The sources and modes of infection

UNIVERSITY OF CALIFORNIA

wCFARTMENT OF CIVIL ENGINEERING

BERKELEY. CALIFORNIA

Civil Engineering Dept,

Engineering
Library

RERKEUF-Y, CAU

THE

SOURCES AND MODES

OF

INFECTION

BY

CHARLES V. CHAPIN, M.D., Sc.D.

\\

SUPERINTENDENT OF HEALTH, PROVIDENCE, R. I.
AUTHOR OF "MUNICIPAL SANITATION IN THE UNITED STATES"

SECOND EDITION REVISED AND ENLARGED

TOTAL ISSUE FOUR THOUSAND

NEW YORK

JOHN WILEY & SONS, INC.
LONDON: CHAPMAN & HALL, LIMITED

1916

CiT

Engineering
Library

COPYRIGHT, 1910, 1912,
BY CHARLES V. CHAPIN

Stanbopc ]pte«ft

». H. OILSON COMPAMV
BOSTON. U.S.A.

PREFACE.

THIS volume is intended to indicate the principles which
should guide sanitary practice, and to show how recent labo-
ratory work and the epidemiological study of disease have
modified these principles. When I began work as health offi-
cer in 1884 the filth theory was still in favor, 'and it was
generally believed that the germs of disease commonly grew
in decaying organic matter. Yet contagion was recognized
as an important factor in the spread of disease, and the isola-
tion of the sick was more and more insisted upon. Fifteen
years ago probably most health officials believed that the
contagious diseases could be completely stamped out if only
all persons sick with them could be isolated. The air was
thought to be the chief medium for their transmission, and
fomites the mechanism for their passage from place to place.
Sanitary practice was based on these premises.

My own views concerning these matters became greatly
modified year by year, partly owing to the rapidly accumu-
lating knowledge of bacteria and other disease-producing
organisms, and partly owing to direct observations on the
manner in which the infectious diseases are disseminated, and
on the effect of preventive measures.

It now appears that the growth of disease germs outside of
the body is not frequent enough to be an important factor in
the causation of disease, but their growth in the body with-
out causing sickness, their latency as it were, often for many
months, is a factor of very great significance. We know now
that direct contact with the sick, or with healthy carriers of
disease germs, is an exceedingly frequent mode of transmis-
sion, and that infection by means of the air, or from infected
articles, is not nearly as common as was formerly believed.

iii

iv PREFACE

We are now better able than ever before to attribute to water
and milk their proper share in the distribution of infection.
The recent discovery of the transmission of disease by insects
gives us entirely new and most effective means of combating
disease. It is time that sanitary measures directed against
the infectious diseases should be modified to correspond with
existing knowledge. Present-day theories and present-day
practice are maintained largely by tradition, and to facilitate
the adaptation of practice to the facts as we now know them,
is the purpose of this book. Some modifications of sanitary
practice are suggested, but no attempt is made to discuss
details; rather are general principles presented, which it is
believed ought to guide administrators in their work.

While some of the following pages may seem rather radical
to many, I believe that practically all laboratory workers will
agree with the contents of the first chapter, and that a large
number of bacteriologists and health officers are convinced of
the great importance of " carriers " and mild unrecognized
cases. The tendency among many, too, is to lay less emphasis
on infection by fomites, though perhaps few are ready to
give up routine terminal disinfection for the common infec-
tious diseases. So also there are very many careful observers
who are attributing more and more importance to what is
generally called contact infection.

The public health administrator is placed at great disad-
vantage because he is obliged to base his acts on knowledge
which is far from exact. The laboratory workers have accu-
mulated a vast mass of quite exact data in regard to the caus-
ative relation of bacteria and protozoa to disease, and no one
appreciates this more than the writer, but there are many
problems which the laboratory men cannot solve, and many
others which they have failed to solve. The epidemiologist
must study in the field the way in which disease is caused.
He must use the statistical method, and the application of
statistical methods to epidemiology is more difficult and
attractive than laboratory experiment.

PREFACE V

We need to measure more carefully the relative importance
of different sources of disease and different modes of infec-
tion. It is not so important to know that typhoid bacilli
live in water for weeks, as it is to know that 99 per cent die
in one week. It is not enough to discover that diphtheria
bacilli can be recovered from articles in the sick-room; we
must learn how often they are found and how often disease
is traced to such a source. We have for years been much
alarmed because tubercle bacilli are found in milk, but since
a serious effort has been made to measure the actual danger,
the alarm has greatly diminished. Doubtless the house fly
has been the cause of typhoid fever, but in what percentage
of cases we are profoundly ignorant. Healthy carriers of
diphtheria have certainly transmitted the disease to others,
and we should earnestly try to determine the amount of
diphtheria caused in this way. The attempt is made in the
following pages to estimate roughly, with the very imperfect
material now available, the relative importance of different
factors in the extension of infectious diseases. The conclu-
sions must to a large extent be merely tentative, and as
indicating lines for further study.

I am under great obligations to my friends Dr. H. W. Hill
and particularly Prof. F. P. Gorham for many suggestions
and criticisms, but neither is to be considered at all
responsible for any of the views presented.

The book is intended primarily for health officers and phy-
sicians, but it is hoped that many others will find some parts
interesting and suggestive.

CHARLES V. CHAPIN.
PROVIDENCE, April, 1910.

PREFACE TO SECOND EDITION.

So much new material has accumulated since the appear-
ance of the first edition that it has been thought best to re-
write several parts of the book. Thus, recent experiments
and particularly the observation that bacteria fresh from
the body are usually less resistant than are " cultured"
germs, indicate that the life of bacteria outside- the body is
even shorter than was supposed. It is in regard to carriers
that literature has been accumulating most rapidly and
several comprehensive articles, as well as very numerous
reports of carrier infection, leave no room for doubt that in
many diseases, as diphtheria, cholera, typhoid fever and
cerebro-spinal meningitis, the carrier is a very important,
if not the most important, factor in the spread of the disease.
Also recent work has shown that contact with carriers is
quite likely to be the key to the epidemiology of poliomyelitis.
New evidence has been presented concerning the carriage
of infection by milk and also by water, but I cannot see that
the reality of the Mills-Reincke phenomenon, which depends
upon the influence of water on so many forms of disease,
has yet been established. Much work during the past two
years has been devoted to insects as carriers of infection and
several additional diseases have been shown probably to be
so transmitted. One of the most interesting of these is
typhus fever, and, if recent work is substantiated, much
light will be thrown on its epidemiology. While there is
more evidence than there was that the fly is a factor in the
spread of the fecal-borne diseases, there does not as yet seem
to be much warrant for the rather sensational literature
with which the public is deluged. Studies in hospitals
and elsewhere have confirmed the belief that air is of minor

vi

PREFACE TO SECOND EDITION vii

importance in the spread of disease. The views presented
in the first edition in regard to the inutility of isolation
under many conditions, and in regard to the small importance
of fomites, were somewhat novel and it was suspected that
they might be subjected to considerable criticism, but such
does not seem to have appeared, and I still believe that,
while isolation and bedside disinfection will in the future
continue to be, when scientifically applied, of the utmost
importance, much of the routine practice of health officials
needs to be profoundly modified.

CHARLES V. CHAPIN.
Providence, July, 1912.

CONTENTS.

CHAPTER I.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY.

PAGE

Review of evidence of growth of disease germs outside of the
body, anthrax, black leg, tetanus. — Typhoid bacilli in soil, feces,
water, ice, oysters, milk. — Epidemiological evidence relating to
typhoid fever. — Evidence relating to cholera, Mediterranean
fever, plague, dysentery, bacteria of suppuration, diphtheria and
various other diseases. — Reasons for former belief in " filth theory "
of disease. — Epidemiological evidence against soil infection. —
Conclusions. — Relations of these views to public sanitation 1

CHAPTER II.

CARRIERS AND MISSED CASES.

Importance of the subject. — Evidence of the occurrence of
carriers and missed cases, and reference to disease caused by
them in typhoid fever, cholera, dysentery, cerebro-spinal menin-
gitis, diphtheria, glanders, influenza, pneumonia, gonorrhea,
tuberculosis, leprosy, suppuration, tetanus, scarlet fever, small-
pox, measles, protozoan diseases such as cattle fever, malaria,
sleeping sickness, nagana, syphilis, amebic dysentery, poliomye-
litis and yellow fever. — Review of evidence. — Conclusions. ... 33

CHAPTER III.

LIMITATIONS TO THE VALUE OF ISOLATION.

The number of carriers and missed cases. — Not realized by
health officials. — Failure of isolation in Providence. — Hospital-
ization has not materially lessened infectious disease. — Less
isolation followed by less diphtheria in Providence. — Failure of
isolation in institutions. — Failures in the isolation of measles, of
cerebro-spinal meningitis, of smallpox. — Failure due to carriers
and missed cases. — Difficulties in the isolation of diphtheria. —
Absolute isolation not possible. — Moderate isolation sufficient.

— Infection not so easy as was believed. — Isolation effective at
the beginning of an outbreak, rarely later. — The more carriers
the less effective is isolation. — True value of isolation hospitals.

— Details of home isolation. — Summary 133

ix

X CONTENTS

CHAPTER IV.

INFECTION BY CONTACT.

PAGE

Transmission of gonorrhea by indirect contact. — Of syphilis
in the same way. — Typhoid fever spread by contact. — Reasons
why contact infection has been disregarded. — Mode of contact
infection in typhoid fever. — Contact infection in dysentery. —
Cholera. — Anchylostomiasis. — Gonorrhea. — Transfer of nasal
and oral secretions by contact. — Presence of germs in secretions
and on various objects. — Diseases do not spread from family to
family in tenements. — No cross infection in hospitals except by
contact. — Contact infection in tuberculosis. — Protection from
contact infection a personal matter. — Need for urging personal
cleanliness 164

CHAPTER V.

INFECTION BY FOMITES.

Definition. — Fomites and yellow fever. — Examples of alleged
foinites infection. — Infection by clothing, rooms, rags, money. —
Fomites infection in tetanus, anthrax, typhoid fever, diphtheria,
plague. — No evidence that fomites are of much importance. —
Reasons for belief in fomites. — Effect of drying on bacteria of
different kinds. — The finding of bacteria on fomites. — Drying
of vaccine and smallpox virus. — Summary of bacteriological evi-
dence. — Experimental work with yellow fever, plague. — Results
of abandoning disinfection after scarlet fever and diphtheria in
Providence. — Disinfection in other diseases. — School disinfec-
tion. — Views of other writers. — Conclusions 212

CHAPTER VI.

INFECTION BY AIR.

Alleged aerial transmission of smallpox, scarlet fever, diph-
theria, typhoid fever, influenza, measles. — Hospital experiments
in aerial transmission. — Transmission of typhus fever and plague.

— History of surgical technique shows that air-borne infection
is of little moment. — Air-borne anthrax. — Bacteria not given
off from moist surfaces. — Bacteria carried in dust. — Dust and
the germs of typhoid fever, diphtheria, plague. — Dust and
tubercle bacilli. — Finding of tubercle bacilli in dust. — Other
bacteria in dust. — Droplet infection. — Finding of bacteria in air.

— Experimental work with tuberculosis, Mediterranean fever,
anthrax, plague. — Conclusions 259

CONTENTS XI

CHAPTER VII.

INFECTION BY FOOD AND DRINK.

PAGE

The Broad Street well. — North Boston well. — Typhoid fever
due to water. — Cholera, dysentery, diarrhea. — The Mills-
Reincke phenomenon. — Malaria and yellow fever. — Purification
of water. — Ice. — Milk, number of outbreaks. — Mode of infec-
tion of the milk. — Typhoid fever, scarlet fever, diphtheria. —
Ice cream. — Butter. — Streptococcus sore throat and milk. —
Tuberculosis and milk, evidence of infection. — Bacilli in market
milk. — Amount of tuberculosis due to milk. — Mediterranean
fever, anthrax, foot-and-mouth disease, rabies. — Diarrhea. —
Infection by meat. — Infection by shellfish. — Crawfish. —
Watercress. — Celery 316

CHAPTER VIII.

INFECTION BY INSECTS.

Insects as biological and as mechanical carriers. — Pioneer
work of Smith and Kilborne in cattle fever. — Malaria. — Modes
of control: quinia, isolation, screening; mosquito reduction;
practical results. — Yellow fever carried by mosquitoes, not by
fomites. — Control of yellow fever; quarantine, isolation; mos-
quito control; practical results. — Filariasis, sleeping sickness,
kala-azar, dengue, pellagra, opilacoa, pappataci fever, oriental
sore, typhus fever, relapsing fever, Rocky Mountain fever. —
Bubonic plague and fleas. — Anthrax. — Carriage of bacteria on
the bodies of insects. — Cockroaches and other insects. — Flies,
experimental work. — Bacteria found on flies. — Habits and
species. — Transmission of disease, murrina, tuberculosis, cholera,
dysentery, diarrhea. — Flies and typhoid fever, local and seasonal
distribution. — Evidence against theory. — Epidemiological evi-
dence in favor of causal relation. — Privies, flies and typhoid
fever. — Conclusions 380

THE SOURCES AND MODES OF
INFECTION.

CHAPTER I.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY

Former Theories. — From time immemorial miasms, ma-
larias, vapors and emanations, gaseous or otherwise, have
been believed to be the frequent cause of disease. These
miasms were thought to arise from stagnant marshes, decay-
ing vegetation, putrid animal matter, and indeed filth
of every kind. This belief in the extra-corporal origin of
disease reached its widest acceptance about the middle of
the nineteenth century. The rise of the germ theory greatly
strengthened it. The discovery of bacteria and of their
wide distribution and almost universal growth in dead
organic substances, and the theory that these bacteria were
the real cause of disease, led men to look for the source of
disease outside of the body, and chiefly in dead animal and
vegetable matter. With the passing of the germ theory as
a theory, and with the demonstration of the parasitic nature
of so many of our most important and dreaded diseases, the
opportunity was afforded for studying in detail the bacteria
and protozoa which are the specific causes of these diseases.
Much has been done by laboratory workers to unravel the
life history of these minute forms, and it is well for us to
examine the knowledge thus gained, and also the newer
epidemiological observations on the spread of the infectious
diseases, and in the light of these data question the belief
that these diseases have their origin in the outer world rather

1

2 THE SOURCES ''AN'ti tiOpES OF INFECTION

than in th& bodies (jf men or animals. Let us consider some
of these diseases in detail.

Anthrax is not common in the United States, but it is of
much interest from a scientific standpoint. It was the first
disease definitely proved to be caused by bacteria. It is
of particular interest in this connection because a relation
to the soil has been better established for this than for any
other disease. Practically all writers are agreed that the
soil may become infected with anthrax, and remain so for
a long time, and that animals pastured upon such soil con-
tract the disease by taking the bacteria in with the food,
or inspired air, or through abrasions of the skin. That this
soil infection is not the sole, or perhaps even the most com-
mon source of infection, and that danger from this source
has perhaps been somewhat exaggerated, is probably true.
Thus Delepine 1 from studying recent outbreaks in Great
Britain is convinced that the disease is perpetuated by
a more or less direct contact, chiefly with unrecognized or
concealed cases, and that there is no necessity for supposing
a long continued soil infection, and no direct evidence for
it as the principal source of the disease. McFadyean 2 can-
not trace anthrax in Britain to the soil, and thinks it im-
probable that it grows there, or it would be more common.
Moreover it does not grow well under 60°. Legge 3 notes that
animal anthrax does not increase in the summer as it would
be likely to if it were due to growth in the soil. Moore 4 does
not think the bacillus maintains a saprophytic existence.
In Louisiana 5 the extensive outbreak in the latter years of
the nineteenth century was believed to be due to some extent
to food infection, and to a large extent to direct inoculation

Delepine, Pub. Health, 1904-5, XVII, 491.

McFadyean, J. Comp. Path, and Therap., Edinb. and Lond., 1903,
XVI, 35.

Legge, Lancet, Lond., 1905, I, 695.
Rep. Comm. of Agric. N. Y., 1907.
Louisiana Agricultural Experiment Sta., Bull. No. 60, 2d. s.f 1900.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 3

by a species of horsefly, Tabanus lineola. Outbreaks of
anthrax have occurred every few years in Louisiana for over
half a century, and persistent soil infection has been alleged
as their source. Very likely it is so to a certain extent, but
on the other hand there is no doubt that the interval be-
tween the outbreaks may well be bridged over by a more or
less direct connection between sporadic and unrecognized
cases occurring in the interval. That such cases really occur
is shown by Delepine's investigations of similar conditions
in England. Nevertheless almost all veterinary and medical
writers are agreed as to the long continued soil infection of
certain areas. Evidence of this is forthcoming from France,
Germany, England 1 and the United States. In England such
infective areas are said to be the most numerous where refuse
from mills using foreign wools is used for manure. In this
country anthrax is believed to have been traced to morocco
factories on the Delaware River using large numbers of
foreign skins.2 Some of the infected Delaware farms had new
tenants with new cattle each year, but infection recurred.
All the evidence pointed to persistent infection of the soil.
Similarly infected farms or fields are reported from New Jer-
sey,3 from the Genesee valley4 and from Louisiana.5 Law
reports that 200 cases in cattle, and 3 in human beings re-
sulted in the space of two weeks, from the soil infection of a
limited area. Dr. Leonard Pearson wrote me that the evi-
dence is conclusive that soil infection with anthrax has existed
in a number of places in Pennsylvania. Two instances have
recently been reported which seem to show pretty conclu-
sively that anthrax bacilli do grow in small ponds under

1 Poore, The Earth in Relation to the Preservation and Destruction
of Contagia, Lond., 1902, 9-21.

2 Delaware Agricultural Experiment Sta., Bull. No. 32, 1896, 6.
» Rep. St. Bd. Health, N. J., 1904, 5.

4 Law, Text-Book of Veterinary Medicine, Ithaca, 1902, IV, 195.
6 Louisiana Agricultural Experiment Sta., Bull. No. 60, 2d s., Insert
opp. 345, and 357.

4 THE SOURCES AND MODES OF INFECTION

natural conditions. Dr. J. Sinclair Holden, health officer of
Sudbury, Co. Suffolk, England, writes that in 1905 the waste
water from a horsehair factory was discharged into a small
pond. There was evidently some seepage from this pond to
another about 20 feet distant. The second pond, in the fol-
lowing year, was found to be so abundantly filled with the
bacilli of anthrax that it seemed that there must have been
free reproduction. Hastings l also reports that he examined
a pond which had received anthrax-infected tannery refuse,
and that this was teeming with the vegetative forms of the
bacillus.

While there is a good deal of evidence, apparently conclu-
sive, that soil may remain infected with anthrax for years,
there also is evidence that the infection after a time disap-
pears. Pasteur records instances of the infection dying out
after a lapse of some years,2 and in Delaware infection did
not persist on all the infected farms.3 The fact that anthrax
has appeared at so many isolated points in England and the
United States, and though in the majority of cases soil inocu-
lation must have taken place, nevertheless the fact that the
disease has never become widespread or long persistent locally
is sufficient reason for the conclusion that its virus does not,
in these countries at least, find a suitable soil. If it increases
at all in the soil it is only for a time, and the tendency is for
it to die out. In other words, the history of this disease is
best explained on the hypothesis that the soil is infected
chiefly if not exclusively by the spores, which may retain
their virulence for years, but which rarely germinate in the
earth.

Charbon Symptomatique. — Another animal disease known
as black-leg, or in Europe as charbon symptomatique, is

1 Hastings, Paper read at meeting of Society of American Bacteri-
ologists, 1908.

* Poore, The Earth in Relation to the Preservation and Destruction
of Contagia, Lond., 1902, 13.

3 Delaware Agricultural Experiment Sta., Bull. No. 32, 1896, 7.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 5

quite prevalent among cattle in this country. Like anthrax
it is caused by a bacillus that forms spores. Veterinarians
are agreed that soil areas become infected with black-leg and
may remain so for some time. Undoubtedly this infection
may be explained, as is the infection by anthrax, as due simply
to the resistance of the spores, and does not necessarily re-
quire the actual multiplication of the germs in the soil.

