organism corresponding to the cultural and morphological
characters of B. putrificus appears in great abundance
in week-old cultures of the mixed flora from some per-
sons suffering from subacute or chronic intestinal symp-
toms at present not satisfactorily definable. As these
organisms regularly appear after inoculation with the

1 "Bacillus Putrificus," Ann. de VInst, Pasteur, xx, p. 407, 1906.
o

194 INFECTIONS OF THE DIGESTIVE TRACT

flora of these people, it seems fair to assume that they
are in these instances regular inhabitants of the digestive
tract, although they may not appear hi their character-
istic sporulating form in the microscopical fields. The
flora of persons free from symptoms of intestinal derange-
ment has not abundantly given us B. putrificus in our
bouillon-calcium-carbonate flasks. This is in harmony
with the statement of Bienstock that he failed to find
B. putrificus in normal persons. It is true also of the
intestinal flora of persons to whom the spores of B.
putrificus had been given by the mouth.

One reason why it is difficult to define with con-
fidence the part of B. putrificus in intestinal processes
of putrefaction is that it is apt to be associated with
other microorganisms which may be concerned with the
processes in question, such as B. aerogenes capsulatus and
some coccal infections. The association of excessive
numbers of these three organisms in the large intestine is
not uncommon in cases of chronic excessive intestinal
putrefaction.

In every instance in which B. putrificus is present hi
bouillon flasks prepared by growing the mixed faecal
flora from cases of intestinal putrefaction, there is found
also methyl mercaptan. This observation corresponds
to the fact that B. putrificus in pure culture in peptone
bouillon is capable of making mercaptan. It has not
always been possible to grow B. putrificus from cases
in which a methyl mercaptan reaction was obtained,
and for this reason and others I believe that methyl
mercaptan may be produced by other intestinal organ-

INFECTIONS OF THE DIGESTIVE TRACT 195

isms than B. putrificus. Nevertheless in my experience
the strongest methyl mercaptan reactions have been
obtained from those cases in which B. putrificus was
present. B. putrificus grown in peptone bouillon makes
more mercaptan than any other anaerobe with which
we have experimented in my laboratory. The mer-
captan reactions obtained from bouillon cultures made
with the mixed flora from some healthy young children
I am disposed to attribute to B. putrificus. This
organism is very widespread and is present in most
samples of dust. It is non-hsemolytic and probably
makes no toxins. Considerable numbers of the typical
bacillus may be injected intraperitoneally into guinea-
pigs without causing noteworthy damage.

I believe B. putrificus to be the most energetic an-
aerobe now known in respect to the ability to proteo-
lyze native proteids. Bienstock's organism is probably
representative of a group whose members vary con-
siderably in biochemical properties. We have in our
laboratory a strain of B. putrificus of intestinal origin
which makes very large quantities of indol in peptone
bouillon. It is evident that the presence of such an
organism in the intestine might contribute in an impor-
tant way to the production of indol.

In a recent study of the anaerobic bacteria of the
mouth, Dr. Antonio Rodella 1 has ascribed importance
to B. putrificus as a cause of caries of the teeth and
inflammation of the pulp. He finds the substance of the

'"Ueber anaerobe Mundbakterien und ihre Bedeutung,"
Archiv f. Hyg., liii, p. 329, 1905.

196 INFECTIONS OF THE DIGESTIVE TRACT

teeth in caries to be infiltrated by this organism appar-
ently in almost pure culture. Moreover, he has been able
to induce a putrefactive decomposition in the substance
of teeth previously decalcified by the use of acid. While
the evidence thus far brought forward in regard to the
role of B. putrificus in causing disease of the teeth is not
complete (owing to the difficulty in excluding some
symbiotic action of other kinds of bacteria), it appears
in the highest degree probable that B. putrificus is
an important factor in setting up caries. As this
organism is an anaerobe, it is easy to see how scrupulous
cleanliness of the teeth and gums must operate to pre-
vent it from gaining a foothold in the mouth.

Bacillus aerogenes capsulatus. — B. aerogenes cap-
sulatus is a strictly anaerobic, difficultly spore-bearing,
Gram-positive, gas-making bacillus. It is found in
small numbers relatively to other bacteria in the intes-
tines of the majority of normal adults and in still smaller
numbers in normal children. There are many healthy
breast-fed and bottle-fed children hi whom one fails
to find evidence of the presence of this microorganism
in the intestinal tract, either through careful study of
the Gram-stained microscopical faecal field or through
ordinary plating methods. I will not say that the organ-
ism is actually absent from the intestine in these cases,
but it is certain that if present it is in such small numbers
as to play no appreciable part in the biochemical pro-
cesses that occur in the digestive tract. Our knowledge
of the part played by this anaerobe in human pathol-
ogy has until very recently been limited to certain

INFECTIONS OF THE DIGESTIVE TRACT 197

surgical and obstetrical diseases. Some confusion has
arisen from the fact that the organism is known in
different countries under different names. In Ger-
many it is known as the gas-phlegmon bacillus (Fraen-
kel) ; in France, as B. perfringens (Veillon and Zuber) ;
and Dr. Welch was able to obtain the organism
first described by him as B. aerogenes capsulatus
from a culture of B. enteritidis sporogenes sent to
him by Dr. Klein. The same organism as B. aerogenes
capsulatus (but in a non-virulent form) has been de-
scribed by Grassberger and Schattenf roh under the name
of granulo-bacillus immobilis liquefaciens.1 It occurs
commonly in market milk.

Opinions differ rather widely as to the pathogenic
properties of B. aerogenes capsulatus for man and animals.
Professor Welch,2 who first discovered the organism in
connection with human surgical infections, was disposed
to regard it as not being ordinarily highly pathogenic.
In reference to healthy rabbits he says we cannot regard
this bacillus as being pathogenic under ordinary con-
ditions. Dunham,3 who, in the course of his clever
investigations on B. aerogenes capsulatus, discovered

1 Professor Welch and some other bacteriologists regard the
bacillus of Grassberger and Schattenf roh as identical with B.
aerogenes capsulatus in everything except pathogenicity. Other
writers dissent from their view. The two anaerobes are surely
very closely related if not identical.

'"A Gas-producing Bacillus (Bacillus Aerogenes Capsulatus,
Nov. Spec.) capable of Rapid Development in the Blood-vessels
after Death," by Wm. H. Welch and Geo. A. F. Nuttall, Bull, of
the Johns Hopkins Hosp., No. 24, July-August, 1892.

3 " Report of Five Cases of Infection by the Bacillus Aerogenes
Capsulatus (Welch)," Bull, of the Johns Hopkins Hosp., No. 73,
April, 1897.

198 INFECTIONS OF THE DIGESTIVE TRACT

its ability to make spores, ascribes a greater degree of
pathogenic action to the microorganism with which
he worked — these bacteria also having been derived
from cases of surgical infection. Grassberger and
Schattenfroh, whose granulo-bacillus saccharo-butyricus
immobilis liquefaciens was derived from milk and is
regarded by Welch, Kamen, and others as identical
with B. aerogenes capsulatus, found it to be non-toxic
for guinea-pigs. Kamen,1 although unable to obtain
powerful toxins on ordinary culture media, regards
B. aerogenes capsulatus as capable, by itself, of exciting
inflammatory purulent processes. It is certain that
there are various strains of B. aerogenes capsulatus as
regards pathogenicity and that the different results
obtained by different investigators with respect to that
feature are due to this fact. Cultures of B. aerogenes
capsulatus made by Mr. Ward from the faeces of a young
man with slight digestive derangement and from material
derived from a case of pernicious ansemia were injected
into the breast muscles of pigeons (according to a sug-
gestion from Dr. Flexner, who had found these animals
to be especially susceptible). The cultures set up a
localized necrotic inflammation with gas production in
the connective tissues. Death occurred, apparently
from toxaemia, within twenty-four hours. The applica-
tion of these facts to capsulatus infection of the intestine
is not now clear. It is certain that the pathogenicity
of B. aerogenes capsulatus is different for different
strains isolated from the human digestive tract. Thus

1 "Zur Aetiologie der Gasphlegmone," Centralbl. /. Bakt., I
Abt., Orig., xxxv, pp. 554, 686, 1904.

INFECTIONS OF THE DIGESTIVE TRACT 199

Professor Theobald Smith sent us an organism (anaerobe
xxxi) which he isolated from the stool of a patient with
pernicious ansemia and which proved to be much less
pathogenic for guinea-pigs than the typical form of B.
aerogenes capsulatus. This organism differed but slightly
from the typical bacillus in its morphology, but showed
the important peculiarity of being non-haemolytic. It
fermented the various sugars, but the gas production
was less abundant than in the case of the typical gas-
bacillus. The gas production and decomposition in-
duced in an incubated rabbit were also less pronounced
than in the case of the typical B. aerogenes capsulatus.

A single observation was made by Mr. Ward on the
blood serum from a patient with pernicious anemia
(and capsulatus infection of the intestine) with respect
to a possible agglutinative action. The results were
entirely negative.1 It is interesting to note in this con-
nection that Kamen 2 obtained no agglutinative action
from the serum of rabbits which had been immunized
with the gas-bacillus. Positive results have, however,
been recently obtained by Werner,3 who employed a
special technical procedure in the immunization of the
rabbits which served as experimental animals. The
immune serum caused agglutination of the homolo-
gous gas-phlegmon bacilli (derived from a gas liver found
in a fatal case of wound infection) in a dilution of 1 : 1000.

1 Professor Theobald Smith tells me that he obtained negative
results in agglutinative tests made with the blood of a patient
with pernicious anaemia who -showed large numbers of B. aerogenes
capsulatus in the faeces. 2 Loc. tit.

3 " Die Agglutination bei Gasphlegmonbacillen," Archiv f. Hyg.,
liii, p. 128, 1905.

200 INFECTIONS OF THE DIGESTIVE TRACT

Passini 1 obtained positive results in some of his cases
not only with homologous strains but also with unre-
lated ones. In this case, however, the agglutinative
action of the immune serum was less marked than in the
case of the sera obtained by Werner. It is evident that
there is still much to be learned in relation to the im-
munizing action of B. aerogenes capsulatus and that such
an action hi man may yet be discovered.

The organisms we have classed as belonging to the
B. aerogenes capsulatus type are large, plump, usually
straight bacilli, which as they occur in the faeces, can
usually be shown to be provided with a capsule. Organ-
isms which have developed in a living or dead rabbit
always acquire capsules. The organisms occur very
often in pairs, end to end, sometimes singly ; sometimes
in chains; sometimes as threads, which may be nearly
straight or sharply bent on themselves. The ends of
adjacent bacilli are slightly rounded or squared, though
not so sharply squared as in the case of anthrax bacilli.
They are immobile when viewed in hanging drops.
Spore formation occurs with difficulty ; i.e. chiefly under
very special conditions, such as on a medium containing
blood serum or within the body of an animal. Occa-
sionally spore formation is seen in blood-agar colonies,
the bacilli from which in other respects conform to the
characters of B. aerogenes capsulatus. On sugar bouillon
gas formation is abundant and rapid, twice as much gas
(or more) being formed in twenty-four hours as is usually

1 "Variabilitat der Bakterien und Agglutinatinosphanomen,"
Munch, med. Wochenschr., li, p. 1283, 1904.

INFECTIONS OF THE DIGESTIVE TRACT 201

formed by organisms of the B. coli group. The gas
consists of from one-third to one-half carbon dioxide,
the remaining gas consisting mainly of hydrogen.1 The

1 Professor Theobald Smith gives the following gas formula for
B. aerogenes capsulatus : —

H 2 =3
CO2 1 2

The following table illustrates the approximate ratio of hydro-
gen and carbon dioxide : —

Source of Microorganism

Medium

Height of Gas
Column in Milli-
meters (in 24
hours)
Height of anaero-
bic limb 95 mm.

H
C02

1 . Young man with slight di-
gestive derangement

Milk

87

I

-

Dextrose-
bouillon-
blood

7

f

2. Same case as No. 1

Milk

75

f

« „

Dextrose-
bouillon-
blood

80

f

3. Pernicious ansemia

Milk

60

f

«

Dextrose-
bouillon-
blood

75

f

4. Pernicious ansemia

Milk

70

1

5. Pernicious ansemia

Milk

100

f

«

M

85

f

6. Pernicious ansemia

Milk

90

f

7. Same case

Milk

84

I

8. Aerobic capsulate bacillus

Milk

87

f

9. Milk

Milk

90

\

10. Milk

Milk

85

|

202 INFECTIONS OF THE DIGESTIVE TRACT

gas production in incubated rabbits is very rapid and is
associated with a characteristic sweetish, sickening odor of
butyric acid mixed with some unknown constituent or
constituents. The gas obtained from the peritoneal cav-
ity and connective tissues gives the hydrogen " bark" and
burns with a blue flame. The liquefaction of muscles,
liver, etc., is remarkably rapid in such incubated rabbits.
Grown on pasteurized milk, the bacteria in question induce
rapid gas formation ("stormy fermentation") with
disruption of curds into small masses and peptonization
of casein. They do not quickly produce hydrogen
sulphide or methyl mercaptan on ordinary sugar-free
media, but may perhaps make these sulphur compounds
more readily on milk. Grown in fermentation tubes
containing blood bouillon, they rapidly liberate haemo-
globin. The organisms are strictly anaerobic and many
of their colonies on blood agar appear after two or three
days as minute points which lie beneath the surface and
develop into fuzzy spheres. These spherical colonies
often have dark centers. The microorganisms induce
inflammatory necrotic changes with gas formation when
injected into susceptible animals, such as pigeons.

As first shown by Professor Theobald Smith, B. aero-
genes capsulatus usually grows readily on bouillon in the
closed arm of the fermentation tube provided small
bits of sterile, fresh tissue are introduced into it. The
liver of the guinea-pig may advantageously be used.
The presence of the tissue probably favors the growth
in two ways : by furnishing a constituent of the medium
necessary for the growth of the organism, and by in-

INFECTIONS OF THE DIGESTIVE TRACT 203

ducing a more strict condition of anaerobiosis by the
reducing activity of the cells.

If we cultivate B. aerogenes capsulatus on sugar
bouillon we find that it is a large producer of gas (mainly
hydrogen and carbon dioxide) and that it makes butyric
and closely related acids in abundance, while the forma-
tion of lactic acid is small. On media which contain
very little sugar but much proteid, the organism is
still able to make gas in considerable amounts, though
less freely than on a sugar medium, in which the libera-
tion of gas is remarkably rapid. In nearly sugar-free
media the gas bacillus produces butyric acid, and the
quantity of this hi old cultures may be surprisingly
great. Ammonia is formed at the same time and serves
to neutralize at least in part the acid which is simul-
taneously made. Most varieties of the organism are not
indol producers when grown alone in blood bouillon.
There are strains which make indol.

More important for the pathologist than any of these
substances is the formation of a moderately hsemolytic
substance (or substances) by the gas-bacillus. Evidence
of such substances was obtained in a five-day culture
of capsulatus in blood bouillon. One-half of one cubic
centimeter of the filtrate from this culture induced
haemolysis in a suspension of rabbit's red cells prepared
by Ehrlich's method, the filtrate having been carefully
neutralized to the litmus point. The same result was
obtained in the case of red cells from a large Rhesus
monkey. Treatment of this filtrate in an exhaustion
apparatus very slightly reduced the hsemolytic action;

204 INFECTIONS OF THE DIGESTIVE TRACT

heating to 70° C. for one hour reduced it still further;
but even boiling did not wholly destroy it.

In order to determine whether this hsemolytic action
was dependent in part on volatile ammonium com-
pounds, the capsulatus filtrate was rendered distinctly
alkaline with sodium carbonate and concentrated under
reduced pressure at a low temperature for the removal
of ammonia. The filtrate was then restored to its
original volume. It was found that the hsemolytic
action of the fluid was somewhat diminished but not
lost.

The hsemolyzing action of B. aerogenes capsulatus is
very clearly shown in fermentation tubes containing
sugar-blood bouillon which have been inoculated with
pure cultures. A free liberation of haemoglobin occurs in
twenty-four hours or less. A similar result is seen in the
case of an aerobic organism closely resembling B. aero-
genes capsulatus in morphology. The bacillus of malig-
nant oedema does not exert a hsemolyzing action under
similar conditions. B. putrificus was found to reduce
hemoglobin, but this change is much less marked in the
case of B. aerogenes capsulatus.

There are other indications that B. aerogenes capsulatus
makes a hsemolytic substance or ha3molytic substances.
Rabbits injected with pure cultures of B. aerogenes capsu-
latus and then incubated at 37° C. soon show indications
of haemolysis, whereas control animals subjected to the
same procedure do not exhibit an equal degree of hee-
molytic change in the same period of time. This corre-
sponds with the observation that advanced haemolysis

INFECTIONS OF THE DIGESTIVE TRACT 205

is usually noted in persons who at autopsy show signs of
general invasion of the gas-bacillus.

The capacity of B. aerogenes capsulatus to form poisons
is apparently not limited to haemolytic and proteo-
clastic substances. Kamen * states that when the sporo-
genic form2 of the organism is grown on a suitable
medium it forms soluble poisons (obtainable in the
filtrate) which are capable of inducing in rabbits a state
of nervous excitation followed by general convulsions
and paralysis of respiration. He found that one-half to
one and one-half cubic centimeters is fatal for one kilo-
gram rabbit in one-half to one minute. Kamen likens
the material studied by him to Faust's "sepsin," hi that
it acts as a respiratory poison and induces vomiting,
bloody diarrhoea, tenesmus, and death. These gastro-
enteric symptoms are associated with oedema and hy-
persemia of the gastro-enteric tract. The poisonous
substances in question are said to be dialyzable and
not destroyed by heating for fifteen minutes at 60° C. I
have been unable to confirm these observations by means
of filtrates prepared from cultures of the vegetative form
of B. aerogenes capsulatus, but it is distinctly stated by
Passini that the products of the sporogenic and asporo-
genic forms of the organism are different.

I think it important to determine whether B. aerogenes
capsulatus is able to produce substances injurious to the

1 "Zur Etiologie der Gasphlegmone," Centralbl. f. Bakt., Orig.,
xxxv, pp. 554, 686, 1904.

2 This form of the organism was obtained by growing the com-
mon vegetative form on an egg-and-bouillon medium.

206 INFECTIONS OF THE DIGESTIVE TRACT

nervous system. Persons suffering from severe forms of
chronic infection by this organism almost always show
signs of intoxication of the nervous system, but such
manifestations are very different in different individuals.
It is not yet clear whether the gas-bacillus makes
a substance capable of exciting an acute inflamma-
tion of the ileum or colon or whether preceding
mechanical or chemical irritation is necessary to enable
the organism to multiply rapidly and excite further
inflammation. Healthy monkeys may be fed consider-
able numbers of capsulati without developing signs of
inflammation in the intestine, although such feeding is
followed by an increase in these organisms in the faeces.
Two monkeys fed on gas livers from incubated rabbits
infused with pure cultures of B. aerogenes capsulatus
developed soft stools temporarily. Such experiments
are, however, quite different from the experiment of
introducing capsulati into a digestive tract already
somewhat inflamed and irritable in consequence of pre-
ceding infections. The ability of B. aerogenes cap-
sulatus to cause an inflammatory necrotic process in the
muscles of guinea-pigs and pigeons, which was noted
by Dr. Flexner many years ago, is of interest in this
relation. It appears probable that B. aerogenes cap-
sulatus is often the cause of slight inflammatory or per-
haps even necrotic changes in the mucous membrane
of the intestine. Howard * has described instances of

1 "Contributions to the Science of Medicine, Dedicated by his
Pupils to William Henry Welch on the 25th Anniversary of his
Doctorate," Baltimore, p. 461, 1900.

INFECTIONS OF THE DIGESTIVE TRACT 207

superficial necrosis of the mucous membrane of the
stomach and intestine, associated with the presence of
capsulatus in abundance. These necrotic areas most
often lie beneath the folds of the valvulse conniventes and
may occur with gas cysts. It does not seem likely that
B. aerogenes capsulatus is responsible for severe acute
inflammatory lesions of the intestine, but it is probable
that its activities will account for the subacute enteritis
that is so often present in cases that show large numbers
of the bacilli in the stools. It is certain that there are
instances of acute diarrhoea associated with very free
capsulatus multiplication and such diarrhoeas are com-
mon in persons with severe primary anaemia.1

In 1905 Tissier 2 described an acute intestinal affection,
especially frequent in nursling children, which he ascribed
to the presence of almost pure cultures of B. perfringens.
He claimed that in the treatment of these perfringens diar-
rhoeas which ordinarily ran a mild course the most effica-
cious therapeutic agency was the employment of cultures
of a lactic acid bacillus (B. paralacticus) . I have in a few
instances met with acute and subacute diarrhceal affec-
tions of children in which B. aerogenes capsulatus may per-
haps have been the cause of the intestinal derangement,
although in these cases pathogenic varieties of colon
bacilli have not been absolutely excluded as causal agents.

1 1 am disposed to attribute this diarrhrea to an associated
streptococcus infection in these cases. One may observe cases
of advanced infection with the gas-bacillus without any diarrhoea;
indeed obstinate constipation is found in some of them.

2 "Etude d'une varie"te" d'infection sur le nourrisson," Ann.
de I'lnst. Pasteur, xix, p. 273, 1905.

208 INFECTIONS OF THE DIGESTIVE TRACT

In this connection it should not be forgotten that Klein
isolated from the fseces in an outbreak of diarrhoea an
organism which very probably is identical with B.
aerogenes capsulatus. This organism was given the name
B. enteritidis sporogenes. The position of the organism
described by Klein is open to some suspicion because
while it is stated by him to produce gas and butyric acid
in the very characteristic manner observed by Welch
and Nuttall for B. aerogenes capsulatus, it differs from
the latter in having motility and apparently in sporulat-
ing much more readily. Cultures sent by Dr. Klein to
Professor Welch contained bacilli which agreed in every
detail with pure cultures of B. aerogenes capsulatus.1
As, however, the faeces sometimes contain bacteria
having the morphology of B. aerogenes capsulatus but
differing from this organism in forming spores much
more readily, it is possible that Klein was in reality
dealing with impure cultures of B. aerogenes capsulatus.
I am disposed to believe, after a careful review of the
available evidence on the subject, that Klein had an
organism distinct from B. aerogenes capsulatus, though
he may have had this also.

B. aerogenes capsulatus belongs in the group of anaerobic
putrefactive microorganisms capable of making butyric
acid and hydrogen. If careful biochemical tests be made,
there is no difficulty about distinguishing the gas-bacillus
from other anaerobes that occur in the human diges-
tive tract. The bacillus of malignant oedema is some-

1 See "Manual of Bacteriology by Muir and Ritchie," American
Edition, p. 354, 1904.

INFECTIONS OF THE DIGESTIVE TRACT 209

times found in the human intestine. Whether it is
ever found in large numbers or ever possesses patho-
logical significance is at present unknown. The bacil-
lus of malignant oedema resembles the gas-bacillus in
morphology, but sporulates more readily, is motile, and
forms less gas on sugar media. Its behavior on milk is
wholly different from that of the gas-bacillus. When
injected into a living rabbit which is subsequently killed
and incubated it is found that there has been little or
no liberation of gas in the liver and elsewhere — a
feature which easily distinguishes it from B. aerogenes
capsulatus.1 An organism which closely resembles B.
aerogenes capsulatus in many respects is the bacillus of
rauschbrand or symptomatic anthrax (the cause of the
quarter-evil or black-leg). This organism is, however,
frequently Gram-negative under conditions of staining
in which the gas-bacillus is Gram-positive; it is also
motile. It resembles B. aerogenes capsulatus in being
an abundant gas producer, the gas formed showing a
relation between hydrogen and carbon dioxide closely
resembling that obtained from the gas-bacillus. This
organism is the only one of the anaerobes which I have
found to be able to give rise to a large production of gas
in an incubated rabbit. It might be supposed that this
organism could be mistaken for the gas-bacillus, but I
think the difference in the conditions of sporulation would

1 Ghon and Mucha gave a comparison of the gas-bacillus with
the bacillus of malignant redema and a third organism (derived
from a case of peritonitis) closely related to these ("Beitrage
zur Kentniss der anaeroben Bakterien des Menschen," Cen-
tralbl. f. Bakt., Orig., xl, p. 37, 1905-06).

210 INFECTIONS OF THE DIGESTIVE TRACT

suffice to make the distinction. It was found that in the
rabbit test the rauschbrand bacillus sent me by Dr.
Theobald Smith sporulated very freely and characteris-
tically in the blood and spleen and liver. We have
never observed sporulation on the part of B. aerogenes
capsulatus within the body of a rabbit incubated accord-
ing to the Welch-Nut tall method. Whether the bacillus
of rauschbrand ordinarily occurs in the human intestine
is questionable. It seems unlikely that it should be
found there except in the case of persons living in asso-
ciation with infected cattle.

Bacillus botulinus. — Botulism is a variety of meat
poisoning in which the pathological effects are due to a
powerful soluble poison made by a strict anaerobe, these
effects being exerted mainly on the central nervous
system. The organism in question resembles in some
respects B. aerogenes capsulatus} and in others the bacillus
of tetanus.1 It was isolated in 1896 by Van Ermengen
from a sample of ham, the eating of which in a raw state
had caused many instances of poisoning and some with

1 B. botulinus possesses a varied morphology, sporulates ter-
minally, makes little gas on sugar-blood bouillon, and is motile.
The specimen sent to me by Krai as Van Ermengen 's B. botulinus
was introduced into a rabbit and subjected to the incubation test.
It was found that the liver was in an advanced state of putrefac-
tion, but neither this organ nor any other part of the body showed
even the slightest indications of gas formation. It is noteworthy
that the vegetative form of the organism (which is the form in
which the botulinus bacillus was introduced) showed evidence of
sporulation in the spleen but not in the blood or the liver. An
absorption test made by Mr. Ward showed that the gas formed by
this bacillus on dextrose-blood bouillon consisted of four parts
carbon dioxide to one part of hydrogen.

INFECTIONS OF THE DIGESTIVE TRACT 211

fatal ending. The symptoms closely resembled those of
the so-called cases of sausage poisoning which had been
reported from time to time. Very recently the interest-
ing fact has developed in a Darmstadt outbreak that
the same type of poisoning may occur from eating vege-
table food ; namely, in this instance, beans which were
imperfectly canned.1 It appears probable that the beans
were infected through the use of. manure obtained from
pigs.

The following are the leading symptoms of classical
botulism: first, disturbances of the external muscles
of the eyeball, especially ptosis, abducens paralysis, and
disturbances of associated movement with nystag-
mic contractions; second, disturbances of the internal
muscles of the eye. In the majority of instances there
have been paralyses of accommodation. In the Darm-
stadt outbreak from infected beans an enlargement and
rigidity of the pupils was never observed, but in the out-
breaks from meat poisoning these pupillary conditions
have been usually noted. Third, there are usually
derangements referable to the region of the pons and
medulla oblongata, including disorders of the tongue,
such as swelling and paralysis, pharyngeal and laryn-
geal paralyses and disturbances of the heart and respira-
tion. Fourth, weakness and even paralysis of motion
may appear. These disturbances have almost always
been unaccompanied by changes in sensibility and

1 Fischer, " Ueber eine Massenerkrankung an Botulismus in-
folge Genusses verdorbener Bohnenkonserven," Zeitschr. f. klin.
Med., lix, p. 58, 1906.

212 INFECTIONS OF THE DIGESTIVE TRACT

consciousness ; fever has been absent and there has been
an absence of disturbances of the stomach and intestine.
The symptoms seldom come on before twelve to twenty-
four hours after the ingestion of the poison. Very
similar symptoms were produced by Van Ermengen in
certain animals by giving them watery extracts of the
infected ham or cultures either by the stomach or by
subcutaneous injections. It is noteworthy that in these
animal experiments there was observed a period of in-
cubation of not less than six to twelve hours before the
onset of symptoms. The characteristic paralytic effects
can be induced by means of the filtered toxin, and it is
therefore clear that the phenomena of botulism are to
be ascribed, like the nervous phenomena of tetanus, to a
powerful toxin. Indeed, the properties of the toxin of
B. botulinus have been studied and have been shown to
resemble closely the toxins of tetanus and diphtheria
with respect to instability, conditions of precipitation,
etc. An antitoxin1 has been prepared by Kempner.
It has the power of neutralizing the toxin and has con-
siderable therapeutic effect when given a short time after
the toxin.

It is practically very important to realize that meat
may be extensively infected with the B. botulinus and
may contain relatively large quantities of the toxins with-
out showing the ordinary signs of decomposition. We

1 For a discussion of the conditions under which antitoxin
formation occurs see J. Forssman, "Studien iiber die Antitoxin-
bildung bei aktiver Immunisierung gegen Botulismus," Cen-
tralbl. f. Bakt., Orig., xxxviii, p. 463, 1905.

INFECTIONS OF THE DIGESTIVE TRACT 213

have therefore in B. botulinus, as in the bacillus of teta-
nus, an example of strict anaerobes which have acquired
the ability to make intense poisons but have apparently
incidentally lost in a measure the common putrefactive
properties ordinarily possessed by spore-bearing anae-
robes. There is reason to think that isolated cases of
botulism are not extremely rare in the United States.
The recognition of these cases by purely clinical means
may be very difficult when the typical symptoms are
not fully developed. In foreign countries, where meat
is distributed by a different system from that prevailing
in the United States, the outbreaks of botulism are apt
to involve many people in one locality, whereas in our
country the cases are apt to occur in small numbers,
owing to the greater opportunities for the consumption
of portions of the carcass hi different localities.