Tetanus or lockjaw, even before it was known to be due
to a bacillus, was believed to occur with exceptional fre-
quency in limited areas. It has been stated by numerous
writers that there is a strip of land near Red Bank, New
Jersey, where tetanus is decidedly endemic. The disease was
also said to be formerly extremely common, especially
among animals, on the eastern end of Long Island, but that
it has now become quite rare in that locality. This alleged
local prevalence on Long Island has been, in this country at
least, one of the most frequently used arguments in sup-
port of soil infection, but Overton1 has shown that the
published statements cannot be verified, and that the disease
has not been especially prevalent in that locality. I have
corresponded with various officials in New Jersey, but have
never been able to obtain any evidence of the alleged local
infection at Red Bank. Tetanus is more common in the
southern than in the northern parts of the United States,
and is a very important cause of death in the West Indies.
Before the organization of the present efficient health depart-
ment in Havana there were often 200 deaths or more annually
from infantile tetanus, due to infection of the navel. It has
long been recognized that tetanus follows wounds in which
dirt is forced deep into the tissues, and that garden earth is
especially dangerous. Even well-established evidence of per-
sistent local infection does not prove growth in the soil. It
might be due to a great variety of causes.

Thus it has been explained that the bacilli of tetanus are
very widely distributed because these bacteria are natural

1 Overton, Long Island M. J., Brooklyn, 1907, I, 176.

6 THE SOURCES AND MODES OF INFECTION

inhabitants of the healthy intestines of domestic animals,
particularly the horse. Hence they are found in profusion
wherever the manure from these animals falls, and soiled skin
and clothing are not likely to carry them. Theobold Smith,1
however, says that there is no evidence that the tetanus
bacilli are normal inhabitants of and multiply in the intes-
tines of animals. He inclines to the view that their home is
in the soil. Vincent2 after introducing tetanus spores into
the stomach of a rabbit could find no evidence of multipli-
cation, and he, too, thinks a saprophytic existence probable.
The tetanus bacillus forms spores which may retain their
vitality for 16 years, so that it is not surprising that lands
have been known to remain infected for several years.3 These
spores, or the bacilli, are said to have been found in gelatine,4
in blank cartridges,5 and on balls of lamp wick used in Havana
for tying the umbilical cord.6 While the soil, and dirt gener-
ally, contain tetanus bacilli or their spores, there is no direct
evidence to show that they are propagated outside of the body.
The fact that they do not grow in the presence of air would
indicate that this is not the case, and the distribution of the
disease and its comparative rarity would also lead to this
conclusion. It is not, of course, to be denied that the tetanus
bacillus may lead a saprophytic existence. It is very pos-
sible that it may do so, but it certainly can be affirmed that
at present we have no proof that it does so, and all observed
facts relating to the bacillus, or the disease caused by it, may
be explained without assuming any such hypothesis.

It is suggestive that the pathogenic bacteria which are
ofteaest assumed to grow in the soil are the very ones the

Theobold Smith, J. Am. Ass., Chicago, 1908, L, 929.
Vincent, Compt. rend. Soc. de biol. Par., 1908, LXV, 12.
Villar, J. Comp. Path, and Therap., Edinb. and Lond., 1897, XX.
Tuck, J. Path, and Bacteriol. Edinb. and Lond., 1904, IX, 38.
Dolley, J. Am. M. Ass., Chicago, 1905, XLIV, 466.
Junta Superior de Sanidad de la Isla de Cuba, Suplemento y Noto
Adicional, 1902-3, 4.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 7

spores of which may retain their vitality for years. Is it not
more likely that it is persistence of spores, rather than growth
of the bacilli themselves, that in most instances maintains
the soil infection ?

Typhoid Bacilli in Soil. — It has been amply demonstrated
that water, milk, soil and various other materials are, when
sterile, suitable media for the growth of the typhoid bacillus.
But this fact is of little practical importance, as sterile mate-
rials are not ordinarily found in nature, but on the contrary
almost everything which could possibly be considered a cul-
ture medium for typhoid and other disease germs is swarming
with bacteria, mostly of entirely harmless varieties. There
has been much painstaking work to determine whether ty-
phoid bacilli actually do grow, or even retain their vitality,
in or on a great variety of substances. The difficulties in
this sort of experimentation are considerable, and not the
least is that of picking out the typhoid bacillus from among
other forms. Robertson1 and Firth and Horrocks2 seem to
have made the most elaborate experiments in regard to its
growth in soil, and to have worked under more natural con-
ditions than most observers. Robertson found that by
moistening soil from time to time with bouillon he could
keep the bacillus alive for 11 months, and even cause it to
grow. Firth and Horrocks did not find any evidence of
increase in soil under a great variety of conditions. When
conditions were favorable it could be recovered up to 74 days.
In peat it could only be recovered after 13 days. More
recently Mair3 has been able to recover the bacillus from un-
sterilized soil in large numbers, for 20 days, and in small
numbers, up to 70 or 80 days. He found no evidence of
increase. Great care was taken not to introduce any nutri-
tive medium with the bacilli. Unlike most observers, Mair
found that in sterile soil the bacilli disappear more rapidly,

1 Robertson, Brit. M. J., Lond., 1898, I, 69.

1 Firth and Horrocks, Brit. M. J., Lond., 1902, II, 936.

3 Mair, J. Hyg., Cambridge, 1908, VIII, 37.

8 THE SOURCES AND MODES OF INFECTION

in 11 days in fact. He believes that this is due to the chem-
ical composition of the particular soil used. Smith1 working
with similar soil, unsterilized, could not recover the organism
after 25 days, and the average duration in the soil was 15
days. Calvagno and Calderini 2 spread upon the soil some
typhoid excreta. It was found that the bacilli could be
recovered from the surface of the soil for 12 to 20 days
and from the deeper portions, 20 centimeters, for 40 days.
Most observers, as Koch,3 Karlinski,4 Uffelmann,5 Martin,6
Pfuhl 7 and others agree that it does not grow in soil,
though it may retain its vitality at times for months.
Savage 8 found that it died rapidly in tidal mud, though a
few bacilli could be recovered after five weeks. Klein9
could not find the organism in dead animals buried in earth,
after 20 days, but Loesner 10 found it after 96 days.

Typhoid Bacilli on Vegetables. — Recently Creel n has
investigated the chance of vegetables becoming infected
with typhoid bacilli. He planted radishes and lettuce in
soil which was watered two or three days later with a
fecal emulsion mixed with a 24-hours-old agar culture of
the typhoid bacillus. Some of the plants were grown in-
doors and some in the open air more or less exposed to
sunshine. The leaves and stems were examined every 3

1 Smith, Rep. on Occurrence of Typhoid Fever in Belfast, 1902,
quoted by Mair.

2 Calvagno and Calderini, Ztschr. f. Hyg. u. Infectionskrankh.,
Leipz., 1908, LXI, 188.

Koch, Die Bekampfung des Typhus, Berl., 1903, 14.
Karlinski, Arch. f. Hyg., Munchen u. Leipz., 1891, XIII, 302.
Uffelmann, Centralbl. f. Bakteriol. [etc.], Jena, 1894, XV, 133.
Martin, Rep. Med. Off. Local Gov. Bd., Lond., 1900-1901, XXX,
508.

Pfuhl, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1902, XL, 555.

Savage, J. Hyg., Cambridge, 1905, V, 146.

Klein, Rep. Med. Off. Local Gov. Bd., Lond., 1898-9, XXVIII, 363.

10 Loesner, Arb. a. d. k. Gsndhtsamte, Berl., 1896, XII, 448.

11 U. S. Pub. Health and Mar. Hosp. Serv., Pub. Health Rep., 1912,
XXVII, 187.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 9

days for the bacilli, and they were found, but not constantly,
for periods varying from 10 to 31 days. The author con-
siders that these experiments demonstrate the possibility
of the transmission of typhoid fever by means of uncooked
vegetables grown in infected soil, and so they do. Yet it
often happens that the soil is manured a long time before
the seed is planted, so that the bacilli have a chance to die
out, and doubtless often the conditions are less favorable
than in the experiments, owing to excessive rains, or pro-
longed drought, or very hot weather; and it is also true
that the use of fresh night soil as a fertilizer in the more
civilized countries is rapidly diminishing, so that it is not
unlikely, after all, that raw vegetables are only occasionally
a source of typhoid fever. If, too, as now seems probable,
bacilli in feces are less resistant than those from cultures,
the danger would be less than at first sight appears from
these experiments. There can, however, be no doubt that
this is a real source of danger, and that human excrement
should not be used for fertilizing vegetables and low-growing
fruits which are to be eaten raw. The futility of attempting
to cleanse them by ordinary washing or rinsing is shown by
Creel, who without success tried to cleanse a lettuce leaf by
three washings with a pipette and by stirring in water.

Typhoid Bacilli in Feces. — According to Park,1 typhoid
bacilli soon die out in feces, usually in a few hours, but he
has recovered them up to the tenth day. He suggested that
this variation may depend on the constitution of the feces.
On the other hand, Levy and Kayser 2 note the persistence
of typhoid bacilli in a cemented privy vault up to 5 months,
and Pfuhl 3 recovered them after 3 months from feces mixed
with garden earth. Dele*pine4 states that typhoid bacilli

1 Park, J. Am. M. Ass., Chicago, 1907, XLIX, 852.

2 Levy and Kayser, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena,
1902, XXXIII, 489.

» Pfuhl, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1902, XL, 556.
4 DelSpine, Rep. Health of Manchester, 1907, 82.

10 THE SOURCES AND MODES OF INFECTION

will survive in a privy for a year. His investigation was
made in 1898, 13 months after the use of the privy by a
patient. Meanwhile the vault had been disinfected several
times. Apparently, however, its use by a carrier was not
excluded. Calvagno and Calderini recovered typhoid ba-
cilli from a privy vault for 30 days and from a barrel for
25 days. Morgan and Harvey l could not recover typhoid
bacilli from a privy vault later than 18 days. Semple and
Greig 2 found that urine containing 60,000,000 bacilli per
cubic centimeter kept at 80° F. was free from them in 72
hours, and that feces under similar circumstances lost them*
in 9(Thours. Mosebach 3 sought for typhoid bacilli in privy
vaults belonging to houses where carriers resided but where
there had been no frank case of typhoid fever for years, and
had no difficulty in recovering the germs. Johnstone 4 in
studying the Janet Hill outbreak, which was probably due
to carriers, could find no bacilli in 6 samples of soil from
the yard of a house where there had been persistent typhoid
fever. Rogers 5 found that the bacillus lived only a few
days in filtered septic tank effluent.

Typhoid Bacilli in Water. — Jordan and Russell6 imitated
natural conditions by enclosing inoculated water in colloidal
sacs to permit of osmosis, and these were placed in the
Chicago River, a sewage-polluted stream. They could
recover the bacilli for from 3 to 7 days only. Russell and
Fuller 7 repeated these experiments with substantially the
same results, though they kept the bacillus alive in lake

1 Morgan and Harvey, J. Roy. Army Med. Corps, 1909, XII, 587.

2 Semple and Greig, Sc. Memoirs, Med. and San. Dept., Gov. India,
1908, XXXII, 40.

3 Mosebach, Centralbl. f. Bakteriol. [etc.], I Abt., Jena, 1909, LII,
Orig., 170, 773.

4 Johnstone, Rep. Med. Off. Local Gov. Bd., Lond., 1909-10,
XXXIX, 166.

6 Rogers, Brit. M. J., Lond., 1903, II, 639.

6 Jordan and Russell, J. Infect. Dis., Chicago, 1904, I, 641.

7 Russell and Fuller, J. Infect. Dis., Chicago, 1906 [Suppl. No. 2], 40.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 11

water from 8 to 10 days. Houston1 has made some careful
quantitative studies of the life of the typhoid bacillus in
raw London tap water. In eighteen series of tests the
average reduction during the first week was 99.9 per cent,
but a few could be recovered up to the eighth week.

More recently Houston2 has compared the bacilli fresh
from a carrier with the same strain after cultivation and he
found that while the former usually lived over 5 weeks, the
latter in 9 tests died in 1 week, in 3 tests in 2 weeks and in 1
test in 3 weeks. He also drank, without bad results, on 10
different days a half pint of water to each half pint of which
had been added, from 23 to 28 days previously, fresh urine
from the carrier containing 218,000,000 bacilli.

Morgan and Harvey3 came to a similar conclusion as
regards the comparative viability of typhoid bacilli from
cultures and from excreta. They consider that cultures are
worthless for practical tests of the viability of the bacillus.
As most of the tests have been made with cultures, it seems
highly probable that the persistence of the typhoid bacillus
outside of the body is not as great as many of the experi-
ments would indicate. Houston 2 has made careful search
for typhoid bacilli in raw Thames and Lea waters which
are more or less polluted with sewage. From 215 samples
of water aggregating 116,900 cubic centimeters, 20,771
colonies were isolated for study, but only 2 proved to be
typhoid bacilli. These examinations continued through
the year.

Wilson and Dickson4 report that they have developed a
new method for isolating the typhoid bacillus from water
by which they can recover a single bacillus from 100 c.c. of
water. They claim that they can, by their method, show
the presence of the bacillus when other methods fail to

1 First Rep. on Research Work, Met. Water Bd., Lond.

2 Seventh Rep. on Research Work, Met. Water Bd., Lond.

3 Morgan and Harvey, J. Roy. Army Med. Corps, 1909, XII, 587.

4 Wilson and Dickson, J. Roy San. Ins., Lond., 1911, XXXII, 9.

12 THE SOURCES AND MODES OF INFECTION

reveal it and that conclusions as to the rapid disappearance
of the bacillus, based on other methods, are consequently
erroneous. They found that the typhoid bacillus could
be recovered from water up to the twenty-third day.
But it may be that a single bacillus in 100 c.c. is so
extremely unlikely to cause infection that somewhat
coarser methods may yield results more indicative of the
practical danger.

Field 1 found that typhoid bacilli would survive in sea-
water for from 6 to 8 days, but that 50 per cent died in 24
hours. Herdman and Boyce 2 found no evidence of the mul-
tiplication of the typhoid bacillus in sea-water but, on the
contrary, it entirely disappeared in about 3 weeks, the
larger part of the reduction taking place during the first
few days.

Recent careful observations have shown that in potable
waters typhoid bacilli are able to survive much longer in
winter than in summer. Ruediger3 had noticed that the
number of colon bacilli in the river supplying Grand Forks,
N. D., was 4 or 5 times greater in winter than in summer.
Believing that the excess in winter was due to the longer
life of the bacilli at that season he tested the question by
hanging in the river at different seasons celloidin sacs con-
taining large numbers of typhoid bacilli. It was deter-
mined that in the winter, when the river was covered with
about 30 inches of ice, the typhoid bacilli lived 5 or 6 times
as long as in the summer. In summer he floated down the
river two dialyses containing large numbers of bacilli and
after 54 hours only 0.11 per cent was left in one and 0.013
in the other. Ruediger attributes the greater longevity
in winter in part to the inter-relation of different organisms
and in part to the cutting off of the sun's rays by the ice.
Houston4 placed typhoid bacilli in tanks at different temper-

1 Field, Rep. Dept. Health, City of New York, 1904, I, 451.

2 Herdman and Boyce, Rep. Thompson Yates Lab., 1898-9, I.

3 Ruediger, J. Am. Pub. Health Ass., 1911, I, 411.

4 Seventh Rep. on Research Work, Med. Water Bd., Lond.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 13

atures and found that a reduction which took place in 2
weeks at 64.4° F. required 3 weeks at 50°, 4 weeks at 41°
and 5 weeks at 32°.

Wheeler finds that in well water with considerable pollu-
tion, at room temperature, and with the exclusion of light,
a considerable increase of typhoid organisms may take
place. Konradi also claims that this bacillus can main-
tain a saprophytic existence in water, but his methods have
been criticised, and, in some experiments at least a good
deal of nutrient material was added to the water with the
organisms. The report on typhoid fever in the District of
Columbia 1 quotes from Kubler and Neuf eld, and Stroezner
and Tavel, instances of alleged longevity of the typhoid
bacillus in well water or, in Tavel's case, in tap water, but
secondary contact infection was not in any instance abso-
lutely excluded. On the other hand, the infection in a
reservoir in Scranton was proved to have died out within 8
weeks.2

Pf uhl 3 found bacilli in tap water after 28 days, but not
after 31 days. In artificially inoculated seltzer water it
lived for 27 days. Hill,4 however, could not recover it from
various carbonated " soft drinks " after 14 hours.

Typhoid Bacilli in Ice. — Various writers have studied the
life of typhoid bacilli in ice, and Prudden,5 Winslow,6 Park,7
Jordan, Russell and Zeit,8 Clark,9 Smith and Swingle10 and
Wheeler11 have shown that they tend to disappear gradually,

U. S. Pub. Health and Mar. Hosp. Serv., Hyg. Lab. Bull. No. 35, 178.

N. York M. J. [etc.], 1907, LXXXV, 1025.

Pfuhl, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1902, XL, 55.

Hill, Rep. Bd. of Health, Bost., 1904, 53.

Prudden, Med. Rec., N. Y., 1887, XXXI, 341.

Winslow, J. Mass. Ass. of Bds. Health, Bost., XI, 133.

Park, J. Bost. Soc. M. Sc., 1899-1900, IV, 213.

J. Infect. Dis., Chicago, 1904, I, 660.

Clark, J. Mass. Ass. Bds. Health, Bost., XI, 124.
10 Smith and Swingle, Science, N.Y., 1905, n. s. XXI, 481.
u Wheeler, J. Med. Research, Bost., 1906, XV, 269.

14 THE SOURCES AND MODES OF INFECTION

somewhat as they do in the water from which the ice is
obtained. It must also be remembered, when considering
the possible danger from ice, that the experiments of
Sedgwick and Winslow and Clark show that when water
freezes 90 per cent of the contained bacteria are excluded
from the ice.

Typhoid and Oysters. — Herdman and Boyce 1 found that
in oysters artificially infected typhoid bacilli could be re-
covered up to the tenth day. The bacilli did not increase
in the oyster but probably perished in the intestine. In
clean water the oysters freed themselves from typhoid bacilli
in from 1 to 7 days.

Gorham 2 has made extensive studies of the contamina-
tion of oysters in Narragansett Bay which is considerably
polluted by sewage. No attempt has been made to isolate
the typhoid bacillus but attention has been devoted to B.
coli. It has been found that while there is not much sea-
sonal variation in the colon content of the water there is
a very great variation in the colon content of the oysters
which are quite free from these bacilli during winter
weather. The same phenomenon has been noticed by
Pease in New York and by Freeman in Virginia. Gorham
believes that during the winter the oyster assumes a con-
dition approaching hibernation and during this time the
ciliary movement ceases, and with it the current of water
over the gills, and feeding stops. No water is taken in
from the outside and the bacteria in the oyster are grad-
ually eliminated. Is it not safe to assume that the oyster
is equally free from typhoid bacilli in the winter, at least
in the northern United States?

Field 3 found that when typhoid bacilli were planted in
living oysters they rapidly died and none could be re-
covered after the ninth day. When the oysters were dead

1 Herdman and Boyce, Rep. Thompson Yates Lab. 1898-9, I.

8 Gorham. Am. J. Pub. Health, 1912, II, 24.

« Field, Med. News, N. Y., 1904, LXXXV, 571.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 15

or dying, there was a very considerable increase. Klein 1
found that in oysters kept in sea water typhoid bacilli would
live from 6 to 7 days, but if kept out of the water, for 1 1 days.
In other shellfish their life was longer.

Typhoid and Milk. — Sterile milk serves as an excellent
culture medium for the typhoid bacillus, but ordinary market
milk is not favorable for its growth, owing to the rapid pro-
duction of lactic acid. Bassenge 2 says that when milk has
soured to the extent of 0.3°-0.4° Soxhlet, and has continued
in this condition for 24 hours, the bacilli are destroyed. Neu-
feld 3 states that they usually disappear from ordinary milk
in from 2 to 3 days. Pfuhl 4 found the bacillus persisting in
the milk for 13 days. Rosenau and McCoy have studied
this question and reviewed the literature.5 They find that
raw milk, when first drawn, has a feeble antiseptic action,
and typhoid and dysentery bacilli, when added to it, decrease
slightly at times, but within 48 hours their numbers increase
enormously. Eyre 6 also states that the typhoid bacillus
may increase in milk to enormous numbers, but as the
milk he experimented with was drawn under careful aseptic
precautions, it is quite likely that his findings would not
obtain in ordinary milk, owing to the hostile influence of
lactic-acid and other bacteria.

If typhoid bacilli increase in number in ordinary market
milk, extensive outbreaks ought to be expected in our large
American cities, where the milk is handled by large dealers
drawing their supply from many producers situated at long
distances, so that the milk is from 48 to 72 hours old before

1 Klein, Tr. Path. Soc. Lond., 1905, LVI, 23; Med. Press & Circ.,
1905, LXXIX, 264.

2 Bassenge, Deutsche med. Wchnschr., 1903, XXIX, 675, 697.

1 Neufeld, Kolle u. Wassermann, Handbuch [etc.] Jena, 1903, II, 213.

4 Pfuhl, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1902, XL, 555.

5 U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab. Bull. No. 41,
449.

6 Eyre, J. State M. Lond., 1904, XII, 728.

16 THE SOURCES AND MODES OF INFECTION

it reaches the consumer. Any dilution ought to be more
than balanced by the alleged increase in the bacteria. But
nearly all of the American milk outbreaks reported in
Bulletin 41 of the Hygienic Laboratory were on small
routes where the interval between infection and delivery was
short. No outbreaks due to railroad milk were reported
from Boston, New York, Philadelphia, Chicago, Buffalo,
Baltimore, or St. Louis, and of one hundred and twenty-
nine outbreaks in American cities only two instances were
reported, namely in Washington, D. C., in which typhoid
infection was brought in over a railroad.

Boers,1 Bruck2 and Pfuhl3 have demonstrated the persist-
ence of typhoid bacilli in butter up to 27 days, but few if
any outbreaks have been traced to this article of food.

Mayer4 states that paratyphoid bacilli will retain their
vitality in dried human feces for a year and a half.

No bacteria can grow except in the presence of moisture,
so an increase of typhoid bacilli on clothing, furniture, wood-
work, etc., is not to be looked for. The duration of life under
such conditions is sometimes shorter and sometimes longer
than it is in the presence of moisture. This will be discussed
further in another connection.

It must be confessed that the experimental evidence relat-
ing to the growth and vitality of typhoid bacilli outside the
body is by no means conclusive. The evidence seems to be,
however, that they rarely if ever increase in numbers, and in
most instances they tend to die off, and that quite rapidly,
often in a few days, or even hours.

Epidemiological Evidence. Soil. — There is not much
epidemiological evidence that typhoid bacilli retain their
vitality outside of the body for more than a few weeks or

1 Boers, cited in U. S. Pub. Health and Mar. Hosp. Serv. Hyg.
Lab. Bull. No. 41, 24.

2 Bruck, Deutsche med. Wchnschr., 1903, XXIX, 460.

3 Pfuhl, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1902, XL, 555.

4 Mayer, Munch, med. Wchnschr., 1908, LV, 2218.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 17

perhaps occasionally for a few months, and there seems to
be no such evidence of their increase, except sometimes in
milk. Westcott x reports an instance where a well continued
infected for 20 months. It was claimed that the conditions
were such that continuous infection could not have taken
place, but this does not appear to have been the fact.
The numerous instances given in the Report on Typhoid
Fever in the Spanish War2 as showing the growth of typhoid
bacilli in the soil, are by no means conclusive. Infected
soil was supposed to have caused outbreaks in various
army corps, but other sources could not be in any case
excluded, and the chances of other modes of infection in
an army are very great. The infection was supposed to
have remained many months, and it is possible that it may
have in some instances done so. As a matter of fact 90 per
cent of the volunteer regiments in the Spanish War sooner
or later became infected whether encamped on a polluted
site or not. Koch 3 believed from epidemiological as well as
from bacteriological evidence that it is very rare for an out-
break to be due to long continued soil pollution, and that the
possibility of the growth of the bacillus outside the body may
be neglected. In 1902, Koch undertook the investigation of
the typhoid fever which had prevailed for some time in sev-
eral villages in Trier. As a result of his labors, every typhoid
case and typhoid carrier in four of the villages was isolated,
and the outbreak ceased, showing that it was due entirely to
contact infection and not to soil infection.

Epidemiological Evidence. Water. — Numberless out-
breaks of typhoid fever have been traced to infected waters.
In some instances the pollution of rivers is continuous, and
the cities supplied from them suffer from a uniformly high

1 Westcott, J. State M. Lond., 1899, VII, 104.

2 Abstr. of Rep. on the Origin and Spread of Typhoid Fever in
U. S. Military Camps during the Spanish War of 1898, Wash., 1900,
20&-209.

1 Koch, Die Bekampfung des Typhus, Berl., 1903, 14, 19.

•18 THE SOURCES AND MODES OF INFECTION

death rate from this disease. In most instances water out-
breaks are of an explosive character, the onset is sudden and
they often end suddenly, though sometimes the decline is
gradual owing to the development of a certain number of
secondary cases due to contact infection. A few days* or
sometimes a single day's pollution is sufficient to account
for the whole outbreak. Three outbreaks of this kind have
occurred in Providence, and in one instance it was clearly
traced to the throwing upon the banks of the river of the
excreta from an infected family. The Providence watershed
covers about ninety-six square miles, and had upon it in 1900
a population of some 35,000, a large part in villages along the
banks, but some scattered in hamlets and farms over the
whole area. The death rate from typhoid fever in Rhode
Island is not less than 20 per 100,000 in the country districts,
which means seven deaths per annum on the watershed. As
the fatality is not over 10 per cent, there must be seventy
cases, and if we add the "carriers/' the number of persons
each year distributing typhoid bacilli on the watershed must
be at least one hundred, and perhaps very much greater. It
is also certain that disinfection of excreta is practiced to such
a slight extent as to accomplish very little. The fact that
for years the city has had no outbreak of disease, and no
excess due to the water, as is shown by the typhoid death
rate which for several years has varied from seven to twenty-
six per one hundred thousand, indicates that the typhoid
bacilli, which are being continually deposited on the water-
shed, fail to multiply. Exactly similar conditions prevail on
the watersheds of Pawtucket, Newport, Woonsocket, Hart-
ford and New Haven, only to mention those cities in my
neighborhood with which I am personally familiar. And the
general testimony of all epidemiologists is that municipal
water supplies are never continuously infected unless com-
paratively fresh excreta from typhoid-infected persons pass
directly and continuously into them. Nevertheless it is cer-
tain that typhoid bacilli must be continually discharged onto

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 19

the soil, and we are justified in considering the freedom from
infection of the surface waters coming from such areas a
strong evidence that the growth of the typhoid bacillus out-
side the V>r>Hy_ flnpg nnf. fftrnmonlv occur, and is a negligible
factor in the causation of the disease.

Cholera. — Early investigators, as Nicati and Rietsch,
working with sterilized soil and water, found that cholera
spirilla would live outside the body sometimes as long as 2
months. But all the more recent workers agree that under
natural conditions, in unsterilized materials, the life of the
organism is quite short. Loesner1 recovered the germs from
dead bodies, which had been artificially injected, as late as
the twenty-eighth day. Houston2 says that they usually die
off in the surface layers of the soil in 12 days, though they
may be kept alive longer if the soil is watered with liquid
manure. Though Heiser3 states that the spirillum was found
in the quiet water in the bends of the Passig river, no evi-
dence was presented to show that it grew there. Gotschlich4
states that the spirillum is rarely found in feces for more than
three days, and quotes Abel and Draer, Claussen and Dun-
bar, and refers to Koch, as stating that it dies in dirty canal
water in 24 to 30 hours. In unsterilized milk it may live
from 1 to 2 days, but dies as soon as the milk becomes sour.
All these agree that there is not the slightest evidence that
the cholera spirillum can increase in numbers outside of the
body. On the other hand, Emmerich and Gemiind5 claim
that it does increase in numbers in the soil, and may be found
for two and one-half months. Paladino-Blandini6 also states

Loesner, Arb. a. d. k. Gsndhtsamte. Berl., 1896, XII, 448.

Houston, Rep. Med. OS. Local Gov. Bd., Lond., 1898-9, XXVIII,
413

Heiser, Philippine J. Sci. (Med.), 1908, III, 92.

Kolle u. Wassermann, Handbuch [etc.], Jena, 1904, IV, 108.

Emmerich and Gemiind, Miinchen med. Wchnschr., 1904, LI, 1089,
1157.

6 Centralbl. f. Bakteriol. [etc.], I, Abt. Ref., Jena, 1905, XXXVI, 53.

UNIVERSITY OF CALIFORNIA
«LF>,et TIWENT OF CIVIL ENGINEER

20 THE SOURCES AND MODES OF INFECTION

that it may grow in polluted soil. Koch1 says that it grows
only in the human body.

Cholera and Soil. — There is certainly a great deal of clin-
ical evidence that it does not increase in the soil, but on the
contrary speedily dies out. The epidemic of cholera in
Europe in 1885 was very widespread in Italy and Spain.
How extensively it prevailed and what a great number of
towns and villages were infected are well shown in the excel-
lent report prepared by Shakespeare.2 Nevertheless by the
succeeding year it had entirely disappeared from Spain, and
largely from Italy. Turkey had a similar experience.3 We
know that with the sanitary conditions prevailing in those
countries at that time the soil must have been infected with
cholera in countless places. But cholera rarely recurred, and
when it did, it was in large cities, where, the most probable
explanation is, it was maintained during the interval by mild
unrecognized cases or latent infections. The extensive epi-
demic in the United States in 1873 was not followed by a
recurrence in the succeeding year. Not only is epidemiologi-
cal evidence strongly against the saprophytic existence of
cholera in temperate climates, but it is equally so for tropical
regions. The great outbreak in the Philippine Islands in
1902-3 attacked hundreds of villages, and soil infection was
universal, yet the disease speedily died out all over the islands.4
In 1905 there was a similar experience.

In the Philippine Islands during the latter outbreak
it was believed that the cooked food offered for sale in the
streets was a frequent vehicle of cholera germs, and a num-
ber of samples, particularly of boiled rice, were found to con-
tain the spirilla.5 The rice was probably contaminated by the

1 Koch, Die Bekampfung des Typhus, Berlin, 1903, 14.
1 Shakespeare, Rep. on Cholera in Europe and India, U. S. Gov.
Print. Off., Wash., 1890.

1 Clemow, Tr. Epidemiol. Soc., Lond., 1904, n. s., XXIII, 223.
4 Woodruff, J. Am. M. Ass., Chicago, 1905, XLV, 1160.
8 Maus, Med. News, N. Y., 1902, LXXXI, 318.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 21

hands of sellers or purchasers. Whether the spirilla increased
in numbers is not known.

Mediterranean Fever. — Mediterranean or Malta fever
has long been recognized as endemic at various points on the
shores of the Mediterranean, particularly on the island of
Malta. It shows all the characters of " endemic " disease,
being confined to certain areas and exhibiting little tendency
to pass to other parts of the world, or to extend by contagion.
The micrococcus, M. melitensis, which is its cause, was dis-
covered by Bruce in 1887 and has been very carefully studied
by an English commission acting under the supervision of
the Royal Society. This commission, as have independent
observers, has given much study to the vitality of the spe-
cific organism of the disease outside of the body. The evi-
dence seems to be that it tends, like other disease germs, to
perish when removed from the body and deprived of nourish-
ment and moisture and exposed to light and heat, but it may
retain its vitality for weeks and even months under certain
conditions, behaving in this respect much like the typhoid
bacillus.1 No direct experiments seem to have been made to
determine whether it ever maintains a saprophytic existence
in the soil, but its habits of life are such that it is a fair in-
ference that it can rarely do so, and that such saprophytic
growth, if it ever takes place, is of no greater practical impor-
tance than is the saprophytic growth of typhoid germs. The
micrococcus was carefully sought for in water and in dust,
but was never found.2 The disease is referred to in this con-
nection because, on account of its localization and slight
apparent contagiousness, it was believed by many medical
men, as well as the laity, to spring from the soil. Yet the
work of the English investigators has conclusively shown
that the chief source of the disease is the herds of infected
goats. It is by the infected milk of these that the disease
is transmitted to human beings. It is barely possible that

1 Report of Commission of Royal Society, 1901, Pts. I and II.
» Bruce, Nature, Lond., 1908, LXXVIII, 40.

22 THE SOURCES AND MODES OF INFECTION

the fever may be spread to some extent by contact with
infected urine either of goats or of men, just as typhoid
fever frequently extends by contact infection, but the fact
that thousands of infected men have been invalided home
to England without any extension of the disease in that
country would indicate that such occurrence is extremely
rare.

Bubonic Plague. — The germ of bubonic plague is not so
resistant as is that of typhoid fever, nor yet is it of such feeble
vitality as that of cholera. It is rather susceptible to disin-
fectants, to high temperature and to drying, but in a moist
condition, particularly at low or moderate temperatures, may
remain alive for some months.1 The endemicity of the dis-
ease in many localities has led some to assume that it devel-
ops in the soil, but the most careful students see no necessity
for assuming soil infection to account for its diffusion, and
there is ample positive evidence that plague is derived from
other sources. Yet, in view of the fact that soil infection
has been so much discussed, it is rather remarkable that so
few actual experiments have been made to test the theory.
Perhaps it is because such experiments are difficult and
those who are most competent to make them have thought
their time better occupied with work giving better promise
of positive results. Elliot 2 found that soil naturally infected
would cause the disease in rats after an interval of a month,
and Watkins-Pitchford 3 in some careful experiments found
that inoculated soil retained its virulence for four weeks, but
not for five weeks. The soil was not sterilized. Gladin4
found the bacillus alive in unsterile moist earth after 2

1 Rosenau, U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab.
Bull. No. 4, 1901. See also Simpson, A Treatise on Plague.

2 Elliot, Lancet, Lond., 1905, I, 1562.

1 Watkins-Pitchford, Rep. Gov. Bacteriologist, Pietermaritzburg,
1903 [Report on Plague, 31].

4 Gladin, Centralbl. f. Bakteriol. [etc.], I, Abt. Orig., Jena, 1898,
XXIV, 588.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 23

months, and Rosenau l kept it alive a long time in cool moist
garden earth, and the Indian Plague Commission, 1901, did
the same in moist sterile cow dung. Other experiments by
Mackie and Winter in Bombay, cited in the Journal of Hy-
giene,2 were with grossly infected cow dung from the floor of a
native house. No pest bacilli were recovered by culture on
inoculation after 96 hours. Further careful experiments have
been made by the Indian Commission of 1905. Their con-
clusions are as follows:

"Floors of cow dung grossly contaminated with the bacillus
of plague remain infective for 48 hours ; floors of a sort of na-
tive cement for 24 hours, the infectivity being tested in each
case by inoculation. The floors were infective to animals
allowed to run on them for only half the above time."

Thus there appears to be no bacteriological evidence that
the bacillus of plague grows outside of the bodies of living
animals, and a great deal of evidence that when separated
from the body it tends to die off more or less rapidly and fre-
quently very rapidly. The Indian Plague Commission con-
siders that reports of soil infections are unworthy of credence
unless continuous and careful observations on the presence
of rats and fleas have been made.

Dysentery Bacillus. — One form of dysentery is caused by
a bacillus belonging to the colon group, and it has a number
of sub-varieties. It is not quite so resistant as the typhoid
bacillus, but it has been known to survive all winter in damp
earth.3 It is said that in Japan local outbreaks often persist
longer than do outbreaks of cholera, perhaps due to the higher
resistance of the germ.4 The bacilli appear to be easily de-

1 Rosenau, U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab. Bull.
No. 4, 1901, 9.

2 J. Hyg., Cambridge, 1906, VI, 511.

3 Schmidt, Centralbl. f. Bakteriol. [etc.], I, Abt. Orig., Jena, 1902,
XXXI, 522.

* Eldridge, U. S. Pub. Health and Mar. Hosp. Serv. Pub. Health
Rep., 1901, 1.

24 THE SOURCES AND MODES OF INFECTION

stroyed by other bacteria, for they can rarely be found in
feces after two days.1

Dysentery Amebae. — One form of tropical dysentery is
caused by an ameba, a protozoan, not a bacterium.

There has been much dispute concerning these amebse,
but it is now, however, believed by the best observers that
Entameba coli, commonly found in the intestines, and else-
where, is a harmless parasite, but that E. histolytica and
E. tetragena, and perhaps E. minuta and E. nipponica, are
true parasites and pathogenic. Craig 2 in recent work
shows that in all probability the amebae which cause dis-
ease in man do not grow readily, if at all, outside of the
body. As most convincing evidence he alleges the ina-
bility to grow in cultures the pathogenic forms in regions
where there are no saprophytic forms to contaminate the
cultures and deceive the observer.

Bacteria of Suppuration. — The formation of pus in
wounds, abscesses, or elsewhere, is practically always the
result of infection by bacteria. Many varieties of bacteria
may cause suppuration, but a few species such as the Micro-
coccus aureus, M. albus and M. citreus, and Streptococcus
pyogenes, are by far the most common cause of this proc-
ess. According to a re*sume* given by Gotschlich3 they are
constantly found growing in the skin and on the mucous
surfaces. References are of course given in his article to
numerous original investigations, but among more recent ob-
servations may be mentioned those of Ruediger,4 Gordon5
and Hess.6 These bacteria are also found in the tonsils and

Kruse, Deutsche med. Wchnschr., 1901, XXVII, 370, 386.

The Parasitic Amoebae of Man, Phila., 1911, 58.

Gotschlich, Kolle and Wassermann, Handbuch [etc.], Jena, 1902,
I, 147.

Ruediger, J. Am. M. Ass., Chicago, 1906, XLVII, 1172.

Gordon, Rep. Med. Off. Local Gov. Bd., Lond., 1904-5, XXXIV,
387.

6 Hess, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1907,
XLIV, 1.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 25

lymph glands, apparently remaining latent for long periods
of time, i.e., not causing suppuration. Bacteria such as the
above, which are the constant parasites, or perhaps rather
commensals of man, are naturally to be looked for in the
vicinity of man and on the surfaces of the body, on clothing,
utensils, furniture; and the dust and dirt of all places fre-
quented by human beings are found to contain more or less of
them. They may also be found in polluted waters. But
whether under natural conditions they are commonly able
to maintain a saprophytic existence is another matter. Judg-
ing from what I have read in the text-books I should suppose
that the pus organisms are not so limited as to the conditions
of their growth as are most disease-producing bacteria. They
are not so dependent on a high and even temperature or on
the composition of the medium on which they grow. I should
suppose that they would be more likely to maintain a sap-
rophytic existence than most other pathogenic organisms, yet
I do not know that such existence for them has ever been
demonstrated. In fact Gotschlich1 says that they are not
saprophytes. Bacillus pyocyaneus has, however, been found,
by Gorham, growing in a heap of moist rags at a paper mill.
Diphtheria. — Diphtheria was not so very long ago believed
to be a " filth disease," that is, its germs were supposed to
have a habitat outside of the body in various forms of " dirt."
This theory was common during my medical-school days, and
when I began health-department work in 1884, I tried to fit
the facts as I saw them to this theory. But they did not
fit, and the impression continued to grow that diphtheria
was a purely contagious disease. The life habits of the diph-
theria bacillus indicate that while it is more resistant than
some other disease-producing organisms next to be men-
tioned, and somewhat easier to cultivate, it is very unlikely
that it is able to propagate itself outside of the body, except
at times in milk. Houston,2 while he does not consider his

1 Gotschlich, Kolle and Wassermann, Handbuch [etc.], Jena, IV, 173.
* Houston, Loc. Gov. Bd. Rep. of Med. Off., 1898-9, XXVIII, 413.

26 THE SOURCES AND MODES OF INFECTION

experiments conclusive, believes that the bacillus of diph-
theria dies off very quickly in unsterilized soils of various
kinds. Graham-Smith1 cites a number of writers concerning
the vitality of the organism in water and milk. According
to Seiler and Stoutz, it multiplies in sterilized water for a
while, but Montefusco and D'Espine and Marignac found
no evidence of multiplication even in distilled water, and in
polluted water it dies in 6 days. Schottelius reported, con-
trary to general experience, that the bacillus multiplies more
rapidly in raw than in sterile milk, while Montefusco found
no multiplication in raw milk after 3 days, and Rubinstein
found that the bacilli died in 24 hours. Eyre showed that
in milk drawn in as sterile a condition as possible the diph-
theria like the typhoid bacillus undergoes rapid multiplica-
tion. Kersten,2 on the other hand, reports that diphtheria
bacilli will persist in raw milk for 72 days, and though
they undergo no increase at first, do so later. Montefusco
found that they died in fresh bread in 24 hours. Except in
fairly fresh milk at room temperature, it is unlikely that the
bacilli of diphtheria multiply outside of the body.