THE FERMENTATIVE AND PUTREFACTIVE PRO-
CESSES FROM THE STANDPOINT OF THEIR
PRODUCTS

IT is helpful to review the fermentative and putrefac-
tive processes in the digestive tract of man from the
standpoint of their products although such a course may
somewhat temper enthusiasm by revealing our ignorance
in many important directions. I shall use the word
"fermentative" to designate the decompositions of car-
bohydrate and fatty substances and the word "putre-
factive" as applied to the cleavages of proteid and allied
substances. The distinction is important, for while the
products of fermentation are in themselves usually
unimportant as agents of intoxication, the products of
putrefaction include substances containing sulphur or
nitrogen or both sulphur and nitrogen, and thus frequently
derive a chemical basis for exerting toxic effects. While
making this distinction, however, I do not lose sight of
the fact that fermentative and putrefactive processes
overlap in the sense that they furnish some products
in common, such as carbon dioxide and volatile fatty
acids ; they are moreover closely linked by the fact that
excessive fermentation in the digestive tract nearly
always leads to excessive putrefaction, for reasons which
I hope to make clear.

The products of fermentative decomposition in the

214

INFECTIONS OF THE DIGESTIVE TRACT 215

digestive tract may be dismissed with few words. The
carbon dioxide which results from the breakdown of
sugars is of interest mainly as a cause of gastric flatu-
lence or small-intestinal flatulence. The alcohol formed
is probably in too small amount to be toxically significant
even when fermentation of sugars is greatly excessive.
Of the acids formed we have to think of lactic, acetic,
propionic, and butyric, but mainly of lactic and acetic.
The higher volatile fatty acids come especially from
spore-bearing anaerobes (although other types of bacteria
may produce them) and therefore interest us also as
putrefactive products. All these acids, however, are
irritants by virtue of their acid properties. If present
in considerable concentration in a healthy digestive
tract or in more moderate concentration in the tract of a
person with an irritable stomach or small intestine, they
may be efficient factors in exciting vomiting or diarrhoea.
Nervous disturbances resulting in delayed absorption
may be responsible for the accumulation of considerable
acid fluid in the stomach or small intestine. There is
also a quite different aspect to the excessive production
of acid ; namely, the withdrawal of excessive amounts of
alkali from the organism. All the acids mentioned are
readily burned in the body, but if their absorption occurs
with uncommon rapidity, they may be excreted unburned
and combined with alkali. This robbing of the organism
-of alkali by acid is under some conditions a grave matter,
leading to a definite form of intoxication, but probably
the acids of fermentative origin are always associated
with those of putrefactive origin in bringing about this

216 INFECTIONS OF THE DIGESTIVE TRACT

kind of intoxication, where this is of digestive rather
than of metabolic origin. It appears to me very ques-
tionable whether one ever meets with cases of marked
acid intoxication referable to the absorption of excessive
quantities of organic acids formed in the digestive
tract, but there is some reason to think that in young
children a moderate degree of acidosis sometimes has
its main origin in fermentative disturbances in the
intestine.

Oxalic Acid and Oxaluria. — It is now well established
that various molds and bacteria are capable of acting
upon media containing sugar, in such a manner as to give
rise to the formation of oxalic acid. The thought that
considerable quantities of oxalic acid may arise in the
digestive tract as the result of fermentative decomposi-
tion led Baldwin l to make studies with a view to learn-
ing whether oxaluria can be experimentally induced in
dogs by feeding large amounts of sugar together with
food (such as meat) which contains no oxalic acid. It
was found in these experiments that oxaluria was not
readily induced in this manner in healthy dogs, but that
prolonged feeding of sugar in large quantities eventually
led to a state of oxaluria. This experimental oxaluria
was associated with a disturbance of function in the
gastric mucous membrane during which large quantities
of mucus were formed. The contents ob tamed from
the stomach under these conditions failed to show the
presence of hydrochloric acid. It seems clear that the

1 "An Experimental Study of Oxaluria, with Special Reference
to its Fermentative Origin," Journ. of Exper. Med., v, p. 27, 1900.

INFECTIONS OF THE DIGESTIVE TRACT 217

oxaluria and the accompanying gastritis were referable
to pathological fermentative processes induced by the
excessive feeding with sugar. It was further found in
these experiments that there had been a production of
oxalic acid in the contents of the stomach.

It was further noticed in the experiments just men-
tioned that oxalic acid may be formed in a medium
containing sugar and beef extract, when this me-
dium is inoculated with material obtained from the
stomach of human patients showing a marked grade
of oxaluria.

Whether it is safe to conclude from the foregoing facts
that oxaluria in human beings ever arises from the fer-
mentation of carbohydrates in the digestive tract is not
entirely clear. It appears probable that there are con-
ditions of carbohydrate dyspepsia in which not incon-
siderable quantities of oxalic acid are formed through
fermentation of carbohydrates. The absorption of this
oxalic acid would necessarily add to the quantity
which must be burned in the organism in order to keep
the excretion of oxalic acid within normal limits. That
excessive excretion of oxalic acid by the urine is ever
due solely to the production of oxalic acid in the digestive
tract appears to me unlikely. Modern studies indicate
that there are many ways in which oxalic acid can be
formed in the organism, and the greatest significance
that can be claimed for the oxalic acid produced within
the digestive tract is that it may sometimes be formed
in such amounts as to make it one factor in bringing
about a pathological condition of oxaluria. Perhaps

218 INFECTIONS OF THE DIGESTIVE TRACT

the occurrence of carbohydrate dyspepsia which is so
common in cases of oxaluria may influence the occur-
rence of oxaluria in an indirect manner through its
effect upon proteid metabolism, but of such an influence
we at present know nothing.

Of the intermediate stages by which sugars are de-
composed with the production of oxalic acid we are at
present in ignorance, but there is some reason to suspect
that glyoxylic acid forms an intermediate product of
decomposition. This, however, does not make it neces-
sary to assume that any considerable quantity of glyox-
ylic acid is present at one time.

Propionic, butyric, and higher volatile fatty acids are
common products of putrefaction. They are formed
by putrefactive anaerobes not merely from sugar-holding
media, but from proteid media in which there is no free
sugar, although carbohydrates may be essential to ini-
tiate decomposition. Among the most important anae-
robic bacteria capable of forming butyric acid during
putrefaction are the motile organisms known as the
amylobacteria, the granulo-bacillus immobilis liquefaciens
of Grassberger and Schattenfroh (probably a non-patho-
genic form of B. aerogenes capsulatus) , the bacillus of
symptomatic anthrax (rauschbrand), the bacillus of
malignant oedema, and the B. putrificus of Bienstock
(cadaver bacillus).

Acetone. — The interesting observation has been
made that the mixed faecal flora growing upon a peptone-
bouillon medium for a period of a week at the body
temperature in certain instances gives rise to the forma-

INFECTIONS OF THE DIGESTIVE TRACT 219

tion of acetone. This appears, however, to be an
exceptional occurrence. Up to the present time it has
been observed only in the case of material derived
from two individuals. In one of these the bacteria with
which the medium was inoculated came from a child
suffering from well-marked chronic intestinal putrefac-
tion of the marantic type with the development of a large
abdomen.

The other instance in which the flora gave rise to the
formation of acetone was one of advanced anaemia.
Why acetone was obtained from a culture in dextrose
bouillon and calcium carbonate but not in dextrose
bouillon without calcium carbonate was not clear. The
flora from this case also gave rise to the production of
acetone in milk to which calcium carbonate had been
added. The presence of sugar is not necessary to the
production of acetone by intestinal flora, as is shown
by the fact that it has been obtained from plain
bouillon.

It is not at present clear what significance should be
attached to the ability of the intestinal bacteria in cer-
tain cases to produce acetone. The observations here-
tofore made do not furnish us with any evidence that
acetone is ever produced in the human organism in the
intestinal tract through the activity of microorganisms.
Nevertheless the finding of acetone under the conditions
just mentioned suggests the possibility that under some
conditions this ketone may be produced in the digestive
tract. I do not know of any observations that have
heretofore been made which have shown the formation

220 INFECTIONS OF THE DIGESTIVE TRACT

of acetone by intestinal bacteria. Schardinger1 found
an organism which he called B. mascerans which gave a
mixture of acetone and alcohol during the fermentation
of various carbohydrates.

An organism was also found by Br£audat 2 which was
able to produce acetone in a peptone solution. This
bacterium was derived from the drinking water of a town
in France, which came from a subterranean well. It was
a Gram-negative facultative aerobe capable of liquefy-
ing gelatin. Under certain conditions this organism
developed spores. On potato it gave rise to a violet
color. The organism was described under the name of
B. violarius acetonicus.

When we turn to the consideration of the nitrogen-
holding and sulphur-holding products of putrefactive
cleavage, the scantiness of our knowledge comes into view
with almost discouraging clearness. That putrefactive
processes are attended by the formation of bases, such
as ammonia, amines, diamines (such as putrescin and
cadaverin), cholin, neurin, sulphur compounds, and va-
rious aromatic bodies, has been known many years,
and something has been learned, though by no means
enough, about the media and the bacteria which deter-
mine the presence and proportions of these substances.
When, however, we ask ourselves what we can safely
say of the conditions under which such substances arise

1 " Mitteilung aus der staatlichen Untersuchungsaustalt fiir
Lebensmittel," Wien. klin. Wochenschr., xvii, p. 207, 1904.

J"Sur un nouveau microbe producteur d'acetone," Ann. de
Vlnst. Pasteur, xx, p. 874, 1906.

INFECTIONS OF THE DIGESTIVE TRACT 221

in the human intestine and of their pathological effects,
we are able to give in most instances only very inade-
quate answers.

BASIC SUBSTANCES

In the course of putrefaction of proteids in the
intestine ammonia is regularly formed. It does not
appear, however, that it is formed in quantities or
under conditions which render it toxic to the organ-
ism, although, as elsewhere stated, it is possible that
ammonium butyrate acts as a local irritant. The
organism is well adapted to care for moderate quanti-
ties of ammonia, which, as is well known, is united with
carbon dioxide in the liver and elsewhere to form urea.
The splitting-off of ammonia from amino-acids formed
in the course of normal digestion is a physiological pro-
cess of fundamental importance and it is probably
owing to the physiological arrangements necessary for
the disposal of ammonia in a relatively non-toxic form
that the organism is able to tolerate the digestion of
large quantities of ammonia-forming foods. Likewise
we know nothing of any toxic action from methylamin
or other alkyl amines. It has been supposed that
guanidin1 and methyl guanidin sometimes arise in the
course of certain kinds of putrefaction. While it cannot
be denied that this may be the case, we have at present
no knowledge that it is the case, and are unable to
correlate any clinical conditions with the formation and

1 Kutcher showed that guanidin may be an end-product of
prolonged pancreatic digestion. He found the base in the arginin
fraction.

222 INFECTIONS OF THE DIGESTIVE TRACT

absorption of guanidin or methyl guanidin. Methyl
guanidin arises from the oxidation of creatin (or
creatinin), and it is conceivable that under some con-
ditions of decomposition of meat these bodies are formed.
Since, however, there is no definite evidence on this
point in the case of man, I shall not enter into a discus-
sion of mere possibilities.

In reference to the bases cholin and neurin the case is
somewhat different. Of these bases neurin is much more
toxic than cholin, its action being directed especially to
the nervous system. That cholin is toxic when absorbed
from the intestinal tract is not wholly clear although it
is well known that the subcutaneous injection of cholin
gives rise to nervous symptoms and a decline hi blood
pressure. The most definite results that have been ob-
tained by experimental methods in respect to the pro-
duction of cholin are those of Beattie Nesbit.1 Nesbit
started from the fact that lecithin may be decomposed
with a yield of cholin. It has been assumed by Bocai
that lecithin is decomposed during the process of diges-
tion into glycerophosphoric acid and fatty acids and cho-
lin, and that these products are individually absorbed.
This view suggests that it might be dangerous to con-
sume very large quantities of lecithin as, for example, in
the form of eggs, as cholin cannot be regarded as a wholly
harmless substance. It must be admitted, however,
that there is evidence to support the idea that some

1 " On the Presence of Cholin and Neurin in the Intestinal Canal
during its Complete Obstruction. A Research on Autointoxi-
cation," Journ. Exper. Med., iv, p. 1, 1899.

INFECTIONS OF THE DIGESTIVE TRACT 223

lecithin may be absorbed without decomposition.1 It
may thus be a matter of importance to the organism
whether the lecithin introduced into the digestive tract
is split through the decomposing action of putrefactive
bacteria or whether it undergoes absorption. The mere
quantity of lecithin introduced would therefore not be a
measure of the quantity of cholin formed and absorbed.
In the experiments of Nesbit dogs which had been fed
for several days upon the yolks of eggs were subjected to
an operation in which a ligature was placed about the
small intestine just above the ileocsecal valve. In ex-
periments of this sort in which the animal lived for several
days after operation the intestinal contents were removed
and studied for the presence of cholin. On the basis of
various reactions for cholin chloride it was determined
that the intestinal contents of the animal contained this
substance. Some evidence also was obtained to indi-
cate that neurin was also formed. I think the evidence
on this point is not quite convincing. Nesbit also found
evidences of a ptomaine associated with cholin and
neurin, the chemical nature of which was left in doubt.
These observations render it probable that, provided the
intestine contains any considerable quantity of lecithin,
cholin, neurin, and perhaps other bases are formed as a
result of intestinal obstruction. It must be stated, nev-
ertheless, that Nesbit failed to obtain definite evidence
that the bases found by him in the intestinal canal of the
dog were really toxic. It is clear that much more work

1 P. von Walter, "Zur Lehre von der Fettresorption," Archiv
f. Anat. u. Physiol, Physiol. Abt., p. 329, 1890.

224 INFECTIONS OF THE DIGESTIVE TRACT

needs to be done before the relation of cholin and neurin
to intoxications from the digestive tract is placed on a
satisfactory basis.

There are some instances of a remarkable idiosyncrasy
on the part of certain persons toward eggs. I recall a
child in whom the taking of an egg was regularly fol-
lowed by prostration and slight fever and headache.
Whether such phenomena as these bear any relation to
the formation of cholin or neurin is of course an open
question.

Putrescin and Cadaverin. — Putrescin (tetramethyl-
endiamin) and cadaverin (pentamethylendiamin) are
products of proteid decomposition which are formed at
times in the human intestine under conditions at present
obscure. It seems not improbable that peculiar states
of bacterial activity may have an influence on the pro-
duction of these bases, and it is very desirable that the
bacterial flora should be studied with great care in those
cases of cystinuria in which putrescin and cadaverin can
be obtained from the intestinal contents. According
to Brieger * the peculiar odor of cholera stools is due
principally to pentamethylendiamin, but strict proof
of this has not yet been furnished.

Baumann and Udranzky2 found about 0.5 gram per

1 Brieger and Stadthagen, "Ueber Cystinurie," Berl. klin. Wo-
chenschr., xxvi, p. 344, 1889.

2 "Das Benzoylchlorid als Reagens," Bericht. d. Chem. Ge-
sellsch., xxi, p. 2744, 1888; "Ueber die Identitat des Putrescins
und des Tetramethylendiamin," ibid., p. 2938; "Ueber das
Vorkommen von Diaminen sogenannten Ptomainen bei Cystinu-
rie," Zeitschr. f. physiol Chem., xiii, p. 562, 1899.

INFECTIONS OF THE DIGESTIVE TRACT 225

day of diamines (principally tetramethylendiamin) in
the faeces of a cystinuria patient. This quantity is
equal to the amounts found in the urine from the same
case. In the urine, however, the cadaverin constituted
about sixty per cent, of the diamines ; in the faeces only
from ten to fifteen per cent, of the bases consisted of
cadaverin. Neither Brieger nor Baumann and Udranzky
was able to find diamines in normal fasces. Moreover,
Baumann and Udranzky were not able to find these
bases in examples of disease other than cystinuria.
Roos l in a case of combined malaria and dysentery with
fever and enlargement of the spleen found in the faeces
a small quantity of pentamethylendiamin. There was
very little indican in the urine of this patient. He was
unable to find diamines in cases of malaria alone. In a
case of fever following gonorrhoea he was able to obtain
small quantities of dibenzoylcadaverin.

Although the study of the conditions under which
putrescin and cadaverin are formed in the intestinal
tract is of much biological interest, there is at the present
time little evidence that these diamines are ever formed
in sufficient quantities in the human intestine to con-
stitute in themselves factors in the production of states
of intoxication. The association with cystinuria is a
striking fact, and the further investigation of this disease
will doubtless give us an explanation of the relationship
between the production of diamines and the formation
of cystin, if indeed there be any necessary relation.

1"Ueber das Vorkommen von Diaminen bei Krankheiten,"
Zeitschr. f. physiol. Chem.t xvi, p. 192, 1892.

q

226 INFECTIONS OF THE DIGESTIVE TRACT

Ellinger J found that putrescin may be obtained from
the bacterial decomposition of ornithin. He found also
that cadaverin results from the decomposition of lysin.
Under the designation " ptomaines" Selmi 2 described a
number of basic substances resembling plant alkaloids
in their reactions and liable to be mistaken for morphine,
conine, etc. He obtained them from cadavers. That
such bodies are sometimes formed in the human digestive
tract appears likely, but has not been satisfactorily
established. In fact, the " ptomaines" are in most in-
stances not so clearly defined in their chemical characters
as could be desired. Moreover, their pharmacological
characters are also insufficiently definite. For these
reasons any discussion of these bodies in the present
connection would be purely academic.

SULPHUR COMPOUNDS

Mercaptan. — The sulphur compounds resulting from
putrefactive decomposition in the intestine have re-
ceived little attention from the standpoint of their
pharmacological action. It is therefore difficult at
present to form a just estimate of their importance for
intestinal intoxications. It was claimed by Nencki
that methyl mercaptan is one of the gases formed in the
intestine during putrefaction. He based this contention
on a study of the faeces, but did not take the precaution
to examine the material promptly after collection.

1 "Die Constitution des Ornithins und des Lysins," Zeitschr. f.
physiol. Chem., xxix, p. 334, 1900.

2"Sulle ptomaine ed alcaloidi cadaverici e lore imputanza in
toxicogia," Bericht. d. deutsch. chem. Gesellsch,, ii, p. 808, 1878.

INFECTIONS OF THE DIGESTIVE TRACT 227

It appears to me, therefore, not clear that the methyl
mercaptan found by Nencki was not formed during
putrefaction outside the body. My scepticism on this
point rests partly on the fact that although I have
made many examinations of the fresh faeces from per-
sons with excessive intestinal putrefaction, I have never
been able to find methyl mercaptan by means of the
isatin-sulphuric-acid reaction. It appears improbable
that methyl mercaptan is produced in the human large
intestine in the course of ordinary putrefactive troubles,
although it must be admitted that it may perhaps arise
under some exceptional conditions. It may be stated
before leaving the subject of mercaptan that experi-
ments were made upon dogs and monkeys in which ethyl
mercaptan in a solution of one part in one thousand
parts of water failed, when injected into the large
intestine, to give rise to definite evidence of intoxica-
tion unless very large quantities were employed. Thus
during a period of ten days a dog of medium size re-
ceived 20, 30, 50, 56, 110, 110, 110, and 110 c.c. of such a
solution on successive days without exhibiting symp-
toms. The injection of 50 c.c. of a solution two and one-
half parts per one thousand parts was not well retained.
It was only when a dose of 120 c.c. of this concentration
was introduced into the stomach that vomiting occurred.
The rectal injection of an ethyl mercaptan solution of
1 : 1000 strength in quantities of 10 to 30 c.c. daily in a
medium-sized monkey failed to give rise to symptoms.

Hydrogen Sulphide. — There is reason for thinking that
the production of hydrogen sulphide in the digestive

228 INFECTIONS OF THE DIGESTIVE TRACT

tract is of more importance to the organism than the
formation of mercaptan. That this gas is regularly
formed in at least moderate quantities under normal
conditions is apparent from the fact that the intestinal
contents are rendered dark in color after the administra-
tion of iron salts owing to the formation of sulphide of
iron. This occurs, as is well known, in well persons.
Moreover, the presence of hydrogen sulphide in the
freshly voided fseces can be readily demonstrated. It
is also true that the mixed faecal flora from the human
intestine, both in health and disease, is capable of
producing hydrogen sulphide upon a medium containing
proteid that has been partially hydrolyzed. This is
true, for example, of the growths upon bouillon.
Hydrogen sulphide is produced early in the course of
putrefaction, and its presence in flasks containing
peptone bouillon inoculated with f secal flora is regularly
demonstrable within the first twenty-four hours. It is
difficult to form a judgment as to how much hydrogen
sulphide is produced through intestinal putrefaction in
the course of a given period or how much is absorbed.
Assuming that the production is the same in a given
series of individuals, it is highly probable that the ab-
sorption would be variable, and it is of course upon the
absorption of the gas that its importance to the organism
chiefly depends.

It seems probable that in health the formation of
hydrogen sulphide is limited to the large intestine and
perhaps a short extent of the adjoining small intestine.
Under pathological conditions, hydrogen sulphide is

INFECTIONS OF THE DIGESTIVE TRACT 229

formed in other parts of the digestive tract. There are
cases of dilatation of the stomach from which one may
remove material which is undergoing putrefaction with
the formation of hydrogen sulphide. The production
of hydrogen sulphide in this situation is certainly to be
regarded as pathological. I am not clear, however, that
it points to the presence of pathological bacteria, for
it is easy to show that typical organisms of the B. lactis
aerogenes type are able to form hydrogen sulphide when
growing in bouillon containing cystin. Cystin is a decom-
position product of proteids, and it is conceivable that in
a stagnant stomach proteid digestion may go on to the
formation of cystin among the cleavage products and that
this is then decomposed through the action of B. lactis
aerogenes. The ability to form hydrogen sulphide from
cystin is widespread among bacteria. The colon bacilli
liberate it abundantly when growing in a cystin medium.1

Among persons who take organic iron it is usual to
note, after a little time, a dark brown color on the tongue
which appears to be dependent on the formation of iron
sulphide in this situation. The sulphide production
is referable, doubtless, to the formation of hydrogen
sulphide, but the conditions of its production are not
I known. It does not appear that the blackening of the
, tongue after taking iron can be regarded as pathological,
|as it may be observed in persons apparently free from
i disorders of digestion.

I have made some observations to determine at what

1 This is true also of the bacilli of dysentery and of typhoid
fever.

230 INFECTIONS OF THE DIGESTIVE TRACT

level of the digestive tract one finds bacteria capable
of forming hydrogen sulphide when grown in peptone
bouillon. These observations were made upon the
digestive tract of infants. The mixed flora from
various portions of the large intestine and lower part of
the ileum regularly gave hydrogen sulphide. Where the
inoculated material was derived from children dying
of pneumonia or some condition unconnected with de-
rangements of digestion, it was unusual to obtain hydro-
gen sulphide from bacteria derived from portions of the
tract lying above the lower part of the ileum. It was
frequently noted, however, in children dying from
marasmus that the flora obtained from the stomach and
from portions of the small intestine above the ileum
gave hydrogen sulphide when grown on peptone bouillon.
In some instances it was observed that more hydrogen
sulphide was formed from the bacteria derived from the
stomach of such children than from the bacteria obtained
from the rectum.

We have at present very little satisfactory knowledge
of the influence of hydrogen sulphide upon the organism
in cases where the gas is liberated in the intestine. Fre-
quent reference is made in works dealing with auto-
intoxications to a publication by Senator,1 concerning
what he regarded as a case of pure hydrothionsemia.
The patient in this instance was a man supposed to have
been well previously and who after an error in diet be-

iaUeber einen Fall von Hydrothionamie und iiber Selbstin-
fection durch abnorme Verdauungsvorgange," BerL klin. Wochen-
schr., v, p. 254, 1868.

INFECTIONS OF THE DIGESTIVE TRACT 231

came ill with a gastro-enteric catarrh, presumably located
mainly in the caecum. On the third day after the illness
there suddenly occurred a severe seizure of vomiting
accompanied with an intense odor of hydrogen sulphide.
Simultaneously the patient suffered from dizziness and
a little later passed into a state of collapse without
loss of consciousness, the eruction of gas continuing mean-
time. The first urine passed gave a definite reaction for
hydrogen sulphide. On the following day the patient
recovered. Other somewhat similar instances have been
described by Betz,1 Stefanio, and Emminghaus.2 Among
the symptoms that have been met with in such cases
there have been prominent those pointing to disordered
function of the central nervous system, including head-
ache, dizziness, delirium, mental depression, drowsiness,
stupor, and collapse. Somewhat similar manifestations
have been observed in experimental poisoning by
hydrogen sulphide in animals and men.3 The nervous
symptoms observed under these conditions have fre-
quently been more severe than those arising in sponta-
neous hydrothionsemia, probably because the quantity
of gas absorbed is greater in the former cases. Appar-
ently somewhat at variance with the results of poisoning
in dogs is the statement that many persons subjected

lUUeber Hydro thionammoniaemie, " Memorabilien, Ix, p. 145,
1864.

2"Zwei Falle von mehrfacher Perforation des Verdauungs-
kanals und H2S-Gehalt im Urin," Berl. klin. Wochenschr., ix, pp.
477 and 491, 1872.

3 See especially K. B. Lehmann, " Experimentalle Studie liber
den Einfluss technisch und hygienisch wichtiger Gase und Dampfe
auf den Organismus," Archiv /. Hyg., xiv, p. 135, 1892.

232 INFECTIONS OF THE DIGESTIVE TRACT

to the Burgeon treatment for pulmonary phthisis have
been able to take considerable quantities of hydrogen
sulphide administered by the rectum without injurious
effects. It appears to me extremely doubtful, however,
from the results of experiments made upon dogs whether
it is really possible to bring about the rapid absorption
of more than small quantities of hydrogen sulphide in
man without giving rise to collapse. In experiments
conducted on dogs it is necessary to use the utmost
caution in administering the gas by the rectum in order to
prevent fatal collapse. Nevertheless, small quantities
of the gas may almost certainly be absorbed from the
human intestine without giving rise to noticeable dis-
turbance of nervous function or to the presence of
hydrogen sulphide in the urine. Friedrich Miiller *
showed that in various disorders of the intestine hydrogen
sulphide is liberated in the lungs, but without simulta-
neous appearance of hydrogen sulphide in the urine,
owing probably to the small quantity of gas formed.
In some cases of extreme dilatation of the stomach the
formation of hydrogen sulphide is so abundant as to
make it seem strange that the patients escape without
symptoms of intoxication. Their immunity in these
cases appears to depend on the fact that there is an enac-
tion of the gas when it accumulates in considerable
volume.

It has come to be well recognized that local inflam-
matory conditions of the bladder may be associated with

'"Ueber Schwefelwasserstoff im Ham," Berl. klin. Wochen-
schr., xxiv., p. 405, 1887.

INFECTIONS OF THE DIGESTIVE TRACT 233

a putrefaction of proteid material with liberation of
hydrogen sulphide. This condition is, of course, one
quite distinct from true hydrothionsemia, and it runs
its course without symptoms of general intoxication.

The toxic character of hydrogen sulphide for dogs is
plainly seen from a typical record of experiments in
which a saturated aqueous solution of hydrogen sulphide
was introduced into the rectum. Fifteen cubic centi-
meters of such a saturated solution were introduced
daily into the rectum of a dog weighing sixteen and one-
half pounds. No effects were observed. After one
week the volume of the enema was increased to thirty
cubic centimeters. After a daily administration of this
amount during two weeks, the volume was increased to
forty-five cubic centimeters. The animal immediately
passed into a state of collapse. It was found that by
injecting forty-five cubic centimeters very slowly the
enema was tolerated, although it gave rise to prostration.
After a period of five days, in which this quantity was
daily administered, the volume was increased to sixty
cubic centimeters, but with the result of bringing on
collapse which proved nearly fatal. The volume of the
enema was then reduced to thirty cubic centimeters.
After a few days it became necessary to discontinue
the injections of hydrogen sulphide, as the dog failed
to eat. The nutrition of the animal suffered in the
course of these infections. There appeared to be a
slight reduction in haemoglobin in the course of the
experiment. A pronounced ansemia was, however, not
induced.

234 INFECTIONS OF THE DIGESTIVE TRACT

Hydrogen Sulphide and its Relation to Enterogenic
Cyanosis. — In 1902 Stokvis 1 described a peculiar and
apparently new clinical state under the name of " auto-
toxic enterogenic cyanosis." His observation had refer-
ence to a man fifty-eight years of age who suffered from
a severe enteritis associated with pronounced cyanosis
of the skin and visible mucous membranes, together with
a slight swelling of the terminal phalanges. On making
a spectroscopic examination of the skin and mucous
membranes of his patient, Stokvis was easily able to
make out in addition to the two feeble oxyhsemoglobin
bands a small absorption band in the red, corresponding
to the absorption spectrum of methsemoglobin. It is
of course well known that the experimental methse-
moglobinuria which may be induced by a variety of
poisons, such as potassium chlorate, nitrobenzol, and
various derivatives of anilin, gives rise to a characteristic
cyanosis of the skin and mucous membranes. Stokvis
was therefore inclined to believe that the cyanotic
discoloration of the skin of his patient could be ascribed
to the methsemoglobin present in the blood. He assumed
that some poisonous substances had been formed in the
intestine which gave rise to a transformation of a portion
of the haemoglobin into methsemoglobin. Three similar
cases were soon after reported by Talma,2 who reached
the same conclusions as Stokvis ; namely, that the methse-

1 " Zur Casuistik der autotoxischen enterogenen Cyanosen
.(Methaemoglobinaemia (?) et enteritis parasitaria)," Festschr. f. v.
Leyden, i, p. 597, 1902.