Other Diseases. — Tuberculosis, pneumonia, influenza,
cerebro-spinal meningitis, gonorrhea and syphilis are caused
by organisms which are difficult to cultivate, and I think all
bacteriologists agree that it is futile to seek for their habitat
outside of the bodies of men or other animals. As for the
protozoan diseases, such as malaria and sleeping sickness,
such a habitat is still more improbable. It is only recently,
and after laborious experiment, that any of this class of
organisms have been cultivated in the laboratory. That they
grow outside of the body under ordinary conditions is in the
highest degree improbable. There is one protozoan, however,
the dysentery ameba, which, if recent observations are cor-
rect, does live outside of the body. But this organism belongs

1 Nuttall and Graham-Smith, The Bacteriology of Diphtheria, Cam-
bridge, 1898, 171.

2 Kersten, Arb. a. d. k. Gsndhtsamte., Berl., 1909, XXX, 341.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 27

to an entirely different class from the blood parasites, and
while the facts so far known render it not improbable that
the ameba does grow outside of the body, the facts in regard
to blood parasites are against any such hypothesis.

Summary of Laboratory Evidence. — It appears, then, that
so far as experimental evidence is concerned there is no war-
rant for assuming a source for the common infectious diseases
outside of animal bodies. It is only with extreme difficulty
that a few of the blood parasites belonging to the protozoa
can be cultivated, and the cultivation of many bacterial
forms is strictly limited, so that it is hardly possible to ima-
gine their maintaining a saprophytic existence. It is true
that the bacteria of typhoid fever and perhaps cholera, dysen-
tery, and diphtheria may be conceived of as growing outside
of the body under natural conditions, but such growth, if it
ever occurs, must be rare.

Lack of Epidemiological Evidence. — Nevertheless during
the larger part of the nineteenth century it was common to
seek such an outside source for most of the infectious diseases.
The filth theory of disease, the vogue of which was largely
due to Chadwick, Murchison and Pettenkoffer, assumed that
the specific poison of many of our common infectious diseases,
and particularly of typhoid fever and cholera, developed in
a contaminated soil, or in other forms of filth. There was at
that time no knowledge of the micro-organisms which cause
disease, and the theories of the origin of disease, so far as
they had any basis at all, depended on clinical or epidemio-
logical evidence. There was, it is true, some epidemiological
evidence for believing that typhoid fever and cholera could
develop in filth, for both of these are excrement-borne dis-
eases, and it was difficult to decide without any knowledge
of the bacteria which cause them whether they developed
in filth or were merely transmitted in filth. But as regards
typhus fever, smallpox, scarlet fever, measles, diphtheria
and similar diseases, there is really no epidemiological evi-
dence to suggest that they develop outside of the body. If

28 THE SOURCES AND MODES OF INFECTION

a disease does have its source in the outer world, we should
expect to find it localized, attached to a definite locality,
endemic, as it was formerly called. The somewhat well-de-
fined infected area we should expect to maintain its infection
for some time, and we should expect persons coming into the
area to become infected. But if we watch the outbreaks,
especially the smaller outbreaks, of such diseases as measles,
smallpox, scarlet fever, diphtheria, etc., which occur in dif-
ferent sections of cities, or in villages, we find nothing to
indicate place infection. A " pin map " of these diseases in
a city shows first one or two pins, indicating infected houses,
then more appearing day by day in the surrounding section,
until there are ten or twenty or a hundred cases within
the radius of a block or a few blocks or half a mile. The
scattered groups of houses are of a somewhat circular form,
a little denser nearer the center. The outbreak lasts a few
weeks, or two or three months, and then disappears, only to
reappear in another part of the city. In village outbreaks
contagion is usually more clearly traced. The importation
of the disease and the sequence of the earlier cases are often
made out. The outbreak lasts only for a moderate time, and
then the disease disappears, usually not to return for a con-
siderable time, often for years. There is nothing to indicate
soil infection. If these diseases really come from privy
vaults, sink drains and garbage heaps, we should expect a
very different distribution in the house from what is actually
found. The ground floor is not oftener infected than the
upper floor, nor (for some might say that rising currents carry
disease germs to the top of the house) the upper than the
lower. But what is most important, when one family in a
house is attacked with such a disease as diphtheria, we should
expect the other families to be usually attacked also, if the
disease is a disease of locality. But in Providence statis-
tics for the last twenty years show that in scarlet fever and
diphtheria in only about seven per cent of the houses does
the disease extend from one family to another. In most of

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 29

these cases of extension, direct contact between members of
the families is shown to have taken place.

In such important diseases as smallpox, measles and scarlet
fever, the germs of which have not been isolated, as well as
in typhus fever, diphtheria and whooping cough, epidemio-
logical evidence of an extra-corporal origin is entirely lack-
ing. Epidemiological and laboratory evidence are against
the growth of disease germs outside of the body under
ordinary circumstances. The notion still common, even
among physicians and health officers, that these infectious
diseases are filth diseases, as that term is ordinarily under-
stood, is absolutely without foundation.

Typhoid Fever and Cholera. — Typhoid fever and cholera
not infrequently appear in small contact outbreaks very
similar to those of scarlet fever and diphtheria, but being
excrement-borne, contact infection is not so common as in
the latter, while infection by water and food is more common.
Some outbreaks of typhoid fever may undoubtedly be inter-
preted as due to soil infection, but they can also be explained
otherwise, and the bacteriological evidence is so strongly
against the soil hypothesis that we are not justified at
present in assuming it. There is no evidence that cholera
finds a habitat in the soil of temperate climates, and the
most careful observers in the tropics are, I think, agreed
that it is not different there. The same is true of bubonic
plague.

Yellow Fever and Malaria. — There is a class of diseases
for which there seemed, at one time, to be very strong evi-
dence that the cause which produced them had its origin out-
side of the body. I refer to yellow fever, malaria, sleeping
sickness and the blood-parasite diseases of cattle. Malaria
is one of the best defined, oldest, and best understood of
diseases. I suspect that for twenty-five hundred years what
has been known of malaria has decidedly colored prevailing
views as to the nature and source of many other infectious
diseases. Malaria is a typical endemic disease. Its localiza-

30 THE SOURCES AND MODES OF INFECTION

tion can scarcely be better illustrated than in my own neigh-
borhood. During the latter half of the nineteenth century
the southern part of New England suffered from extensive
outbreaks of malarial disease. The whole country was by no
means affected, but well-defined limited areas of infection were
to be noted all over the region, and in different portions of
some of the cities. Infection in many of these persisted from
year to year. People moving into them became sick, and
malaria was escaped by moving away. Most of these areas
were in close proximity to swampy land and marshes, others
were some little distance removed, but usually in the direc-
tion of the prevailing winds. Thus there was presented a
perfect picture of place infection. We were justified, in the
then existing state of knowledge, in assuming that the mala-
rial poison developed in the wet and marshy places which
were closely connected with these infected spots. We now
know that this assumption was incorrect, though it was very
near the truth, and that the germs of malaria do not
develop in the marshes, but that the mosquitoes which
carry the germs from one person to another do breed
there. As far as sanitation was concerned the error was
not serious. The new facts have simply enabled us to
work more economically, by determining accurately just
what wet places are dangerous. But the apparent certainty
that the malarial virus developed outside of the body had
enormous influence in encouraging the belief that other
disease poisons also had an extraneous origin. The history
of yellow fever is somewhat similar to that of malaria. It is
eminently a place disease. So are sleeping sickness and Texas
cattle fever and a number of other diseases. All of this group
we now know are transmitted by insects, and it is the insects
which have an extraneous existence and not the parasite of
the disease. The mico-organisms which cause these diseases
do not grow outside of the bodies of human beings or other
animals which serve as hosts, or of the insects which serve as
carriers.

LIFE OF DISEASE GERMS OUTSIDE OF THE BODY 31

Review of Evidence. — In reviewing this subject we are
forced to the conclusion that while it is possible that the
anthrax and tetanus bacilli and the pus-forming bacteria
may develop in the soil, there is no evidence that they com-
monly do so. It is also possible that the typhoid bacilli, and
to a still less extent the bacteria of cholera, dysentery and
plague, maintain a limited saprophytic existence, but this is
probably very unusual. There is ample epidemiological evi-
dence that in temperate climates such a source for these
diseases must be an almost infinitesimal factor in their
development. Probably the diphtheria bacillus never has a
saprophytic growth of any significance, unless possibly very
rarely in milk. As for tuberculosis, pneumonia, influenza,
cerebro-spinal meningitis, scarlet fever, typhus fever, small-
pox, whooping cough, gonorrhea and syphilis, malaria, yellow
fever and sleeping sickness, there is not the slightest reason
for supposing that they ever develop outside of the bodies of
animals.

Changes in Present Theories and Practice. — If these facts
are correct — and I can scarcely believe that any will seriously
contend that we have any evidence that an appreciable
amount of our common infectious diseases arises in the exter-
nal world — prevailing notions as to the sanitary functions of
the state should be decidedly modified. The laity and the lay
press still believe that most of the infectious diseases have
their origin outside of the body, in filth, or if admitting con-
tagion, attach equal importance to external sources of infec-
tion. And even very many health officials and some teachers
and writers on sanitation hold the same view. Municipal
sanitation and municipal cleansing are still synonymous
terms to many health officers. It is true enough that two or
three diseases, as typhoid fever and cholera, the germs of
which are found in human excrement, have markedly dimin-
ished as a result of improved methods of excrement disposal,
or because of the introduction of water supplies uncontami-
nated with human excrement. It is also doubtless true that

32 THE SOURCES AND MODES OF INFECTION

whatever promotes municipal cleanliness tends to promote
personal cleanliness, and real personal cleanliness, as we
shall see, is doubtless an important factor in the preven-
tion of contagion. Intelligent and effective control of
excreta disposal, of the mosquito nuisance in certain regions,
of the rat nuisance in the presence of plague, and perhaps
of the fly nuisance, are of great importance in the preven-
tion of certain diseases. In the northern parts of the
United States typhoid fever and dysentery and perhaps
diarrhea are the only diseases likely to be effected by
nuisance inspection and only when it is directed along cer-
tain lines. Except for a few diseases or except for very
indirect effects, the cleansing of streets, alleys, and back
yards, of dwellings and stables, the regulation of offensive
trades, and the prevention of nuisances generally, have, so
far as we can see, no effect on the general health, nor any
value in the prevention of specific diseases. While munici-
pal improvements such as the above are desirable, there
is little more real reason why health officials should work
for them than there is that they should work for free
transfers, cheaper commutation tickets, lower prices for coal,
less shoddy in clothing or more rubber in rubbers — all good
things in their way and tending towards comfort and health.

CHAPTER II.

CARRIERS AND MISSED CASES.

A Recent Discovery. — That there are occasionally seen
mild cases of the infectious diseases difficult or impossible to
recognize, has long been known. That such cases are rare
has always been generally believed. That the germs of dis-
ease can maintain themselves and increase in number in a
person without causing any symptoms at all, was until
recently scarcely thought possible, and the idea that such
latent infections are extremely common would have been
scouted as preposterous. Even to-day the facts are denied
by many sanitary officials, and there are comparatively few
who recognize the frequency with which mild atypical forms
of disease and healthy " carriers " of germs are found, or
realize the tremendous importance which such cases have in
the spread of the contagious diseases. Undoubtedly the most
fruitful medical discovery of the last century, and perhaps of
all time, was the discovery of the parasitic nature of the
infectious diseases. Probably the most important discovery
bearing on preventive medicine since the demonstration of the
bacterial origin of disease, is that disease germs frequently
invade the body without causing disease. The succeeding
pages will be devoted to a consideration of some of the data
available concerning the existence of mild cases and carriers.
The term ' ' carrier " is applied to those persons in whom patho-
genic micro-organisms exist, but who, nevertheless, show no
symptoms. Such carriers are rarely found by the health offi-
cer, and the very mild cases also naturally escape notice and
are hence called by the English " missed cases," i.e., cases

which fail of recognition.

33

34 THE SOURCES AND MODES OF INFECTION

Typhoid Fever not an Intestinal Disease. — Bateriolo-
gists and pathologists now consider typhoid fever essen-
tially an infection of the blood, rather than a disease of the
intestines. Houston l urged this view in his report of a
urinary carrier in 1899, and it was also set forth by Horton-
Smith in 1900.2 The latter considers that the bacilli proba-
bly pass through the intestinal wall without causing changes
there and proliferate in the mesenteric glands, whence they
pass into the circulation. Pyer's patches are secondarily
infected from the blood stream. A number of instances are
recorded in which the bacilli pass through the blood into
the fetus, causing a systemic infection but without intes-
tinal lesions. It is a fact that in adults the intestinal
lesions are frequently slight, and often the bacilli are few in
number in the feces, and sometimes they cannot be found
at all. Semple and Greig 3 report a case of typhoid fever
in which the bacilli were found in the blood from July 20th
to September 20th, but were never once found in the feces or
urine.4 So far as we know, the typhoid bacillus may enter
the lymph and blood from any portion of the alimentary
canal, and Semple and Greig, Lentz, Forster, Kayser and
others believe that it frequently enters through the tonsils;
and Pratt, Peabody and Long 5 say that there is no more
evidence of entrance through the intestines than through
the tonsils. Indeed, Lentz 6 says that he has shown, by the
finding of bacilli in the tonsils, that the latter may be the
portal of infection. He states that tonsillitis is of common
and early occurrence in typhoid fever and quotes Drigalski
as finding it in 40 per cent of all cases.

1 Houston, Brit. M. J., Lond., 1899, I, 78.

2 Horton-Smith, Brit. M. J., Lond., 1900, I, 827.

* Seinple and Greig, Sc. Memoirs, Med. and San. Dept., Gov. of
India, 1908, XXXII, 9.

4 See also Opie and Bassett, cited by Pratt, Peabody and Long.
6 Pratt, Peabody and Long, J. Am. M. Ass., Chicago, 1907, XLIX,
846.

• Lentz, Brit. M. J., Lond., 1910, II, 1501.

CARRIERS AND MISSED CASES 35

In any event, the bacillus is soon found in the blood,
and continues in this fluid through the acute stages of the
disease. Typhoid fever is, then, essentially a bacteremia.
Audibert l has recently shown that this at times presents a
subacute type. The bacillus may migrate from the blood
to any organ. As has been shown by Pratt, Peabody and
Long, a favorite habitat is the gall bladder. Pratt found it
in the gall bladder in 21 out of 30 cases. Kelly 2 says that
typhoid bacilli were found in 7 of 74 gall-bladder opera-
tions, and he states that in many cases there is no evidence
of intestinal infection. Primary cholecystitis has also been
reported by many others. Recently Elmer 3 reports a case,
occurring in a small milk outbreak, in which the only symp-
toms were cholecystitis. The distended gall bladder was
drained with immediate relief of the symptoms. The
bacillus may also infect the bones,4 kidneys,5 ovaries and
cerebro-spinal fluid.6

It would appear from the pathology that typhoid infection
deviating from the intestinal type is by no means uncommon.

Typhoid Bacilli in the Feces. — Typhoid bacilli were first
sought for and found in the feces, and it was then thought
that they freely developed in the intestinal contents. It
is now believed that, while some of the bacilli in the feces
may have their origin in the intestine itself or its contents,
the greater number come into the gut with the bile. It
was long known that they could be found in the bile, but
the importance of this fact and the relation of these bac-
teria to the gall bladder was not recognized until the car-
rier question came to the front. From the evidence fur-
nished by a considerable number of animal experiments, as

Audibert, Le Processus Eberthien, Masson et cie, Paris, 1911.
Kelly, Am. J. M. Sc., Phila., 1906, n. s., CXXXII, 447, 744.
Elmer, Arch. Pediat., N. Y., 1911, XXVIII, 217.
Sultan, Deutsche med. Wchnschr., 1894, XX, 675.
Greaves, Brit. M. J., Lond., 1907, II, 75.

Lavenson, Univ. Penn. Med. Bull., 1908-9, XXI, 55. Silber-
berg, Berl. klin. Wchnschr., 1908, XLV, 1354.

36 THE SOURCES AND MODES OF INFECTION

well as by pathological studies on human beings, it seems
probable that the bacilli reach the gall bladder by way of
the blood stream. Once there, they may merely lead a sap-
rophytic existence, increasing to enormous numbers in the
bile and passing with it into the intestine, whence they are
discharged in the feces. At other times the mucosa of the
gall bladder is attacked and becomes inflamed, and nests
of bacilli are found in the tissue. Such nests are not con-
fined to the gall bladder but are found in the hepatic
ducts, which fact must be taken into account when it is
proposed to cure intestinal carriers by the washing out or
removal of the gall bladder. Indeed, Ledingham refers to
cases in which this operation did not prevent the subse-
quent discharge of bacilli in the feces.

Loele 1 opened the gall bladder in a convalescent typhoid
case which was excreting bacilli, but no trace of them could
be found in the gall bladder.

Typhoid bacilli are not rarely found in gallstones and it
is thought by some that they make their way into the stones
after they are formed, but the general opinion is that they
are the nuclei around which the stones develop.

The bacilli are by no means constant in the intestines,
even during the course of the disease, and indeed they are
sometimes entirely absent, as in Semple and Greig's case
previously mentioned. With improvements in technique
they are more frequently found than formerly. The fol-
lowing figures from Gaehtgens and Bruckner in 1910 2 give
the highest percentages.

Cases.

Bacilli found.

Percentage.

1st week . .

21

12

57

2nd week . . .

32

17

53

3rd week . .

13

10

77

4th week

4

2

50

5th week

2

2

100

1 Loele, Deutsche med. Wchnschr., 1909, XXXV, 1429.

» Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1910, LIII, 559.

CARRIERS AND MISSED CASES 37

Typhoid Bacilli in Urine. — Typhoid bacilli were noted
in the urine as early as 1886 by Hueppe,1 and similar find-
ings were reported in succeeding years by many other ob-
servers. The first extensive study of this condition was
made by Richardson in 1898.2 The bacilli are usually not
found in the urine until the later stages of the disease, but
are occasionally observed earlier, as by Connell,3 on the
seventh day. They are usually in large numbers and often
in pure culture. Richardson found them present in about
23 per cent of all cases examined, and earlier and later
observers give about the same figures. Thus Connell tab-
ulated 621 reported examinations, of which 24 per cent were
positive. The bacilli generally continue in the urine for
a while after recovery, but usually disappear within three
months. Oftentimes the presence of the bacillus is not
accompanied by symptoms referable to the urinary tract.
Often, however, cystitis develops, which may persist an
indefinite time, accompanied by the presence of the bacilli
in the urine. The pelvis of the kidney may also be attacked
by the bacilli, and foci of infection may be found in the
kidneys themselves. As early as 1889 Konzajeff 4 reported
finding in the kidneys infarctions containing typhoid bacilli.

Marchildon 5 has recently reported two cases in which
the presence of typhoid bacilli in the urine was due to their
growth in the seminal vesicles and the prostate, and Gould
and Quails6 found the bacilli in the prostatic fluid of 1 of 19
convalescents.

Typhoid Bacilli in the Sputum. — Pratt, Peabody and
Long, previously quoted, cite several observers as finding

1 Hueppe, Fortschr. d. Med., 1886, IV, 448.

2 Richardson, J. Exper. M., N. Y., 1898, III, 349; 1899, IV, 19.

• Connell, Am. J. M. Sc., Plula. and N. Y., 1909, n. s., CXXXVII,
637.

< Konzajeff, Centralbl. f. Bakteriol. [etc.], Jena, 1889, VI, 672.

6 Marchildon, Am. J. M. Sc., 1910, CXL, 74.

6 Gould and Quails, J. Am. M. Ass., Chicago, 1912, LVIII, 542.

38 THE SOURCES AND MODES OF INFECTION

the bacillus in the bronchial secretions of typhoid fever
patients, and state that Dieudonne found it for seven
weeks after recovery. Richardson l also reports finding it
in the sputum on three successive days. That this is not
a common condition is probable from the investigations of
Tenney,2 who examined 53 typhoid fever patients without
finding the bacilli in the sputum. A few of the cases had
bronchial symptoms. He found that the bacilli would
grow and survive in sputum for 125 days.

Typhoid Carriers. — Though typhoid carriers seem to
have been first observed in England, their epidemiological
importance was not realized until, under the direction of
Koch, the active campaign against the disease in south-
west Germany was begun in 1903. During the next few
years extensive observations of carriers were made in this
and other portions of Germany.

Since the first edition of this book appeared, J. C. G.
Ledingham 3 has published a most valuable study of ty-
phoid carriers, and in addition to the data presented in
my first edition, I have made free use of his material in the
following pages.

A typhoid " carrier " is a person without signs of illness
in whom typhoid bacilli are living and increasing in num-
bers. Some of these carriers have had typhoid fever, others
so far as can be learned have not. In a large proportion of
cases the carrier condition follows convalescence and lasts
only for a few weeks or months. Some cases, on the other
hand, become chronic carriers. Again, persons may be car-
riers for a period, perhaps usually short, before the disease
develops. Various names have been applied to the differ-
ent kinds of carriers, both by the Germans and by the
French, and a formal classification has been suggested. In

1 Richardson, J. Bost. Soc. M. Sc., 1897, II, 21.
1 Tenney, Bost. M. & S. J., 1910, CLXIII, 124.
8 Ledingham, Rep. Med. Off. Local Gov. Bd., Lond., 1909-10,
XXXIX, 249-384.