2 Tijdschrift voor Geneesk., ii, p. 721, 1902.

INFECTIONS OF THE DIGESTIVE TRACT 235

moglobin observed in his cases was the result of intoxi-
cation through the intestinal tract, due to poisonous
substances formed there. In 1905 Hi j mans van der
Bergh l reported two cases of enterogenic cyanosis which
at first sight appeared to have the same pathological
basis as those reported by Stokvis and Talma. He found,
however, on careful investigation that the substance
contained in the blood of one of his patients could not
have been methaemoglobin.2 The absorption spectra
observed by him showed the closest similarity to the
spectrum of sulphhsemoglobin. Attempts were made to
determine the presence of hydrogen sulphide or rather
of sulphhsemoglobin in the blood, but these were not
wholly successful, although he was able by means of the
Caro-Fischer reagent (a solution of pure amidodimeth-
ylanilin with ferric chloride) to obtain a blue color
due to the formation of methylene blue and indicative
of the presence of hydrogen sulphide. Van der Bergh
concludes that his patient suffered from the passage of
hydrogen sulphide into the blood from the intestine.3

lf'Enterogene Cyanose," Deutsch. Archive /. Tdin. Med.,
Ixxxiii, p. 86, 1905.

2 This writer has lately reported one instance of enterogenic
cyanosis associated with methaemoglobinsemia, in which the blood
was shown to contain nitrites. This observation appears of
much importance for the doctrine of intestinal intoxication, as it
is well established that nitrites readily cause methaemoglobinsemia
under experimental conditions. A. A. Hymans van der Bergh
und A. Grutterink, " Enterogene Cyanose," Zweite Mittheilung,
Berl. klin. Wochenschr., No. 1, p. 7, 1906.

3 Several of these cases of sulphhsemoglobinsemia have been
associated with obstinate constipation, the relief of which has been
followed by rapid improvement in the blood state.

236 INFECTIONS OF THE DIGESTIVE TRACT

The clinical conditions present in one of Van der
Bergh's cases of enterogenic cyanosis deserve mention.
This case was a child nine years of age in which there were
marked digestive disturbances with diarrhoea. It had
been observed by the mother that the child passed very
little urine. This anuria was probably owing to the
constant diarrhoea. For two years the mother had ob-
served that her child readily became bluish on slight
excitement. This condition gradually became more
pronounced, and finally the child grew very cyanotic
even when at rest. The ends of the fingers were also
somewhat clubbed. When the child was three or four
years of age the abdomen became much distended and
this distension gradually increased. There was no evi-
dence whatever of any cardiac affection. The urine was
very much reduced in amount and contained no albumin
and no sugar, but is said to have contained a little indol.
The sediment contained many bacteria, leucocytes, and
calcium oxylate. The movements were very thin, had
a strongly alkaline reaction, and smelled of ammonia.
The movements were sometimes yellow, sometimes dark
brown or nearly black, and always became very dark on
standing in the air. No protozoa were found. The
author states that a fistula existed between the bladder
and rectum, and this was probably the cause of the indol
found in the urine. The child was not anaemic. An
endeavor was made to control the cystitis by means of
urotropin and the proteid food was greatly reduced in
order to diminish the hydrogen sulphide formed in the
intestine. This resulted in the disappearance of indol

INFECTIONS OF THE DIGESTIVE TRACT 237

from the urine. The development of hydrogen sulphide
in the urine, however, could not be wholly checked. After
a time, nevertheless, there was distinct improvement.
The size of the abdomen became reduced and, strikingly
enough, the blue color of the skin and mucous membranes
in the course of a few weeks became much less marked
and ultimately scarcely any cyanosis could be detected.
Nevertheless by spectroscopic examination the charac-
teristic sulphhsemoglobin band in the spectrum remained
visible although apparently somewhat weakened.

AROMATIC PRODUCTS OF PUTREFACTIVE DECOMPOSITION

Phenol and Cresol. — Among the aromatic products
of putrefaction are the phenols. In health the intestinal
contents contain only small quantities of these sub-
stances. In some pathological conditions attended by
excessive putrefaction in the intestine they are found in
the intestinal contents in amounts considerably above
the normal amount, which is always small. The quan-
tities found are, however, never large — never so large,
for example, as in the case of indol. What are the con-
ditions of bacterial decomposition which lead to this
production of phenol it is at present impossible to state
with confidence. Phenol is doubtless derived from the
breaking-down of tyrosin, and it is therefore in decom-
positions in which tyrosin-containing foodstuffs are in-
volved that we would expect to find the phenols in
greatest amount. I am not able to say whether the facts
of experience correspond to this view. Of the common
proteid foods, milk is one which yields considerable

238 INFECTIONS OF THE DIGESTIVE TRACT

quantities of tyrosin from its casein (the yield reaching
as high as four and one-half per cent.).1 Phenol is paired
in the body, as is well known, with sulphuric acid and is
excreted as phenol-potassium-sulphate.

Although the quantity of phenol found in the fa?ces
at any one time is small, the quantity excreted in twenty-
four hours in the urine may be fairly high. Cresols are
included with phenol in these estimates. The reason
for this considerable excretion is in part the fact that
phenols are produced in the organism in the course of the
metabolism of normal cells. The quantity thus formed
is, however, small, and the excess above this amount
which we find in certain putrefactive cases may there-
fore be properly attributed to the phenols produced by
putrefaction in the intestine. I have found the highest
values for phenols in the urine in cases of chronic intes-
tinal indigestion characterized by marasmus and disten-
sion of the abdomen in children.

It does not appear that the phenols can be regarded as
important toxic agents, since the maximal quantities
which we find in the urine in cases of disease may be
considerably exceeded during the prolonged therapeutic
administration of phenols without giving rise to any
symptoms whatever. Nevertheless, it is likely that the
continued absorption of moderate quantities of phenols

1 Among other substances which have been found to yield
relatively large proportions of tyrosin are fibrin, 3.82 per cent.;
zein from corn, 10.06 per cent.; glutenfibrin, 4.43 per cent.; con-
glutin from the seeds of lupins, 3.2 per cent.; horn, 4.58 per cent.;
histone from thymus, 6.31 per cent.; protamin of .the seehare, 8.4
per cent.

INFECTIONS OF THE DIGESTIVE TRACT 239

from the intestine over a long period of time may prove
injurious to the cells of the liver and to other cells con-
cerned with the pairing of phenol to phenol-sulphuric
acid. This deleterious effect is probably of greater
importance in persons in whom the cell protoplasm of
the liver has been somewhat damaged than in the case
of normal persons.

Skatol. — This substance is formed in very small
quantities from day to day in some normal persons, and
in persons suffering from excessive intestinal putrefac-
tion it may be formed in larger quantities than is nor-
mally the case. As compared with the quantity of indol
found in the intestinal tract, the quantity of skatol is
almost always small. I have occasionally met with in-
stances in which the f seces contain relatively large quan-
tities of skatol, eight to ten milligrams in one hundred
grams of fresh material. This has occurred only in
persons with marked signs of intestinal or nervous dis-
order. Very rarely I have observed much skatol in the
intestinal contents with only a mere trace of indol.1

Like indol, skatol is derived from tryptophan, but
what are the conditions that determine its formation
rather than the formation of indol we do not at present
know. I have given this subject considerable attention,
but without being able to determine the bacteria con-

1 The most extreme instance of this sort was observed in a
patient suffering from multiple neuritis and the psychosis charac-
teristic of alcoholic intoxication. As this patient had not taken
alcohol, the nervous condition was attributed to an obscure type
of intestinal intoxication. A study of the urine and faeces gave
some ground for this suspicion.

240 INFECTIONS OF THE DIGESTIVE TRACT

cerned in the formation of skatol rather than indol.
Suspicion rests on a slender, curved, Gram-positive organ-
ism about four microns in length which I have several
times found associated with the production of skatol,
but which has not yet been obtained in pure culture.1

I found that the administration of skatol to monkeys
by the mouth and by subcutaneous injections has been
followed by the appearance of a substance in the urine
giving the Ehrlich dimethylamidobenzaldehyde reaction
of the urine, and the administration of 0.1 gm. of skatol
to man has been followed by the appearance of a stronger
Ehrlich reaction than was previously present.2 I am
inclined therefore to ascribe the intensification of the
reaction in these cases to the administration of skatol.
The direct evidence, however, is lacking to show that
the reaction thus induced or accentuated is dependent
on the presence of the same substance in the urine that
gives rise ordinarily to the Ehrlich aldehyde reaction.

1 The ground for this suspicion is that flasks containing pu-
trefactive media inoculated with faecal bacteria have shown the
presence of this organism in very large numbers in a case where
skatol has been present, whereas flasks prepared with the same
media and the same bacteria plus magnesium carbonate have
failed to show this microorganism and have also failed to show
skatol. In these cases the difference between these flasks as re-
gards other microorganisms than the one mentioned have been
slight. The subject is difficult of investigation. One might sup-
pose that the faecal bacteria from a case in which the intestine
contains considerable skatol would form skatol when grown in
bouillon. I have never found this to be the case. Indol is formed,
not skatol.

2 " On a Relation between Skatol and the Dimethylamidobenz-
aldehyde Reaction of the Urine," Journ. BioL Chem., i, p. 251,
1906.

INFECTIONS OF THE DIGESTIVE TRACT 241

In most cases in which the faeces contain considerable
skatol the urine gives a strong reaction with dimethyl-
amidobenzaldehy de .

Skatol behaves in the organism much like indol as
respects its toxic properties, although it is somewhat
less poisonous. There is seldom reason to attribute to
skatol any definite pathological effects, as it is formed in
quantities too small to fall under suspicion. It is pos-
sible, however, that like phenol it may, under some con-
ditions, play an auxiliary part in association with other
substances in damaging living cells.1

Indol. — It has long been known that indol is a
product of putrefactive decomposition of proteids, and
Baumann showed many years ago that this substance is
formed in the large intestine in the course of putrefac-
tive processes occurring there. He showed, moreover,
that the absorption of indol from the intestine is followed
by the appearance of indican in the urine. Some ob-
servers have concluded that the indican of the urine
may in part depend on the liberation of indol from the
breaking down of body cells. I regard the evidence in
favor of this view as unsatisfactory and believe that the
indican of the urine depends exclusively on the resorp-
tion of indol from the intestinal tract excepting in those
cases in which pathological processes such as a putrid
abscess are associated with the formation of the base.

The observation has been repeatedly made by clini-

1 There is an excellent discussion of the fate of skatol to be found
in a recent paper of Porcher and Hervieux, entitled "Untersuch-
ungen uber das Skatol," Zdtschr. f. physiol. Chem., xlv, p. 487,
1905.

R

242 INFECTIONS OF THE DIGESTIVE TRACT

cians that persons in whom a very strong indican reac-
tion can be obtained in the urine during a long period
of time invariably suffer from nervous or dyspeptic dis-
orders, and many careful physicians have believed that
there is some causal connection between the absorption
of indol from the intestine and the development of func-
tional nervous or nutritional derangements. I believe
that in recent years the grounds for this conviction have
been considerably strengthened. Nevertheless, it can
hardly be said that the relation between the absorption
of indol and the symptoms of intoxication in man has
ever been placed upon a firm scientific basis. I believe
that I have myself underestimated the importance of
indol as a toxic agent in man, and desire to present here
evidence which has lately come to light in regard to the
influence of indol on the animal organism. For this rea-
son I shall discuss the subject of indol poisoning some-
what fully.

Indol is not a product of the tryptic digestion of proteids,
and probably cannot be formed in the course of physio-
logical processes without the intervention of organized
ferments such as bacteria. In the human intestine the
presence of indol is dependent on the action of living
bacteria, although it is likely that the action of the diges-
tive juices may prepare the way for the attack of bacteria
on proteid material. The indol produced in the intes-
tine is, like skatol, dependent on the production of a
more complex substance known as tryptophan. A very
careful study of the chemical constitution of tryptophan
was lately made by Gowland Hopkins of Cambridge

INFECTIONS OF THE DIGESTIVE TRACT 243

University.1 He ascribed to it the constitution of
skatol-amido-acetic acid, but it has been shown by other
workers that there is evidence for considering trypto-
phan as an isomer of skatol-amido-acetic acid; namely,
indol-amido-propionic acid.

It has been known since the days of Claude Bernard
that tryptophan arises at an early period in the putre-
faction of proteids. It was further observed that while
the tryptophan color reaction is almost always observed
in the early stage of putrefaction, it later disappears.
The cause of this disappearance is the further cleaving
action of bacteria or unorganized ferments. Hopkins
was able to show that the action of bacteria upon trypto-
phan may lead to the formation of indol, skatol, indol-
acetic acid, and indol-propionic acid. An endeavor was
made to determine the influence of individual types of
bacteria upon the cleavage of tryptophan, and it was
found that B. coli is capable of giving rise to considerable
yields of indol at the same time that it produces indol-
acetic acid. Observations with cultures of symptomatic
anthrax representing a typical spore-bearing anaerobe,
showed that this organism is able to make indol-propionic
acid. At present, however, there is no evidence that
more than a small proportion of tryptophan in a medium
is converted into indol through the action of putrefactive
anaerobes. I studied many years ago the influence

1 Hopkins, F. Gowland, and Sydney W. Cole, "Contributions
to the Chemistry of Proteids," Ft. II, "The Constitution of Tryp-
tophane and the Action of Bacteria upon It," Journ. of Physiol.,
xxix, p. 451, 1903.

244 INFECTIONS OF THE DIGESTIVE TRACT

of introducing large numbers of B. coli communis,
proteus vulgaris of Hauser, and B. acidolactici into the
intestinal tract of dogs, with a view to determining the
effect upon indol production. It was found that the in-
jection of pure cultures of colon bacilli into the jeju-
num of dogs was followed by an increase in the indican
reaction of the urine and an increase in the output of the
ethereal sulphates. In order to exclude the introduction
of indol into the intestine in the course of these experi-
ments large numbers of colon bacilli were grown on agar-
plates and washed in salt solution. The introduction of
the living bacteria was followed by an increase in the
indican of the urine, whereas the introduction of sterile
cultures prepared in the same way showed at most only
a slight increase in the indican excretion. The inference
was drawn that the small amount of indol introduced in
feeding experiments made with colon bacilli does not
account for the marked increase in indican which was
noted in these experiments and that this increase is to
be attributed to the action of the colon bacilli themselves
growing upon a suitable medium. Similar experiments
were made with pure cultures of proteus vulgaris with neg-
ative or slight effects as regards the excretion of indican.
The lactic acid bacilli employed in similar experiments
in which cultures were injected directly into the small
intestine showed a tendency to cause a reduction in the
excretion of indican and of the ethereal sulphates.1

1 Herter, " On Certain Relations between Bacterial Activity in
the Intestine and the Indican of the Urine," Brit. Med. Jour.,
II. p., 1847, 1897.

INFECTIONS OF THE DIGESTIVE TRACT 245

There are many persons from whose intestines it is
impossible to recover indol at all or in more than a mere
trace. This is particularly the case with young children
and young adults. There are, however, some older
persons who, although suffering from disorders of diges-
tion, do not form indol in the digestive tract. On the
other hand, the production of considerable quantities of
indol in the large intestine is a feature of many instances
of intestinal putrefaction, and in some cases the quantity
formed is large. One may find as much as fifty to sixty
milligrams of indol by the naphthaquinone method in
one hundred grams of the fresh stool. This probably
approaches the maximal amount present at any one
time. It is of course clear that such indol production
is distinctly pathological in occurrence. Its production
in smaller amounts permitting the recovery of five
milligrams in one hunndred grams of moist faces is
no uncommon occurrence, and such indol production
may or may not be associated with the development
of intestinal or nervous or other disturbances. The
significance of the indol produced during putrefaction in
the intestine depends upon its absorption into the organ-
ism through the intestinal wall. That this occurs in con-
siderable amounts is shown by the appearance of large
quantities of indican in the urine of persons in whom
the intestine contains large amounts of indol. A close
relationship between the quantity of indican in the urine
and of indol formation in the intestine is not always
demonstrable. The faeces may contain little indol while
the urine holds much indican. Conversely, the fseces

246 INFECTIONS OF THE DIGESTIVE TRACT

may contain a considerable quantity of indol; and owing
to imperfect absorption the urine may contain only
moderate quantities of indican. Upon the whole, how-
ever, if one considers the findings from day to day over
a considerable period of time, it holds true that there is a
rough relation between the indol formed in the intestine
and the quantity of indican excreted. In conditions of
health the absorption of moderate quantities of indol
from the intestine is followed by the rapid oxidation of
indol to indoxyl or some indoxyl compound, and this
oxidation is associated with or followed by a synthesis
with sulphuric acid, occurring mainly in the liver and
partly in the muscles, and which results in the produc-
tion of the indoxyl-potassium-sulphate of the urine. It is
the indoxyl-potassium-sulphate (or indican) of the urine
which on further oxidation yields indigo. Something is
now known in regard to the fate of indol in the organism.
While it is true that in general the aromatic compounds
of putrefaction are resistant to oxidation, it is probable
that indol when introduced in moderate quantities into
the organisms of carnivora and omnivora suffers a
break-down both in the pyrrol ring and the benzene
ring. That is to say, a portion of the indol is burned
completely in the organism. This view is based partly
on what has been found to be the case for phenols and
especially for cresol,1 and partly as direct observations
on the facts of ingested indol.

1 Blumenthal, " Biochemische Untersuchungen iiber Vergiftung
und Entgiftung bei der Lysolvergiftung," Biochem. Zeitschr., i,
p. 135, 1906. The author found that only from twenty to
twenty-five per cent, of the cresol introduced into the animal

INFECTIONS OF THE DIGESTIVE TRACT 247

It was shown by Herter and Wakeman * that the living
cells of the body, especially the hepatic and renal cells
and the epithelial cells of the intestinal tract, have the
power of absorbing considerable quantities of indol as
well as of phenol and of tying them loosely in such a way
that these bodies cannot be recovered by distillation.
Owing to this property of the cells, by which they hold
these aromatic bodies while subjecting them to oxidation
and pairing, the nervous system is screened from their
action. The importance of this screening action is
considerable, for the presence of indol or phenol in very
slight concentration in the blood of the carotid artery
suffices to induce violent nervous excitation followed
by a depression of nervous function. It is very notice-
able that the capacity of different animals to remove
indol from the circulation after intravenous injections
differs greatly in individuals of the same species. What
is particularly striking is the fact that in those animals
which do not possess livers capable of promptly removing
the greater part of the indol, the nervous system falls a
prey to the action of the poison. Conversely, it is also
true that in those animals in which nervous symptoms are
very pronounced after such injections of indol, it is found
that the blood and brain hold considerable indol, whereas
the liver may be shown to have fallen far below its nor-
mal capacity in the removal of the poison from the blood.

organism could be found again in the urine when the doses are
moderate in amount.

1 Herter and Wakeman, "The Action of Hepatic, Renal, and
Other Cells on Phenol and Indol," Journ. of Exper. Med., iv,
p. 307, 1899.

248 INFECTIONS OF THE DIGESTIVE TRACT

It may be regarded as settled that the liver, muscles,
intestinal epithelium (and other cells) normally exert a
protective action to the nervous system in screening it
from the effects of an injurious percentage of indol in
the blood by the ability of these structures to quickly
bind any indol which comes to them. It may, moreover,
be regarded as established that the same dose of indol
administered to two human beings of about equal
weight may regularly give rise to more pronounced ner-
vous manifestations in one than in the other. While
these inequalities may be due partly to differences in the
rapidity of absorption, no striking differences due to this
factor were noticeable in the excretion of indican in the
urine, and it appears more probable that the differences
in the observed toxic effects were dependent on inequali-
ties in different persons in respect to their ability to
oxidize indol and to pair it with sulphuric acid. The
probability that individual differences in the oxidizing
capacity of the tissues of different persons might play a
part in determining the toxic effects of indol made it
desirable to get experimental evidence as to the in-
fluence of the imperfect oxidizing action of the cells on
the fate of indol in the body. Such experiments have
been planned and made by Dr. A. N. Richards and Dr.
J. Rowland, and their results are of such interest for the
general pathology of intoxications as well as for the
question of indol poisoning, that I shall speak of them at
some length.1

1 The details of these observations have not yet been published,
and I feel under obligations to Dr. Richards and Dr. Rowland for
permitting me to use their notes.

INFECTIONS OF THE DIGESTIVE TRACT 249

To reduce the oxidizing processes in the cells, animals
were poisoned with hydrocyanic acid, a substance which
possesses a high degree of power to depress the ability
of animal cells to take up oxygen from the blood. The
effect of dyspnoea from tracheotomized animals was
also investigated to some extent and similar observa-
tions were made upon the influence of chloroform given
to the point of narcosis upon the fate of indol.

In general it may be said that rats, mice, guinea-pigs,
and dogs subjected to subcutaneous injections of potas-
sium cyanide too small in themselves to cause marked
symptoms, were later subjected to subcutaneous or peri-
toneal injections of indol or phenol in watery solution
or in oil. The symptoms observed in such cases were
compared with those obtained from the injection of
indol or phenol without any preceding treatment with
potassium cyanide. The result almost regularly ob-
served was that the convulsive twitching which is char-
acteristic of the action of the phenol and of indol was of
greater intensity and longer duration in the animals
subjected to potassium cyanide than in the case of the
control animals. For example, into a guinea-pig weigh-
ing 370 gm. was injected 0.0003 gm. of indol per gram
of body weight. The characteristic twitching came on
in eight minutes and lasted ninety-seven minutes. It
was mild in character. Into another guinea-pig, weigh-
ing 380 gm., 0.005 mg. of potassium cyanide per gram
of body weight was injected together with 0.0003 gm.
indol per gram of body weight. The twitching began
in two minutes, was much more violent than in the

250 INFECTIONS OF THE DIGESTIVE TRACT

control animal, and lasted for five hours. Experiments
were made which showed that the lengthening effect is
not due to a decreased rate of absorption dependent on
the depression of the circulation by potassium cyanide.
Experiments were also made to determine whether
after subcutaneous injections of indol any uncombined
indol may be excreted by the gastro-intestinal tract, and
if so, under what conditions such excretion occurs. For
example, in a dog weighing eleven kilos the small intestine
in one foot of its extent (in the region just above the
ileocsecal valve) was severed by Dr. Maury from the
main gut and the ends sewed into the abdominal wall.
The cut ends of the main gut were joined by end-to-end
anastomoses. Complete recovery occurred and the
health of the dog was apparently perfect. The mucous
membrane at the fistulous openings was normal.
One month later an experiment was made in which
0.5 gm. of indol dissolved in ten cubic centimeters of oil
was injected subcutaneously. No symptoms developed.
The loop of gut was washed out at intervals of one-half
hour during the following nine hours, and no indol could
be detected in the distillate by means of the naphthaqui-
none reaction. The urine obtained by catheter forty-five
minutes after injection of indol gave an extremely intense
indican reaction. Repeated examinations showed little
falling off in the intensity of the reaction until the lapse
of twenty-two hours. After twenty-eight hours the
urine showed a normal indican reaction. After an in-
terval of two days, in which the animal was apparently
normal, he received successive injections of a two per

INFECTIONS OF THE DIGESTIVE TRACT 251

cent, solution of the potassium cyanide. The doses
given sufficed to cause vomiting and some muscular
weakness, but no convulsions. The injections were
given during a period of two and one-half hours. At the
end of this 0.5 gm. of indol was injected subcutaneously
as in the first experiment. No further vomiting was
noted, but twitching of the muscles of the face and
legs was more marked than before. One and one-half
hours after the indol injections the animal was lively
and seemed to be recovering. Soon after he received
more cyanide, in consequence of which there developed
a few mild convulsions. There was much prostration
and dyspnoea and some fall in temperature. The odor
of indol was not detected at the fistulous opening.
Washings made from the gut were distilled and gave a
slight reaction for indol. It was found that the indican
reaction in the urine was delayed in its appearance as
compared with the previous experiment and that the
quantity of indican as judged by the intensity of the
color was considerably less developed in the Obermeyer
reaction. The quantity of indol excreted was thus
distinctly less than during the previous experiment. The
most striking feature of this experiment is the fact that
on the day following the administration of indol and
potassium cyanide the animal remained all day in a
condition of stupor interrupted by periods of frantic
and poorly coordinated movements in which he recog-
nized none of the attendants and did not respond to
calls. These seizures could be brought about by local
stimuli and resembled the behavior seen in dogs etherized

252 INFECTIONS OF THE DIGESTIVE TRACT

or chloroformed during the period of lessening conscious-
ness which precedes that of complete anaesthesia. The
animal refused food and water and showed signs of in-
creased peristalsis. During the second day prostration
and stupor were less marked, but uncoordinated move-
ments were more frequent and associated with a state of
delirium at times very violent. During the third day
the condition was the same. During the fourth day the
condition was the same, but the animal was weaker.
Warm diluted milk given by a tube was instantly vomited.
On the sixth day egg- water was retained. On the
seventh day death occurred. The brain was found to
be soft and congested. The cortical nerve cells were
found to be oedematous and showed excessive chroma-
tolysis. The mucous membranes of the loop and of the
duodenum and jejunum were considerably congested.
The ileum was slightly congested. The total amount
of potassium cyanide given to this animal was 0.061 gm.
in six doses in three hours during the first period, that
is before giving the indol ; after the indol, that is in the
second period, 0.085 gm. was given in four doses in three-
quarters of an hour.

The leading effects of the poisoning in this case were,
first, a prolongation and intensification of the direct
indol effects upon the nervous system, for example, the
convulsive twitching; secondly, a delayed transforma-
tion of indol and a delayed excretion in the form of
indican; third, gastro-intestinal symptoms; fourth,
blindness ; and lastly, various mental symptoms. Other
similar experiments were performed which clearly showed

INFECTIONS OF THE DIGESTIVE TRACT 253

that when indol poisoning is preceded by the adminis-
tration of physiological quantities of potassium cyanide
the indol poisoning takes a wholly different course from
that observed when indol alone is injected. As it is
known that potassium cyanide depresses the ability of
the cells to take up oxygen, it is reasonable to attribute
to this depression in the oxidizing capacity of the body
cells the failure of the organism normally to oxidize the
injected indol. It is not unreasonable to suppose that in
chronic disease in which the cells most concerned in
oxidation, such as the liver cells, have been structurally
damaged, that the ability to transform indol is less than
in a state of health. Such alterations are common in
human beings as a result of infections, alcoholism, and
intoxications, etc., and the experiments just quoted
enable us to understand how under such pathological
conditions the absorption of a moderate quantity of
indol from the intestine may be more harmful to the
central nervous system than is the case in persons whose
cells have suffered little deterioration as the result of
disease.

As to the effects of absorbed indol upon the organism
in disease it is necessary to speak with caution, since there
is no evidence that indol is the only toxic substance
absorbed in those cases where it enters the organism
from the gut. Some light is thrown on the question by
experiments made through administering indol to normal
men by the mouth. In one of the cases, a robust young
man whose urine had been free from indican, felt no effect
from large doses given until after several days' adminis-

254 INFECTIONS OF THE DIGESTIVE TRACT

tration, when the influence on the nervous system be-
came distinct and led to irritability, headache, flight of
ideas, etc. It should be noted that the quantity of indol
administered in this case was probably in excess of any
amount that would be absorbed from the intestine, even
in the most pronounced pathological condition. It
would not, however, be safe to infer from this that smaller
doses would have been harmless ; for while such smaller
doses might have produced little or no effect if given
during a few days to a normal person, it is not unlikely
that its long-continued administration would have led
to symptoms. But here again it must be emphasized
that the significance of indol absorption for the organism
must depend largely on the ability of the organism to
quickly oxidize and pair the indol to indoxyl-potas-
sium-sulphate.

The idea that the circulation of free indol in the blood
may act in a depressing manner upon muscular struc-
tures is suggested by the rapid muscle fatigue which
comes on in some persons who have suffered for a long
period of time from a high grade of indicanuria. It is
also suggested by the observations which I have made
that moderate doses of indol by the stomach may be
followed in normal persons by a sense of muscle fatigue,
which has worn off rapidly on prolonged exercise. Pro-
fessor Fred. S. Lee of Columbia University was so kind
as to undertake experiments upon the muscles of cold-
blooded animals and mammals with a view to determining
the influence of indol upon the onset of muscle fatigue.
His results are expressed in the f ollowing notes : —

INFECTIONS OF THE DIGESTIVE TRACT 255

"Experiments have been performed on the muscles of frogs
and cats. The method has been to irrigate corresponding muscles
for a given length of time — the one with physiological-salt solu-
tion, the other with physiological-salt solution plus a small quan-
tity of the drug, and then to stimulate the muscles and record the
contraction and the amount of work done before exhaustion sets in.

"Indol in 0.05 per cent, solution — that is, one part to 2000

— causes an early fatigue of the muscles and a diminution in the
amount of work of which the muscle is capable. In a specific
experiment the indolized muscles performed 24,320 g. mm. of
work before it became exhausted, while at the same time the cor-
responding normal muscle performed 44,000 g. mm. of work, and
was still capable of doing more. The work actually accomplished
by the two muscles was 100 (normal) : 55 (indol) .

"Phenol in 0.05 per cent, solution — that is one part to 2000

— usually puts the muscle into a better working condition than
that of the normal muscle. In a specific experiment the phenolized
muscle performed 7550 g. mm. of work before it became ex-
hausted, while at the same time the corresponding normal muscle
performed only 4400 g. mm. of work. The work actually ac-
complished by the two muscles was 100 (normal) : 171 (phenol).
This favorable action of phenol is followed, however, by the re-
verse effect, the phenolized muscle usually becoming exhausted
before the normal muscle. In 0.1 per cent, solution — that is,
one part to 1000 — the effect of phenol is like that of indol ; in
other words, the phenolized muscle performs less work and be-
comes fatigued more readily than the normal one."

Still more important than the foregoing experiments
is the following observation by Dr. Lee : —

"The two muscles of a cat were irrigated for half an hour, the
one with whipped blood, the other with whipped blood to which
indol had been added in 0.004 per cent, solution; in other words,
one part of indol to 25,000. The normal muscle then performed
34,400 g. mm. of work, while the indolized muscle performed only
12,880 g. mm. — the ratio of normal work to indolized work is
100:37."

This shows that indol in very minute quantity acting
for a long time on the muscle diminishes its working

256 INFECTIONS OF THE DIGESTIVE TRACT

power ; an observation of much significance, since it
gives an experimental basis to the suspicion that indol is
capable of acting as a depressant to the muscular
mechanism. The most pronounced clinical indications
of such action are perhaps those derived from some
cases of myasthenia gravis. Dr. Tuttle of the Pres-
byterian Hospital tells me that he has lately had under
observation a patient whose clinical history corresponds
to that of myasthenia gravis and in whom the only
objective sign of a pathological condition was an intense
and persistent indicanuria which was not much modi-
fied by any treatment which was instituted. It is clear
that cases of this type should be most carefully scruti-
nized from the standpoint of the possibility that the
depression of muscle function is dependent on the action
of aromatic products upon the muscles.1

The rapid onset of fatigue observed experimentally
in muscles that have been irrigated with indol and the
curves which such muscles show are presumably not
specific for indol. They derive their significance for the
human subject from the fact that indol is the only
aromatic product which is known to be absorbed in
quantities sufficient to render it probable that the indol
has a toxic influence upon the neuro-muscular system.
The absorption of skatol may act in an accessory manner
to heighten the effect induced by indol poisoning, but the

1 1 am told that there are cases of myasthenia gravis in which
the indican of the urine is not excessive. This fact is not neces-
sarily a valid argument against the idea that indol acts as a muscle
poison in myasthenia gravis, for it is true that the given derange-
ments of function may arise from more than one poison.

INFECTIONS OF THE DIGESTIVE TRACT 257

absorption of considerable quantities of skatol must,
I think, be an exceptional phenomenon in man. In
some cases of excessive intestinal putrefaction in child-
hood associated with a retardation in growth and ab-
dominal distension there is clearly a poisoning of the
muscular system. These children show signs of fatigue
very rapidly and in some cases, where the condition has
come on in early life, they are slow in learning to walk.
In such cases there is not only a large amount of indican
in the urine but a considerable excretion of phenol. It is
quite likely that the phenol in these cases plays a part
in bringing about the depression of muscular action.
Perhaps in some instances it is as much a factor in
inducing fatigue as is indol.1

INDICANURIA

The term " indicanuria" is used to designate the pres-
ence of an abundance of indican in the urine as demon-
strated by the presence of strong reactions (with the form-
ation of indigo) on the use of Obermeyer's test or similar
tests. It has already been mentioned that in normal child-
hood little or no indican appears in the urine. There are
also many adults who seldom show the presence of indican
in the urine. There are also many adults who show a

1 The primary beneficial effect of phenol on muscle action
observed by Dr. Lee does not speak against this view, as this
effect is one that soon gives way to fatigue.

The very interesting observation has been made by Dr. Lee
that a temporary beneficial action is exerted on muscle by all the
common fatigue products studied by him.

258 INFECTIONS OF THE DIGESTIVE TRACT

moderately strong reaction for indican in the urine
during a long period of time (that is, most of the time
during many years) and still retain good health and do
not suffer obtrusively from digestive disorders. This
is true of many students who live sedentary lives. On
the other hand, there are persons with digestive disorders
who show little or no indican in the urine. Finally,
there are not very rare instances in which during a short
or a considerable period the urine contains very large
quantities of indican so that the reaction with Ober-
meyer's reagent is intense. I think it safe to say that
such persons are seldom free from clinical evidences of
intestinal disorder and that often there are also in these
cases some indications of intoxication.

Almost every physician who has observed considerable
numbers of patients with reference to the presence or
absence of indican in the urine, has been puzzled to
interpret his results. For he has no sooner begun to
attach importance to the presence of a strong indican
reaction in certain cases than his faith in this as a sign of
disease has been shaken by the fact that he has found a
strong indican reaction in the urine of some person who
is apparently well. I think it would greatly aid us in the
interpretation of the meaning of indicanuria if we took
into consideration the important factor of the ability
of the organism to defend itself against the toxic action
of indol by quickly disposing of it. It is of course
difficult to fashion standards for the judgment of this
factor of defense. But there are certain obvious con-
siderations which bear on the judgment of the defensive

INFECTIONS OF THE DIGESTIVE TRACT 259

powers of the organism for oxidizing and pairing indol.
It is certain that the oxidizing powers of the cells in
early life are greater than in later life and that on this
account organisms are likely to suffer more from the
absorption of equal amounts of indol between the fiftieth
and seventieth years of life than between the fifteenth
and thirtieth. This is doubtless true quite aside from the
degenerations that occur in organs which have been sub-
jected to the action of disease. In cases where the ele-
ment of disease has entered to injure the liver, kidney,
intestinal epithelium, muscles, and other structures,
the organism naturally becomes more sensitive to the
action of indol than before its powers of oxidization and
synthesis have become impaired. There are very many
conditions in which the liver undergoes cirrhotic or
fatty changes or in which both of these processes are
marked or in which there are well-defined parenchyma-
tous changes. It is certain that such persons will be
especially sensitive to the action of all kinds of poisons,
including that of indol. We have therefore in judging
of the significance of any case of pronounced indican-
uria to consider whether it occurs in a young person or in
an older one ; in a healthy individual or in one whose cells
have been damaged by disease. In youth it is com-
paratively easy to find a diet and mode of living under
which the indican can be made to disappear largely from
the urine. In aged persons the task is usually more
difficult, especially if the indicanuria has been of long
duration.
Aside from cases of indol absorption from the presence

260 INFECTIONS OF THE DIGESTIVE TRACT

of a dead foetus or an abscess, I believe that indicanuria
always points to absorption of indol from the intestinal
tract. There is no convincing evidence that the autoly-
sis of organs during life yields indol, and I think that
such a source may be regarded as a negligible factor in
dealing with cases of indicanuria. Admitting, however,
that indicanuria is of intestinal origin, it remains to be
explained why it is present in some cases of intestinal
disturbance and not in others. A wholly satisfactory
answer to this question is not at the present time possible.
Some facts which bear on the problem maybe stated here.
It has been claimed by Pizenti and others that the pres-
ence of pancreatic juice is essential to the production of
indicanuria. This claim is based on the fact that indol is
a product of proteolytic cleavage. The proteid food en-
tering the intestine is supposed to be attacked vigorously
by the abundant tryptic ferment of the pancreatic
juice, and this decomposition is presumed to reach so
advanced a stage as to lead to the indol production from
which the indicanuria arises. While it is true that
indol is a product of proteid cleavage, I think it ex-
tremely doubtful whether there is an opportunity for the
process of cleavage to go so far as to yield indol, unless
the quantity of proteid food is largely excessive and thus
delays absorption. Ordinarily absorption occurs in the
peptone and amino-acid stage of proteolytic digestion.
It is not conceivable that in the absence of putrefactive
bacteria from the small intestine a rapidly proteolyzed
meal should lead to the formation of indol in the intestine,
since indol does not arise from tryptic digestion. It has

INFECTIONS OF THE DIGESTIVE TRACT 261

been held that persons with pancreatic disease who are
unable to secrete pancreatic ferments into the intestine
do not develop indicanuria. This is surely a mistaken
view. I have observed indican reactions to persist in
dogs from which the pancreas had been removed, and I
have met with some of the most marked instances of
indicanuria in persons in whom autopsy or operation
showed both the pancreatic and the biliary duct to be
occluded.

Intestinal indicanuria must be regarded as an evidence
of intestinal putrefaction. It is commonly easy to pro-
duce it experimentally in dogs by feeding a greatly
excessive quantity of meat. Not only may an habitual
indicanuria, such as is common among dogs, be increased
by excessive feeding of meat, but the condition may be
induced by excessive feeding in cases where the urine
was previously quite free from indican. The explanation
of this fact appears to me to depend on the presence of
putrefactive bacteria in the ileum and large intestine of
the dog. If we examine any portion of the large intestine
or the lower ileum, we shall usually find there moderate
or considerable numbers of anaerobic, spore-forming,
butyric-acid-producing bacteria as well as colon bacilli.
These anaerobes have the power of attacking native
proteids which, under suitable conditions, they may
energetically hydrolyze. Some anaerobes such as B.
aerogenes capsulatus are usually unable to carry the de-
composition as far as indol production. The colon bacilli
normally present in the intestine, however, are able to
make indol from peptones. It is only necessary, there-

262 INFECTIONS OF THE DIGESTIVE TRACT

fore, to feed the animal so excessive a quantity of meat
that small masses of muscle fiber enter the lower ileum
or the large intestine. Here they will be attacked by
anaerobes and colon bacilli with a production of indol and
subsequent development of indicanuria. Naturally any
condition favoring stagnation in the small or large
intestine will help to bring about this condition. It is
possible that moderate quantities of hydrolyzed proteid
reach the lower part of the ileum owing to impaired
absorption. In this region both in man and in dogs the
colon bacilli are numerous. Under these circumstances
it is not necessary for anaerobes to take part in the attack
on the already hydrolyzed proteid, since, as repeatedly
stated, the colon bacillus itself suffices to do this. The
effect of introducing partially hydrolyzed proteids into
the large intestine was very plainly shown in the following
experiment. A healthy dog which had been fed on meat
in moderate quantities gave only a slight reaction for
indican. Fifty cubic centimeters of a concentrated egg
and meat mixture were subjected to short partial tryptic
digestion (without the development of indol). This
material was sterilized in the autoclave and then injected
into the transverse and ascending colon and confined
there by ligature. Two hours later the indican reaction in
the urine was markedly increased and at the end of ten
hours had become intense, this change doubtless having
been due to the bacterial attack to which the material
was subjected in the large intestine.

There are some persons in whom the imperfect action
of a cathartic leads to the development of headache,

INFECTIONS OF THE DIGESTIVE TRACT 263

flatulence, and an increase of the indican in the urine.
The explanation of this is to be sought in the fact that
the cathartic causes the passage of native proteids and
possibly of peptones from the small into the large intes-
tine, where it is possible to attack the former if anaerobes
be present. Presumably in persons harboring few an-
aerobic bacteria this effect would be slight, but in per-
sons in whom organisms of the class of B. putrificus or
B. aerogenes capsulatus are abundant one would expect
the appearance of considerable indican in the urine.

The influence of constipation upon the development of
indicanuria is very different in different persons. Most
children and many adults may go without a movement
for several days and still fail to develop indicanuria if
this has previously been absent. There are other per-
sons in whom constipation is followed by a marked
increase in the output of indican. A wholly satisfactory
explanation of this difference in the influence of constipa-
tion upon indol formation in the intestine is not at pres-
ent possible. Sufficient bacteriological data are lacking.
A probable explanation is the following. The formation
of indol in the intestinal tract depends on the decomposi-
tion of proteids through the activity of microorganisms.
If the colon bacilli have for any reason ascended in large
numbers above their usual upper limit, namely the
ileum, into the jejunum or duodenum, they will act
there upon peptones formed in the course of the normal
digestion of proteids. From such peptones they will
make indol. Under these circumstances the aid of putre-
factive anaerobes is not necessary for the production of

264 INFECTIONS OF THE DIGESTIVE TRACT

indol. If, however, there is no ascent of the colon bacilli
to the region in which these organisms would find pep-
tones to attack, the abnormal production of indol is not
excluded, for, as already stated, the action of the anaerobes
in the ileum and large intestine permits the hydrolyzing
of proteids which then fall subject to the colon bacilli
with a resulting production of indol. Both these methods
of indol production may occur in the same individual,
namely its production in the small intestine through the
action of colon bacilli on peptones, and, secondly, the
formation of indol in the ileum and large intestine through
the combined action of putrefactive anaerobes and colon
bacilli acting upon such native proteids as may have
escaped digestion at higher levels. The occurrence of
constipation may be considered as favoring both these
methods of indol production, but only under abnormal
conditions of digestion. The stagnation of food rem-
nants and bacteria in the large intestine would cause no
indol production according to the hypothesis here set
forth unless masses of undigested proteid or hydrolyzed
proteid material had entered the region of the large
intestine. The stagnation of food in the small intestine
would not lead to the production of indol there except in
the presence of colon bacilli or other microorganisms
capable of making indol from peptones or albumoses.
Of the actual ascent of such bacteria from the upper level
of their normal habitat into the higher levels of the small
intestine we know as yet very little that is definite.
The invasion of the gall-bladder by colon bacilli under
conditions which make it improbable that there has been

INFECTIONS OF THE DIGESTIVE TRACT 265

any passage of colon bacilli from the intestine into the
blood (and thus through the liver into the gall-bladder)
makes it likely that there is in these cases an ascent of
the colon bacilli to the level of the common bile duct.
The colon bacilli travel so rapidly after death that it
is only through very early autopsies that one could ob-
tain evidence of the presence of these organisms at high
levels of the intestinal tract. Reliable data relating to
this point are at present lacking. There is, however,
a probability that in altered states of secretion in the
small intestine incidental to subacute or chronic enteritis
the colon bacilli do ascend above the level of their normal
habitat and thus become factors in the production of
indol and hence of indicanuria.

That the conditions which lead to pathological indi-
canuria are not always the same is rendered probable
by the marked differences in the amenability of this
condition to treatment which have been observed by all
clinicians who have attempted to modify this condition
in a variety of patients. There are instances in which
a reduction in proteid diet is in itself sufficient to greatly
diminish the quantity of indican excreted. Reduction
in the quantity of meat which is daily taken by a patient
is often efficacious in effecting this end. The explanation
of the improvement following a diminished use of meat
perhaps lies in the fact that on the more limited diet
smaller quantities of undigested meat pass into the region
of the associated anaerobes and colon bacilli. A change
from a diet containing much meat to one in which milk
is substituted for meat sometimes leads to a prompt

266 INFECTIONS OF THE DIGESTIVE TRACT

reduction in the output of indican. The improvement
is probably due to the more ready digestion and absorp-
tion of the milk. But there are instances in which the
substitution of milk for meat does not materially influence
the output of indican as judged by the Obermeyer
reaction. The invasion of the small intestine by colon
bacilli (and perhaps putrefactive anaerobes) would
satisfactorily explain this phenomenon, since in such a
case even the more rapid absorption incidental to a
milk as compared with a meat diet would not suffice to
elude the action of the colon bacilli upon peptones formed
in the course of digestion. There are also cases in which
no dietetic measures have a marked influence in reducing
the excretion of indican. That is to say, the indicanuria
remains well marked whether proteid food be employed
in the form of meat or milk or eggs or cereals. Dr. George
A. Tuttle of the Presbyterian Hospital has made the
observation that there are cases of indicanuria which
though persisting on a milk diet are susceptible of
striking improvement or complete disappearance through
the agency of potassium iodide. The cases of this type
have always been cases of marked arterial sclerosis
and some of them have had a history of syphilitic
infection. The explanation of this noteworthy fact
is obscure. The possibility occurs to one that the
potassium iodide influences favorably the circulation in
the intestine and this aids rapid absorption, but it is
also possible that the iodide may hi some way increase
the oxidizing activity of the cells mainly concerned with
the composition of indol in the body, notably the

INFECTIONS OF THE DIGESTIVE TRACT 267

liver cells.1 The increased oxidation of indol would
lead to a diminished excretion of indican and perhaps
to its complete disappearance. Dr. Tuttle tells me also
that he has had patients in whom the administration of
iron in the form of Basham's mixture has soon been
followed by a marked lessening of the indicanuria. As
this has occurred in cases where the diet has been care-
fully studied and dietetic measures have failed to give
the desired relief, Dr. Tuttle has attributed the thera-
peutic action to the iron.

It is worth noting that wherever there is a persistent
indicanuria which has not yielded to a change in the
character of the proteid food nor to a diminution of it
nor to a restriction of the carbohydrates, one may
rationally employ gelatin as a nitrogenous substance
capable of replacing in some degree the proteids of the
diet. Gelatin does not contain a tryptophan nucleus
and hence cannot yield indol. It has been shown by
physiologists that the metabolic needs of the organism
for nitrogen may be in part met by the use of gelatin,
although as is well known gelatin alone is incapable
of maintaining the protoplasm of cells where proteids
are wholly withdrawn from the dietary. I consider it
entirely safe to replace a portion of the nitrogen of the
food by gelatin (in the form of jellies) and believe I
have observed a falling off in the indican of the urine in
consequence of this substitution. Clinically, experience

1 The presence of iodides in a urine containing indican inter-
feres with the Obermeyer reaction, but this source of error appears
to have been eliminated in the procedure followed by Dr. Tuttle.

268 INFECTIONS OF THE DIGESTIVE TRACT

shows plainly that in the treatment of indicanuria one
has to consider not only the proteid food which is given,
but also the carbohydrates. With many persons the
excessive use of carbohydrates is habitual and the sub-
stitution of small quantities of quickly digested carbo-
hydrates like rice for large quantities of bread or sugars
may make a considerable difference in the amount of
indican excreted. Diastatic enzymes can be made to
render good service in this connection. Assuming that
the persistent indicanuria in the cases in question is due
at least in part to the upward extension of colon bacilli
and certain putrefactive anaerobes into the small
intestine, it seems reasonable to believe that the with-
drawal of excessive carbohydrates is equivalent to a
diminution in the pabulum on which the microorganisms
multiply so readily with the production pf gas and acid
and indol.

As will be seen in speaking of the treatment of chronic
excessive intestinal putrefaction, the use of cathartic
medication is usually followed very promptly by a
diminished absorption of indol from the intestine.
Persons suffering from great indicanuria are apt to show
a marked diminution in the excretion of indican after the
administration of calomel or cascara or a saline or other
cathartic. The indican may in rare instances entirely
disappear temporarily. This effect is, however, only
of very short duration. The diminished excretion of
indican is due of course to the mechanical removal of
indol-forming and indol-containing contents of the
intestine. A diminished absorption of indol is associated

INFECTIONS OF THE DIGESTIVE TRACT 269

frequently with the relief of certain symptoms, such as
headache and dizziness. It is not clear that these
symptoms are due wholly to the excessive indol absorp-
tion. Unfortunately the relief is temporary and is
followed by a continuance of the original putrefactive
conditions and the resumption of indicanuria. Indeed,
in some persons constipation following catharsis is
associated with a still greater excess of indican in the
urine than has been habitual. The repeated use of
cathartics daily or at somewhat longer intervals for the
purpose of diminishing the absorption of toxic sub-
stances is inadvisable, for it ultimately leads to a loss
in weight and strength. It seems unnecessary to
further discuss here the subject of indicanuria, which
is, as a rule, linked with other intestinal conditions
requiring consideration and which therefore it is not
usually advisable to single out as the sole object of
treatment.

THE POSSIBILITY OP THE OCCURRENCE OF INDOLEMIA
AND INDOLURIA

It naturally occurs to any one who considers the
phenomenon of indicanuria and the conditions leading
to it that in extreme instances of this phenomenon,
where relatively large quantities of indol have to be
oxidized, and paired, these preparatory processes may
sometimes be insufficiently active to prevent the cir-
culation of indol in the blood in such concentration as
would lead to the appearance of free indol in the urine.

270 INFECTIONS OF THE DIGESTIVE TRACT

The occurrence of free indol in the urine is certainly
a very exceptional phenomenon even hi cases of ex-
perimental poisoning with indol. In the combined
cyanide and indol experiments of Richards and How-
land it was found that a trace of indol sometimes appeared
in the urine. It is, however, noteworthy that in the
experiments in which examinations were made for the
detection of free indol in a loop of intestine (after poison-
ing with considerable quantities of indol) much more
indol was found to be excreted by the intestinal loop
than was excreted into the urine. This fact points
decisively to a relatively difficult excretion of indol by
the kidney. The tests employed for the identification
of the indol included Ehrlich's dimethylamidobenzalde-
hyde and the /3-napthaquinone-sodium-monosulphonate
reactions. The latter reaction, as already explained,
is absolutely distinctive for indol.

Although it is thus apparent that the conditions re-
quired to bring about the excretion of free indol in the
urine must be such as seldom occur spontaneously in
human beings, it cannot be denied that traces of indol
may sometimes appear in the urine in the course of ex-
treme indicanuria. In a few instances distillates from the
urine contained a body which strongly suggested indol.
But the actual proof of the occurrence of free indol in the
urine in these cases of extreme indicanuria. does not, I
think, exist at present. It seems probable that a con-
siderable concentration of indol in the blood is nec-
essary to bring about the phenomenon of indoluria, and
such a concentration as will insure this leakage into the

INFECTIONS OF THE DIGESTIVE TRACT 271

urine must be of very exceptional occurrence under
clinical conditions. This fact does not seem to me to
weaken the probability that it is not very uncommon
for the blood to contain traces of indol in some cases
of extreme intestinal putrefaction. The proof of the
presence of indol in the blood in experimental poisoning
cases is easily obtained, but I know of no instances of
extreme indicanuria in which any attempt has been
made to obtain free indol from the blood. The demon-
stration of the occurrence of indol in the blood even in
the minutest trace would be of especial interest in those
cases of indicanuria in which nervous symptoms are
extremely pronounced.

In a recent paper dealing with the subject of indigouria,
Porchet and Hervieux have reached the conclusion that
indol is in itself devoid of toxic properties, but that
indoxyl, its oxidation product, is a comparatively poison-
ous agent. They prepared their indoxyl from indoxyl
carbonic acid and injected it into rabbits but fail to
give the doses in which they found the indoxyl fatal.
It does not appear to me that the opportunity for the
occurrence of free indoxyl in the blood or tissues can
ever be good, since the process of pairing with sulphuric
acid is so intimately connected with the oxidation of indol
to indoxyl that it is difficult to believe any significant
amount of free indoxyl can ever be present at one time.
One may, however, admit the possibility that in con-
ditions of extremely abundant absorption of indol from
the intestine there might be formed an amount of indoxyl
larger than usual, within a time insufficient to secure

272 INFECTIONS OF THE DIGESTIVE TRACT

its complete pairing. Under such .conditions it might
conceivably be a factor in the production of symptoms.
As compared, however, with the quantity of indol
existing as such in the organism, it seems to me that the
indoxyl must be insignificant. Evidence has already
been offered to show that while indol is not a highly
toxic substance for normal carnivorous or omnivorous ani-
mals, it becomes much more toxic when conditions arise
under which there is defective oxidation. This aspect of
the question of the toxicity of indol is one which has not
been touched upon by any writers upon this subject.

Indigouria. — It has long been known that there
occasionally occur cases of extreme indicanuria in which
the urine upon standing in contact with the air gradually
becomes blue and may liberate a pelicle of indigo upon
the surface of the urine at the same time that a precipi-
tate of fine particles of indigo occurs in the urine. Such
cases are, I believe, very exceptional and do not arise
except where the indicanuria has been of long standing.
The subjects of this indigouria are almost invariably
badly nourished and in poor, often precarious, health.

Procher and Hervieux 1 have lately studied experi-
mental indigouria, which they found to arise when larger
quantities than one or two grams of indol are administered
by the stomach to dogs. The phenomena as they appear
in the urine of these experimental cases are essentially
the same as those observed in man. In the majority

1 " Recherches experimentales sur les chromogfenes urinaires
du Groupe de ITndol' (5* M6moire) "De ITndigourie," Journ. de
Physiol. et de Path, gen., viii, p. 841, 1906.

INFECTIONS OF THE DIGESTIVE TRACT 273

of cases the pigment which is separated from the urine
is blue in color. In the case of rabbits the substance
separated from the urine is red in color and represents
the presence of indirubin, a substance closely related to
indigo. It often happens in indigouria that both these
substances are present together. The onset of the indigo-
uria in dogs has been observed after three or four hours
from the time of the administration of the indol, but the
phenomenon is at its height after six or eight hours have
gone by.

Porcher and Hervieux claim to have detected the
presence of indoxyl in the urine of dogs which had been
given experimental indigouria. It is thought by them
that the indoxyl-potassium-sulphate (which they call
the indigouria chromogen) is decomposed with the lib-
eration of indoxyl. In herbivorous animals, such as the
horse and the rabbit, the phenomenon of indigouria is
said not to be a rare occurrence. In the horse the pigment
consists of a mixture of indigo blue and indigo red, or
indirubin, whereas in the rabbit the coloring matter
is indigo red. It is stated that the occurrence of the
phenomenon of indigouria in these herbivorous animals
is not necessarily connected with any evidence of illness.
I believe there is at present not the slightest evidence that
the phenomenon of indigouria occurs in man in a state of
health. There seems to me no evidence that indigouria
depends on the presence of free indoxyl in the blood and
urine. It is much more likely that it is a phenomenon
dependent on the spontaneous oxidation of indoxyl-
potassium-sulphate in the urine in those cases in which

274 INFECTIONS OF THE DIGESTIVE TRACT

the quantity of this substance is very extreme. I
believe also that the phenomenon of indigouria never
arises except where indol is being produced in the intes-
tine in uncommonly large amounts and is subsequently
absorbed in uncommon abundance.

INDIVIDUAL SUSCEPTIBILITIES TO DIFFERENT ENTEROG-
ENOUS POISONS AS POSSIBLE FACTORS IN DETERMIN-
ING CLINICAL TYPES

Instances are many in which clinical experience has
made it clear that two persons of approximately the same
weight react differently to the same drug, and do so
regularly. This is true of commonly used drugs, such
as strychnine, morphine, cocaine, digitalis, and antipyrin.
A scientific explanation of these differences is for most
cases not now possible. That they do not depend on
gross differences in rapidity of absorption is shown in
the case of some of these drugs by the fact that sub-
cutaneous injections give rise to the same qualitative
differences. The individual variations are to be at-
tributed rather to idiosyncrasies of the neuro-muscular
system — the mechanism through which functional
derangement most readily finds expression in the form
of symptoms. The nervous system may prove uncom-
monly susceptible either on account of inherent physico-
chemical peculiarities of its constituent nervous ele-
ments or because it is not adequately protected from the
action of poisons by hepatic and other cells which have
the power to bind, or bind and oxidize, or bind and oxi-

INFECTIONS OF THE DIGESTIVE TRACT 275

dize and pair, many substances injurious to the nervous
elements. An inherent susceptibility of the nervous
elements themselves probably has its physical sub-
stratum in side chains or receptors (according to the
conception of Ehrlich) possessing affinities for sub-
stances possessing a definite chemical constitution.

Of individual human susceptibilities and reactions
to the action of enterogenous poisons almost nothing
is now known. Nevertheless one cannot fail to recognize
the possibility that such individual susceptibilities and
reactions may play an important part in determining
the clinical manifestations of intoxications. It is well
known to clinicians that there are some persons who
promptly develop a widespread urticaria after indulgence
in certain foods or drinks, as shellfish or strawberries
or champagne. Sir A. E. Wright tells me that a patient
of his develops an urticarial eruption after ingesting
organic acids. I had a patient who suffered frequent
urticarial seizures attributable to somewhat injudicious
living. She had marked indicanuria and excessive
putrefactive decomposition in the intestinal tract. A
rigid diet maintained more than a year led to a cessa-
tion of the urticarial seizures and to a diminution in
the indican excreted. In a mood of experiment she de-
cided to test the action of champagne and found that
one glass was followed within three hours by a violent
outbreak of general urticaria. Here the susceptibility
was highly pronounced. The urticaria may be connected,
as Wright has pointed out, with a diminished coagu-
lability of the blood, but this is in itself not a full ex-

276 INFECTIONS OF THE DIGESTIVE TRACT

planation of the locally deranged function, since it fails
to take account of the nervous mechanisms involved in
the localization of the urticarial wheals.

In some persons the indulgence in a single glass of
champagne is followed within twenty-four hours by
manifestations of gout. In still others champagne
causes headache and the excretion of increased amounts
of uric acid. The explanation of these different effects
is to be sought in the individual cellular reactions of the
patient rather than in the poison. There are probably
many similar examples of individual susceptibility, but
when we come to study the question in relation to pro-
cesses found in the digestive tract we cannot make
close comparisons between different persons because we
cannot say what substances are being absorbed. We
may know that patients of a certain group are alike in
having intense indicanuria, but we cannot say that the
intoxications may not be different in these cases owing to
differences with respect to the absorption of other sub-
stances than indol. Careful research may help us to get
much closer to the resemblances and differences in such
cases. In the meantime we must content ourselves with
the suspicion that chronic intoxications through the
absorption of similar quantities of the same poisons
produce different effects in different persons. Among
half a dozen persons suffering from extreme indicanuria,
one suffers from headaches (sometimes migrain-like),
another is prone to lumbago, another, perhaps, has
epileptic seizures, another mental depression, another
progressive muscular atrophy, and still another suffers

INFECTIONS OF THE DIGESTIVE TRACT 277

from cyclical vomiting. It would be well worth while
to learn in how far these different manifestations of
intoxications are dependent on common factors or in how
far on different agencies. There is good reason for
suspecting that very similar bacterial processes in the
digestive tract lead in one case mainly to digestive dis-
orders and in others (owing to a lesser sensitiveness of the
digestive tract itself) to better absorption of poisons and
the development of more remote consequences such as
gout, arthritis, anaemia, or nervous disorders. While
it is possible that these very different manifestations
are always connected with different and perhaps spe-
cifically different types of gastro-enteric infection and
intoxication, the possibility is not excluded that even
such very different disorders may have much in common
in their etiology.

That the mental and emotional peculiarities of indi-
viduals have a large part in fixing the type of nervous
reaction that occurs in consequence of intoxications has
become apparent to careful students of pathological
conditions. It is not unreasonable to believe that it
will well repay close observers of the phenomena of dis-
ease to take into consideration the individual nervous
reactions much more than is at present customary.

TYPES OF CHRONIC EXCESSIVE INTESTINAL
PUTREFACTION

i

THE considerable vacations in the clinical manifes-
tations and pathological accompaniments of chronic
excessive intestinal putrefaction suggest that the etio-
logical conditions vary considerably in different cases. I
shall endeavor in the following pages to show that there
are grounds for the separation of different types of
intestinal putrefactive processes and that these grounds
are based on differences in the character of the putre-
factive products and in the bacteria concerned with these
processes. The classification which I offer is a tentative
one which may in time have to give way to one founded
on a fuller knowledge of the significant intestinal con-
ditions than we at present possess. In any event it
must undergo modification. It is, however, proposed
here in the belief that it will serve a useful purpose in
helping to classify cases of disease and in giving direction
to systematic study.