CARRIERS AND MISSED CASES 39

Germany carriers who excrete bacilli for less than 3 months
are generally called Bazillentrdger, and those who remain
carriers over that time are Dauertrdger. The English
equivalents are " temporary " or " transitory " carriers,
and " chronic " carriers. Sacquepee l names the excretors
of bacilli in the incubation stage as " precocious " carriers,
those who have had the disease and who continue to ex-
crete bacilli for less than 3 months as " convalescent "
carriers, those who excrete them over three months as
" chronic " carriers, and those who have never been sick as
" healthy carriers," or " well carriers."

For the present purpose interest attaches chiefly to the fre-
quency of the carrier state, the length of time it continues
and the liability of carriers to cause disease in others. Some
carriers excrete bacilli in the feces, some in the urine, and
some in both. Only one sputum carrier has been reported.

Carriers during Incubation. — Cler and Ferazzi,2 during
a food outbreak, found 6 of 39 persons who had eaten the
food and who, though exhibiting no symptoms, were carry-
ing the bacilli in their intestines. These " precocious "
carriers may become sick later, and sometimes the period
of incubation, so called, may be three or four weeks, or even
longer.3 Indeed, it is not rare to find typhoid bacilli in the
feces during the period of incubation. Conradi noticed
this, and considers it an important factor in the spread of
the disease.4 Ravenel and Smith 5 have reported an out-
break of forty cases due to contact with a case before the
symptoms had developed. Prigge 6 discovered 3 carriers
who developed the disease, 18, 19 and 20 days afterwards.

1 Sacquepee, Bull, de 1'Inst. Pasteur, Paris, 1910, VIII, 1, 49.
1 Cler and Ferazzi, Centralbl. f. Bakteriol. [etc.], Jena, I Abt., Ref.
1905, XXXVI, 479.

J Griffith, Med. Press and Circ., 1905, LXXIX, 208.

4 Conradi, Deutsche med. Wchnschr., 1907, XXXIII, 1684.

8 Ravenel and Smith, J. Am. M. Ass., Chicago, 1909, LII, 1635.

6 Prigge, Klin. Jahrb., Jena, 1909, XXII, 245.

40 THE SOURCES AND MODES OF INFECTION

G. Mayer l reports 3 cases. One of these was a boy whose
stools contained the bacilli 8 days before the symptoms
arose. According to Ledingham, Klinger concluded that of
812 cases of contact infection studied by him, 33 acquired
the disease from a case in the first week of incubation and
150 during the second week. He assumes two weeks as
the average period of incubation, though he found it in 60
cases, in which its duration was pretty well determined, to
vary from 5 to 45 days, the average being 16 days. During
the Spanish war many cases of typhoid fever were attrib-
uted to exposure to cases in the incubation stage.2

Convalescent Carriers. — Drigalski 3 was the first to
study the persistence of typhoid bacilli in the feces of con-
valescents. Of 64 patients, he found that 7, or 11 per cent,
continued to excrete them from 8 to 10 weeks, and 3 for
over 3 months. One of these was later, at 9 months,
found to be still a carrier. Klinger 4 at Strassburg examined
482 cases of typhoid fever during convalescence and 63, or
13.1 per cent, were carriers, of whom 8, or 1.7 per cent, con-
tinued so for a period of over 6 weeks. Later 5 he reported
that of 604 convalescents, 80, or 13.2 per cent were tempo-
rary carriers, 70 intestinal and 10 urinary. G. Mayer6
found that 232, or 24.9 per cent, of 930 typhoid fever cases
became carriers during convalescence. Graham, Over-
lander and Dailey7 found the bacilli in the feces of 11, or
16.9 per cent, of 65 patients after defervescence and pre-
vious to their discharge from the hospital. Including the

1 Mayer, Centralbl. f. Bakteriol. [etc.], Jena, I Orig., 1910, LIII, 234.

5 Abst. of Rep. on Origin and Spread of Typhoid Fever in U. S.
Military Camps during Spanish War of 1898, Wash., 1900, 178.

8 Drigalski, Centralbl. f. Bakteriol. [etc.], I Abt., Jena, 1904,
XXXV, 776.

4 Klinger, Arb. a. d. k. Gesundhtsamte., Berl., 1906, XXIV, 91,

6 Klinger, Arb. a. d. k. Gesundhtsamte., Berl., 1907, XXV, 214.

6 Mayer, Centralbl. f. Bakteriol. [etc.], Jena, I Abt. Orig., 1910,
LIII, 234.

7 Graham, Overlander and Dailey, Bost. M. & S. J., 1909, CLX, 38.

CARRIERS AND MISSED CASES 41

urinary carriers there were 15, or 23 per cent. Semple and
Greig 1 report that 16, or 18.8 per cent, of 86 typhoid con-
valescents continued to excrete bacilli after 'their tempera-
ture became normal.

Ledingham 2 gives a table and chart taken from Klin-
ger's account (1909) of the anti-typhoid work in Germany,
which shows the age and sex distribution of transitory and
chronic carriers. Of 211 of the temporary carriers, 35
were between 5 and 10 years of age, the period showing the
maximum number of cases. The age period of greatest
incidence for the 220 chronic carriers was between 40 and
45, at which period there were 30 cases. Among the trans-
itory carriers the proportion of females to males was 1.4
to 1, while among the chronic carriers it was nearly 5 to 1.
Prigge 3 shows two diagrams, on one of which the maximum
age of female cases is given as 15 to 20 years and the aver-
age age of female carriers as 40 to 45 years.

Chronic Carriers. — The larger number of convalescent
carriers cease excreting bacilli at varying intervals after
their recovery. A certain number remain carriers for an
indefinite period, sometimes for many months, sometimes
for years, and perhaps occasionally for a long lifetime.
The earliest and most extensive series of observations as to
the number of typhoid fever patients who become chronic
carriers was made by the men carrying on the campaign
against this disease in Germany. According to Ledingham,
of 6708 typhoid fever cases observed during the years 1904
to 1906, 166, or 2.47 per cent, became chronic carriers. Of
3867 cases in 1906 and 1907, 38, or 1 per cent, proved to
be chronic carriers. In the Bavarian Pfalz from 1904 to
1907, there were observed 930 cases, of which 78, or 8.1 per

1 Semple and Greig, Scient. Mem., Med. and Surg. Dept., Gov. of
India, 1908, XXXII, 9.

2 Ledingham, Rep. Med. Off. Loc. Gov. Bd., Lond., 1909-10,
XXXIX, 267.

8 Prigge, Klin. Jahrb., Jena, 1909-10, XXII, 245.

42 THE SOURCES AND MODES OF INFECTION

cent, continued to excrete bacilli for over three months.
Park l examined the feces of 52 cases eight months after
recovery and found bacilli present in 2; in one of them,
however, it was present in only one of three tests. Of 16
other persons who had the disease six months previously,
2 were carriers. He thinks that 2 per cent of all typhoid
fever cases became permanent carriers, and these may be
found in the population at the rate of about one to five hun-
dred. Bruckner 2 states that of 316 persons who had the
disease, 12, or 3.8 per cent, became carriers, or omitting the
104 children, only 1 of whom was a carrier and who rarely
become such, 5.2 per cent continued to excrete bacilli.
Semple and Greig, before referred to, found nearly 4 per
cent of typhoid patients remained carriers for over three
months. Aldridge,3 also working in India, reports that 6
of 190 convalescents, or 3.1 per cent, remained carriers for
longer than six months. Tsuzuki 4 found 3 of 51 convales-
cents, or 5.8 per cent, to be chronic carriers. Recently
Hamilton,5 following out the suggestion of Gaehgtens, tested
the opsonic index of 25 persons who had gall-bladder
trouble and found 7 with an abnormally high index. All
7 proved to be carriers. She hopes that this will prove a
simple means of discovering carriers.

In Washington 6 380 persons who had had typhoid fever
during the years 1904 to 1909 were in 1909 examined and
8, or 2.8 per cent, found to be carriers.

Klinger found that of 220 chronic carriers 30, or 13.6 per
cent, had gallstones. There is no doubt that the presence
of typhoid bacilli in the bladder and gallstones are defi-
nitely associated.

1 Park, J. Am. Ass., Chicago, 1908, LI, 981.

8 Bruckner, Arb. a. d. k. Gesundhtsamte., fieri., 1910, XXXIII, 435.

* Aldridge, J. Roy. Army Med. Corps, Lond., 1909, XIII, 221.

4 Tsuzuki, Arch. f. Schiffs-u. Tropen-Hyg., Cassel, 1910, XIV, 147.
6 Hamilton, J. Am. M. Ass., 1910, LIV, 704.

• U. S. Pub. Health and Mar. Hosp. Serv., Hyg. Lab. Bull. No. 77, 171.

CARRIERS AND MISSED CASES 43

There is little doubt that the percentage of carriers both
convalescent and chronic is considerably larger than is
indicated by the above figures. Owing to the very marked
intermittency with which bacilli are excreted in the feces
of many carriers, an intermittency which is also shown by
the urinary carriers, though to a lesser extent, it is certain
that more extended examination of the excreta would dis-
cover many more carriers. It must be remembered that
most studies of this subject have been based on not more
than two or three examinations.

Carriers among Contacts. — Persons brought into inti-
mate relation with the sick may become infected without
exhibiting any symptoms whatever. Drigalski and Con-
radi *• found the infection in 4 well persons in contact with
typhoid cases, Liefmann and Nieter 2 found 7 carriers out
of 252 persons examined in an insane asylum, some of whom
it was believed were carriers and the causes of the out-
break in the institution, but some of them, however, were
true contacts. Scheller 3 examined 40 persons who drank
milk which had been handled by a typhoid carrier. Of these
5 were sick, and 13 others, who had no symptoms, were yet
found to be excreting typhoid bacilli in either feces or urine
or both. All of them became free from germs within a few
weeks.

Ledingham states that at the St. Brieuc garrison in 1909,
Billet and others found 1 carrier among 53 men who lived
in rooms where the cases had been most numerous.

Typhoid Bacilli in Persons Never Sick. — Not only are
persons who have had typhoid fever found to be " carriers "
of the germs, but persons who have never had the disease,
and who give no history of contact, may be infected.

1 Drigalski and Conradi, Ztschr. f. Hyg. u. Infectionskrankh.,
Leipz., 1902, XXXIX, 283.
*2 Liefmann and Nieter, Miinchen med. Wchnschr., 1906, LIII, 2097.

3 Scheller, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1908,
LXVI, 385.

44 THE SOURCES AND MODES OF INFECTION

Allowance must, however, be made for failure to recognize
and remember mild atypical attacks, and some of the
chronic carriers who are said never to have been sick may
have forgotten an attack of " grip " or " malaria " years
before. Soper's case so far as known had never had the
disease. Houston's case, which had the bacilli in the urine
for three years, had not been sick. One of the carriers, the
source of several cases, discovered by Semple and Greig,
had never had typhoid fever so far as known. Roscoe 1
reports an outbreak of twelve cases in an insane asylum due
to contact with a carrier who had never had the disease.
Jundell 2 notes a series of twenty-two cases on an estate
in Sweden occurring at intervals since 1854. The grand-
mother of the family, who had never had typhoid fever, was
shown to be excreting the bacilli from 1904 until the time
of the report, 1908. According to Ledingham, Prigge found
that 8 of 84 carriers studied by him had never had the dis-
ease. Of 431 carriers reported by Klinger, 163 gave no
history of having had typhoid fever.

Carriers among the Public. — There have been a few
studies to determine the number of carriers in the general
population. Minelli 3 found 1 carrier in 250 prisoners in
Strassburg. The most extensive investigation has recently
been carried on in Washington,4 where the feces from 986
well persons were examined bacteriologically. Only one
test was made for each person, and 3 carriers were found.
These were subsequently re-examined and two proved to be
negative.

Much more evidence is needed as to the number of car-
riers to be found, both among the general public and among

1 Roscoe, Lancet, Lond., 1909, II, 1137.

8 Jundell, Abst. J. Am. M. Ass., Chicago, 1909, LII, 388.

3 Minelli, Centralbl. f. Bakteriol. [etc.], Jena, I Abt. Orig., 1906,
XLI, 406.

4 U. S. Pub. Health and Mar. Hosp. Serv., Hyg. Lab. Bull. No. 52,
145.

CARRIERS AND MISSED CASES 45

contacts. Extensive investigation ought to be made in dif-
ferent parts of the world and among different classes of people.

Duration of Infection. — The bacilli may be found in the
urine and feces of carriers, often in enormous numbers, for
years. Not so very many cases have been followed bacte-
riologically for a great length of time, though in one in-
stance the positive examinations lasted four and one-half
years. Tsuzuki followed several cases over a year. Mayer,
before referred to, reports the following duration of infec-
tivity: for 6 months, 56; 6 months to 1 year, 38; 1 year
to 2 years, 16; 2 years to 4 years, 17.

There is, however, epidemiological evidence for assuming
a much longer continuance of the infection. Dean 1 reports
the case of a medical man who had had typhoid fever
twenty-nine years before, and had since then frequent
attacks of biliary colic. Typhoid bacilli were recovered
from his feces. It was believed that no one had contracted
the disease from him, but he had always been very careful
in his personal habits. Huggenberg 2 noted thirteen cases
in a household extending over a period of thirty-two years.
One woman who had the disease in 1877 was shown to be
a carrier in 1908. Scheller 3 reported thirty-two cases ex-
tending over a period of fourteen years, all probably due
to a carrier who had been sick seventeen years before.
Gregg 4 found a woman whose blood gave a positive Widal
reaction, and in whose feces bacilli were found, and who
had had typhoid fever fifty-two years before. She had
presumably infected seven persons. JundelPs case reported
above was infectious perhaps for fifty-four years. Chal-
mers' case 5 had had the disease sixteen years before.

Dean, Brit. M. J., Lond., 1908, I, 562.

Huggenberg, Cor.-Bl. f. Schweiz. Aerzte., 1908, XXXVIII, 622.
Scheller, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1908,
LXVI, 385.

Gregg, Boston M. & S. J., 1908, CLIX, 80.

Chalmers, Rep. of Med. Off. Health, Glasgow, 1907, 61.

46 THE SOURCES AND MODES OF INFECTION

Frosch l reports that evidence was presented to the com-
mission appointed by the Prussian Government to study
this subject as follows: That fourteen carriers had been in-
fective four to nine years, six for ten to twenty years, and
five for from twenty-one to thirty years. Soper's case has
now been infectious for ten years, and a number of other
writers report instances of carriers who were presumably
excreting bacilli more or less constantly for periods of from
four to eight years.

Intermittent Excretion. — From what has been said about
the carrier state and the existence of nests of typhoid bacilli
in the tissue of the gall bladder, the hepatic ducts and in the
tissues of the urinary tract, it might be surmised that ex-
cretion of bacilli is not in all cases continuous. That there
have been reported considerable periods in which the feces
and urine of carriers remain free from bacilli is not sur-
prising. G. Mayer 2 claims to have been the first to note
this intermittent excretion in 1905, but the matter did not
receive much consideration until Davies and Hall 3 called
attention to the marked intermittency of bacillus excretion
in their case, reported previously by Davies, and which will
be again referred to. This patient had been infectious at
times for four years, particularly in the spring, and on one
occasion she was herself sick.

Davies and Hall laid considerable stress on this presumed
seasonal intermittency, but Ledingham, who with Thompson
afterwards followed up their case and also six others, con-
siders that the evidence is not convincing, though certainly
their carriers gave many more positives during the first
than during the last half of the year. Semple and Greig
report 18 instructive cases which they followed daily for a
considerable period. One of their cases gave only negative

1 Frosch, Klin. Jahrb, Jena, 1908, XIX, 537.

8 Mayer, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1909-
10, LIII, 234.

8 Davies and Hall, Lancet, Lond., 1908, II, 1585.

CARRIERS AND MISSED CASES 47

tests for thirty-one days, followed by a positive, another
carrier remained free from bacilli, as shown by daily exam-
ination, for a period of seventy-five days, and other cases
for lesser periods. Scheller noted considerable intermit-
tency in the excretion of bacilli. Of 108 examinations of
urine and feces from 18 carriers, 48 only were positive.

Instances of marked intermittency have been noted by
others, as Bruckner, Kayser, Nieter & Liefmann, Eccard
and Prigge. The latter, according to Ledingham, noted
intermittent periods lasting from one to two and one-half
years. Eccard, in an effort to check typhoid fever prevail-
ing in an asylum, discovered three carriers who were effec-
tually isolated. The disease, however, continued, and on
three subsequent examinations, made several months apart,
fresh carriers were found who before had given only nega-
tive results, in one instance five times. The disease ceased
on the isolation of this carrier. This experience of Eccard
illustrates how greatly intermittency interferes with all pre-
ventive work based on the discovery and control of carriers.
An instructive table showing the marked intermittency of
carriers in an insane asylum is given by Neisser.1

Carriers Cause Disease by Contact. — Perhaps one of
the most convincing instances of contact infection from a
carrier is the celebrated case of " Typhoid Mary," so well
investigated by Soper.2 Between August 27 and Septem-
ber 3, 1907, 6 cases of typhoid fever developed in a banker's
family of 11 persons at Oyster Bay. All the usual routes
of infection were most carefully investigated by Soper and
discarded because of the lack of evidence. Convinced that
some peculiar event must have occurred in the family
about August 1, he sought for it, and the only change that
he could discover was a change in cooks. But the cook had
left and no trace of her could at the time be found. Soper
concluded that she was, in view of the evidence, the most

1 Neisser, Berl. klin. Wchnschr., 1910, XL VII, 2142.
8 Soper, J. Am. M. Ass., Chicago, 1907, XLVIII, 2019.

48 THE SOURCES AND MODES OF INFECTION

likely source of the trouble, and he made every effort to
locate her, but was unsuccessful until she had figured in
two more outbreaks. No information was ever obtained
from her as to her wanderings, but in various ways Soper
learned that a case developed in a family in Mamaroneck
where she lived in 1900, another case in a family in New
York in 1901, 7 in a family of 9 persons at Dark Harbor,
Me., in 1902, 4 in a family at Sands Point in 1904, and 7
cases at Oyster Bay in 1906. After leaving Oyster Bay
she went direct to Tuxedo, N. J., a locality free from ty-
phoid fever for several years, but where the laundress de-
veloped the disease fourteen days after the cook's arrival.
" Mary " went to a family in New York in December and
within a few weeks 2 cases developed in this family. No
wonder that with this evidence extraordinary means were
taken to obtain specimens of her excreta, and it can be no
surprise to any epidemiologist that she proved to be a
carrier. She was confined in a hospital in New York for a
long time and her release was refused by the courts, but
she was after a while discharged, though still a carrier, and
has recently brought suit against the city for $50,000. I
am inclined to the opinion that few even of the scoffers at
the " carrier theory " would care to employ Mary Wallon
as cook.

Hilgermann l narrates an instance in which new servants
coming to a certain house soon developed typhoid fever.
Cases were reported in 1895, 1900, 1902, 1903, 1905 and
1907, in all 15 cases. Suspicion fell upon a woman of 71
years who had had typhoid fever in 1894. She was found
to be a carrier. This is the complement of Soper's case,
where a moving carrier infected persons in successive houses.
Here a stationary carrier infected a series of arrivals at her
home. G. Mayer, before quoted, shows a genealogical tree
of 195 cases occurring in a certain district in Bavaria during
the course of about five years. In this tree are 13 carriers,
1 Hilgermann, Klin. Jahrb., Jena, 1908, XIX, 463.

CARRIERS AND MISSED CASES 49

and from 8 of them the disease extended, that is, without
them the " tree " would have had many less branches.
Of these 13 carriers 2 had a recurrence of the disease.
During the years in which his observations were made, of
495 endemic cases of typhoid fever 160 were traced to 91
darners. O. Mayer x has traced several " contact chains "
from carriers. Sumacher,2 after a careful investigation in
the village of Crov, reported that he could trace 26.6 per
cent of the cases to direct contact with carriers and 44.4
per cent to indirect contact.