It is a striking fact that while many cases of excessive
intestinal putrefaction are associated with the presence
of indican in the urine in large amount, there are other
cases of considerable gravity clinically which fail to
show much indican in the urine or may indeed be entirely
free from indicanuria during a long period of observa-
tion. I propose to divide cases of chronic excessive

278

INFECTIONS OF THE DIGESTIVE TRACT 279

intestinal putrefaction into those which show an exces-
sive indicanuria and those which do not. The proposed
classification recognizes three types of putrefaction which
are common : the first may be called the Indolic Type,
and is characterized by indicanuria dependent chiefly
upon excessive decomposition induced through the co-
operation of members of the B. coli group (and probably
B. putrificus) ; the second type of intestinal putrefaction
may be designated the Saccharo-butyric Type, and is
initiated especially by anaerobic organisms in the digest-
ive tract. In its simplest form it is associated with
very little production of indol in the intestinal tract
and hence with slight indicanuria or even an entire
absence of indican from the urine. In the third group
of cases we find associated the characters of the indolic
and the saccharo-butyric types of decomposition.1 I
will describe these types briefly according to my
present conception of their etiology and pathological
characters.

1 1 think it will become desirable to distinguish a skatolic type
of putrefactive decomposition, i.e. a condition in which the in-
testinal bacteria persistently form skatol in amounts in excess
of the indol made. As mentioned in speaking of skatol, I have
found this product of decomposition abundant and dominant
over indol only under conditions of clinical deviation from normal
standards. But as I am still in doubt as to the bacterial conditions
that determine skatol production, I hesitate at present to make
skatol prominent in a plan of classification. Moreover, I do not
yet know whether excessive skatol production corresponds to
characteristic manifestations.

280 INFECTIONS OF THE DIGESTIVE TRACT

I. THE INDOLIC TYPE OF EXCESSIVE INTESTINAL
PUTREFACTION

It has already been stated that there is some reason
for believing that organisms of the B. coli type which
are capable of forming indol abundantly may invade the
small intestine and there give rise to bacterial cleavages
which largely replace the normal try p tic digestion. The
proof that the small intestine is invaded by some variety
of indol-producing colon bacilli in cases of marked indi-
canuria is still lacking, for the reason that it is ordinarily
impossible to obtain evidence on this point from a living
subject, whereas postmortem studies with reference to
this point are open to criticism on account of the rapid
wandering of the colon bacilli which occurs after death.
Of all the organisms of the digestive tract which we have
individually studied, we have found none that will make
a larger amount of indol than certain strains of colon
bacilli when grown on peptone bouillon. There are,
moreover, cases of excessive intestinal putrefaction in
which the dominant organism in the intestinal tract (if
we may judge from the material obtained through the
use of cathartics) is some variety of colon bacilli. In such
cases, the anaerobes of the intestine may be present in
very small numbers, so that it seems safe to exclude
them from a significant part hi the formation of putre-
factive products. It seems reasonable hi these cases
to attribute the formation of indol to an unusual activity
on the part of the colon bacilli. There are, perhaps, also
instances in which the presence of an indol-producing

INFECTIONS OF THE DIGESTIVE TRACT 281

proteus organism is a factor. But I am not inclined
to ascribe great importance to the proteus organism in
this connection, for aerobic liquefiers of gelatin have
generally been lacking or few in number. If we assume
that the indol produced in these cases of indicanuria is
made largely though the action of colon bacilli/ we are
brought to the question, In what part of the intestinal
tract does the indol production chiefly occur ? We know
that the colon bacilli have only a very feeble action upon
native proteids such as are used for food. We are there-
fore forced to the conclusion that the indol produced by
the colon bacilli must have been formed through the
cooperative agency of the digestive juices or putrefac-
tive bacteria. The anaerobic putrefactive bacteria
may be excluded from the present discussion, since the
cases here in question are those in which these bacteria
are present in only very small numbers in the digestive
tract.2 So we have to fall back on the pro teoly zing action
of the digestive juices to explain that preparation of the
proteid food which is necessary for a successful attack
by organisms of the B. coli type. Where only a moder-
ate amount of food is being used, and this in a form
readily digested and absorbed, such as egg-white or
milk, it is unreasonable to suppose that the foodstuff

1 1 am confident that an indol-making strain of B. putrificus is
sometimes an important agent in establishing indicanuria, but do
not know any instance where the influence of colon bacilli could be
excluded.

2 1 have noticed that certain Gram-positive diplococci from the
small intestine peptonize casein with moderate rapidity, and it is
possible that such organisms, when excessive, prepare the way for
the putrefactive action of colon bacilli.

282 INFECTIONS OF THE DIGESTIVE TRACT

is attacked by the colon bacilli mainly in the large in-
testine. It seems more legitimate to assume that they
are attacked in the small intestine in the manner already
mentioned. It is, of course, likely that in some cases
hydrolyzed proteids find their way into the ileum and
large intestine, and here it is necessary to assume that
the colon bacilli have migrated upward from their usual
habitat in order to explain the phenomenon of excessive
indican production. I may mention an illustration of
the indolic type of excessive intestinal putrefaction. The
patient was a lady seventy-five years of age who had the
peculiarity of having lived since childhood on a diet
consisting almost exclusively of potatoes, hominy, and
bread, and from which meat had been rigorously excluded.
She ate little food, but was well nourished and uncom-
monly robust. . She ultimately fell ill from a subacute
gastro-enteric disturbance in which there was loss of
appetite, irregularity of the bowels, and some signs of
gastritis. The faeces at this time showed fields consisting
mainly of Gram-negative bacilli mostly of the B. coli
type.1 These fields contained extremely few putrefac-
tive anaerobes. Nevertheless the urine contained ex-
ceptionally large quantities of indican, and this was
apparently a persistent feature.

The indolic type of chronic excessive intestinal putre-
faction is common in cases of complete or partial occlu-
sion of the common bile duct with or without jaundice,

1 Two kinds of colon bacilli and a positive diplococcus were iso-
lated by Mr. Ward. One of these strains of colon bacilli was a
strong indol maker.

INFECTIONS OF THE DIGESTIVE TRACT 283

and is probably a common association of functional pan-
creatic achylia,1 a state which is probably not uncommon.
Some of the most pronounced instances of indicanuria
with which I have met have been in cases of organic
structural occlusion of the common duct. There must
be in these cases a great diminution in the proteolytic
action of the intestinal digestive juice, and putrefactive
cleavages are doubtless due very largely to the action of
bacteria and especially the colon bacilli. I have noticed
in several cases of jaundice that the fseces held mainly
Gram-negative organisms of the B. coli type. In some
autopsies such organisms have been dominant through-
out the digestive tract. Bacteriological studies for the
identification of the organisms and their biochemical
characters in such cases have not yet been made. I men-
tion these appearances, however, as being of interest
in the present connection, since they were characterized
by the absence of putrefactive anaerobes in more than
very small numbers. In some instances cultural obser-
vations on the mixed faecal flora have borne out the view
that the colon bacilli were the dominant type and the
putrefactive anaerobes were few in number.

Occlusions of the small intestine are almost invariably
followed by intense indicanuria even when the occlu-
sion is in the jejunal region. The bacteriological ex-
planation in such cases is not clear. The reversal of
peristalsis, above the level of obstruction, must serve
to distribute colon bacilli from the ileum or jejunum

1 A. Schmidt, " Functionelle Pankreasachylie," Deutsch. Archiv
/. klin. Med., Ixxxvii, p. 456, 1906.

284 INFECTIONS OF THE DIGESTIVE TRACT

throughout the small intestine, between the stomach
and the obstruction. This dissemination of colon
bacilli above the occlusion may be an important factor
in favoring indol production. Where the obstruction
lies high, as in the upper jejunum, indicanuria is said
not to occur.

The development of a complete obstruction of the
small intestine is followed by obstinate constipation,
and this may influence the putrefactive conditions below
the obstruction. But I have already mentioned that
mere constipation does not necessarily lead to indican-
uria. On this account it is difficult to determine the part
played by bacterial conditions below the obstruction, in
contributing to the formation of indol. It appears, on
the whole, that it is mainly the bacterial conditions above
the obstruction that determine the presence or absence
of an extreme indicanuria. This view is not inconsistent
with the fact that indicanuria may persist after the small
intestine has been emptied of food through repeated
vomiting, for it is known that indol often continues to
be found in the digestive tract of starving animals.
Small quantities of proteid (perhaps derived mainly from
desquamated epithelial cells) apparently suffice to insure
indol formation when the bacterial conditions are favor-
able.

A well-marked group of cases which may provisionally
be classed with the simple indolic type is that which
may be characterized as the marantic, large-belly type
of chronic intestinal indigestion in children. This
condition has long been known to clinicians as an

INFECTIONS OF THE DIGESTIVE TRACT 285

extremely obstinate form of digestive derangement, mani-
festing itself in a distinct retardation of growth (usually
without much mental retardation), in distension of the
abdomen by gas, flatulence, intolerance of carbohydrates,
and voluminous, light-colored, usually gray and fatty
stools containing much gas. In many of these children
there is excessive sweating of the head. I have studied
with considerable care several children corresponding to
this clinical type (most of them patients of Professor L. E.
Holt, who has in every way aided me in the effort to
learn something about the underlying conditions present
in this striking derangement). It was found that it is
a characteristic in these cases that the urine contains
a large amount of indican and usually an excessive
amount of phenol. The phenol may amount to more
than one hundred milligrams in twenty-four hours, in the
urine from a child not more than three or four years of
age. The Ehrlich aldehyde reaction of the urine may or
may not be excessive. After a long period of careful diet-
ing in which carbohydrates are very much restricted the
indican may disappear from the urine in such cases
although the patients show but little gain in weight or
strength. This falling off in indican of the urine is to
be regarded as a favorable indication and often precedes
by many months more obvious clinical indications of
recovery. In some instances, however, there occurs a
slow gain in strength and in weight and in the ability
to digest carbohydrate food despite the fact that the
improvement in the excretion of indican and phenol is
only slight. Under the most favorable circumstances

286 INFECTIONS OF THE DIGESTIVE TRACT

improvement is very slow, and slight errors in diet are
followed by prompt reappearance of flatulence and per-
haps diarrhoea. Fats and proteids are well tolerated.

A feature in this form of chronic excessive intestinal
putrefaction is the readiness with which fatigue comes
on. A child of six or eight years of age will show fatigue
after a very short walk and will manifest a want of in-
terest in exercise and play. There is in fact a state of
chronic fatigue, which I attribute to the absorption of
putrefactive products. If the trouble comes on during
infancy, there may be considerable delay in the develop-
ment of the ability of the child to stand and walk without
support, and a child of five or six years or over may be
unable to go upstairs except by a kind of creeping
motion. A certain emotional irritability is common in
these children. As a rule they are not much retarded
mentally except in so far as they have missed opportuni-
ties for conventional learning. On the contrary, one not
infrequently is struck by the contrast between the sharp
wits of these children and their physical retardation.

The ethereal sulphates are regularly increased in the
urine in cases of the kind under consideration, that is to
say, there is not merely a relative increase of the ethereal
as compared with the preformed sulphates, but usually
also an actual excess in the absolute amount of the
ethereal sulphates. The proportion of ethereal to pre-
formed sulphates may be one to seven, or one to five,
or even one to four. It can usually be shown that the
stools contain considerable quantities of indol and phenol.

A noteworthy peculiarity of these cases is the presence

INFECTIONS OF THE DIGESTIVE TRACT 287

of a very large proportion of Gram-positive bacilli cor-
responding closely to the morphology of the organism
described by Tissier as B. bifidus. It has already been
pointed out that the fseces of breast-fed children consist
very largely of Gram-positive organisms of the bifidus
kind. When children reach the age of three or four
years, the bifidus bacilli are usually comparatively few
in number and the dominant type is some form of colon
bacillus. In children from four to ten years of age the
Gram-negative bacilli of the colon type are much in
excess normally as compared with the organisms repre-
sentative of bifidus. In the cases of chronic intestinal
indigestion which are under consideration I have met
with a great preponderance of bifidus-like organisms in
children as old as seven or eight years. The cultural
proof that the organisms to which I refer are in reality
the same as those normally present in the faeces in in-
fancy is lacking, as I have been unable to adequately
study the biochemical characters of these organisms.
In sugar-bouillon fermentation tubes, however, the
resemblance is a very close one ; that is to say, the growths
observed in lactose or dextrose bouillon show many
instances of the bifid form of bifidus and also many
instances of so-called punctate forms in which Gram-
positive portions of the bacillus alternate with segments
of the cell body which fail to take the Gram-stain. As-
sociated with these bifidus forms in the faeces of the
marantic cases one finds moderate numbers of colon
bacilli, or colon bacilli may be wholly absent from the
fields. In the lactose-bouillon fermentation tubes one

288 INFECTIONS OF THE DIGESTIVE TRACT

may find excessive numbers of positive diplococci. It
should be especially noted that the common anaerobic
putrefactive organisms are present in very small num-
bers. It is not certain that the bifidus-\ike organisms
found in these cases of chronic indigestion in children
represent the preponderance of a type which should,
under normal conditions, gradually be replaced by colon
bacilli, but the idea is one which forces itself strongly
upon one. It is possible that the organism may have
acquired other characters than those pertaining to the
bifidus bacilli of the nursling.

Associated with the Gram-positive bifidus-\ike or-
ganisms in the faces one may find moderate numbers
of positive diplococci or short streptococcal chains.
These diplococci may appear in considerable abundance
in the fermentation tubes or may give rise there to an
abundance of streptococcus forms. In one apparently
typical case which was under observation for a long time,
the child at the height of the disease (when indicanuria
was very marked) showed in the faecal fields considerable
numbers of these Gram-positive diplococci. After a
period of six months, upon a strict proteid and fat diet,
the diplococci were no longer present in the fields. It
was also noticeable that indican had disappeared from
the urine although the patient had gained very little in
weight and strength.

Whether in these cases the production of indol is due
to the colon bacilli present in the upper part of the tract
or whether it is due in part to other types of organisms
is by no means clear. B. bifidus in these cases certainly

INFECTIONS OF THE DIGESTIVE TRACT 289

does not occupy the entire intestinal tract, as may be
shown in diarrhoeal stools which contain colon bacilli
as well as B. bifidus. It is uncertain in how far the re-
tardation in growth and the myasthenia which are
characteristic of these cases are explicable through the
intoxication induced by the absorption of excessive
quantities of indol or phenolic substances. I am dis-
posed to think these bodies are an important factor in
determining the undue fatigue. The retardation in
growth may have hi part a similar origin, but it is prob-
able that this is partly the effect of a failure of the or-
ganism to utilize food normally, owing to the excessive
decompositions that occur in the intestinal tract. No
matter how much food is given, the excessive bacterial
changes rob the organism of enough caloric potential
to make normal growth impossible.

In the treatment of these cases it is necessary to be
patient. It is essential to improve the character of the
digestive processes before one can reasonably hope to
secure a gain in weight and strength. Carbohydrates
should be permitted in only very small amount and
chiefly in the form of well-cooked rice or hominy or
Huntley and Palmer's breakfast biscuits. It may be
advantageous for a time to peptonize the milk in order to
secure early absorption. Moderate quantities of chicken
or beef or mutton may be permitted, but all meat should
be given in a finely divided form. It sometimes has
seemed best, in children as old as five or six years of age,
to allow only two meals in the course of the day, these
being rather generous. It is helpful to practice high irri-

290 INFECTIONS OF THE DIGESTIVE TRACT

gation in order to avoid as far as possible putrefactive
decompositions in the large intestine. A great deal of
rest is necessary for these children and they should not
be permitted to play or walk to the point of fatigue.
Patients of the class under consideration often do badly
in cold weather, probably because they cannot afford
the caloric loss to which they are then subjected. Where
this is the case a period of residence in a mild winter
climate is helpful. Improvement is sometimes so slow
that every one concerned with the case may become dis-
couraged, but it is certain that a high degree of improve-
ment can occur even after several years of extremely
slow progress. I have several times known this to be
the case. In order to secure this improvement, however,
it is impossible to make concessions in regard to diet, and
the policy of largely excluding carbohydrates must be
enforced. It cannot be denied that even in those cases
where the best results are obtained there is commonly
a retardation in growth which is still manifest at the
time of puberty. Moreover, there is no doubt that per-
sons who have suffered from the conditions under consid-
eration during the period of childhood become especially
susceptible to intestinal disorders in later years and sel-
dom attain robust health. There are probably many
persons in the community whose digestive derangements
date from early life, and in many of these cases the less
pronounced forms of the type of chronic intestinal indi-
gestion which has just been described have doubtless
constituted the foundation of invalidism in adult life.

INFECTIONS OF THE DIGESTIVE TRACT 291

H. THE SACCHARO-BUTYRIC TYPE OF CHRONIC EXCES-
SIVE INTESTINAL PUTREFACTION

This form of intestinal derangement is characterized
by a chronic putrefactive process (having its seat mainly
in the large intestine and lower ileum) and due to the
action of very large numbers of strictly anaerobic
butyric-acid-producing bacteria capable of multiplying
by means of spore formation. A thorough study of the
anaerobes concerned with this type of decomposition
has not yet been made, but it may confidently be stated
that the organism most prominently concerned in at
least a large number of the cases is B. aerogenes capsu-
latus (B. Welchii). The characters of this organism
have already been elsewhere described. Associated
with B. aerogenes capsulatus may be found B. putrificus
and possibly in some cases the bacillus of malignant
oedema. In many cases, however, B. putrificus has not
been found in the fermentation tubes or in the flasks
containing various media into which suspensions of
the mixed faecal flora have been inoculated. Anaerobic
plates made from the suspensions, heated for twenty
minutes to 80° C., have shown B. aerogenes capsulatus to
be the dominant or even the exclusive strict anaerobe
growing upon sugar-blood agar or blood agar. In a
certain proportion of cases the fsecal fields have shown
B. aerogenes capsulatus to be associated with large
numbers of Gram-positive diplococci, and these Gram-
positive diplococci have been grown very abundantly
in the fermentation tubes; that is, they have assumed a

292 INFECTIONS OF THE DIGESTIVE TRACT

prominence there which is not often observed in the case
of persons in good health. It seems not improbable
that the presence of these diplococci are of importance
to the process, but their precise relation to it cannot be
at present confidently stated. In some instances these
Gram-positive diplococci appear in the fermentation
tubes in great abundance, although they are not seen
in the faecal fields in such number as to attract special
attention.

A very common phenomenon in cases of the saccharo-
butyric type of chronic excessive intestinal putrefaction
is the formation of only small volumes of gas in sugar-
bouillon fermentation tubes after these have been inocu-
lated with the mixed faecal flora. As already explained,
it has been our custom in routine work to study the
gas production on four sugar-bouillon tubes containing
dextrose bouillon, levulose bouillon, saccharose bouillon,
and lactose bouillon. The gas production in these
tubes has often fallen to one-half that observed ordinarily
in health, and the volume may even fall to one-quarter
or one-fifth the total gas production observed under
normal conditions. In a few cases some of the fermen-
tation tubes have contained no measurable amount of
gas. This is, however, an exceptional condition. The
explanation of this phenomenon of small gas production
in the fermentation tubes appears to lie in the diminu-
tion hi the number of living organisms of the B. coli
type that are present in the faeces. That the colon
bacilli are much reduced in numbers is evident in some
of the cases of saccharo-butyric putrefaction from the

INFECTIONS OF THE DIGESTIVE TRACT 293

microscopical examination of the Gram-stained faecal
fields, even in fields from diarrhceal stools. It is pos-
sible that in some of the cases that have come under
observation the typical colon bacilli have been replaced
by some allied form of Gram-negative organism capable
of forming less gas than is ordinarily found as the prod-
uct of normal colon bacilli on sugar bouillon. We have
not yet had an opportunity to make a careful study
of our cases of saccharo-butyric putrefaction from the
standpoint of the characters of the colon bacilli present.

While the reduction in the total number of living colon
bacilli has been shown to exist in many of our cases of
saccharo-butyric putrefaction by means of plating on
litmus gelatin, there are exceptional instances in which
there is no evidence that the number of colon bacilli
has been reduced. The evidence, however, in these
cases is confined to the appearance of acid-forming
colonies on litmus gelatin and agar and to the occur-
rence of normal gas production in the fermentation
tubes.

The presence of large numbers of putrefactive anae-
robes in the intestinal tract and especially of B. aero-
genes capsulatus, gives to the intestinal contents a pecul-
iar character. These organisms, as already mentioned,
are capable, under anaerobic conditions, of attacking
carbohydrates and proteids in a vigorous manner. In
either case there is a formation of butyric acid, often in
considerable amount. This may be associated with the
production of propionic, caproic, or valeric acids. The
odor of the movements from patients in whom this

294 INFECTIONS OF THE DIGESTIVE TRACT

form of putrefaction is pronounced is often intense and
characteristic, suggesting butyric or caproic acid. When
proteids are attacked by B. aerogenes capsulatus there is
not only a production of butyric acid, but also a libera-
tion of gas, although there is less gas liberated than when
the organisms attack carbohydrates. The gas liberated
consists in part of hydrogen and in part of carbon
dioxide. Perhaps some methane is also formed. As a
result of the excessive production of gas the faeces have
a low specific gravity and float on the surface of water.
Frequently small bubbles of gas may be seen to be
liberated. The presence of considerable quantities of
gas in the form of small bubbles in the intestinal contents
contributes to give them a light color. Another factor
which enters into the light color often observed in the
faeces in cases of saccharo-butyric putrefaction is a re-
duction of bilirubin and other coloring matters by the
anaerobic bacteria. Reducing processes normally go
on in the intestinal tract but they are usually of moder-
ate intensity in conditions of health. In excessive sac-
charo-butyric putrefaction, on the contrary, the processes
of reduction are extremely active, owing doubtless largely
to the liberation of free hydrogen. The faeces of cases
of saccharo-butyric putrefaction when subjected to an
appropriate oxidizing agent, such as mercuric chloride
in concentrated solution, usually give a strong pink or
red color which is rendered more striking when the
material which has already been reddened in this way is
thrown into water, which then dissolves out the newly
formed coloring matter and reveals its intensity. This

INFECTIONS OF THE DIGESTIVE TRACT 295

reaction is probably due to the presence of an excessive
amount of hydrobilirubin.

In consequence of the presence of the higher fatty
acids in the faeces in considerable amount, there may be
imparted to the contents of the lower bowel a slight acid
reaction. The reaction, however, may be neutral or
even alkaline. The fatty acids in question are neutral-
ized by bases formed in the course of putrefactive
processes and especially by ammonia, ammonium buty-
rate sometimes being formed in considerable amount.
It is probable that the production of ammonium buty-
rate in excessive amount acts as an irritant to the in-
testinal tract and may be the cause of a certain softness
of the movements 1 in cases of excessive saccharo-butyric
putrefaction or may even be responsible for a diarrhoaal
condition. It is, however, not clear that this is the sole
cause of such evidence of irritation, as other irritative
products may be associated with ammonium butyrate.

The faeces in the condition under consideration usually
show the presence of little or no indol. Indol may be
wholly absent. More commonly there is a small amount
of indol, — perhaps two or three milligrams in one hun-
dred grams of the intestinal contents. The quantity of
phenol may be slightly in excess of that normally found,
but this is, I think, not usually the case. In con-
sequence of the small absorption of indol the urine con-

1 The consistence of the stools may be influenced by the liquefy-
ing action of the anaerobes on proteids. A normal stool may
stand in the laboratory for days without changing in consistency;
a stool containing a great excess of putrefactive anaerobes may
spontaneously liquefy in the course of a week.

296 INFECTIONS OF THE DIGESTIVE TRACT

tains little or no indican. The cases in which large
quantities of indican are found in the urine must be
regarded as belonging in a different category from the
cases which are now under discussion. But it must be
admitted that a patient may show considerable ^changes
in regard to the indol production; a person who at one
time produces little intestinal indol will at other times
produce a considerable amount. Although the indican
of the urine is usually small or may be wholly absent,
the ethereal sulphates are often somewhat in excess.
The explanation of this fact is not clear. It indicates
that other putrefactive products than indol have been
absorbed from the intestine and have paired with sul-
phuric acid. In some cases perhaps phenolic bodies are
responsible for the increase in the ethereal sulphates,
but I have observed cases in which the increase in the
ethereal sulphates appeared to me to be out of propor-
tion to the amount of phenol recoverable from the urine.
It is quite possible that substances heretofore over-
looked have been largely responsible for the increase of
ethereal sulphates observed. There are, however, cases
of excessive saccharo-butyric putrefaction in which one
cannot say that the ethereal sulphates are in excess.

Of the formation of sulphur compounds hi the intesti-
nal tract during the process of saccharo-butyric putre-
faction little is yet known. There is no evidence that
hydrogen sulphide is produced in excessive amount in
these cases. As regards the formation of mercaptan,
the evidence is against the view that anything more
than a trace of methyl mercaptan is liberated from the

INFECTIONS OF THE DIGESTIVE TRACT 297

putrefying material in the intestine, but it must be ad-
mitted that the facts bearing on this point are not con-
clusive. Outside the body the mixed fsecal flora from
cases of excessive saccharo-butyric putrefaction is usu-
ally capable of forming mercaptan on peptone bouillon
and in some cases the amount produced hi twenty-four
hours is sufficient to give a very strong reaction in the
isatin-sulphuric-acid tube. This condition is unphysio-
logical, as it is not observed in most adults in good
health. A moderate mercaptan reaction is often ob-
tained, as already stated, through the action of the
mixed fsecal flora derived from apparently healthy
young children. That the reaction is abnormal is
shown not merely by the infrequency of its occurrence
in healthy adults, but also by the fact that a patient
when improving from conditions of marked saccharo-
butyric putrefaction will usually lose the flora capable
of making methyl mercaptan in peptone bouillon. A
satisfactory explanation of the formation of methyl
mercaptan by the mixed fsecal flora cannot yet be offered.
It is easy to show that B. putrificus is able to form methyl
mercaptan, and in some cases of saccharo-butyric pu-
trefaction B. putrificus occurs abundantly, especially in
peptone bouillon to which has been added magnesium
or calcium carbonate. In such instances it would
appear that the methyl mercaptan is formed as a result
of the activity of B. putrificus. There are other cases,
however, in which, although methyl mercaptan is
formed by the mixed fsecal flora, one fails to obtain
B. putrificus, and hence the reaction cannot confidently

298 INFECTIONS OF THE DIGESTIVE TRACT

be ascribed to the presence of this organism. I think
it probable that B. aerogenes capsulatus is sometimes
capable of making methyl mercaptan, but apparently
it usually does not make this gas as abundantly as
does B. putrificus, and there are strains which do not
form it under the conditions in which the organism has
been grown on blood bouillon.

Chronic excessive saccharo-butyric intestinal putre-
faction is a widespread condition among adults. In
children it is, I believe, a relatively uncommon occur-
rence. Although there is evidence that in children
some acute disturbances of short duration are dependent
upon this process, as people grow older they are in
general more subject to the occurrence of this type of
putrefactive disorder. The grade in which the dis-
turbance is present varies within wide limits. B.
aerogenes capsulatus is present in the intestine of nearly
all adults, and there are probably few persons who from
time to time do not suffer slight temporary derangements
of intestinal digestion connected with the temporarily
excessive multiplication of these anaerobes. There are
probably persons who during an entire lifetime have
digestive disturbances, usually slight but sometimes
more marked, which are dependent on this form of
putrefaction. Under such circumstances the duration
of life may not be appreciably shortened. Persons who
have long had disturbances of this sort may attain to
seventy or seventy-five years of age. They, however,
suffer in various ways a diminution in efficiency and are
subject to various obvious disturbances. It sometimes

INFECTIONS OF THE DIGESTIVE TRACT 299

happens that people have a mild grade of saccharo-
butyric putrefaction for a long period of time without
being conscious of any distinct impairment of powers
and without being conscious of anything approaching
invalidism. On the other hand, there are some persons
who develop this type of putrefactive disorder in a high
degree. In them the evidences of intestinal indigestion
are pronounced, and there is, after a time, a decline in
the capacity of the organism to perform work. Such
persons develop a condition of distinct invalidism, and
life may be considerably shortened in consequence of
the intoxications arising from this condition. As a
rule, however, the severest cases of chronic saccharo-
butyric putrefaction are not of the simple type at
present under discussion, but are rather of that form
which is associated with the production of an excess
of indol in the intestinal tract.

Some of the immediate and remote consequences of
chronic excessive saccharo-butyric intestinal putrefac-
tion may be briefly enumerated. As already stated,
the excessive production of ammonium butyrate is
apt to lead to irritability of the digestive tract, and this
is the more pronounced if there is also an abundant indol
production. There is a tendency to desquamation of
the epithelium of the digestive tract in the mouth and
stomach and probably throughout the entire digestive
tract. The indications of this desquamation in the
mouth and on the tongue are often plain. A frequent
sign of such desquamation in the intestinal tract is
the presence of an excessive quantity of epithelium or

300 INFECTIONS OF THE DIGESTIVE TRACT

epithelial nuclei in the faeces. The patients who suffer
from this condition usually do not tolerate well either
carbohydrates or acids. The intolerance of carbohy-
drates is shown by the occurrence of excessive intestinal
flatulence and sometimes slight diarrhoea after the use
of considerable meals of cereals or starchy foods. Sugars,
especially, are apt to be badly borne.1 Probably this
intolerance is due in part to the production of acids.
There are cases in which the acid production from sugar
begins very rapidly in the mouth, and a certain amount
of butyric acid may be formed even in this situation,
owing either to the presence of B. aerogenes capsulatus
in the mouth or to the presence of aerobic butyric-acid
producers. Probably the tendency to diarrhoea ob-
served in some instances is due to the irritative action
of the acids formed from the decomposition of the
sugars in the small intestine. This intolerance is
naturally most marked hi those persons in whom evi-
dences of epithelial desquamation in the mouth and
elsewhere are most pronounced. The mucous mem-
branes of the digestive tract are almost constantly in a
state of excessive irritation, so that stimuli which in
normal people would meet with little response give
rise to an excessive peristalsis and diarrhoea. I think
it not improbable that in advanced cases of saccharo-
butyric putrefaction B. aerogenes capsulatus invades the

1 There are instances in which excessive indulgence in carbohy-
drates is followed by mental depression and muscular prostration.
These symptoms appear to be simply a great exaggeration of
what is often noticed in healthy persons after similar indulgences.