The following case came under my own observation in
1911: A girl 15 years old, whom we will call A, was re-
ported as going to bed with typhoid, March 3. On April
12 the family went to their farm for a few days, and again
they were at the farm on April 29 and 30. At this time
two friends of A, Miss B and Miss C, were also of the
party. There were no servants at the farm and the family
prepared the food. On May 7 two other members of the
family were taken sick with typhoid fever, and on May 6
Miss C went to bed with the same disease and on May 9
Miss B. On May 17 the father came down with typhoid
fever, and a few days later the mother. The evidence
pointed to infection at the farm, and of course the family
suspected milk and water. The former was from the cow
of a neighbor, where there had been no sickness, and the
latter was found to be excellent. Further evidence acquit-
ting the farm and farmer's family was furnished by the fact
that another member of the family, a student at Bryn
Mawr, had 6 of her college friends at the farm from March
22 to April 2 and none of them developed typhoid fever.
On the other hand, every one of the party who went to the
farm with the convalescent, April 29-30, later had the dis-
ease. The city milk supply was beyond suspicion. I nat-
urally suspected that the girl A, the one first sick, was a

1 Mayer, Miinchen. med. Wchnschr., 1908, LV, 1782.
» Sumacher, Klin. Jahrb., Jena, 1909-10, XXII, 263.

50 THE SOURCES AND MODES OF INFECTION

carrier, and had infected the food at the farm on April 29
or 30, thus causing the sickness of her brother and sister
and two friends. The sickness of father and mother might
also have been derived from A, but more likely from the
later cases which they helped to care for. No examina-
tions were made of excreta until June 28, when A was
found to be a carrier, as I at first suspected. Typhoid
bacilli were again found in the excreta on July 29.

Davies and Hall l report the case of a cook who, after
discharge from a hospital, where she had had typhoid fever,
went to her home, where 4 cases developed in the family.
Besides these, 4 other cases occurred in three other loca-
tions; in all 8 cases in about three years.

Carriers Cause Disease in Institutions. — Typhoid fever
not infrequently appears in institutions; and in insane asy-
lums especially, owing to the difficulty in controlling the
habits of patients, such outbreaks are apt to occur. Some-
times the source is to be found in a sick person and some-
times in a carrier. A. & J. C. G. Ledingham 2 report an
instructive and long-standing institutional infection of this
kind. The asylum in 1908 held 92 male and 53 female
patients. Cases of typhoid fever had developed in every
year between 1893 and 1907, except four. In all there
were 31 cases, of which 24 were women. Drainage and
water supply could be eliminated as causes. All of the
female patients were examined and 3 carriers were found.
These were isolated in November, 1907, and since then no
cases of the disease have developed in the institution..

Ledingham quotes Eccard in regard to typhoid fever in
a large asylum at Frankenthal. Most of the cases, how-
ever, occurred in a female block containing 110 " unclean,
stupid and noisy women." From 1901 to 1906 there had
been 21 cases. Repeated search for carriers was made in

1 Rep. Med. Off. Health, Bristol, 1909, 61.

- A. & J. C. G. Ledingham, Rep. Med. Off. Local Gov. Bd., Lond.,
1909-10, XXXIX, 304.

CARRIERS AND MISSED CASES 5]

1903 and in 1906 and 1907. In the latter years some were
discovered whose examination in earlier years had been
negative, owing probably to intermittent excretion. Since
1906 no cases of typhoid fever have developed. Other in-
stances of asylum typhoid due to carriers have been re-
ported by Liefmann and Nieter, Levy and Kayser, Friedel,
Dehler,1 C. Neisser,2 Roscoe 3 and Ledingham.4 Dehler \
removed the gall bladder from two carriers who were found
to be the cause of the outbreak, and their feces were after-
wards free from bacilli.

A very convincing instance of carrier infection was re-
ported by Davies 5 of Bristol, England. A woman, Mrs. H,
50 years old, had had typhoid fever in 1901. In May,
1904, she went to a home for girls at Brislington to work
in the kitchen. From that time until September, 26 per-
sons in the institution developed typhoid fever. From
September, 1904, to March, 1905, Mrs. H was in private
service. She was then cook in a children's home for some
months. The one case which developed here may have
had no significance. In April, 1906, she was employed in
the kitchen of the Inebriate Reformatory in Bristol. In
the autumn of that year 4 cases developed among users of
institution milk. There was no typhoid fever in the neigh-
borhood, nor had there been any in the institution since it
was opened. There was another outbreak in May, 1907,
and up to November 4 of that year there had been 23
cases. Apparently the milk was at fault, and if so was
infected after sterilization. Evidence pointed to Mrs. H,
who handled the milk after it had been sterilized, and it
was discovered by Davies that she was a carrier. She was
then removed from contact with the food and there were

1 Dehler, Munchen. med. Wchnschr., 1907, LIV, 779, 2134.

2 C. Neisser, Psychiat.-neurol. Wchnschr., Halle, 1908-9, X, 37.
» Roscoe, Lancet, Lond., 1909, II, 1137.

4 Ledingham, Brit. M. J., Lond., 1908. I, 15.

6 Davies, Proc. Roy. Soc. Med., 1908, I, Epidemiol. Sec., 175.

52 THE SOURCES AND MODES OF INFECTION

no more cases. Davies and Hall l by following up this
case showed that the excretion of bacilli was markedly
intermittent. They report also that the woman had what
might be called a slight relapse in May, 1908, probably a
cholecystitis, and in July a woman with whom she shared
her lunch developed typhoid fever.

Carriers Cause Disease through Milk. — In typhoid out-
breaks due to milk, water or food, the large number of
cases, massed closely together in point of time, often fur-
nish far more certain evidence as to causation than do
smaller and less marked outbreaks due to contact infec-
tion. There have now been quite a number of milk out-
breaks reported as due to carriers, and some of these are
very striking. The following outbreak was reported by Lums-
den and Woodward.2 Fifty-four cases of typhoid fever were
reported on the routes of two milk dealers in Washington
during the autumn of 1909. The outbreak was localized
in that part of Washington still known as Georgetown.
This section of the city was served by about thirty milk
dealers. Of the 54 patients, 33 took milk from dealer A
and 21 from dealer B. Dealer A, it was found, received
40 gallons of milk daily from the farm of Mrs. X, and
dealer B received 20 gallons daily from the same source.
The rest of Mrs. X's milk was peddled direct to eleven
families nearby. Among these were 3 cases, not seen by a
physician, which quite likely also were typhoid fever. The
age distribution of the cases and all the other features of
the outbreak pointed to milk infection, with its source on
the farm of Mrs. X. No recognized cases of typhoid fever,
or illness likely to be mistaken for it, had occurred on the
farm, and naturally a search was made for carriers. Mrs.
X herself was found to be excreting typhoid bacilli in her
feces, and conditions were such that infection of the milk
from her was very possible. If a well-defined case of

1 Davies and Hall, Lancet, Lond., 1908, II, 1585.

2 Lumsden and Woodward, J. Am. M. Ass., Chicago, 1909, LII, 749.

CARRIERS AND MISSED CASES 53

typhoid fever had been found on this farm, it is improb-
able that any epidemiologist would for a moment hesitate to
attribute the outbreak to that case, even if no examination
were made to show the presence of the bacilli in the ex-
creta. When now a person is found actually discharging
bacilli in large numbers, I can see absolutely no reason for
not considering her in all probability the source of the
trouble. The fact that the woman had typhoid fever as
long as 18 years before, and had not, so far as known, previ-
ously been a cause of disease, is urged by some as a reason
for not admitting her to be the source of infection at this
time. We know in the first place that many carriers are
markedly intermittent in the excretion of bacilli, and, what
is more to the point, it is by no means as easy to cause
infection of persons or things as has generally been be-
lieved. Plenty of cases of typhoid fever have occurred on
milk farms without doing harm, and numberless cases of
scarlet fever, diphtheria and smallpox, as well as of typhoid
fever, are most carelessly cared for, yet fail entirely to
spread disease.

Bigelow 1 reports that in Worcester in 1910 there were
>nly 295 reported cases of typhoid fever. Of these, 204
were the customers of a single dealer. The facts pointed
to four farms as possible sources. No cases of the disease
were found on the farms, and blood tests gave a positive
Widal test from only one person, who daily assisted in the
milking. He had had typhoid fever 26 years before, and
about two weeks before the outbreak he had a slight attack
of diarrhea with a little blood in the stools, and accompanied
by some headache. He was on two occasions shown to be
a urinary carrier.

During February, 1910, there was an outbreak of typhoid

fever in New York City in a section corresponding to the

route of a large milk dealer. Of the 48 cases, 41 used this

milk. The infected milk was traced to a distant farm in

1 Bigelow, J. Am. M. Ass., Chicago, 1911, LVII, 1418.

54 THE SOURCES AND MODES OF INFECTION

Vermont, where a carrier who was doubtless the source of
the trouble was found. It is worthy of note that a labora-
tory assistant while pipetting a culture of the bacillus from
this carrier drew some of it into her mouth and two weeks
later developed a typical attack of typhoid fever.1

Bolduan and Noble 2 report an outbreak of several hun-
dred cases in New York City in 1909 which was traced to
that portion of the milk of a large dealer which came from
Camden, N. Y. There had been a case of typhoid con-
nected with a creamery there, and also a case on farm X,
the milk from which was supposed not to go to the cream-
ery, both cases becoming sick at about the time of the city
cases. But it was found that farmer X did send a part of
his milk to the creamery and thence to New York. It was
also learned that there had been much typhoid in Camden
for several years, and that of 27 cases in 1908-09, 20 were
on the milk route of X. It was also found that on the
farm of X during the years 1878 to 1909 there had been
7 cases of what was probably typhoid fever. Mr. X him-
self had had typhoid fever in 1863-64. He was, on two
occasions, a month apart, found to have typhoid bacilli in
his feces.

The Abundance of Evidence. — Only a very few though
striking instances of the rise of sickness from carriers have
been selected for the purpose of illustration. Within the
past three or four years a great mass of similar material has
appeared in the medical press, — material already too abun-
dant for satisfactory compilation. Articles are constantly
appearing in the medical journals of all lands reporting
instances of apparent infection by carriers. It is true that
in many instances the evidence is by no means conclusive,
often, indeed, only warranting a surmise. In many cases, on
the other hand, — and the number is rapidly increasing, —
the evidence which leads to the discovery of the carrier is

1 Bolduan and Noble, N. York M. J.( 1911, XCIV, 1313.

* Bolduan and Noble, J. Am. Med. Ass., Chicago, 1912, LVIII, 7.

CARRIERS AND MISSED CASES 55

as good, and often better, than that on which, with un-
questioning confidence, outbreaks are traced to recognized
cases. The evidence which implicates the carrier is the
same, and is as convincing as that which determines the
contagiousness of the disease.

Evidence against Carriers. — While the evidence is thus
seen to be very strong that typhoid carriers are an impor-
tant factor in the spread of this disease, some facts are pre-
sented which apparently point the other way. Thus Linos-
sier1 says that at least 10,000 persons with biliary lithiasis
must visit Vichy each year. From what is known of this
condition, it seems probable that a large proportion of
these must be typhoid carriers, though no systematic exam-
inations have been made to determine this. If this is so,
typhoid fever ought to prevail in Vichy, but as a matter
of fact there is very little. Linossier explains this as
perhaps due to the fact that most visitors to Vichy are
beyond the age of marked susceptibility to the disease,
and in part also because they are mostly well-to-do and
of cleanly habits.

T. Thomson 2 also finds the conditions after extensive
water outbreaks somewhat puzzling. He says that in
Worthing after the outbreak of 1893, in which 8 per cent
of the population of 17,000 were attacked, there was, for
some years, less typhoid fever than before. After the out-
break in Maidstone in 1897, involving 6 per cent of 33,000
persons, there was somewhat more typhoid fever for about
four years; and in Lincoln, after the outbreak in which 2
per cent of 52,000 persons were sick, the amount of typhoid
fever remained about the same. As we could expect the
outbreaks to have left 39, 57 and 30 chronic carriers in the
three towns respectively, Thomson says that one would
naturally look for an increase in the disease over the pre-
epidemic years. Such, however, does not seem to have

1 Linossier, Bull. Acad. de m6d., Par., 1909, 3 S. LXII, 627.

1 Proc. Roy. Med. Soc., Lond., 1910-11, IV, Epidemiol. Sect., 1.

56 THE SOURCES AND MODES OF INFECTION

taken place. While this is somewhat paradoxical, the
problem is so complicated by the development of acquired
immunity, and the generally falling incidence of typhoid
fever, that the apparent innocuousness of hypothetical car-
riers under such conditions should not be allowed to weigh
too heavily against the very clear and abundant evidence
on the other side.

Neisser 1 calls attention to the danger of attributing too
much importance to the reports from certain insane asy-
lums that endemic typhoid fever has ceased after the iso-
lation of carriers. He shows that in other institutions
typhoid-free years have occurred without any control of
carriers. He also states that in four institutions where
there were outbreaks involving 79 cases, 21 attendants
were attacked, while of the attendants on carriers he can
find only 3 attacked. He considers the danger from car-
riers less than is generally believed, yet he thinks it suffi-
cient to warrant the examinations of all newcomers at
Benzlau and their isolation under the care of immune
attendants.

Atypical Typhoid Fever. — There can be no question
that mild unrecognized cases of typhoid fever are extremely
common. Greater care in diagnosis has made an apparent
increase in the number of reported cases of this disease, and
a decrease in fatality. Thus in my own city the reported
fatality from typhoid fever has during the last twenty
years been reduced from 50 to 12 per cent; and it is prob-
ably really considerably less than that. I have noticed
that in milk and water outbreaks, when public attention
is directed strongly towards the disease, the case fatality
is often very low, which merely means that most of the
cases have been recognized. In the Spanish War the com-
mittee of investigation believed that the number of cases
actually existing amounted to 20,738, while the number
reported by the army surgeons was only 10,428.

1 Neisser, Berl. klin. Wchnschr., 1910, XL VII, 2142

CARRIERS AND MISSED CASES 57

Recently Bates l has reported a series of mild atypical
typhoid fever cases in Panama Canal Zone, and states that
such cases are quite common there, and are the chief factor
in the extension of the disease. At the time Koch made his
investigation of the four Trier villages there were 8 recog-
nized cases, but a thorough bacteriological investigation of
suspects discovered 64 more. Of these 49 were children.2
These mild unsuspected walking typhoid cases not rarely
result in death. Velich 3 notes 36 such cases, and a number
of others are reported by Curschmann, and they also have
been seen by the writer. I have recently investigated two
milk outbreaks, in one of which the cause was apparently
a mild unrecognized case, and in the other either a con-
valescent or a carrier associated with him. Neufeld 4 de-
votes considerable space to this class of cases.

Lemoine 5 for three years examined the blood of every
case'of gastro-intestinal disturbance and jaundice in his ser-
vice at the military hospital at Val-de-Grace and found ty-
phoid bacilli in 40 per cent, although there was little to lead
him to suspect typhoid fever. Ledingham quotes Billet,
etc., who report an outbreak of typhoid fever of 142 cases
in a regiment at St. Brieuc. Besides these, there were 57
other atypical cases which doubtless would not have been
recognized except for the epidemic. Bruckner 6 reports
three outbreaks, all originating in mild missed cases. He
finds that children are very susceptible, the most suscep-
tible age being 11 to 15 years, and that in them the disease
is apt to run an atypical course. More recently 7 he has
reported an outbreak in an institution for boys in which

1 Bates, J. Am. M. Ass., Chicago, 1908, L, 585.

8 Koch, Die Bekiimpfung des Typhus, Berl., 1893, 14-15.

3 Velich, Arch. f. Hyg., Munchen u. Leipz., 1904, XLIX, 113.

4 Kolle u. Wassermann, Handbuch [etc.], Jena, 1903, II, 271.

6 Lemoine, Presse meU, Par., 1910, XVIII, 113.

« Bruckner, Munchen med. Wchnschr., 1910, LVII, 1213.

7 Bruckner, Munchen., med. Wchnschr., 1911, LVIII, 1008.

58 THE SOURCES AND MODES OF INFECTION

three-fourths were attacked, many cases running an atyp-
ical course and only discovered by the use of the ther-
mometer. Chamberlain1 says that a third of all cases
among both American and Philippine soldiers can only be
detected by laboratory methods.

The Spanish War investigation showed that most of the
volunteer regiments were infected when they came to camp,
that is, they must have contained carriers or mild cases, and
it was by extension from these that most of the true typhoid
fever later developed.2 The distinction between a case of
true typhoid fever of mild type and a carrier often cannot in
practice be made. There is no sharp line of demarcation,
but infection by typhoid bacilli may result in a series of
cases presenting gradation from the most severe symptoms
to none at all. It is most unwarranted to assume, as some
appear to do, that a mild unrecognized walking typhoid case
may start up an outbreak, but that a true carrier cannot
do so.

Paratyphoid Fever. — It is generally admitted that ag-
glutination does not afford a reliable means for differenti-
ating the various members of the typhoid group of bacilli.
Recent English writers, however, as Dean, Bainbridge and
Firth, believe that this can be satisfactorily accomplished by
means of complement fixation. According to Bainbridge,3
B. suipestifer, B. enteritidis (Gartner), B. partyphosus A
and B. paratyphosus B can be differentiated in this way.
The first two of the above-mentioned bacilli are connected
with disease in the lower animals or are at least found in
such animals and have been definitely connected with out-
breaks of sickness in man due to the use of infected food.
B. paratyphosus A, on the other hand, is said by Bainbridge

1 Chamberlain, Philippine J. Sc., 1911, VI, Med., 299.

2 Abst. of Rep. on the Origin and Spread of Typhoid Fever during
the Spanish War of 1898, Wash., 1900, 168-175.

8 Bainbridge, Proc. Roy. Soc. Med., Lond., 1911. IV, Epidemiol.
Sect., 51; Lancet, Lond., 1912, I, 705.

CARRIERS AND MISSED CASES 59

to be a human parasi-tc, just as is B. typhosus (the ordi-
nary typhoid bacillus). The disease caused by B. paratypho-
sus A is said by Firth l to be very common in India. It is
milder in type, with a more irregular fever and a more
sudden onset, and is accompanied by more headache. It
may very frequently be recognized clinically. Grattan and
Wood 2 say that one-third of all cases of uncertain fever are
due to this bacillus. This type of enteric is not, according
to these authors, and to Bainbridge, common either in Eng-
land or the United States, though perhaps its apparent
rarity is due partly to lack of careful investigation. Grat-
tan,3 however, could not find this bacillus in 48 cases of
enteric fever studied by him in London. But in India
Grattan and Wood studied 157 cases of this disease. Of
these 10 became carriers, but in only one case did the bacilli
persist for as long as 5 months. They believe that these
acute carriers are a greater factor in the spread of disease
than are typhoid carriers in ordinary typhoid fever.
These authors 4 report an outbreak of 9 cases of this
type of paratyphoid fever in barracks at Benares, prob-
ably due to a man who was discovered to be a carrier.
A similar outbreak of 8 cases due to a carrier was reported
by Grattan.5

B. Paratyphosus B. — Bainbridge and O'Brien 6 consider
that B. paratyphosus B as well as A has its habitat in
man. Carriers are frequently found, and such may infect
food and thus cause food outbreaks of the disease. They
did not find this bacillus in 300 typhoid convalescents, but

1 Firth, Roy. Army Med. Corps., Lond., 1911, XVII, 136.

2 Grattan and Wood, J. Roy. Army Med. Corps., Lond., 1911,
XVII, 143.

3 Grattan, J. Roy. Army Med. Corps., Lond., 1910, XIV, 385.

4 J. Roy. Army Med. Corps., Lond., 1911, XXII, 131.

* Grattan, J. Roy. Army Med. Corps, Lond., 1911, XVI, 9.

* Bainbridge and O'Brien, J. Hyg., Cambridge, 1911, XI, 68; also
Brit. M. J., Lond., 1910, II, 1503.

60 THE SOURCES AND MODES OF INFECTION

they did find it in 6 cases of fever in which it was appar-
ently the causative agent, also in 4 convalescent carriers
and in 3 healthy persons. According to Bainbridge, car-
riers of B. paratyphosus B were first noted by Lentz and
later by Hamilton, Gaehtgens, Bruckner, Mayer and Prigge
and Sachs-Miike. But if Bainbridge's contention is cor-
rect, and if these Germans, as he states, have not made use
of complement fixation to differentiate the bacilli, their
data cannot be relied upon. It is not unlikely, however,
that some of these were really carriers of B. paratyphosus B.
Of those noted by Prigge and Sachs-Miike,1 4 were in con-
nection with a food-poisoning outbreak. One of the car-
riers was under observation in a hospital for 2 years.
Another, a well person, was found to be a carrier during a
food-poisoning outbreak, and 6 months later, after an in-
termission of excretion for 3 months, finally became sick
and the bacilli were found in the feces.

Bainbridge and Dudfield 2 report a contact outbreak of
13 cases in a boarding house at Paddington. It was thought
probable that it was due to a carrier, but none was found.
At Wrexham in 1910 over 100 cases of food poisoning oc-
curred which were traced to pork pies. The meat was not
infected when received, as other portions sent to other
places did no harm. Apparently infection took place dur-
ing the making of the pies. It was found that the head
cook, who did not eat the pies and who was not sick, had
the bacilli in the feces and was therefore presumably the
carrier who infected the pies, though she claimed not to
have had anything directly to do with the making of the
pies. She had another typhoid-like bacillus in her urine,
and these too were found in the pies. Sacque*pee* and
Bellot 3 also report the case of a cook who had an abortive
attack and while continuing at work infected 19 persons in

1 Sachs-Miike, Klin. Jahrb., Jena, 1909-10, XXII, 237.

2 Bainbridge and Dudfield, J. Hyg., Cambridge, 1911, XI, 24.

» Sacque>ee and Bellpt, Prog. meU, Pax., 1910, 3. s. XXVI, 25.

CARRIERS AND MISSED CASES 61

a garrison of 250. G. Mayer J writes of a man who became
a carrier from eating meat and later caused the infection
of another. An outbreak of 38 cases was due to eating
vegetables fertilized with the contents of a privy vault used
by a man with " liver trouble," who had paratyphoid B
bacilli in his feces.

Cholera Spirilla in Convalescents. — Usually the germs
of cholera disappear from the feces early in convalescence,
and until recently chronic carriers were unknown. Pfeiff er 2
reviews the literature, and cites Simond's observation that
the average duration of infection is only about 6 days, and
that the longest seen by him was 18 days. Of 117 cases
reported to Rumpel, not one carried the germs over 24
days. Abel and Claussen found the average of 17 cases to
be 5 or 6 days, and Pfeiffer the average of 39 cases 10 days,
though in 2 the infection persisted 23 days. Other writers
have made similar observations.

Zirolla 3 found 29 convalescents excreting bacilli from 6
to 40 days. Zlatogoroff 4 followed 255 cases until three
negative examinations were made. In 134 the spirilla dis-
appeared by the fourteenth day, and in 22 they persisted
after 21 days, in one case lasting for 56 days. Burgers 5
found the average duration of infection in a small outbreak
was about 3 weeks from the beginning of sickness, but in one
case the spirilla persisted for 69 days. According to Kolle,6
cholera spirilla are sometimes found in the intestines of con-
valescents as long as 48 days. Rommelaere 7 reported a

1 G. Mayer, Centralbl. f. Bakteriol. [etc.], Jena, I Abt. Orig., 1910,
LIII, 234.

2 Pfeiffer, Klin. Jahrb., Jena, 1908, XIX, 483.
8 Abst., Med. Officer, 1911, VI, 84.

4 Zlatogoroff, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1911,
XLVIII, 14.

6 Burgers, Hyg. Rundschau, Berl., 1910, XX, 169.

6 Kolle, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1895, XVIII.

» Rommelaere, J. de meU, Brux., 1892, XCIV, 837.

62 THE SOURCES AND MODES OF INFECTION

case retaining the infection 47 days. Forrest x found the
germs remaining 6 weeks. Kirchner 2 says that the spirilla
are sometimes carried for weeks or months.

Zlatogoroff says that Jakowleff found the germs in the
feces 56 days after the attack, and Zeidler 93 days. Zlato-
goroff himself found them for 56 days.

Chronic and Intermittent Carriers. — Until recently it
was thought that cholera carriers were always of the
" temporary " type, and that propagation of the spirilla
was confined to the intestinal contents. It was believed
that they did not invade the gall bladder or other organs
as do typhoid bacilli. Observations by Kulescha, however,
show that the spirilla can be found in the gall bladder in
10 per cent of the cases, sometimes producing lesions and
propagating in the biliary passages. Consequently it is
not surprising that Kulescha reports a case which was dis-
charged from the hospital in January, 1909, after having
had the spirilla in the feces for 57 days, and which re-entered
the hospital in October of the same year with hepatic
trouble and soon died. Cholera germs were found in the
biliary passages. McLaughlin,3 who takes these facts from
Kulescha, states that Gaffky reports a cholera convalescent
who was a carrier for 6 months. Adami, Vallee and Mar-
tineau 4 have published notes of a case which arrived in
Quebec from Russia 16 November, 1910. The man had
perhaps had a slight attack of cholera on the steamship.
At any rate cholera germs were found in his feces, and he
remained a carrier until the next May. The spirilla were
identified by agglutination.

Intermittent Excretion of Spirilla. — Intermittent excre-
tion, which at first was not suspected, is now known to
occur. Zirolla states that 2 cases observed by him showed

1 Forrest, J. Trop. M. [etc.], Lond., 1908, XI, 321.

- Kirchner, Klin. Jahrb., Jena, 1908, XIX, 483.

8 McLaughlin, Boston M. & S. J., 1911, CLXV, 561.

* Adami, Valle"e and Martineau, Canad. M. Ass. J., 1910, I, 697.

CARRIERS AND MISSED CASES 63

a germ-free period of 1 month and 20 days jrespectively.
During such a period attacks of indigestion or the admin-
istration of salts caused the spirilla to reappear in the feces.
Creel 1 reports the case of a carrier under observation at
quarantine in New York for 54 days. Of 12 examinations
of the feces, 6 were positive and 6 negative.

Spirilla in Healthy Persons. — Dunbar2 was the first to
note the occurrence of cholera spirilla in the feces of healthy
persons. He discovered 28 healthy carriers in Hamburg
in 1892-93. Rommelaere noted such a carrier in 1892, and
carriers are by Koch considered an important factor in the
spread of the disease. In 1905 cholera seemed to have
been brought down the rivers from Russia into Germany
by raftsmen, and Kirchner says that of 27 persons on one
raft 2 were carriers. Pfeiffer states that in this outbreak
there were 174 cases of the disease, and that 38 other car-
riers were discovered. He cites Frosch as discovering 16
carriers, of whom 12 were children, and Friedheim as find-
ing 51. In one family observed by Pfeiffer 4 died, 1 was
very sick, 1 was mildly sick, and 2 were carriers. He says
that there are very many mild cases of the disease which
can only be recognized bacteriologically. Burgers3 dis-
covered 6 carriers connected with an outbreak of 30 cases
in East Prussia. In Madeira in 1910,4 of 600 contacts, 37
proved to be carriers. Of these, 12 later developed the dis-
ease. The average duration of infection was 6 to 8 days.
In a convalescent it continued for 5 weeks. McLaughlin5
found 17, or 6.44 per cent of 264 prisoners in Manila, to be
carriers, and in the city 27, or 7.18 per cent of 376 persons
examined.

During the summer of 1911 there was some cholera in

Creel, J. Am. M. Ass., Chicago, 1912, LVIII, 187.
Dunbar, Mod. Med., Osier, Phila. & N. Y., 1907, II, 720.
Burgers, Hyg. Rundschau, Berl., 1910, XX, 169.
Franca, Bull. Soc. path, exot., Par., 1911, IV, 358.
McLaughlin, J. Am. M. Ass., Chicago, 1909, LII, 1155.

64 THE SOURCES AND MODES OF INFECTION

the Philippines but only 1 case in Manila. Of 7 persons
brought in close contact with this case all were shown to
be carriers.1 While during an outbreak, as stated above,
6 per cent of the population may be found to be carriers,
McLaughlin 2 states that between outbreaks thousands of
stools may be examined without finding a single carrier.
That carriers increase as the number of cases increase is a
phenomenon common to other diseases as well as cholera.
It will be shown in the following pages to be true for
cerebro-spinal meningitis and for diphtheria. Jakowleff,
Zabolotny, Zlatogoroff and Kulescha3 state that in St.
Petersburg the feces from 2440 apparently well persons
were examined, all of whom had been more or less in con-
tact with cholera cases. Of these 125 showed the cholera
spirillum, of whom 40 proved to be mild unrecognized
cases, 25 were incubating the disease, and 60 were true
carriers.

During nearly four months of the summer of 1911, while
cholera was prevailing in certain parts of Italy, bacterio-
logical examination of the feces of immigrants was largely
made use of to prevent the introduction of the disease into
the United States. A sharp watch was kept by the steam-
ship companies and by the United States officials on the
other side to prevent the embarkation of possibly infected
persons, and indeed all immigrants from infected districts
were kept under observation for five days. Many also
were subjected to bacteriological examination. According
to a letter from Dr. John F. Anderson, of about 20,000 so
examined in Italy, 41 were found to be carriers. Of about
25,000 immigrants examined at American ports, 27 were
shown to be carriers. All of these were discovered early

1 U. S. Pub. Health and Mar. Hosp. Serv., Pub. Health Rep., 1911,
1493.

* McLaughlin, N. York M. J., 1911, XCIII, 115.

* Jakowleff, Zabolotny, Zlatogoroff and Kulescha, Bull. Soc. path,
exot., Par., 1909, II, 276.

CARRIERS AND MISSED CASES 65

in the season while the control of embarkation was not
so strict.

Gotschlich1 examined pilgrims returning from Mecca,
and though cholera had not so far as known prevailed
among them, he found several Russian and Turkish pil-
grims who proved to be carriers of the spirilla. Accord-
ing to Pfeiffer, these spirilla of Gotschlich have been
carefully studied by a number of observers and show
slight variations from the type, and have probably lost
their virulence.

Atypical Cases of Cholera. — While perhaps most cases
of cholera are readily recognized, atypical and mild cases
occur, and they are particularly frequent among children.
McLaughlin 2 says that cholera in children simulates acute
and chronic enteritis and meningitis, and is often thus
wrongly diagnosed by practicing physicians. By means of
examination of the feces he raised the proportion of cholera
cases reported among children from 22 to 35 per cent of
the total cases.

Cholera Derived from Carriers. — Cholera spirilla from
carriers have according to Zlatogoroff and others been
shown to be as virulent for animals as those from cases,
but this perhaps ought not to be taken as an indication of
their virulence for man. While Pfeiffer gives several in-
stances in which cholera was spread by carriers, the num-
ber of definite cases recorded in medical literature is not
large. A very striking one, however, is reported by Mac-
rae.3 In a hospital in Calcutta, 10 nurses, 3 patients and
a sweeper developed cholera within 4 days. An examina-
tion of 127 articles of food and drink demonstrated spirilla
in 14 samples of water and milk, all of which had been
handled by the servants. An examination of 12 servants

1 Gotschlich, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1906,
LIII, 281.

2 McLaughlin, Philippine J. Sc., Manila, 1909, IV, Sec. B, 363.
» Macrae, Indian M. Gaz., Calcutta, 1909, XLIV, 361.

66 THE SOURCES AND MODES OF INFECTION

showed that 2 were carriers, and spirilla were recovered
from their hands.

Since the possibility of the spread of cholera by carriers
has been recognized, outbreaks of the disease have occurred
in Russia and Germany, and both Russian and German ob-
servers consider that carriers afford the best explanation
of many of the phenomena of cholera outbreaks. The re-
peated occurrence of the disease in the Philippines, since
the occupancy of the islands by the United States, has
given a splendid opportunity for the study of cholera epi-
demiology, which has been well taken advantage of by our
officials. Heiser and McLaughlin are firmly convinced that
carriers are a most important factor in the spread of the
disease. Indeed, McLaughlin says that their importance
can hardly be overestimated. He has no doubt that chol-
era is often spread from village to village by carriers, and
that intervals between local outbreaks are bridged over by
the same means. There is no evidence that the spirilla
maintain a saprophytic existence in the soil or water of the
islands.

Chantemesse 1 says cholera was introduced into Italy in
1910 by a party of gypsies who arrived at Brindisi, July 27,
from Batum. They then went to Trani, where the disease
developed August 7. None of the party were sick. Cases
developed in October in Marseilles after the arrival of im-
migrants from the East, though there was no recognized
sickness among them. On October 7 cholera developed
on a ship which carried Russian emigrants from South-
ampton 30 days after the latter had left Russia. The bag-
gage in all these cases had been disinfected and the emi-
grants themselves had been examined several times by
physicians, who failed to find recognizable cholera.

In 1911 there were five cases of cholera in the United
States, of which only one had any connection with a pre-
vious case. Shall we assume that dried and dying germs
1 Chantemesse, Bull. Acad. de m6d., Par., 1911, LXXV, 113.

CARRIERS AND MISSED CASES 67

on clothing caused them, or living germs from a carrier?
In 1893, at a time when a cholera ship lay in the harbor
several miles from New York, a number of unconnected
cases occurred in that city. Were they caused by flies, by
floating mattresses or clothing, as was then supposed, or
were they derived from a carrier from some other ship?
Similarly, •" sporadic " cases occurred in England at the
same period. In 1873, at least three local outbreaks in the
United States were traced to immigrants recently arrived
from Europe at points in the interior. Then it was thought
that the germs were brought in the baggage. To-day do
not carriers seem a more likely explanation? In 1866,
New York was the starting place of an extensive epidemic.
From May to July there were only a few cases in the city
not directly connected with any imported cases, nor often
with each other. The epidemiologists of the day attrib-
uted these cases to fomites, or the feces of recovered cases.
Our present-day knowledge shows us how much more
likely the latter is as a cause than the former, and vindi-
cates the judgment of the men of that day who without
the help of bacteriology surmised the existence of the
" carrier." McLaughlin x says that in the Bilibid prison at
Manila the food and water were so well controlled that the
only entrance for cholera was by means of carriers. As
soon as carriers were sought for and isolated the disease
was stamped out.

Bacillary Dysentery. — There are two forms of dysen-
tery, one caused by bacilli, the other by amebae. The dys-
entery bacilli form a group of closely allied types, as the
Shiga, the Flexner and the Harris types. Certain of the
group are called pseudo-dysenteric but yet are pathogenic.
What is commonly called dysentery may be caused by a
number of different pathogens, and the dysentery bacilli
may cause diarrheal symptoms quite different from typical
dysentery. For the present purpose the whole dysentery
1 McLaughlin, N. York M. J., 1911, XCIII, 115.

68 THE SOURCES AND MODES OF INFECTION

group of bacilli may be referred to without distinction.
Collins 1 and Goodwin 2 have studied the occurrence of the
bacillus in well persons. According to Collins, Flexner,
Wollstein and others failed to find it in healthy persons.
Duval 3 found it in 2 instances, and Charlton and Jehle 4
in 2 of 10 cases examined. Collins found it in 2 of 30 normal
persons, and in 1 three weeks after an attack of dysentery,
and in another child who had had a few raucous stools.
Goodwin found the Flexner-Manila type of the bacillus in
1 of 59 well persons examined. Kruse 5 says that all types
of the bacilli have been found in well persons, and that these
carriers are an important cause of the disease. Though
carriers have fewer bacilli in their feces, which are also less
in volume, they come in contact with a larger number of
persons than do the sick. There are also many mild
atypical cases. Kruse has found the bacilli in relapses
after two years,6 and says that Drigalski and Lentz have
made similar observations. He also cites Ford as finding
10 carriers among 50 persons examined. In an asylum
outbreak Heuser 7 found 3 carriers. Conradi,8 while study-
ing a contact outbreak near Metz, found several carriers.
Kuster 9 isolated bacilli from a carrier who had probably
been excreting since an attack years before. Kruse cites
Kriege as saying that about 4 of 36 cases of dysentery
become chronic, but usually the infection lasts only 2 to 6
weeks. Conradi found- that in 4 of 27 cases the bacilli
persisted up to the end of the fourth week, while in 11
cases they disappeared by the middle of the second week.

Collins, Rep. Dept. Health, City of New York, 1904, I, 428.

Goodwin, Rep. Dept. Health, City of New York, 1904, I, 423.

Duval, Studies from Rockefeller Inst., 1904, II, 42.

Charlton and Jehle, Tr. Ass. Am. 'Physicians, 1904, XIX, 405.

Kruse, Med. Press & Circ., 1908, LXXXV, 175.

Kruse, Klin. Jahrb., Jena, 1908, XIX, 529.

Heuser, Deutsche med. Wchnschr., 1909, XXXV, 1694.

8 Conradi, Festschrift v. Robert Koch, 1903, 555.

9 Kuster, Miinchen med. Wchnschr., 1908, LV, 1833.

CARRIERS AND MISSED CASES 69

Shiga 1 says that the bacilli generally remain 1 or 2 weeks,
and that perfectly normal persons sometimes harbor the
germs. Aveline, Boycott and McDonald 2 failed to find the
bacillus in 27 contacts in an asylum.

Cameron3 also, and Macalister,4 rarely found bacilli in
perfectly healthy contacts. The latter writer found that
26 per cent of the cases studied by him in a certain asy-
lum developed a mild chronic state or had relapses, and
that these conditions only could be regarded as dangerous.
These he considers as practically carriers, and says their
importance cannot be overestimated.

O. Mayer 5 has reported finding entirely healthy carriers,
and one of these later became sick. The bacillus found
was of the " pseudo- " type Y. In convalescents he found
them persisting up to 202 days. The average persistence
is 3 to 7 months. In healthy persons they were not found
over 30 days. Intermittency of excretion was noted in
both convalescent and healthy carriers.

Chronic Plague in Rats. — Bubonic plague is a disease
which attacks not man alone but many other species of
animals, particularly the rat. In fact it may be considered
primarily a rat disease, and without doubt the rat is the
most important agent in its diffusion. That mild cases
and chronic cases exist among rats which superficially ap-
pear not to be sick, seems to be proved. Simpson 6 says
chronic plague was observed in some of the animals experi-
mented on in Hong Kong, and by Albrecht and Ghon in
guinea pigs, and in rats for months by Kolle and Martini.
The Indian Plague Commission (1905)7 found eleven of the

1 Shiga, Philippine J. Sc., Manila, 1906, I, 485.

2 Aveline, Boycott and McDonald, J. Hyg., Cambridge, 1908, VIII,
309.

3 Cameron, Brit. M. J., Lond., 1911, I, 973.

4 Macalister, Brit. M. J., Lond., 1910, II, 1506.

5 Mayer, Munch, med. Wchnschr., 1910, LVII, 2566.

6 Simpson, A Treatise on Plague, Cambridge, 1905, 129.

7 J. Hyg., Cambridge, 1907, VII, 379.

70 THE SOURCES AND MODES OF INFECTION

rats which they had fed with plague bacilli to be infected,
although they appeared to be perfectly well. While chronic
plague has been seen in laboratory animals, several observers
in Bombay and Sydney have failed to find it under natural
conditions, as also did Blue in San Francisco.1 But Hunter
found rats with chronic plague in Hong Kong, and the In-
dian Plague Commission (1905) found a number of rats in-
fected with plague at a time when no rats with acute plague
could be discovered, and when there was no outbreak among
human beings. The infected rats showed no sign of sick-
ness.2 Later observations discovered similar instances of
chronic rat infection in animals caught in a number of dif-
ferent places. The pathological conditions indicated a
slowly resolving rather than a true chronic process. Wat-
kins-Pitchford3 found the bacillus in convalescent rats and
guinea pigs. It is evident that such chronic " carriers "
may be an important factor in the maintenance and exten-
sion of the disease.

Atypical Human Plague. — Among human beings mild
cases of the glandular type are by no means uncommon.
But fortunately these are rarely dangerous, for without
suppuration there is no escape of bacilli. But in certain in-
stances, later suppuration may take place, or lung symp-
toms develop, so that the individual may become a focus of
infection.4 Rat " carriers/' on the other hand, are always
dangerous, for they may at any time suffer accidental death,
and their carcasses may then readily infect other animals
and even man.

Plague Bacilli in Convalescents. — In the pneumonic type
in human beings the bacilli are thrown off in large numbers

1 Blue, J. Hyg., Cambridge, 1909, IX, 1.

* J. Hyg., Cambridge, 1906, VI, 530-535; 1907, VII, 373.