INFECTIONS OF THE DIGESTIVE TRACT 301

entire extent of the small intestine and is enabled there
to decompose sugar with the formation of butyric and
other acids. Such an opportunity to grow upon a car-
bohydrate medium might lead to a considerable in-
crease in the number of anaerobes. These on passing
into the large intestine would find an opportunity to
set up their putrefactive activities on proteids after
the carbohydrates had been exhausted.

The consequences of chronic excessive saccharo-
butyric intestinal putrefaction are by no means limited
to disturbances of the digestive tract. The formation
of an excessive quantity of the higher fatty acids may
be presumed to lead in some instances to the absorp-
tion of an excessive quantity of these acids. As, how-
ever, the organism is able to oxidize the fatty acids
with great readiness, one would hardly expect that so
long as the oxidizing powers of the organism are un-
impaired there would be any evidence of an increased
excretion of organic acids in consequence of this ex-
cessive absorption. An experiment made at my sug-
gestion by Dr. A. J. Wakeman bears out this view. A
dog weighing about eighteen pounds was fed on a fixed
amount of meat, namely one kilo daily, in order to
study the influence of the feeding of ammonium buty-
rate upon the excretion of the nitrogen of ammonia.
During the preparatory period of nine days determina-
tions were made of the total nitrogen and of the nitrogen
of ammonia excreted by the urine. The total nitrogen
of ammonia excreted daily amounted to between one
hundred and one hundred and fifty-six milligrams

302 INFECTIONS OF THE DIGESTIVE TRACT

during this period of preparation. During the five
days on which the animal received daily four grams
of ammonium butyrate the urine collected failed to show
a distinct increase in the nitrogen of ammonia or in the
proportion between the nitrogen of ammonia and the
total nitrogen. Thus while one would hardly expect
that in normal organisms on a moderate diet of meat
the nitrogen of ammonia would be appreciably increased
through the excessive absorption of fatty acids which
are removed as compounds of ammonia, it is nevertheless
conceivable that in conditions of disease in which the
oxidizing powers of the organism are as a whole dimin-
ished (or taxed to the utmost) there might be an absorp-
tion of fatty acids in such excess as to somewhat increase
the output of the nitrogen of ammonia. In other
words, there would then be a slight degree of acid in-
toxication or acidosis resulting from the increased ab-
sorption of the volatile fatty acids. There are individual
instances in which we have found indications from the
urine that this is probably the case, but the evidence
is not entirely convincing. The subject of the rela-
tion between excessive saccharo-butyric putrefaction
and the excretion of an increased amount of nitrogen of
ammonia calls for further and more careful observations,
both experimental and clinical.

There is some evidence that the faecal extracts
obtained from persons with advanced chronic saccharo-
butyric putrefaction may exhibit a decided hsemolytic
action on the red blood cells of rabbits and monkeys,
and in some instances such an action is pronounced.

INFECTIONS OF THE DIGESTIVE TRACT 303

The methods that have been employed in the study
of the haemolytic action of the faecal extracts in such
cases are not wholly beyond criticism. Nevertheless
the indications seem sufficiently definite that there are
substances formed in the intestine (in extreme cases of
excessive saccharo-butyric putrefaction) that are capable
of haemolyzing more actively than is the case with
material derived from healthy persons on similar diet.
This haemolytic action is perhaps in part referable to
ammonium butyrate, but it has been found to persist
even after the removal of the ammonium salts of the
volatile fatty acids, and is probably due, in part at least,
to some unknown substance. The fact already mentioned
in speaking of the characters of B. aerogenes capsulatus,
that this organism is able to produce haemolytic sub-
stances, is worthy of notice at this point, since in many
people who have long suffered from this type of putre-
faction there are distinct evidences of anaemia.1 The
onset of anaemia in excessive saccharo-butyric putre-
faction is usually very slow. It is first manifested by
indications of a decreased volume of blood without any
decided fall in the percentage of haemoglobin or red blood
cells. After a time, however, there may be a distinct
fall in the haemoglobin and later in the red blood cells,
so that a moderate or considerable grade of secondary
anaemia may be associated with the intestinal condition.

1 1 have discussed at some length the relation of anaemia to
intestinal infections in a paper entitled : " On Bacterial Processes
in the Intestinal Tract in Some Cases of Advanced Ansemia, with
Especial Reference to Infection with B. aerogenes capsulatus (B.
Welchii)," Journ. of Biol Chem., ii, p. 1, 1906.

304 INFECTIONS OF THE DIGESTIVE TRACT

In some instances which I have observed this association
has been apparently a very close one ; that is to say, it
has appeared as if the anaemia in question were secon-
dary to the chronic intestinal process, since other causes
of anaemia, such as malaria, syphilis, intestinal worms
or parasites, or other infections, have been excluded
with a very high degree of probability. In some in-
stances the grade of anaemia is intense and the patients
may present the blood picture and clinical characters
of a progressive pernicious ansemia.

The occurrence of a considerable degree of ansemia
in any case of advanced saccharo-butyric putrefaction
must depend upon an excessive destruction of red blood
cells as compared with the reproduction of such cells.
Individuals are known to possess very different powers
of regeneration, and it is doubtless true that the absorp-
tion of equal quantities of hsemolytic poison by two
individuals might lead to very different results so far
as the blood is concerned, owing to differences in regen-
erating power. The fact that a patient is not anaemic
in the ordinary sense, that is, shows no reduction below
what is considered a normal percentage of haemoglobin
and no fall in the number of red blood cells, is by no means
a positive indication that there is not an excessive blood
destruction going on. In health the ability to regener-
ate blood cells is probably much in excess of the ordinary
requirements. This margin of excess may not be ex-
hausted for some years, notwithstanding a considerably
increased drain is being made upon the regenerative
structures through the excessive destruction of red blood

INFECTIONS OF THE DIGESTIVE TRACT 305

cells. Sooner or later a definite and increasing dispro-
portion arises between the destruction and reproduction
of red blood cells, and under these circumstances there
arises a slowly or more rapidly progressive anaemia.

In some instances which I have classed as falling into
the category of the simple saccharo-butyric type of putre-
faction, depressive mental symptoms have been noted.
In some of these cases an abundant growth of Gram-
positive diplococci or streptococci have been found in
the fermentation tubes. Whether these microorganisms
have any part in the production of neurotoxic substances
is uncertain. It has been mentioned that B. aerogenes
capsulatus is said to make poisons capable of acting on
the nervous system, but the bearing of this fact upon
human symptoms is still obscure. As a rule well marked
nervous symptoms have not been a feature in the class
of cases to which the term saccharo-butyric putrefaction
has been applied.

In the course of time, in well-marked instances of the
saccharo-butyric type of excessive chronic intestinal
putrefaction, there has been observed a considerable loss
in weight and a diminution in muscular powers. In
spite of a fairly good appetite patients are unable to gain
either in weight or in strength. Frequently the indica-
tions of premature senility are evident. This is notice-
able in an atrophy of the subcutaneous fat in premature
wrinkling of the skin. A loss of general vigor is apt to
be associated with a loss in sexual power.

306 INFECTIONS OF THE DIGESTIVE TRACT

III. THE COMBINED INDOLIC AND SACCHARO-BUTYRIC
TYPE OF CHRONIC EXCESSIVE INTESTINAL PUTRE-
FACTION

The association of the features of the two types of
putrefactive decomposition already described is common.
In cases presenting the features characteristic of both
types of putrefactive decomposition the two processes
often vary somewhat independently. It is in a measure
an arbitrary matter to determine what cases fall under
the combined type and what cases belong under the
indolic or saccharo-butyric type. I would include as
falling under the combined type of putrefaction all cases
in which putrefactive anaerobes of the intestine are
regularly very abundant in the faeces at the same time
that there is a persistent indicanuria of high grade — an
indicanuria not immediately and markedly influenced
by even the strictest precautions in regard to the dietary.
In the combined type of saccharo-butyric putrefaction
nervous symptoms are apt to be prominent relatively
early in the course of the affection. These symptoms
are excessive emotional irritability, an inclination to
mental depression, and the early onset of muscular and
mental fatigue. At a somewhat later period in the
course of pronounced cases, the haemic disturbances
become noticeable. The patients grow gradually more
and more anaemic. There may be periods of improve-
ment, both as regards the anaemia and the nervous
symptoms, but on the whole the tendency is a downward
one, and from year to year the patients become a little

INFECTIONS OF THE DIGESTIVE TRACT 307

less robust and less capable of rapid recuperation under
favorable hygienic conditions. After a time — com-
monly, however, not until the process has been a very
marked one for ten or fifteen years — the patients lose
so much strength as to be unable to attend to ordi-
nary business occupations without very great effort. A
period of carefully regulated living, free from anxieties,
may be followed by a considerable degree of improve-
ment, but this improvement is usually extremely slow.
In certain cases the anaemia ultimately deepens, and the
patient may present the picture of a progressive per-
nicious anaemia. In other instances it is especially
the nervous symptoms which increase, and periods of
depression become more and more marked and of longer
duration. The mental depression may become so pro-
nounced as to necessitate a residence of the patient in
a sanitarium or asylum, the conditions being those of
mild or pronounced melancholia. As already mentioned
when discussing the character of the toxic effects of the
various poisons absorbed from the intestine, it is highly
probable that different individuals react differently to
the same toxic agencies. If, for example, there be
absorbed from the intestine substances capable of
damaging both the nervous system and the red blood
cells, it is conceivable (assuming the same proportions
and amounts of these substances to be absorbed in each
instance) that one individual would become invalided
first through damage to the nervous system, whereas
the invalidism of another might come first through dam-
age to the blood. The condition of invalidism in either

308 INFECTIONS OF THE DIGESTIVE TRACT

case would lead to an enforced rest and perhaps also to
improved conditions of diet. The result is that while
the improved hygienic conditions may not lead to a
restoration of normal function, there is for a time a
cessation in the progress of the injurious action of the
intestinal poisons upon those structures that are most
susceptible. Even in cases where there is very little
or no improvement in respect to the especially suscep-
tible tissues, there is also a cessation of the progressive
damage done to the less susceptible cellular elements.
For example, a patient suffering from combined indolic
and saccharo-butyric excessive intestinal putrefaction
may succumb on the side of the nervous system and be
compelled to enter an asylum on account of mental
depression. Under the improved hygienic conditions
there may be only a little improvement in the nervous
manifestations, but there may nevertheless be a cessation
of the slowly progressive loss of balance between the
destruction and the formation of red blood cells. The
patient remains moderately ansemic, but does not grow
extremely so, because there has been a slight mitigation
of the absorption of substances injurious to the blood.
It is in fact true that high grades of anaemia do not
develop in patients who suffer from the combined indolic
and saccharo-butyric type of putrefaction and find their
way into an asylum on account of nervous symptoms.

In general, one may say of the combined indolic and
saccharo-butyric type of chronic excessive intestinal
putrefaction that the persons in whom these processes
are highly developed reach a state of chronic invalidism

INFECTIONS OF THE DIGESTIVE TRACT 309

more rapidly than where either condition is alone present.
Not only do the nervous symptoms come on earlier, but
there is a relatively early loss in weight and power.
The indications of premature senility maybe pronounced.
It is probable that atrophic processes are not limited
to the fatty, muscular, and blood tissues. It seems
probable that the cells of the liver and of the kidney
undergo premature degenerative atrophy in these cases.
This is doubtless the result of chronic parenchymatous
changes due to constant exposure of the cells to cyto-
lytic toxic agents present in very small amounts at any
one time but acting constantly. We are still far from a
full understanding of the effects of enterogenic poisons on
different types of protoplasm. One tends, for example,
to assume that the destruction of red blood cells, in
the course of chronic or acute infections of the intestine
with the gas-bacillus, is due to the direct hsemolytic
action of hsemolyzing agents formed in the intestine.
It is possible, nevertheless, that the destruction of red
cells is sometimes accomplished through a process of
phagocytosis which is made possible by injuries to red
cells inflicted by special opsonins.1

The derangements of the nervous system which have
been mentioned as concomitants of the combined in-
dolic and saccharo-butyric form of excessive intestinal
putrefaction — mental depression and muscle fatigue -

1 Sir A. E. Wright tells me that he has observed pronounced
phagocytosis of red cells in a case of pneumococcus infection
associated with a pernicious type of anaemia. The subject of
phagocytosis of red cells has lately been investigated by Hektoen
(Journ. of Infect. Dis., iii, p. 721, 1906).

310 INFECTIONS OF THE DIGESTIVE TRACT

constitute only a small part of the nervous manifesta-
tions of this pathological process. Almost every known
expression of functional disorder of the nervous system
is noted in some of the subjects of this variety of intestinal
putrefaction. But it would confuse rather than help
us to make a list of these derangements, including as
they do almost all the symptoms encountered in the pot-
pourri of neurasthenia . If we could distribute these
symptoms with confidence into those that are clearly
the result of intestinal intoxication and those that are
due to derangements of circulation or other mechanical or
dynamic departures from normal function, it would
repay us to attempt their classification. At present
such an attempt would be foolhardy and futile. Only
the most thoughtfully and carefully conducted clinical
and experimental studies can help us to determine the
influence of enterogenic poisons on the protoplasm of
nervous structures, and the opportunities for serious
studies of this kind do not now exist anywhere in the
world. Among the problems which must sooner or later
be approached by modern methods are the action of
enterogenic poisons on the motor structures of the central
nervous system, on the sensory paths (both peripheral
and central), on the sympathetic nervous system, and
on the peripheral and central mechanisms that subserve
the special senses. To what extent these various struc-
tures maybe damaged hi consequence of chronic excessive
intestinal putrefaction is uncertain. There are instances
of typical progressive muscular atrophy (due to lesions
of the ganglion cells of the anterior horns) in which there

INFECTIONS OF THE DIGESTIVE TRACT 311

are also pronounced evidences of excessive intestinal
putrefaction, but the relationship between the two
series of phenomena is not now definable. There are
cases of intestinal putrefaction in which B. aerogenes
capsulatus is the prominent microorganism of the di-
gestive tract and in which there occurs a slow muscular
atrophy associated with some fibrillation of the wasting
muscle bundles. Possibly in such instances the resistance
of the motor ganglion cells of the spinal cord is distinctly
below that of normal structures, thus rendering these
cells especially vulnerable to neurotoxic substances
made in the digestive tube. There are also cases of
multiple neuritis, resembling alcoholic neuritis, in which
alcohol can have no etiological part, but in which ante-
cedent gastro-enteric derangements are very prominent.
The probability that these instances of peripheral
neuritis (with the associated psychosis) are in reality
due to intoxications from enterogenic poisons appears to
me considerable, although the data now at my command
do not suffice to establish this view. There is at present
no evidence that infection by B. aerogenes capsulatus
is in itself capable of inducing this type of nervous
disease.

It is essential to realize that the onset of severe clinical
manifestations of excessive intestinal putrefaction may
not coincide with the period of most extreme develop-
ment of the gastro-enteric process. A short time after
therapeutical measures have been commenced the bac-
terial conditions in the tract may have been much
mitigated. It may thus, be unsafe to base a conclusion

312 INFECTIONS OF THE DIGESTIVE TRACT

regarding the etiology of a disease upon the bacterial
conditions that present themselves after the disease has
become established. I observed that it sometimes
happens in the course of a pernicious anaemia that
the numbers of B. aerogenes capsulatus in the faeces
undergo so great a decrease in the course of treatment
that the microscopical faecal fields present a nearly
normal appearance with respect to this organism.
Hence the failure to find the gas-bacillus in large excess
after a patient has for some weeks or months had the
benefit of treatment by diet and rest does not necessarily
exclude a preceding infection.

I believe the most intense period of the saccharo-
butyric process sometimes precedes the development
of a pernicious anaemia by a considerable length of time.
The explanation of this seemingly paradoxical view lies
in the fact that the normal defenses of the organism
must be greatly impaired before a pernicious type of
anaemia can arise. The bacterial process involved in
breaking down these defenses (which include the ability
of cells to bind, oxidize, and pair with enterogenic
poisons) are intensive and of long duration. After a
time there develop symptoms of weakness which make
it necessary for the subject of the gas-bacillus infection
to live with much greater regard for hygienic conditions;
for example, with more rest, more out-of-door life, and
a more restricted dietary. In consequence of these
changes the intensity of the putrefactive process is
mitigated. The quantity of injurious products absorbed,
including those that are haemolytic, may become so

INFECTIONS OF THE DIGESTIVE TRACT 313

moderate that little harm could be accomplished by them
in a person with normal defenses. But in a person with
greatly impaired defenses and diminished hsemopoi-
etic powers (correlative to parenchymatous changes
in the hepatic and other cells, on alteration in the bone
marrow) the absorbed poisons suffice ultimately to make
blood destruction preponderate permanently over blood
formation. A relatively slight putrefactive process in
the intestine thus ushers in the signs of the severe
anaemia.

METHODS RELATING TO THE MODIFICATION
AND CONTROL OF BACTERIAL PROCESSES
CONCERNED IN CHRONIC EXCESSIVE IN-
TESTINAL PUTREFACTION

THE difficulty of formulating measures for the control
of the bacterial processes concerned with the occurrence
of chronic excessive intestinal putrefaction is obvious.
No two cases that come under observation are quite
alike ; hence specific measures which appear appropriate
in one case cannot be recommended without modifica-
tion in another. It becomes, therefore, almost impossi-
ble to describe a method of treatment with all the
detailed modifications that are necessary in order to
cover adequately the specific cases with which one has
to deal in practice. Our store of experience is as yet so
incomplete that it is impossible to predict in many instances
the outcome of the measures proposed. This is equivalent
to the admission that the therapeutical side of the subject
of chronic intestinal putrefaction is., still ^in a largely
experimental stage. It will require many years of
careful observation associated with a large amount of
experimental work to enable the practitioner to place
his methods of treatment on a really scientific basis.
Nevertheless there are certain general therapeutical
guides which, though by no means adequate for the
direction of the practitioner in individual cases, are

314

INFECTIONS OF THE DIGESTIVE TRACT 315

helpful in forming a conception 'of the principles that
must enter into the treatment of chronic infections of
the digestive tract by putrefactive anaerobes. It must
be emphasized that the degree of success with which
these general principles of treatment are applied in
individual instances of disease will depend upon an
actual knowledge of individual peculiarities pertaining to
patients.

The slighter grades of chronic excessive intestinal
putrefaction are as a rule quickly improved by suitable
measures. It is helpful to think of the clinical mani-
festations in these cases as represented graphically by
a curved line running at a variable distance below a
straight horizontal line representing the separation of
physiological from pathological manifestations. A line
traced above this critical line indicates that a person
has no symptoms of diminished function ; and the greater
the distance between the two lines, the greater is the
reserve or margin of resisting power which has to be lost
before symptoms of derangement set in. A line running
below the critical line indicates that the defensive pro-
cesses of the organism are no longer adequate to pre-
vent clinical manifestations of disease. Where derange-
ments are slight it may be assumed that there will be a
prompt disappearance of symptoms as soon as there has
been a restoration of the balance between the powers of
resistance and the agencies underlying disordered func-
tion. The mere disappearance of symptoms, however,
does not mean that a condition of normal health has
been reestablished, for a state of robust health would

316 INFECTIONS OF THE DIGESTIVE TRACT

be indicated by a line running at a considerable rather
than a slight distance above the critical line. The
mere crossing of this line in an upward direction
connotes the disappearance of symptoms. But this
does not necessarily mean health, because there may be
a margin so small that it is quickly exhausted by a detri-
mental influence. On the other hand, very pronounced
symptoms may exist (especially hi cases of short dura-
tion) without these symptoms necessarily possessing a
grave significance. A slight improvement in physiologi-
cal conditions is sufficient to make the line indicative of
derangement in function cross the critical line of demar-
cation between physiological and pathological conditions.
This conception of the part played by symptoms as an indi-
cation of actual states of functional activity is helpful in
explaining why some cases are so much more refractory
than others. In long standing conditions, in which a high
grade of intestinal putrefaction has for some time been
present, improvement is extremely slow even under the
most favorable hygienic conditions. There is a dis-
couraging persistence of symptoms until the critical line
is reached. The depression of function is, however, so
considerable that there may be a real gain in function
without an immediate and corresponding subjective im-
provement. In such cases the patient seldom accumu-
lates an adequate functional reserve even after the
disappearance of all obtrusive symptoms. The line
representing his physiological condition must be traced
close above the critical line, and a slight depression in
function is promptly 'followed by a return of symptoms.

INFECTIONS OF THE DIGESTIVE TRACT 317

Thus it is that in these severe types of intestinal de-
rangement relapses are common and discouragements
frequent, even among persons who from time to time
are quite free from definite symptoms of disease.

The principles entering into the management of cases
of chronic excessive intestinal putrefaction may be
divided into three groups. First, those that relate to
the avoidance of putrefactive bacterial contamination of
the food; second, those that relate to the promotion
of prompt digestion in and absorption from the small
intestine; third, those that relate to agencies designed
to reduce the numbers of putrefactive anaerobes living
in the intestinal tract. It seems hardly practicable to
separate here the indolic and the saccharo-butyric types
of decomposition from the standpoint of the practical
measures to be employed in modifying them, especially
as the subject of indolic putrefaction has been already
considered in part.

The Avoidance of Putrefactive Contamination of the
Food. — As regards the avoidance of bacterial con-
tamination of the food with injurious organisms it is
evident that care in regard to ordinary principles of
cleanliness must be of importance. In a large propor-
tion of cases of chronic excessive intestinal putrefaction
the hydrochloric acid of the gastric juice is secreted in
diminished amount and frequently no free hydrochloric
acid can be detected in the stomach. Often there
is also a moderate degree of dilatation of the stomach
or at least some atony. This results in a delay in the
emptying of the stomach, which in the absence of an

318 INFECTIONS OF THE DIGESTIVE TRACT

adequate secretion of hydrochloric acid is favorable to
fermentative decompositions and even to the beginning
of putrefaction, if putrefactive bacteria are abundantly
present. It is clear that in such cases the greater the
freedom of the food from putrefactive bacteria, the less
will be the liability to putrefaction at lower levels, since
the stomach in the condition under consideration is
unable to extensively destroy such bacteria, even in
their vegetative forms. In the present connection I
lay especial stress on the avoidance of putrefactive
bacteria, but the introduction of pathogenic organisms
related to the colon type, including, for example, organ-
isms capable of inducing dysentery, should also be
avoided, since any injury to the mucous membrane of the
tract tends to impair the physiological inhibitions to pu-
trefactive decomposition. The use of cooked food goes a
long way toward securing conditions of sterility within
the stomach. It is, however, usually impracticable
to live exclusively on food that is sterile. The use of
milk, for example, means the introduction of very large
numbers of bacteria. So long as these bacteria are
lactic-acid formers, no harm is done, but in many
samples of milk there are putrefactive organisms, and
these may be present in large numbers. This is true
of B. putrificus and of an organism closely resembling
B. aerogenes capsulatus, but differing ordinarily from it in
being less pathogenic for experimental animals. Cheese
is another article of food which commonly contains large
numbers of putrefactive bacteria. In normal stomachs
these putrefactive bacteria are quickly disposed of ; in

INFECTIONS OF THE DIGESTIVE TRACT 319

cases of chronic excessive intestinal putrefaction the gas-
tric conditions are usually such as to make the stomach
incapable of adequately coping with the putrefactive
bacteria introduced with the cheese. Cheese is there-
fore an article which should generally be excluded from
the dietary of nearly all cases of excessive chronic
intestinal putrefaction, although some forms of cheese
(especially those which are newly made from milk con-
taining few putrefactive bacteria) may be permitted in
moderate quantities. In eating uncooked fruit it is
desirable to guard against the organisms found on
the skin. Fruits therefore should be always peeled.
Bacillus putrificus and the bacillus of malignant oedema
are often found on the surface of fruits, and Dr. Rettger
has found the latter abundant on the skin of bananas.
The question of the sterilization of milk may be
briefly touched upon in its relation to the dietary of
persons suffering from chronic excessive putrefaction.
So far as the treatment of chronic putrefactive disorders
is concerned, I think it preferable to avoid sterilization
except where the milk contains considerable numbers
of putrefactive bacteria. One can form some judgment
of this point by permitting samples of milk to stand
in fermentation tubes in the incubator after subject-
ing them to a single partial sterilization by the process
of pasteurization. Under these conditions lactic-acid
organisms which restrain putrefactive decomposition are
killed ; the surviving spore-bearing putrefactive organ-
isms of the gas-bacillus type reveal their presence by
setting up a stormy fermentation of the milk. Effective

320 INFECTIONS OF THE DIGESTIVE TRACT

sterilization involves the destruction of the spores of the
putrefactive anaerobes and can only be accomplished
with certainty by the use of the autoclave. It is evi-
dent from what has been just said, that the partial
sterilization of milk containing putrefactive organisms
in abundance is not a desirable procedure, since it
destroys the lactic-acid bacteria which are normally
present and which exert an antiputrefactive action. It
is safer for patients to drink milk containing lactic-acid
organisms together with putrefactive ones than to drink
the latter alone, unless these are very few in number.

An important measure bearing upon the introduction
of putrefactive bacteria into the intestine is the proper
cleansing of the mouth. Putrefactive bacteria such as
putrificus are not uncommon in the mouth, especially
where there is dental caries. They find in cavities
sufficiently anaerobic conditions for their development.
Whether other forms of putrefactive organisms occur in
the mouth in cases of chronic excessive putrefaction
in the intestine is not clear. I have never certainly
found B. aerogenes capsulatus. The growth of this
organism and of allied strict anaerobes could easily
be prevented or minimized by careful cleansing of the
teeth and especially by the avoidance of the lodging of
food between the teeth. The maintenance of the mouth
in a relatively cleanly condition through the intelligent
and free use of the toothbrush and through the employ-
ment of an oxidizing tooth paste, such as Pebeco,1

^his is a paste containing a high percentage of potassium
chlorate.

INFECTIONS OF THE DIGESTIVE TRACT 321

is a material aid in restricting putrefactive conditions in
the upper part of the digestive tract. These measures
are of much more service if supplemented by the process
of flossing the teeth. It is, however, easy to exaggerate
the importance of the oral conditions especially hi the
direction of assuming that a really cleanly mouth will
insure a disappearance of excessive putrefaction in the
intestine. While it is evident that to swallow large
numbers of putrefactive anaerobes which have grown
in the mouth (in cases where the stomach is unable to
destroy these organisms) must tend to increase the
intensity of the putrefactive processes in the intestine,
it does not follow that a patient will promptly convalesce
if this additional increment of bacteria be removed.
Some improvement is perhaps to be expected through
the cleansing of a very offensive mouth, but the per-
manent eradication of the intestinal conditions (when
this is possible) depends on a number of factors and not
upon one. There are cases in which an oral butyric
putrefaction occurs very rapidly after a meal. In such
cases food which is lodged in the teeth will hi the course
of an hour develop an intense odor of butyric acid. It
is not yet certain upon which organisms this putrefaction
depends. It is not impossible that the decomposition
may be due to aerobic producers of butyric acid, but
it seems probable that anaerobes are at least to some
extent concerned in many instances.

The suitable care of the stomach may also be a factor
in controlling putrefactive decompositions in the in-
testines. In persons suffering from atony of the stomach

322 INFECTIONS OF THE DIGESTIVE TRACT

(with or without pronounced dilatation) putrefactive
microorganisms may gain a hold and initiate high up in
the digestive tract a process which normally begins
only in the region of the lower ileum. All measures that
act in the direction of stimulating the stomach to nor-
mal motility must favor the mechanical removal of the
bacteria concerned in putrefaction. But hygienic meas-
ures and drugs may not suffice to free the stomach from
putrefactive organisms. In such cases the practice of
lavage may be very helpful and may give quick relief
from some symptoms of intoxication, especially head-
ache. Even in persons in whom dilatation is not marked
the use of lavage before breakfast may be a very helpful
measure, and the selection of this time has the advantage
of not depriving the patient of undigested food.

The Promotion of Prompt Digestion and Absorption in
the Small Intestine. — Generally speaking it is safe to
say that all those measures which aid in securing prompt
digestion in and absorption from the small intestine will
operate to diminish intestinal putrefaction, since the
outcome of prompt absorption in the region of the small
intestine is equivalent to offering a diminished oppor-
tunity for the attack of putrefactive anaerobes located
in the ileum and large intestine. Probably the restora-
tion of normal gastric secretion and of pancreatic secre-
tion is the most important physiological agency in
securing rapid digestion and absorption of proteids
from the small intestine. An improvement in the
quantity of gastric juice secreted and in the quantity
of the hydrochloric acid which it contains usually goes

INFECTIONS OF THE DIGESTIVE TRACT 323

hand in hand with the betterment of motility. The
influence of imperfect pancreatic secretion may safely
be assumed to be important, but in this case direct
clinical observation is not possible, and one has to infer
the absence or presence of a pancreatic achylia from the
manner in which the proteid food in a test meal is at-
tacked by the digestive juices. The use of small cheese-
cloth bags containing meat in a state of moderate sub-
division may be helpful in measuring the ability of the
intestinal enzymes to carry on the digestion of meat
proteids.1 The use of this method as an index of pan-
creatic digestion can hardly be regarded as free from
error.

The most important physiological factor in the partial
or complete restoration of pancreatic secretion is prob-
ably rest. The influence of fatigue upon the digestive
processes is often very clearly demonstrated in a given
individual, and the removal of opportunity for fatigue
is conversely a factor in determining recovery. The
rest which should be secured in such cases is in part
physical, in part mental and emotional and sexual.
Many patients commit the error of exercising freely when
rest is far more important. The effect of mental, emo-
tional, and sexual fatigue upon the digestive tract has
not been so carefully studied as it deserves, but there is
on doubt that in each case the expenditure of nervous
energy may lead to very similar results as regards the
digestive organs. Rest from work of a taxing character
may be an extremely important factor in the restoration

1 This procedure was lately suggested by A. Schmidt.

324 INFECTIONS OF THE DIGESTIVE TRACT

of normal gastro-enteric secretions. In many persons
emotional fatigue, such as often results from strenuous
business activity or from some kinds of recreation, has a
distinctly harmful influence. Even the interest attend-
ant on free and animated conversation may be in-
jurious. In the case of fatigue of sexual origin great
individual differences in susceptibility are manifested
by different persons. In all cases of long-standing
chronic putrefactive disorder in the intestine it becomes
important to take cognizance of any sexual factor that
may lead to fatigue. This relationship is one of the
utmost importance and is frequently overlooked by
physicians in their regulation of the lives of patients
suffering from chronic digestive disorders. It is a
very important point in this connection to realize that an
intelligent abstinence from sexual excitement should be
practiced before such abstinence is forced by the onset
of impotence. When this has occurred, the conditions
are probably much less favorable for the restoration of
normal secretory activity in the stomach and intestine.
It is customary in cases where the free hydrochloric
acid of the gastric juice is regularly diminished to ad-
vise the use of hydrochloric acid after meals in order to
compensate the diminished secretion of the acid.1 In
some instances the use of hydrochloric acid is distinctly

1 The absence of free hydrochloric acid in the gastric juice is
a feature of nearly all cases of advanced saccharo-butyric putre-
faction. The influence of gastric secretion has recently been studied
and discussed by D. van Tabora in its relation to intestinal pu-
trefaction. "Ueber die Beziehungen zwischen Magensaftsecretion
und Darmfaulniss," Deutsche Archiv f. klin. Med., Ixxxvii, p. 254,
1906.

INFECTIONS OF THE DIGESTIVE TRACT 325

beneficial if one may judge by the feelings of the patient
and by the diminution in the excretion of indican which
is noted. Among the long-standing cases of chronic
excessive intestinal putrefaction, especially in cases of
the combined indolic and saccharo-butyric type, the use
of hydrochloric acid is often attended by little improve-
ment. The stomach, indeed, often irritable or unduly
sensitive, may not accept even moderate quantities of
the acid without showing indications of slight gastritis
or local discomfort. The use of digestive enzymes such
as pepsin, trypsin, pancreatin, etc., has been much lauded.
It is possible that they do good in some instances, but
I have never been convinced that their use is really an
important factor in determining the ultimate outcome
in marked cases of chronic excessive intestinal putrefac-
tion. The use of an efficient diastatic ferment gives, I
think, more tangible results, and there are cases in which
a better utilization of carbohydrates is noticeable in con-
sequence of the use of an active diastase. The criterion
in such cases has been the better tolerance for carbohy-
drates which is manifested by the reduction in habitual
flatulence, especially in the upper part of the digestive
tract. I know of no observations claiming to demon-
strate a direct influence either of diastatic or proteoclastic
ferments upon the excretion of the ethereal sulphates or
indican or upon the numbers of putrefactive bacteria
in the large intestine.

Careful mastication of the food is certainly an important
factor in determining how it shall be utilized. But one
may readily exaggerate the influence of this factor, and

326 INFECTIONS OF THE DIGESTIVE TRACT

this has, I think, been done in some recent researches.
There is, however, no doubt that in cases where there is
a failure of the stomach to secrete free hydrocholric acid
the careful prolonged mastication of the food much
diminishes the opportunity for stagnation in the stomach,
while it subsequently makes it less likely that undi-
gested proteid will appear in the large intestine to be
attacked there by putrefactive anaerobes. The fine
subdivision of the food before it is eaten is often very
helpful, especially in persons whose teeth do not permit
full mastication. I question, however, whether the com-
minution of meat outside the body is a real substitute
for its equally fine subdivision through the mastication
by the patient, since in the latter case there is the ad-
vantage of an ultimate admixture with the enzymes
of the saliva. In most cases benefit is derived both
from a fine subdivision of meats outside of the body
and from their further subdivision through prolonged
mastication. The tolerance for meats as measured by
clinical phenomena is often greatly increased by insisting
that the patient give due attention to mastication.

The avoidance of chemical and mechanical irritants
assumes much importance in some instances of chronic
excessive intestinal putrefaction. Among the chemical
irritants may be mentioned the use of condiments such
as pepper and mustard, and the use of free acids such as
vinegar and lemon juice, and indulgence in an excess of
salt, this being an error often committed by neurasthenic
patients suffering from putrefactive disorders. The ex-
cessive use of salt is stated by some investigators to be

INFECTIONS OF THE DIGESTIVE TRACT 327

operative in bringing about the removal of too much
alkali from the body, thus favoring a degree of acidosis.
Among mechanical factors one has to consider mainly
the consistence of food. The use of raw apples, of un-
cooked celery, and of spinach, and the use of salads may
be detrimental for purely mechanical reasons. Some-
times the detrimental effect shows itself in the produc-
tion of colic, in other cases it is manifested by an undue re-
tention of the food within the stomach, where it excites
discomfort. It has been already stated that there are
long-standing cases of excessive intestinal putrefaction
in which there is evidence of an abnormal sensitiveness
of the partially denuded mucous membrane. In cases
where the stomach is sensitive even to the presence of
moderate quantities of food that has been carefully
masticated, it is desirable to use a demulcent (such as
flaxseed tea) before meals. This is often a highly effi-
cient measure in allaying gastric tenderness. It may be
desirable to give a demulcent drink before each meal
during a long period of time.

It is obvious that in order to secure the best possible
conditions of resorption from the small intestine it is
important not to permit the consumption of excessive
quantities of food. This is especially important in the
case of proteids, particularly those of meat and milk.
The use of an excessive quantity of meat frequently
goes hand in hand with imperfect mastication. The
result is that many masses of muscle fiber find their way
through the small intestine into the lower ileum and
large intestine, where they are attacked by putrefac-

328 INFECTIONS OF THE DIGESTIVE TRACT

tive bacteria. The putrefactive bacteria find in meat
proteid and casein good media for their support.

The occurrence of large numbers of fragments of meat
or smaller aggregations of meat fibers in the fa3ces is
an indication of the imperfect utilization of meat and
should operate as a signal for its more cautious and
intelligent use. It has already been mentioned that
the intestinal contents of carnivora contain many more
putrefactive, spore-bearing bacteria than is the case
with the herbivora. The full meaning of this observa-
tion is not at the present time clear. As certain an-
aerobes, especially the gas-bacilli, are closely associated
with the development of some putrefactive disturbances
of high intensity, it seems worth while to investigate
more carefully than has yet been done the influence of
meat upon the growth of the putrefactive anaerobes.
Until further knowledge has been obtained in regard to
the influence of meat upon the development of putre-
factive anaerobes, in cases of well-marked saccharo-
butyric putrefaction, it seems rational to advise the
practice of moderation in the use of meats in the dietary
of such patients. In practice I restrict the use of meat
to one meal each day and keep its amount moderate.
There may even be conditions in which there is something
to be gained by eliminating meat entirely from the die-
tary for short periods of time. The desirability of this
more extreme measure must be judged partly by the
clinical effects. It may be found that in some persons
who have long been accustomed to the use of large
quantities of meat in the dietary, a complete deprivation

INFECTIONS OF THE DIGESTIVE TRACT 329

would be followed by loss of strength, and this would
hardly be compensated for by a slight gain in the direc-
tion of a diminution of the anaerobes in the large
intestine.

It is often a difficult matter to determine how much
food it is safe to permit a patient with chronic excessive
saccharo-butyric putrefaction. If the patient be weak,
anaemic, and emaciated, or muscularly atrophic, there
is often a strong temptation to feed generously or indeed
excessively. I consider it bad practice to urge the use
of large amounts of food at the beginning of a course of
treatment, since this must result in at least some exag-
geration of the putrefactive process. But if rest can be
secured to the patient and a more thorough digestion
and resorption of food can be achieved, it becomes safe
to increase the food gradually. It is much better to
devote a few weeks to this elevation of the digestive
functions than to proceed hastily to inaugurate a gener-
ous regimen. Even under the most favorable conditions
the effect of the diet must be carefully watched. A dis-
tinct increase in the ethereal sulphates, an increase in
the numbers of gas-bacilli in the movements, an increase
of intestinal flatulence or the development of emotional
irritability or mental depression are signs that the food
should be decreased in quantity. Intestinal flatulence
calls for a diminution in carbohydrate food.

Methods designed to reduce the Numbers of Putre-
factive Anaerobes. — Among the agents most used with
a view to reducing fermentative and putrefactive de-
composition in the gastro-enteric tract are the so-called

330 INFECTIONS OF THE DIGESTIVE TRACT

intestinal antiseptics. A considerable literature exists
in regard to them. Opinion as to their efficacy is di-
vergent. I shall not undertake to discuss here the
literature bearing on the subject, for probably all of the
experiments that have been conducted with a view to
determining the value of intestinal antiseptics have
left out of account a quantitative study of the putre-
factive anaerobes of the f seces — a study which obviously
is beset with great difficulties, and could only be under-
taken under especially advantageous conditions for
work.1 In some of the experiments upon this subject
observations have been made on the excretion of the
ethereal sulphates, the indican, and the phenolic bodies
of the urine. These unquestionably possess a certain
value, and if the observations be sufficiently long con-
tinued and properly controlled, may be regarded as
conclusive indications as to the intensity of putrefaction.
One difficulty in measuring the action of many of the
antiseptic substances which have been used depends
on the fact that they either decompose with the libera-
tion of aromatic substances capable of pairing with
sulphuric acid and forming ethereal sulphates, or they
themselves are aromatic bodies capable of pairing in this
manner. A study of the indican is not open to this
objection, but as already pointed out there are cases of
intestinal decomposition in which the ethereal sulphates
may be considerably above the normal, although little
or no indican is excreted. Again, in such observations

1 Much of the modern literature relative to this subject is given
by von Tabora, loc. cit.

INFECTIONS OF THE DIGESTIVE TRACT 331

it is most important that the quantity of proteid food
(especially of meat) should be regulated in an exact
manner, since an increase of putrefaction in the intestine
through the use of excessive quantities of meat can easily
be demonstrated.

I have made some observations on the action of so-
called intestinal antiseptics, and I have reached the con-
clusion that most of them do very little good in effecting
a dkninution of the putrefactive anaerobes of the in-
testine. Their ability to control fermentative processes
in the stomach cannot, I think, be questioned, and in
cases where such processes are excessive they may
indirectly do good by diminishing the opportunities for
putrefaction in the intestine by placing obstacles in the
way of the development of the sugar-splitting gas-form-
ing anaerobes. It cannot be denied that through the
use of glutoid or other protective capsules antiseptics
can be made to reach the intestine itself before they are
liberated. The tendency of such antiseptics when used
in practical doses to enter into combination with many
substances outside the bodies of the living bacteria
must be regarded as greatly diminishing their ideal
efficiency, estimated on the supposition that they act
simply to kill or arrest the growth of living microorgan-
isms. I have found in certain instances that salicylates,
aspirin, and salol have exerted some action in diminish-
ing the output of indican, but beyond this I have not
been able to satisfy myself that the effect of intestinal
antiseptics is pronounced. In many cases they are not
well tolerated by the stomach after they have been used

332 INFECTIONS OF THE DIGESTIVE TRACT

for a time, and may thus do harm to the structures on
which normal secretion depends. I do not feel that the
subject of intestinal antisepsis has been developed in a
scientific manner, and very carefully planned experi-
ments may yet teach us that certain substances possess
considerable value when used under appropriate con-
ditions. At present antiseptics are used in an empirical
and usually unintelligent way. It seems to me on theo-
retical grounds that agents which exert an oxidizing
action in the large intestine are among those worthy of
careful study. Some of the oxides of manganese liberate
their oxygen slowly and would appear on this account
to be suitable for studies of this kind. There is, indeed,
some evidence that the dioxide of manganese may act
beneficially upon putrefactive processes in the large
intestine, but a scientific demonstration of the fact is,
I believe, lacking.

Experiments made by Dr. Wakeman at my suggestion
in reference to the action of oxidizing substances upon
the contents of the large intestine did not give results
encouraging to the view that such agents are likely to
be in a high degree efficient in mitigating the activities
of putrefactive anaerobes. He experimented with dogs
showing a high degree of indicanuria and injected into
the large intestine at various points a solution of hydro-
gen peroxide. In other experiments ferric sulphate
was used as a catalyzer in recognition of the fact that
the presence of this salt in small quantities renders cer-
tain oxidations of organic material much more efficient.
Different percentages, ranging from one-half of one per

INFECTIONS OF THE DIGESTIVE TRACT 333

cent, to about two per cent, of hydrogen peroxide, were
employed in different experiments. The solution of
hydrogen peroxide was slowly introduced and in large
volume, half a liter or more being injected in the course
of an experiment extending over many hours. In these
experiments no distinct decline in the indican was
observed. It is possible, of course, that different re-
sults may be obtainable with other oxidizing agents or
with oxidizing agents brought into the intestine in a
different way. The results might have been more con-
clusive had the peroxide been introduced into the ileum.
It is well known to clinicians that the use of laxatives
is followed by great temporary benefit in many cases
of excessive intestinal putrefaction associated with con-
stipation, sometimes even if this is not pronounced. Some
advise the use of salts, others of cascara preparations,
and still others are partial to the employment of calomel.
That there is an actual reduction in the excretion of
ethereal sulphates and of various individual products of
putrefaction after the use of a laxative is certain. The
explanation is of course very simple. The reduc-
tion depends on diminished absorption from the intesti-
nal tract owing to the acceleration of the contents of
the digestive tract. A similar phenomenon is witnessed
in cases of even severe diabetes in which the sugar may
drop very much during the period of action of a dose of
calomel, or some other cathartic. Notwithstanding the
immediate benefit to be obtained from cathartics I be-
lieve they must be employed with much discrimination
and caution. It is easy to render the digestive tract

334 INFECTIONS OF THE DIGESTIVE TRACT

excessively irritable through their use, and it often hap-
pens that patients fail in nutrition owing to the diminished
absorption of foodstuffs. The long-continued and fre-
quent use of cathartic remedies has in my experience
nearly always resulted badly. I think laxatives should
be employed mainly for the control of the disturbances
of a subacute or acute character arising hi the course of
chronic derangements rather than for the treatment of
the chronic conditions themselves. If it is possible to
use food containing an abundance of cellulose, this may
have a beneficial effect in the direction of preventing
constipation, but it frequently happens that patients
who suffer from constipation are also excessively sensi-
tive to those foods containing much cellulose and show
gastric disturbances from their use.

The idea that pathological bacterial processes in the
intestinal tract may be favorably modified in a clinical
sense by making use of the antagonistic properties of
certain harmless microorganisms is by no means a new
one. It has long seemed probable that the well-known
beneficial action of fermented milks such as Kefir and
Kumyss is in part due to the restraining action exerted
by microorganisms contained in these drinks upon the
harmful bacteria in the digestive tract. The action has
been attributed in these cases largely to the lactic-acid-
forming bacteria concerned in the fermentation. The
fermented milk known as Matzoon is a similar preparation.
Lately there has been placed on the market a fermented
milk which has been named Badllac. The chief or-
ganism in this fermented milk is a lactic-acid producer

INFECTIONS OF THE DIGESTIVE TRACT 335

which has been studied by Professor Metchnikoff.1 It is
capable of inducing a high grade of acidity in milk, but
does not grow readily on other media. This lactic-
acid producer is said by Professor Metchnikoff to be
entirely harmless and has been used by him and his as-
sociates as a means of lowering the putrefactive processes
in the digestive tract. One feature of superiority claimed
for Bacillac is that it does not contain any yeasts. This
is a point of some importance, since it is probable that
some of the fermented milks do contain yeasts capable
of exerting a detrimental action on the human organism.
Whether the Bacillac preparation will show itself to be
superior in a therapeutic way to other fermented milks
from which undesirable yeasts have been excluded
must be regarded as an open question and one that can
be decided only by extensive and very careful ex-
perimentation. There is certainly much to recom-
mend the view that putrefactive intestinal processes
are favorably influenced in their clinical course by
fermented milk containing an abundance of lactic
acid and lactic-acid organisms. The subject is one,
however, which is very far from resting on a scientific
basis, and a large amount of most careful research is
necessary to establish the exact position of this prin-

1 Dr. Collins found the organism to be Gram-positive, similar
to B. aerogenes capsulatus in morphology, but larger and incapable
of making indol or hydrogen sulphide. The presence and growth
of this organism in milk flasks inoculated with human faeces con-
taining an excessive number of gas-bacilli appeared to exert a
restraining effect on putrefaction. The maintenance of an acid
reaction is essential to this restraint.

336 INFECTIONS OF THE DIGESTIVE TRACT

ciple of treatment. It is important to determine the
influence exerted by the lactic-acid bacteria themselves.
It is by no means certain that most of the beneficial
influence exerted by the fermented milks is not depend-
ent on the lactic acid existing preformed in the milk.
Whether the lactic-acid bacilli in a thoroughly fer-
mented milk go on producing lactic acid in the digestive
tract is uncertain and probably depends on the presence
or absence of fermentable carbohydrates. If it can be
shown that any type of lactic-acid bacilli is capable of
carrying on its activities in the ileum and large intestine
under anaerobic conditions, with a restraining action
upon anaerobic putrefactive spore bearers (such as B.
aerogenes capsulatus), an important contribution will
have been made to rational therapeutics.1

Another therapeutic method which has been syste-
matically employed in recent years in many cases of
intestinal disease is irrigation of the colon. It is even
possible to introduce a tube through the sigmoid flexure
into the descending colon, and in some instances the
tube may be passed still further in the direction of the
ileocaecal valve. By means of physiological salt solu-
tionTne large intestine has in many cases been washed
free of its contents. This procedure has now been
frequently practiced with extremely beneficial results
in the direction of diminishing excessive intestinal pu-
trefaction. I have known it in several instances to be
followed by a reduction in the output of ethereal

1 The bacillus of Bacillac grows well in milk under anaerobic
conditions.

INFECTIONS OF THE DIGESTIVE TRACT 337

sulphates and in the excretion of indican by the urine.
Coincidentally there has been definite relief from pro-
nounced symptoms such as headache, mental depression,
irritability, intestinal flatulence, etc. In several instances
of advanced ansemia (including some showing the blood
changes of the pernicious form) there has been a very
rapid improvement in the haemoglobin and red blood
cells, together with a prompt disappearance of megalo-
blastic and other pathological forms of red blood cells.
There is no room for doubt that the intelligent practice
of high intestinal lavage leads to considerable improve-
ment in many cases of chronic saccharo-butyric putre-
faction. The beneficial effects have been pronounced
and prompt in some cases where an extreme indolic
putrefaction was associated with the saccharo-butyric
type. If the lavage be practiced cautiously with respect
to the avoidance of an excessive volume of fluid, the
method may be applied for a considerable period of time
without discernible harmful effects. It is perhaps better
to employ lavage two or three times in the week than to
use it each day. Where excessive quantities of fluid
are used, distension of the colon occurs and may be
followed by troublesome atony.

In some persons who are victims of a chronic infection
of the intestine with putrefactive anaerobes the intelli-
gent application of the foregoing measures suffices to
bring about a high grade of improvement. The chances
of obtaining a rapid amelioration of symptoms are
greatest in those persons in whom the process has not
been one of long standing. The factor of duration of the

338 INFECTIONS OF THE DIGESTIVE TRACT

infection appears to me fully as important as its intensity.
It is doubtless true that many patients suffering from
neurasthenic symptoms with mental depression or
moderate anaemia, have been benefited by ordinary
therapeutic measures suggested by experience and com-
mon sense, but not consciously directed by the physician
toward the control of the putrefactive conditions in the
intestinal tract. The recoveries in these cases have often
been temporary in character, for, as already mentioned,
there is a strong tendency for relapses to occur. Con-
siderable improvement may follow a second course of
treatment, and a third or fourth, but after a time it is
noticed both by the patient and his physician that
relapses occur more readily than formerly; i.e. in re-
sponse to apparently more trivial errors of living. It
is also noticeable that although the correction of these
errors in living is still followed by improvement, the
improvement is not so rapid and the periods of rest and
recreation and careful living are less beneficial than
was formerly the case. This may lead to discourage-
ment. It is certainly true that the physician must
be extremely cautious about venturing a prognosis in
those cases of chronic excessive intestinal putre-
faction that are not merely of long standing but have
had the benefit of intelligent medical treatment and
long periods of exemption from depressing conditions
of living. The prognosis is best in those persons whose
symptoms have not only been of moderate duration but
who have had little opportunity to care for themselves —
persons who have continued actively at work and have

INFECTIONS OF THE DIGESTIVE TRACT 339

committed gross errors in diet, in sexual life, etc. The
removal of the obviously injurious conditions in these
cases is almost always followed by a quick and satis-
factory improvement, provided the patient be not
burdened by significant neurotic taints or has not
developed a considerable degree of ansemia.1 In persons
who have had the benefit of much rest and recreation
and have cared for themselves under luxurious conditions
of living, the outlook is much less promising because in
many of these cases most of the ordinary therapeutic
measures have already been exhausted. It should be
remembered, however, that even in such cases a period
of rest of one or two years under favorable conditions
may bring about a great change in physical state, while
short periods of rest show very little effect on the patient.
For those who find it very difficult to follow in a rational
way the advice of a physician, owing to inability to exert
the necessary self-control, a period of residence in a good
sanitarium under the immediate supervision of a tactful
and intelligent practitioner may do much more good
than home treatment. Similarly a cure at Carlsbad
under rigid conditions of diet and general hygiene may
be of very distinct benefit in reducing the manifestations
of a chronic anaerobic infection of the intestine.

In extreme cases of chronic excessive intestinal
putrefaction in which nearly everything has been tried

1 1 regard the presence of a high degree of ansemia and the per-
sistence of a high grade of infection with the gas-bacillus as bad
prognostic signs, especially if associated with a persistently intense
dimethylamidobenzaldehyde reaction of the urine.

340 INFECTIONS OF THE DIGESTIVE TRACT

that can be suggested by competent medical men, the
question of surgical interference may arise. The past
ten years have recorded a very large number of surgical
triumphs in connection with the human digestive tract.
Unfortunately in a majority of the instances of disease
successfully relieved by surgical methods little or nothing
is known of the bacterial conditions in the digestive
tract and what modifications they may have undergone
as a result of treatment. There can be no doubt, how-
ever, that in some of the cases that have been success-
fully subjected to operation a part at least of the im-
provement has come about through the mitigation of
putrefactive processes in the large intestine. It is, I think,
extremely desirable that much closer attention should
be given to the study of cases from the bacterial stand-
point than has ever been the case, for the knowledge
which must surely be accumulated in this way must
prove extremely helpful in deciding what shall be done
with certain patients suffering from chronic anaerobic
infections of a severe type. An operation which has been
very often practiced in one or another form is that of
gastro-enterostomy. In cases of dilatation of the stomach
which have resisted ordinary forms of treatment, ad-
mirable results have many times been obtained in the
relief of a great variety of digestive disturbances. It
seems highly probable that the relief afforded in these
cases has been due in most instances in part at least to the
greatly improved conditions of bacterial activity which
followed the removal of a source of stagnation and
putrefaction in the upper part of the digestive tract.

INFECTIONS OF THE DIGESTIVE TRACT 341

The pyloroplastic operations have also frequently been
followed by striking benefit in chronic conditions.
Even in cases where an operation is done, not so much
for dilatation of the stomach as for a spastic or hyper-
trophic condition at the pylorus, the improvement may
have been associated with a diminished excretion of the
ethereal sulphates. I recall one case in which the
ethereal sulphates had regularly run far in excess of
the normal but in which this excess promptly disap-
peared after a pyloroplastic operation done for the relief
of a gradually increasing stenosis of the stomach as-
sociated with hyperchlorhydria and periodical seizures
of vomiting. Here the relief from putrefactive decom-
positions was probably merely one incident of ameliora-
tion of more acute and pressing conditions. Looking at
the pyloroplastic and similar operations from the stand-
point of chronic putrefactive intestinal disorders, one
has to admit that the indications for such operations are
at present not entirely clear.

The operations to which reference has just been made
are designed to secure the prompt emptying of the
stomach into the small intestine. An entirely different
type of operation and one which has as yet been seldom
performed relates to the attempt to exclude a portion of
the large intestine. The fact that the large intestine
is the seat of some of the most distressing derangements
associated with putrefaction strongly suggests the pos-
sibility that these putrefactive disorders maybe mitigated
by surgical procedures which have for their aim the
elimination of a portion of the colon. Operations with

342 INFECTIONS OF THE DIGESTIVE TRACT

this end in view have in a number of instances been
practiced, but not, so far as I am aware, with the primary
object of greatly reducing the opportunities for intestinal
putrefaction. The value of such procedures is therefore
as yet largely a matter of conjecture. That the elimina-
tion of the large intestine would be followed by a very
great reduction in the absorption and excretion of pro-
ducts of decomposition must be regarded as certain.
The experiments which have been made by Baumann
and others on dogs in which a fistula in the ileum had
been established, leave no room for doubt that putre-
faction is greatly decreased by such a procedure.1 The
closure of the fistula and the reestablishment of the
natural paths have been in such cases promptly followed
by a renewed excretion of the ethereal sulphates. Aside
from the surgical difficulties incidental to operations
designed to shorten the large intestine there are two
possibilities of a disadvantageous outcome which will
have to be tested by future experience. One of these
is the possible detrimental influence of such a shortening
of the gut upon the processes of nutrition. It is claimed
by some physiologists that the processes of peptoniza-
tion and absorption continue to be carried on in no
unimportant degree in the large intestine, and hence that
an elimination of a considerable portion of this part of

1 In a dog whose ileum was inserted a few inches above the
rectum the ethereal sulphates were low but the indican reaction
(on a meat diet) was intense. This animal was prepared for use
by Dr. Carrel of the Rockefeller Institute. The fluid faeces con-
tained indol. After the operation there was a considerable gain in
weight on a strict meat diet.

INFECTIONS OF THE DIGESTIVE TRACT 343

the digestive tract must necessarily be followed by some
failure in nutrition. I question whether the latter part
of this contention is justified ; for while some digestive
action cannot be denied to the large intestine, there are
many reasons for thinking that by far the larger part of
the nutritive resorption necessary for the maintenance
of a good state of nutrition occurs above the ileocsecal
valve. Even, however, if it be ordinarily true that a
significant amount of material is absorbed from the
region beyond the ileocsecal valve, there is no reason
why such absorption cannot under special conditions
be relinquished without detriment to nutrition. Even
the partial artificial peptonization of proteids would
do much to secure an adequate absorption of such food
materials above the ileocsecal valve and in cases of long-
continued intestinal intoxication with a grave prognosis,
the resort to permanent peptonization of the food mate-
rials would certainly be the substitution of a lesser for a
greater evil. Although it must be admitted that our
knowledge on this point is not at present decisive, I am
strongly disposed to take the view that nutrition could
be adequately maintained in the absence of the greater
part of the large intestine.

Another possibility of an unfavorable sort relates to the
inconvenience incidental to the rapid passage of semi-
solid or fluid intestinal contents. It seems probable that
if the conditions of diet were carefully studied, any
serious difficulty from this source might be avoided.
Moreover, the objection just mentioned would be greatly
minimized by connecting the ileum, csecum, or trans-

344 INFECTIONS OF THE DIGESTIVE TRACT

verse colon with the rectum instead of making a com-
munication in the abdominal or lumbar region.

It is to be remembered that operations like those just
discussed would seldom be undertaken except where
life is in danger from the persistence of a process which
has been unsuccessfully treated by the usual hygienic
and medical resources of the physician. There are
conditions of mental depression and conditions of
anaemia in which I believe an operation involving the
shortening of the intestine to be justified. There is
much to be said in favor of urging operation before the
establishment of extreme conditions of deranged function
in the nervous system or the development of anaemia
associated with greatly impaired regenerative capacity of
the blood. It is easy to understand that by too long wait-
ing a biological situation, susceptible of great improve-
ment by timely operation, might be rendered so extreme
that the recuperative powers of the damaged cells would
not suffice to restore even a fair degree of health.

The operation of connecting the vermiform appendix
with the outer world through a fistulous opening has
been repeatedly performed in cases of chronic dysentery,
for the purpose of giving patients the benefit of thorough
irrigation of the large intestine. It is possible that this
operation (which might entail fewer risks to the patient
than operations of the class just mentioned) may be
worthy of trial for the relief of chronic intoxications
leading to extreme anaemia or to serious disorders of
the nervous system.

The possibility of checking the progress of specific

INFECTIONS OF THE DIGESTIVE TRACT 345

bacterial processes in the intestinal tract by means of
specific bacterial vaccines or through the use of specific
bactericidal sera has come plainly into view during the
past ten years. Mention has already been made of the
methods that have been employed to check the develop-
ment of infections from organisms of the B. coli group
and closely related bacteria. The work of Sir A. E.
Wright upon the elevation of the opsonic index in cases
of local staphylococcal infections and in the case of
streptococcus septicaemia makes it not improbable that
in the case of intestinal affections clearly shown to be
in part dependent on staphylococcus or streptococcus
infection, the methods of vaccination employed by him
may be effective in increasing the phagocytic defenses
of the body against these infective agents. Whether
similar methods are likely to be of service in relation to
such cases as show the presence of well-defined chronic
infections dependent on anaerobic putrefactive organ-
isms such as B. aerogenes capsulatus it is impossible to
predict. It has been already mentioned that we have
not as yet observed any instance of chronic saccharo-
butyric putrefaction (due to an infection with B. aero-
genes capsulatus) hi which there has been a heightening
of the agglutination of the patient's blood with the infect-
ing organisms. This, however, does not prove that
specific antibodies are never formed through the agency
of these anaerobes in the intestine. There are in fact
some experimental observations which indicate that it
is possible to develop a specific agglutinating power for
B. aerogenes capsulatus. Whether the ability of the

346 INFECTIONS OF THE DIGESTIVE TRACT

leucocytes to take up these organisms suffers any diminu-
tion in chronic infections is not yet known; hence
speculations as to the beneficial influence of the use of
vaccines prepared from B. aerogenes capsulatus are use-
less. It is likely that this phase of study will receive
attention before long.

SOCIOLOGICAL CONSIDERATIONS

IT is hardly satisfying to conclude this consideration
of the infections of the digestive tract without some
reference to their sociological bearings. The medical
practitioner too often is forced to deal exclusively with
damage done — damage of an irreparable character.
He is unable to grapple with the weighty problem of
preventing disease. In modern civilized communities
the task of prevention has fallen into the hands of
health boards. Where these have been guided by
highly intelligent and trained men, unselfishly devoted
to the public good (as has lately been the case hi New
York City), there has resulted a noteworthy diminution
in the death rates. This is true of many kinds of disease,
but applies with great distinctness in the case of the mor-
tality of infants from diarrhoeal disorders. The infant
mortality from such disorders is due mainly to the bac-
terial contamination of milk, and some municipal author-
ities have lately been able to insist on the systematic re-
jection of bad milk. The decline in deaths from summer
diarrhoeas is a result of this improved milk sanitation.

The health boards have, however, only begun their
task. It is certain that the beneficent work of preven-
tion will in time grow so efficient as to make the fatal
milk infections of infancy comparatively rare, provided
something be also done to raise the intelligence of pa-

347

348 INFECTIONS OF THE DIGESTIVE TRACT

rents. Equally preventable are the majority of cases
of typhoid fever. It is not because of any lack of
hygienic knowledge that typhoid fever yearly claims
its hundreds of victims in the United States. It is
because of the semi-criminal indifference of the un-
educated politicians who rule many of our great cities.
Eradicate the sources of typhoid contamination of the
water supplies of cities, and the deaths from typhoid
fever will dwindle to insignificant numbers. One highly
competent and determined publicist who, backed by pub-
lic opinion, should force the rulers of our cities to give
the people uncontaminated water, would do more for his
country than the hundreds of physicians who are con-
stantly engaged in the treatment of typhoid fever.

But the problem of prevention will have been only
partially solved when the infections of cholera infantum,
dysentery, and typhoid fever are broadly checked by
measures based upon what modern medicine has taught
us about the causes of these diseases. There will re-
main the more elusive task of preventing the occurrence
of the severe chronic infections of the digestive tract.
The acute infections like dysentery and typhoid are so
numerous and obtrusive in their manifestations that it
is impossible to overlook them; the chronic infections
are insidious in their onset, obscure in their manifesta-
tions, and seldom involve groups of individuals ; hence
they pass relatively unnoticed. The injurious effects
of the chronic infections are none the less real. They
restrict the fullest development of a nation by shortening
the lives of many useful individuals and by rendering

INFECTIONS OF THE DIGESTIVE TRACT 349

many distinctly less productive than if these persons
attained physical vigor more nearly in accord with their
inherent possibilities. These partial failures of per-
formance entail a severe economic loss to the community
in which they occur. But this loss is one that cannot
be adequately measured in dollars and cents, for it must
make a nation less efficient in art, in science, in letters,
and in all intellectual pursuits. And, finally, it must not
be forgotten that the chronic enterogenic intoxications
of the nervous system may be extremely detrimental
to character and contribute to complicate and mar the
finer human relations.

It is tacitly assumed by many that these various
morbid manifestations are inevitable, are inherited, or
are in some other way an expression of the will of Provi-
dence. The conditions to which I refer are indeed
recognized by few persons to be what they actually are
— the effects of chronic poisoning of gastro-enteric
origin. Nevertheless it is certain that these relatively
obscure physical conditions are in large measure pre-
ventable. But their prevention is not simple; for it
implies, first, the recognition of chronic infections of the
digestive tract (in which the putrefactive anaerobes
are prominently concerned) capable of slowly damag-
ing the cell potentialities upon which health depends,
and, secondly, it involves the recognition and control of a
series of causative or favoring elements, such as habitual
errors in diet, sedentary habits, excessive mental and
physical work, emotional fatigue, errors in sexual life, etc.

The recognition of the presence of a chronic intestinal

350 INFECTIONS OF THE DIGESTIVE TRACT

infection (as by B. aerogenes capsulatus) is not in itself
difficult, but necessitates patient inquiry and a knowledge
of methods of investigation on the part of the physician.
The methods of investigating the digestive tract which
have been briefly outlined in this volume must prove
valuable in enabling the practitioner to determine the
presence of abnormal bacterial processes before the onset
of the clinical signs of incurable or highly refractory
states of intoxication.

The control of the causal and favoring factors in the
establishment of chronic intoxications is partly the duty
of the physician. It is, however, a duty which it is
unreasonable to expect him to discharge fully under the
conditions existing to-day, no matter how capable and
willing he may be. It is not possible for any practitioner
to undertake the entire education of his patients in
respect to knowledge of hygiene, or to develop their
will power to the point where such knowledge becomes
practically efficient. Yet this is what he must do to-
day if he would achieve a high degree of success in the
prevention of the worst consequences of the chronic
infections of the digestive tract. Before he can become
ideally efficient as a physician he must have the intelli-
gent cooperation of parents and educational institutions.

Education in the facts of hygiene and in the practice of
intelligent self-control cannot be acquired in a few hours,
but necessitates years of appropriate teaching in home
and school. So long as the schools and universities
almost wholly fail to fit their pupils to meet even the
most obvious requirements of practical life, so long will

INFECTIONS OF THE DIGESTIVE TRACT 351

it be impossible to avoid great losses of energy and other
unhappy consequences of the chronic intoxications.
At present the schools look to parents to instruct their
children in the supposedly simple matters of regulating
eating and drinking, exercise, habits of work, and sexual
habits, while the parents vaguely hope (if they think
at all about such matters) to be relieved of these embar-
rassing duties through the schools. The truth is that
neither parents nor schools are to-day able to give this
much needed sort of education. The remedy must
be provided by the schools, which in their eagerness
to impart conventionalized facts are now quite blind to
some of the most pressing needs of their pupils. Through
the schools and universities (or other appropriate
organizations) the parents of the future must be educated
both as to the facts and the moral aspects of bodily
hygiene. The physician will thus be enabled to do better
work in the prevention of some of the most distressing
human ailments. And it seems not unreasonable to
hope that some of the lessons now learned only by bitter
experience, after much that is best in life has been
sacrificed to ignorance and uncurbed impulse, will be
assimilated sufficiently early in life to mitigate materially
the lot of a not inconsiderable part of mankind. I
believe the lengthening of the span of human life to be
among the attainable results of such teaching. Is it
not likely that as men grow wiser an increasing number
will deliberately strive so to regulate their lives as to
improve the expectation of crowning well-spent days
with the peculiarly fine satisfactions of old age ?

INDEX

Acetone, 218; formation, in bac-
terial cultures, 136.
Acid, acetic, product of bacte-
rial activity, 8; in bottle-fed
babies, 65 ; in healthy animals,
66.

butyric, production in putre-
faction, 218; by anaerobes,
293; by B. butyricus, 46; from
lactic acid, 22; in mouth, 29,
30, 300 ; in stomach, 30.
caproic, 293, 294.
glycerophosphoric, 222.
indol-acetic, 243.
indol-amido-propionic, 243.
indol-propionic, 244.
lactic, 22.
oxalic, 216.
propionic, 218, 293.
skatol-amido-acetic, 243.
valeric, 293.

Acid formation, in bottle-fed
babies, 67; in fermentation,
215.

Acids, volatile fatty, 21.
Acidophile bacteria, 11.
Adult life, bacteria in intestines, 72.
Aerobic and anaerobic conditions,
in digestive tract, 23; in in-
testine, 31 ; in stomach, 30 ;
tested by methylene blue, 34.
Age, effect on intestinal bacteria,

35.

Aggressins, 160 ; in dysentery, 174.
Alcohol, formation in bacterial

cultures, 8, 136.
Amines, 220.

Ammonia, formation in bacterial
cultures, 136, 220; in bottle-
fed children, 66, 67.
Ammonium butyrate, 295, 299.
Anemia, in saccharo-butyric pu-
trefaction, 303.

Anaerobic bacteria, 191 ; action
on bilirubin, 294; action on
proteids, 24 ; in bottle-feeding,
60, 63 (footnote); with diet
of bread, 87; in caries of
teeth, 29; growth in intestine
of cats, 80, 81; influence of
reaction on growth and prod-
ucts, 21 ; obligate, 24 ; patho-
genicity, 117; putrefactive,
10; reduction of numbers by
treatment, 329; in saccharo-
butyric putrefaction, 279, 291 ;
spore-bearing, 6.

Anaerobic life, definition, 25.

Animals, bacteria in intestines, 80 ;
relative duration of life, 85.

Anthrax bacillus, 162.

Anthrax, symptomatic, formation
of indol, 243.

Aromatic products of putrefactive
decomposition, 237.

Autolysis, 1.

Autotoxins, 18.

Autotoxin theory, 17-19.

Bacillac, 334, 335.

Bacillus, acidolacticus, 207, 244, 335.
acidophilus, 11 ; in bottle-fed
children, 62; in nurslings, 38;
protective action, 11.
aerogenes capsulatus, 2 (foot-
note), 6, 11, 25, 196, 197; in
acute inflammation of the
ileum, 206; in adult life, 74;
agglutinating action, 199; in
anaerobic cultures, 117; in
animals, 82 ; biochemical
properties, 208; in bottle-fed
children, 63; in carnivorous
animals, 80 ; in childhood and
adolescence, 70; cultural in-
vestigation, 116; different

353

354

INDEX

Bacillus —

names, 45 (footnote) ; different
strains, 198, 199; gas for-
mation, 201, 203; growth in
fermentation tubes, 127, 130;
growth in milk, 117; hsemo-
lytic action, 203, 303; mor-
phology, 200 ; in nurslings, 39,
45; in pernicious anaemia,
312; relation to indicanuria,
261, 263; in saccharo-butyric
putrefaction, 291, 293, 294,
298; toxins acting on nervous
system, 205 ; in senescence, 78 ;
studied by incubation method,
118-124; in tiger with osteo-
malacia, 81.

alkaligenes, 24, 28.

anthracis, 162.

anthracis symptomatici, 24, 118,
209.

bifidus, 5, 10, 17, 113, 190;
anaerobic cultures, 116; in
bottle-fed children, 62; in
childhood and adolescence, 70 ;
growth in fermentation tubes,
131 ; infection through anus,
56 ; in marasmus, 288 ; in
meconium, 57 ; morphology, 41
and footnote ; in nurslings, 38.

botulinus, 6, 118, 210.

butyricus (Botkin), 45.

cloacae, 63, 117.

coli communis, 5, 8, 14 et seq. ;
in adult life, 74; in animals,
82; antagonism to other bac-
teria, 7 ; in bottle-fed children,
60; characteristics, 7; de-
fensive action, 10, 12; dis-
appearance from intestine in
disease, 156; effect of dif-
ferent diets, 87; in gastric
ulcer, 154; growth in fermen-
tation tubes, 130; indol for-
mation, 243, 262, 264, 279,
280, 281; in intestinal putre-
faction, 155; in meconium,
56; in nurslings, 49; number
diminished in saccharo-butyric
putrefaction, 292 ; pathogenity,
150; production of hydrogen
sulphide, 229.

Bacillus —

coli group, 9; defensive action,
9, 10; in nurslings, 39; in
senescence, 75, 76.

dysenteries, 6, 101, 172; effect
of toxins on nervous system,
180; immunizing sera, 175;
lesions due to elimination of
poison, 178 ; polyvalent serum,
176.

of Eberth, see B. typhi.

entericus, 184.

enteritidis sporogenes, 45 (foot-
note), 169, 197, 208.

fcecalis alcaligenes, 158 (foot-
note), 169.

flavosepticum, 159.

Flexneri, 15.

fluorescens, 14.

fluorescens non liquefaciens, 28.

of Friedlander, 14.

of Kruse, 15.

lactis aerogenes, 5, 7, 15, 16, 24,
31 ; in bottle-fed children, 62 ;
characteristics, 8 ; hydrogen
sulphide formation, 229; in
nurslings, 49; pathogenicity,
153.

liquefaciens ilei, 70 (footnote),
113.

mascerans, 220.

cedematis maligni, 24, 118, 292,
319.

paralacticus, 207.

paraputrificus, 193.

paratyphoid, 6.

perfringens, 11, 45 (footnote),
197, 207.

of plague, 162.

prodigiosits, 14, 15, 16.

proteus vulgaris, 5, 14, 24, 182,
183, 244.

putrificus, 6, 14, 24, 25, 192, 204 ;
in adult life, 74 ; in bottle-fed
children, 63; in caries of
teeth, 195 ; on fruits, 319 ; in-
dol formation, 279, 282 (foot-
note) ; mercaptan formation,
195, 297 ; in nurslings, 44 ; rela-
tion to indicanuria, 263; in
saccharo-butyric putrefaction,
291, 297.

INDEX

355

Bacillus —

pyocyaneus, 14, 15, 16; growth
in fermentation tubes, 132;
in intestines in health, 103.
of quarter evil, see B. anthracis

symptomaticus.
of rauschbrand, see B. anthracis

symptoma tici.
of Shiga-Kruse, 15.
subtilis, 56.

typhi, 6, 10, 15, 102, 157.
violarius acetonicus, 220.
Welchii, see B. aerogenes cap-

sulatus.

Bacteria, anaerobic, 191.
in excreta, 1 and footnote,
in intestines, in adult life, 72;
in animals, 80; buffalo, 83;
camel, 82; carnivorous ani-
mals, 82; cat, 80; elephant,
82; herbivorous animals, 82;
horse, 82 ; lion, 81 ; wolf, 81 ;
classification, 105 ; autopsy,
normal boy, 70 (footnote) ;
at different ages, 35, 60;
distribution, 7; elective an-
tagonistic action, 14; influence
of food, 86; in health, 1, 4;
obligate, 7; their significance,
5 ; pathogenic forms in health,
102.

liquefying, 181.
in milk, 60.
Bail, 160, 165.
Baldwin, 216.
Basic substances, 221.
Bauer, 146, 147, 149 (footnote).
Baumann, 241.

Baumann and Udranzky, 224, 225.
Baumstark, 146.
Beijerink, 26.
Bergh, 235.
Bergman, 4.
Bernard, Claude, 243.
Betz, 231.

Bienstock, 14, 24, 44, 192, 193, 194.
Bilirubin, reduction by anaerobes,

294.

Bjeloussow, 11.
Blair, 86.
Blumenthal, 247.
B6cai, 222.

Booker, 186.
Botkin, 45.
Botulism, 211.
Boycott, 29.
Breaudat, 220.
Brieger, 224, 225.
Bryant, 99.

Burgeon treatment, 232.
Butyric acid, see Acid, butyric.
Butyric acid bacillus in bottle-fed
children, 63.

Cadaverin, 220, 224, 226.

Caprice acid, 293, 294.

Carbon dioxide in saccharo-butyric
putrefaction, 294.

Caries of teeth, 29.

Carnivorous animals, pathogenic-
ity of faecal flora, 82.

Champagne, producing urticaria
and gout, 276, 285.

Chantemesse, 162.

Charrin and Roget, 95.

Children, bottle-fed, bacteria in
intestines, 59 ; anaerobic
forms, 50, 61 ; distribution, 61 ;
products of decomposition, 64.

Cholin, 220, 222, 223, 224.

Citron, 161.

Classification of bacteria in digest-
ive tract, 105 ; by agglutinins,
110 ; by biochemical properties,
106; by morphology, 106; by
motility, 106 ; by pathogenicity,
108 ; by spore formation, 109 ;
by staining properties, 106.

Clemens, 148.

Clinical types of enterogenous
poisoning, 274.

Collins, 110, 336.

Colon, anaerobic conditions, 33;
number of bacteria, 33.

Colon bacillus, see B. coli.

Colon irrigation, influence on pu-
trefactive processes, 337.

Colon-typhoid-dysentery group of
organisms, 150.

Conradi and Kurpjuweit, 9, 15,
37, 18, 20.

Cresol, 237, 246.

Gushing and Livingston, 31.

Cyanosis, enterogenic, 234-236.

356

INDEX

Cyclical vomiting, 277.
Cystin, 229.
Cystinuria, 224.

Defensive action of digestive
juices, 6.

Diamines, 220, 224.

Diarrhoea, from saccharo-butyric
putrefaction, 301.

Dibenzoylcadaverin, 225.

Digestive juices, defensive action,
6.

Dimethylamidobenzaldehyde reac-
tion, in faeces, 145; in maras-
mus, 286 ; with skatol, 240 ; in
urine, 147.

Diplococci, in marasmus, 288; in
mucous colitis, 101 ; in per-
nicious anaemia, 188; in
saccharo-butyric putrefaction,
291, 292, 305.

Diplococcus intestinalis, 15, 16.

Distribution of bacteria in in-
testines, 7.

Doebert, 158 (footnote).

Dopter, 180.

Dunham, 106, 110, 197.

Duval, 103.

Dysentery bacilli, see B. dysenterice.

Eberle, 51.

Eberth's bacillus, 157.

Ehrlich, 147, 148, 275.

Ehrlich aldehyde reaction, see Di-
methylamidobenzaldehyde re-
action.

Ellinger, 226

Emminghaus, 231.

Enterogenic cyanosis, 234, 236.

Enterogenous poisons, individual
susceptibilities, 274.

Enzymes, in cellulose, 5 ; tryptic, 6.

Epileptic seizures, With indicanuria,
276.

Epithelial cells of digestive tract,
91 ; effect of excessive desqua-
mation, 92 ; protection against
indol, 93.

Escherich, 10, 38, 39, 49, 55, 56,
151, 186.

Ethereal sulphates, 72, 297; in
adult life, 74 ; in childhood, 72 ;

Ethereal sulphates —

influence of colon irrigation,
337; in marasmus, 286; re-
duced by lactic acid, 244;
in saccharo-butyric putrefac-
tion, 296.

Exhaustion, influence on bacterial
penetration in intestinal tract,
95.

Fatigue, in excessive intestinal
putrefaction, 286; in indol
and phenol poisoning, 289;
in marasmus, 289.

Fatty acid formation in bacterial
cultures, 136.

Fatty acids, molecular weights,
136 ; in saccharo-butyric pu-
trefaction, 295.

Ferment, peptic, 6.

Fermentation tubes, study of bac-
teria, 126; study of sediment,
130.

Fermentative processes, 214.

Ficker, 95.

Flexner, 172, 175, 178, 180, 198,
206.

Fliigge, 45.

Food, 184; influence on intestinal
bacteria, 86.

Foster, 139.

Fraenkel, 11, 197.

Gaffsky, 157.

Gartner, 169.

Gas formation, by mixed faecal
flora from nurslings, 66; in
saccharo-butyric putrefaction,
292.

Gasphlegmon bacillus, see B. aero-
genes capsulatus.

Gas production, influence of diet,
129 ; significance of diminu-
tion, 128.

Gelatin, use in indicanuria, 267.

Glycerophosphoric acid, 222.

Gram-negative bacteria, 37.

Gram-positive bacteria, 37.

Gram stain, value of, 111-114.

Granulo-bacillus immobilis lique-
faciens, 45 (footnote), 125,
196.

INDEX

357

Grassberger, 45 (footnote).
Grassberger and Schattenfroh, 125,

197.
Guanidin, 221.

Haemolytic action of faecal ex-
tracts, 302.

Headache, with indicanuria, 277.

Hemiparasites, 162.

Herbivorous animals, bacteria in
intestines, 82 ; development of
anaemia, 85 ; pathogenicity of
faecal flora, 83.

Herter, 245.

Herter and Foster, 139.

Herter and Wakeman, 247.

Hervieux, 271, 273.

Hilgermann, 94.

Hiss, 173, 174.

Holt, 285.

Hopkins, 242.

Howard, 206.

Howland, 248, 270.

Hunger, influence on bacterial
penetration through intestinal
walls, 95.

Hydrobilirubin, 295; Schmidt's
reaction, 144.

Hydrogen formation in saccharo-
butyric putrefaction, 294.

Hydrogen sulphide, 227 ; excretion
in urine, 231 ; formation by
bacteria in animals, 80; in
bottle-fed children, 67 ; in
enterogenic cyanosis, 234-236 ;
in health, 230 ; in marasmus,
230; in stomach, 229; influ-
ence on organism, 230; toxic
action on dogs, 233.

Hydrothionaemia, 230.

Ileum, anaerobic conditions, 33 ;
number of bacteria, 32.

Indicanuria, 241, 246, 257, 278, 279,
281; in adult life, 74; effect
of cathartics, 268; of colon
irrigation, 337; of oxidizing
substances, 332 ; of treatment,
265 ; experimental production,
261 ; individual symptoms,
277; in occlusion of small in-
testine, 283; relation to con-

Indicanuria —

stipation, 263 ; significance,
258.

Indigo, 246.

Indigo blue, 273.

Indigo red, 273.

Indigouria, 271, 272.

Indirubin, 273.

Individual susceptibilities to nitrog-
enous poisons, 274.

Indol, 7, 20, 241 ; action on mus-
cles, 254 ; formation by bac-
teria, from bottle-fed children,
67 ; in cultures, 135 ; in child-
hood and adolescence, 72;
in nurslings, 53; formation
from tryptophan, 243; in
marasmus, 284; method of
determination, 138; in oc-
clusion of common bile duct,
282; in occlusion of small
intestine, 283; in pancreatic
achylia, 283; production by
B. coli, 7, 8; production in
bottle-fed children, 64; in
saccharo-butyric putrefaction,
295, 299; separation from
skatol, 139; in stools in ma-
rasmus, 283; toxic action,
242, 247-257.

Indol-acetic acid, 243.

Indolaemia, 269.

Indol-amido-propionic acid, 243.

Indolic type of intestinal putre-
faction, 279, 280.

Inclol-propionic acid, 244.

Indoluria, 269.

Indoxyl, 271-272.

Indoxyl-potassium-sulphate, 246.

Intestinal putrefaction, see Putre-
faction, intestinal.

Jaffe, 144.

Kamen, 198, 199.

Kedrowski, 27.

Kefir, 334.

Kikuchi, 175.

Klein, 45 (footnote), 52, 115, 197,

208.

Kohn, 56.
Kruse, 62, 173.

358

INDEX

Kruse and Bail, 160.

Kumyss, 335.

Kurpjuweit, 9, 10, 15, 17, 18, 20.

Lactic acid, 22.

Lactic acid bacillus, 207, 244, 335.

Lecithin, 222, 223.

Lee, 254.

Lembke, 88.

Levin, 2, 4.

Libbman, 62.

Livingston and Gushing, 31.

Lubenau, 184.

Lumbago, 276.

Malzushita, 26.

Manteufel, 18.

Marasmus, 17; bacteria in intes-
tines, 287; B. bifidus, 287;
ethereal sulphates, 286; indol
formation, 284; phenol for-
mation, 236; symptoms, 285;
treatment, 289.

Matzoon, 334.

Maury, 250.

Meat, reducing action, 89.

Meconium, 17; bacterial infection,
56; B. aerogenes capsulatus,
57; B. bifidus, 57; B. coli
communis, 56 ; B. subtilis, 56 ;
physiological functions, 59.

Melancholia, from excessive in-
testinal putrefaction, 308.

Mental depression, with indican-
uria, 277.

Mercaptan, 30, 226; formation in
bacterial cultures, 137; in
bottle-fed children, 67; in
carnivorous animals, 82; in
herbivorous animals, 80, 83;
in saccharo-butyric putrefac-
tion, 297; method of deter-
mination, 137.

Mereschkowsky, 11.

Metchnikoff, 78, 335.

Metchnikoff, Madame, 3.

Methaemoglobinuria, 234.

Methods of investigation, 111, 115,
116, 125, 133; cultures in
seven-day flasks, 135; gas
formation, 1 18 ; gas formation
in sugar bouillon, 126 ; growth

Methods of investigation —

on milk, 117; incubation
method of Welch and Nuttall,
118-124; microscopical fields,
111; products of mixed flora,
129; value of Gram stain, 111-
114.

Methyl guanidin, 221.

Meyer, 148.

" Micro-aerophile, " 26.

Microscopical fields, methods of
investigation, 111.

Milk, number of bacteria, 60;
question of sterilization, 319.

Moro, 3, 11, 13, 14, 39, 41, 49, 55, 57.

Moro and Murath, 17, 18.

Mucous membrane of intestine,
permeability for bacteria, 93.

Mviller, 232.

Multiple neuritis in excessive in-
testinal putrefaction, 311.

Murath, 13, 14, 17, 18.

Muscle fatigue in indol poisoning,
254 ; in phenol poisoning, 255.

Myasthenia gravis, 256.

Nencki, 70 (footnote), 227.

Nesbit, 222, 223, 224.

Neubauer, 147.

Neurin, 220, 222, 223.

Nurslings,

bacteria in intestines, 66;
action of mixed faecal flora,
66; cultural peculiarities, 40;
distribution, 48 ; infection
through anus, 55; inhibitory
powers, 13, 14, 15, 16, 17;
numbers, 51 ; origin of bac-
teria in intestines, 53; prod-
ucts, 53 ; spore formation, 58 ;
staining properties, 37, 38, 39.

Nuttall and Thierfelder, 2.

Oebius, 18.

Opsonic index, in staphylococcal

infection, 190.
Opsonins, 309.

Origin of bacteria in intestines, 53.
Oxalic acid, 216.
Oxaluria, 216.
Oxygen, distribution in intestines,

29.

INDEX

359

Paracolon organisms, 14, 158.

Paratyphoid bacillus, 10, 168.

Park, 110, 173.

Passini, 18, 46, 200.

Pasteur, 2, 23, 27.

Pathogenicity of anaerobes, 117.

Pentamethylendiamin, 224.

Peptic ferment, 6.

Pernicious anaemia, B. aerogenes
capsulatus, 312.

Petruschky, 158 (footnote).

Pfeiffer, 162.

Phenol, 20, 237, 247; in bacterial
cultures, 135, 136; in bottle-
fed children, 68 ; in marasmus,
286; influence on muscles,
255.

Phenol-potassium-sulphate, 238.

Phylogenetic significance of large
intestine, 97.

Pizenti, 260.

Plague bacillus, 162.

Porcher and Hervieux, 272, 273.

Progressive muscular atrophy, with
indicanuria, 276.

Propionic acid, 218, 293.

Proteolytic action, 6.

Ptomaines, 226.

Putrefaction, intestinal, aromatic
products, 237; indolic type,
279, 280 ; influence of prompt
resorption, 99; mixed type,
279, 306 ; nervous manifesta-
tions, 309 ; in production of
fatigue, 286; production of
parenchymatous degeneration,
309; prognosis, 337-339; sac-
charo-butyric type, 279, 291;
treatment, 314; types, 278.

Putrefactive processes, 214.

Putrescin, 220, 224, 226.

Reaction, influence on growth and
products of intestinal anee-
robes, 21.

Receptors, 275.

Reducing action of meat, 90.

Resorption from small intestine,
99.

Rettger, 24.

Richards and Rowland, 248, 270.

Rodella, 63, 195.

Roget, 95.
Roos, 225.

Saccharo-butyric type of intestinal
putrefaction, 279, 291 ; anae-
mia, 303 ; carbon dioxide for-
mation, 294 ; characteristics
of faeces, 293, 294; conse-
quences, 299; ethereal sul-
phates, 296; gas formation,
292; hydrogen formation, 294.

Saprophytic bacteria, 7, 162.

Sarcinse in bottle-fed children, 63.

Schardinger, 220.

Schottelius, 3.

Schottenfroh, 45 and footnote, 125.

Schmidt, 283.

Schmidt and Strassberger, 52.

Schmidt's reaction for hydro-
bilirubin, 144.

Schotmuller, 168.

Schottenfroh and Grassberger, 63
(footnote).

Selmi, 226.

Senator, 230.

Senescence, bacteria in intestines,
75, 77, 78.

Shiga, 172, 174, 175, 176, 177, 180.

Simon, 149 (footnote).

Skatol, 239, 243; colorimetric
estimation, 142 ; determina-
tion, 139; Ehrlich aldehyde
reaction, 148, 240; separation
from indol, 139.

Skatol formation, in bacterial
cultures, 135, 136; in bottle-
fed babies, 64.

Skatol-amido-acetic acid, 243.

Skatolic type of intestinal putre-
faction, 279.

Smith, Theobald, 90, 111, 117,
201, 202.

Sociological considerations, 347.

Soor, 15.

Spirillum of cholera, 28.

Spore-bearing bacilli, in bottle-fed
children, 64; in cats, 80; in
meconium, 58.

Staphylococcal infections, 189.

Staphylococci, in bottle-fed chil-
dren, 62; pyogenic, 6; treat-
ment by vaccination, 190.

':Ml

RETURN
TO-—*

BIOLOGY LIBRARY

3503 Life Sciences Bldg.

642-2531

LOAN PERIOD 1 2

ALL BOOKS MAY BE RECALLED AFTER 7 DAYS
Renewed books are subject to immediate recall

DUE AS STAMPED BELOW

-

Sub;ect to

UNIVERSITY OF CALIFORNIA, BERKELEY
FORM NO. DD4 BERKELEY, CA 94720

®s
Berkeley

Y.B

UNIVERSITY OF CALIFORNIA LIBRARY