8 Watkins-Pitchford, Rep. of Bacteriologist, Pietermaritzburg, Natal,
1903 [Rep. on Flague, 31].

4 Gotschlich, Kolle u. Wassermann, Handbuch, Jena [etc.], 1904,
IV1, 69.

CARRIERS AND MISSED CASES 71

from the lungs. Martin,1 Gotschlich 2 and others have found
them in the sputum up to 76 days after the attack, or 42
days after recovery. Gaffky3 cites Vagedes as reporting pul-
monary infection lasting 2 months in a case in Oporto, and
bacilli in an abscess persisting more than 2 months, and Vages
one lasting 4 weeks in Paraguay. The latter also isolated the
bacilli from a man who later became sick with the disease.
Shottelius found the germs in the bronchial secretion of mild
ambulant cases. It is thus very probable that mild cases
and true carriers among both rats and human beings may
play a considerable part in the dissemination of this disease.
White Diarrhea of Chicks. — One form of diarrhea com-
mon among young chickens appears to be due to a bacillus,
B. pullorum. The nature and mode of spread of this dis-
ease were well worked out by Rettger and Stoneburn.4 It
affects young chickens three or four days old and spreads
among them by contact or infection of their food with
excreta. Many of the survivors become carriers of the
bacillus until after they have become mature hens. Such
hens show few or no symptoms. The eggs which these
carriers lay become infected in the oviduct, and chicks
hatched from them develop the disease, though they never
come in contact with frank cases. Similar observations
have been made by Gage.5 Note is made of this disease
here not because it is transmissible to human beings, for
so far as is known it is not, but because it is an excellent
illustration of latency, and shows how efforts to suppress a
disease are doomed to failure unless account is taken of the
part played by carriers. No amount of isolation of chicks
sick with this form of diarrhea can ever be successful in
stamping out the disease.

1 Martin, Ann. de 1'Inst. Pasteur, Par., 1900.

2 Gotschlich, Ztschr. f. Hyg. Infectionskrankh., Leipz., 1899, XXXII,
402.

3 Gaffky, Klin. Jahrb., Jena, 1908, XIX, 491.

4 Rettger and Stoneburn, Storrs, Agric. Ex. Sta., Bull. 60 and 68.
6 Gage, J. Med. Research, Bost., 1911, n. s., XIX, 491.

72 THE SOURCES AND MODES OF INFECTION

It is also interesting to note the analogy to ophthalmia
neonotorum, which in many cases is due to mild, long-
standing and unrecognized infection of the genital tract of
the mother with the gonococcus.

Mediterranean Fever. — Mediterranean fever, like plague,
appears to be a disease of the lower animals, only secondarily
affecting man. The germ which is its cause may, like so
many other pathogenic organisms, develop in the body with-
out giving rise to symptoms. Goats appear to be the chief
source of human infection. In 1905 there were 363 cases
among the garrison at Malta, but in 1906, presumably owing
to the cutting off of the supply of goat milk, there were only
35 cases; * and in 1907 it was practically exterminated.2 An
interesting account is given of an outbreak of the disease on
a steamship, among persons who drank the milk of a herd of
goats that were being brought to America.3 The goats were
not sick. Horrocks4 shows that probably one or more
animals in every herd are excreting the germs in the milk
and urine, and that 50 per cent give evidence by serum
reaction that they are, or have been, infected. Other in-
vestigations show that as high as 10 per cent of milch
goats have the germs in their milk, although they present
no symptoms of the disease. Carriers are also common
among human beings. Shaw5 found that 10 of 525 well
persons were excreting the germs in the urine. Ross,6
of Port Said, found the bacillus not constant in either milk
or blood.

Meningococcus in Nose. — While the germ of epidemic
cerebro-spinal meningitis (Micrococcusmeningitidis) has been

1 Hewlett, Practitioner (Lond.), 1908, LXXX, 222.

2 Bruce, Nature, Lond., 1908, LXXVIII, 39.

3 U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab. Bull. No. 41,
203.

4 Horrocks, Rep. of Commission of Roy. Soc., 1905-6, Pts. Ill, IV.
6 Shaw, J. Roy. Army Med. Corps, Lond., 1906, VI, 638.

• Ross, J. Roy. Army Med. Corps, Lond., 1911, XI.

CARRIERS AND MISSED CASES 73

known for many years, it is only very recently that any ex-
planation has been forthcoming as to the mode of infection.
The finding of the organism in the nose of patients suggests
the possibility that infection may pass to the brain from this
point. It is theoretically possible for infection to take place
through the cribriform plate of the ethmoid bone, or, as sug-
gested by Westenhoeffer,1 by the lymphatics from the pha-
ryngeal tonsils, or as some think, by absorption from the
alimentary canal and passage through the lymph channels or
the general circulation. The organism is certainly found in
the blood in a considerable number of cases.2 What the exact
route to the brain is, however, has not yet been determined.

That the germ of this disease is found in the nose of the
sick is now generally recognized, though considerable doubt
has been thrown on the accuracy of some of the earlier obser-
vations, as this organism is not readily distinguished from
the cocci frequently found in the normal nose. Culture and
agglutination tests are now recognized as the only valid
methods of differentiation. Among those who have certainly
isolated it in a considerable number of instances from the
nasal mucous membrane of cerebro-spinal meningitis cases
may be mentioned Dunham,3 Weichselbaum and Gohn,4 Lin-
gelsheim,5 Goodwin and von Sholly 6 and others.

Causes Rhinitis. — That thismicrococcus is frequently found
in the nose of contacts, and other persons showing no symp-
toms of the disease, is now well established. An interesting
case is that of Kiefer,7 who while working with a culture in
the laboratory developed a severe rhinitis and succeeded in

Westenhoeffer, Berl. klin. Wchnschr., 1905, XLII, 737.

Birnie and Smith, Am. J. M. Sc., Phila., 1907, CXXXIV, 582, and
Simon, J. Am. M. Ass., Chicago, 1907, XLVIII, 1938.

Dunham, J. Infect. Dis., Chicago, 1906 [Suppl. No. 2], 10.

Weichselbaum and Gohn, Wien. klin. Wchnschr., 1905, XVIII, 625.

Lingelsheim, Klin. Jahrb., Jena, 1906, XV, 373.

Goodwin and von Sholly, J. Infect. Dis., Chicago, 1906 [Suppl.
No. 2J, 21.

7 Kiefer, Berl. klin. Wchnschr., 1906, XXXIII, 628.

74 THE SOURCES AND MODES OF INFECTION

recovering the micrococcus from his nose. A similar case was
reported to me by P. E. Rauschenbach, at the time working
in a hospital at Newark. Ford1 also had a case of this kind.
Meningococcus in Contacts. — Among those who have
found the organism in well persons may be mentioned Weich-
selbaum and Gohn,2 who obtained it in 3 of 24 persons exam-
ined. Ostermann3 found it in 17 of 24 contacts in Breslau,
but his methods of employing the agglutination test have
been called in question. One well child in whose nose the
germs were found was taken sick three days later and died
the next day. Many of the carriers had pharyngitis and
rhinitis. Of 49 children in schools where these carriers at-
tended, 2 showed the meningococcus, and one of these had
been playing with a carrier. Ostermann failed to find it in
50 persons who had not come in contact with cerebro-spinal
cases. Goodwin and von Sholly, in New York, obtained it
in 5 of 45 contacts, and a similar coccus which did not agglu-
tinate, in 2 of 55 medical students not contacts. Bolduan4
found the organism in 10 per cent of 150 contacts. Lingels-
heim,5 during an extensive outbreak in Beulen, Prussia, found
the organism in the nose of 26, or 9 per cent of 289 contacts.
Later he found it in 4 of 56 school children, but all 4 came from
families where there had been cases of the disease. In 2 of
the 4 cases no agglutination test was applied. In the cases
of the disease examined by Lingelsheim the micrococcus dis-
appeared by the fifth day in 66 per cent, and by the sixth to
tenth day in 24 per cent. In 4.39 per cent of the cases it per-
sisted over three weeks, and in 1 case it was found three
months from the beginning of the attack.

1 Quoted by Councilman, J. Am. M. Ass., 1905, XLIV, 999.

2 Weichselbaum and Gohn, Wien. klin. Wchnschr., 1905, XVIII,
625.

3 Ostermann, Deutsche med. Wchnschr., 1906, XXXII, 414.

4 Bolduan, Med. Times, N. Y., 1908, XXXVI, 193.

6 Lingelsheim, Deutsche med. Wchnschr., 1905, XXXI, 1017, 1217;
Klin. Jahrb,, Jena, 1908, XIX, 519.

CARRIERS AND MISSED CASES ^5

Infection in the Family. — Bruns and Hohn1 found that the
proportion of carriers decreased as the outbreak decreased.
This is shown in the following table:

Reported
Cases.

No. of
Families

No. of Well
Persons
Examined.

No. of
Carriers.

Per cent of
Carriers.

March

148

7

23

14

61

April

278

39

135

67

50

May

327

42

172

81

47

June

188

23

93

34

36

July .

146

21

67

18

27

August

68

22

119

10

8 5

They found that of the fathers in these families 60 of 113
carried the germs; of the mothers, 39 of 114; of children in the
families, 118 of 360; and of other members of the families
7 of 22 were carriers.

Persistence of Infection. — Bruns and Hohn give for the
duration of the infection the following:

For 8 days 28

2 weeks 18

3 weeks 13

4 weeks.. 10

For 5 weeks 4 cases

6 weeks 3

7 weeks 3

8 weeks 1 case

Selter, in Bonn,2 has observed a very much longer persist-
ence of the infection. In the case of a mother and daughter
who recovered from the disease, the cocci persisted from
February 3 to June 4. The father in this family, who
had not been sick, yielded positive findings in May, June,
July and August, and had in all probability been infective
for seven months. In another family where the disease ap-
peared, the 6 well persons were carriers at one time or
another from February 18 to June 5, during which period

1 Bruns and Hohn, Klin. Jahrb., Jena, 1907-08, XVIII, 285.
» Selter, Klin, Jahrb., Jena, 1908-09, XX, 457.

76 THE SOURCES AND MODES OF INFECTION

they were examined twelve times. Sometimes the examina-
tions were positive and sometimes negative, which is just
what occurs in the examination of typhoid and diphtheria
carriers. This apparent intermittency may be due in part
to faulty technique, and in part to the temporary absence
of the bacteria. In the 69 examinations of the family
above referred to, 49 were positive and 20 negative. Selter
could find no difference between the cocci found in the
sick and in the carriers, but he was not able to trace a
case of sickness to any of his known carriers. Kirchner,1
in Hamburg, found 22, or 9.7 per cent, carriers, out of 237 well
members of infected families, but in 3 families all the mem-
bers were carriers, and in 10 other families 75 per cent were.
Most of these were adults. In two instances the infection
persisted 44 and 66 days respectively. Vagedes2 reports 3
cases in barracks at Charlottenburg. Of 58 hospital attend-
ants, etc., there, 4 were carriers, and of 593 of the soldiers 10
only were carriers. On a second examination five days later
1 only was found, and nineteen days later none. By the iso-
lation of carriers the disease was " stamped out," but it will
be noticed that the infection was not widely diffused before it
was recognized. Bochalli,3 in certain barracks where the
disease prevailed, found 10 of 16 roommates of the sick, or
62 per cent, to be carriers. Of 485 in other companies, 42, or
8.6 per cent, were infected. Usually the germs quickly dis-
appeared, but in one case they persisted for four and a half
months. In another instance a nurse, going to a district
where there was no meningitis, was attacked about one month
later. Similar observations have been made in several places
during the recent outbreak in Scotland. Buchanan,4 in Glas-
gow, found the micrococcus in 81, or 26.3 per cent of 308 con-

1 Kirchner, Klin. Jahrb., Jena, 1908, XIX, 473.

2 Vagedes, Deutsche mil.-arztl. Ztschr., Berl., 1907, XXIII, 647.

» Bochalli, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1908, LXI,
454.

4 Buchanan, San. Rec,, Lond,, 1907 n, s,, XI, 245.

CARRIERS AND MISSED CASES 77

tacts in 74 families. Most of them were over fifteen years of
age. In 14 families he obtained it from more than one
person, in 2 instances from five. In 4 instances it was
found in the nose two years, one year, one year, and three
months, respectively, after an attack. He quotes Arbuckle,
medical officer of health of Partick, as finding 23.1 per cent
of 230 contacts infected. In Leith, Fraser and Comrie 1
found it in 10, or 14 per cent, of 69 contacts. Of these 5 were
adults whose children were sick, and all of whom had worked
on a ship in the air of whose engine room meningococci were
found. Bethge,2 immediately after the appearance of the
first case in a certain institution, found that of 187 persons
66 were carriers.

Found only in Those near Sick. — The micrococcus which
is the cause of this disease, while frequently noted in contacts,
is rarely found in those not exposed to the disease. Oster-
mann, when there were no cases about, failed to find the germ
in 50 children and in many adults. Bolduan did not find it
in 150. Kolle and Wassermann3 recovered the germs from
2 of 114 persons, but one had been in contact with the disease,
and the other shortly became sick. Bochalli found none in
40 men in a non-infected regiment, Lingelsheim4 none in
129 persons otherwise sick, and none in 184 non-exposed chil-
dren, and he cite.s Droba and Kucera as finding none among
210 children living where there was no meningitis. In 23 per-
sons not exposed to the disease and examined by Fraser and
Comrie none of these organisms were found. Arkwright5
failed to find them in 54 well persons, and Flexner6 could
find none in 50 persons in Philadelphia at a time when the
disease was not present in the city.

Fraser and Comrie, Scot. M. & S. J., Edinb., 1907, XXI, 18.
Bethge, Deutsche med. Wchnschr., 1910, XXXV, 66.
Kolle and Wassermann, Klin. Jahrb., Jena., 1906, XV, 507.
Lingelsheim, Klin. Jahrb., Jena, 1906, XV, 373.
Arkwright, J. Hyg., Cambridge, 1907, VII, 145.
Flexner, J. Exper. M., N. Y., 1907, IX, 105.

78 THE SOURCES AND MODES OF INFECTION

As Lingelsheim says, it appears that the nearer we approach
cases of the disease the more numerous carriers are, and the
more extensive the outbreak the more numerous they are.
Adults are more commonly infected than children. Fliigge
thinks that carriers are ten to twenty times as numerous as
cases, and that is doubtless true.

Infection by Carriers. That carriers are the chief source
of the disease is the opinion of those Germans who have lately
had experience in severe outbreaks, and it is shared by the
Scotch health officers. Lingelsheim says that there is no
bacteriological or epidemiological evidence to show that the
sick are more dangerous than the well, otherwise tire cases
would occur in groups. On the contrary the carriers mingling
freely with the public are the most dangerous. Jehle x states
that 23 cases occurred in children in the families of miners.
None of the miners were sick, but when they were transferred
to another mine, children there soon began to develop the
disease. Their parents were doubtless carriers. Meyer2
reports an instance where carriers gave rise to three cases.
At Leith, Buchanan3 says that the first few cases could be
traced to carriers, and Thomson4 makes a similar statement
regarding the early cases in Lanarkshire. Flatten 5 gives the
details of numerous instances of the transmission of the dis-
ease by well persons.

Carriers explain Spread of Disease. — Previous to the dis-
covery of the presence of the specific organism of cerebro-
spinal meningitis in the nose, and particularly in the nose of
contacts, the mode of extension of the disease was a mystery.
Though certainly an infectious disease, and appearing in epi-
demics, often quite severe, it was apparently only slightly if

1 Jehle, cited by Warrington, J. Roy. San. Inst., Lond., 1907, XV, 656.

2 Meyer, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1909,
XLIX, 305.

3 Buchanan, Brit. M. J., Lond., 1907, II, 852.

4 Thomson, Med. Press & Circ., Lond., 1908, n. B., LXXXVI, 344.
6 Flatten, Klin. Jahrb., Jena, 1906, XV, 265.

CARRIERS AND MISSED CASES

79

at all contagious. Personally, out of over a hundred cases, I
have only twice seen two cases in the same family, and have,
except in those instances, never been able to trace connection
between any two cases. Cases are frequently treated in con-
siderable numbers in general hospitals all over the world, and
infection of others has rarely, if ever, taken place. In New
York a greater degree of contagiousness has been noted than
in most places.1 Of 1500 consecutive cases, 112, or 7 per cent,
were secondary cases in the family. It is said that when more
than one case occurs in a family they all are usually taken sick
at about the same time,2 but this was not so in New York.
The following shows the date at which the secondary cases
appeared in that city:

Where First Cases were removed to
Hospital.

Where First Case remained at Home

No. of Secondary
Cases.

Interval.

No. of Secondary
Cases.

Interval.

14

1-7 days

3

Iday

5

1-2 weeks

4

2 days

4

2-3 "

1

3 "

3

3-4 "

1

5 "

2

4r-5 "

1

6 "

3

5-6 "

4

7 "

2

7-8 "

1

9 "

1

3mos.

1

11 "

34

1

21 "

1

30 "

18

The 34 cases which developed after the removal of
the primary case to the hospital indicate either that the
germ of the disease remained in the house, on fomites, which
is highly improbable, as its powers of resistance are extremely

1 Am. Pub. Health Ass. Rep., 1905, XXXI, 359.

1 Bolduan and Goodwin, Med. News, N. Y., 1905, LXXXVII, 1222.

80

THE SOURCES AND MODES OF INFECTION

slight, or that unrecognized carriers, persons with the coccus
growing in the nose, remained behind, a supposition which
is entirely in accord with bacteriological findings. In Glas-
gow,1 of 194 cases, 7, or 4.1 per cent, were secondary cases in
the family. Some of these occurred after disinfection. It is
not unlikely that the more extensive the outbreak the more
numerous will be the " secondary " cases in the family. In
Oppeln, where there was a very severe outbreak, Flatten2
found the number of secondary cases as follows in successive
months of the outbreak:

No. of Families.

No. of Cases.

January

71

79

February

143

163

March

257

296

April

310

368

May

349

414

June

374

439

If, as appears, the meningococcus is frequently found in
the nose of sick persons and of contacts, the mode of exten-
sion of the disease becomes apparent. Observations indicate
that a very considerable number of persons who come in con-
tact with cases of this disease carry the organism on their
mucous membrane. Yet very few of these persons ever be-
come sick. It appears that the development of the cerebro-
spinal symptoms, that is the disease as we know it, is really
a rather unusual accident of a not uncommon harmless infec-
tion of the nasal passage. We can therefore understand how
the disease, though apparently showing little contagiousness,
is nevertheless a strictly contagious disease. It extends almost
entirely through the agency of carriers. Sometimes, as was
previously stated, these carriers can be discovered, particu-

1 Rep. Med. Off. Health, Glasgow, 1906, 89.
• Flatten, Klin, Jahrb,, Jena, 1906, XV, 228.

CARRIERS AND MISSED CASES 81

larly early in an outbreak. The appearance of cases in dif-
ferent localities in a community, and at varying intervals, is
readily explained by, and is almost a proof of, the existence
of a considerable number of undiscovered carriers.

Meningitis an Accident of Infection. — The growth of the
germs in the nose cannot properly be considered the "incu-
bation" of the disease. Inoculation experiments1 indicate
that the nervous symptoms develop in a very few hours after
the specific bacteria are injected into the meninges. There
is much evidence, as given above, to show that the germs may
grow on the surface of the nasal mucous membrane for days,
weeks and perhaps months, but when they once gain access
to the meninges the symptoms of the disease speedily
develop.

Meningitis also occurs as a result of infection by the pneu-
mococcus. The disease thus caused is not essentially different
from the epidemic form. The number of well carriers of the
pneumococcus is enormously greater, and is much more con-
stant than is the number of carriers of the meningococcus,
and the chance of the former invading the meninges and caus-
ing sickness is small, very much less than the chance of the
latter doing so. In either case it may be considered an unfor-
tunate accident of a usually benign infection.

Inutility of Isolation of Meningitis. — If these views are
correct, little can be done by means of isolation to prevent
the spread of the disease. When an outbreak occurs, there
is certain to be a large number of carriers that cannot be
found or isolated. To isolate the sick, and even those in
immediate contact with the sick, will probably never make
any appreciable difference in the progress of the disease, that
is if the findings of to-day are confirmed by more extended
observations. By the time several cases have developed in
a community, there will probably usually be a considerable
number of carriers who cannot be found. Stringent isola-

1 Lingelsheim, Deutsche med. Wchnschr., 1905, XXXI, 1017, 1217;
Flexner, J. Exper. M,, N. Y., 1907, IX, 142.

82 THE SOURCES AND MODES OF INFECTION

tion of the families of known cases will accomplish little,
will work much hardship and injustice. v^iflUit0^^

Diphtheria Bacilli in Well Persons. — Loeffler ^himsel^in
1884 found diphtheria bacilli hi a person who was not sick,
and in 1889 Roux and Yersin2 called attention to the per-
sistence of diphtheria bacilli on the mucous membranes of
convalescents, and their observations were at intervals sub-
stantiated by others. In 1894 Park and Beebe 3 examined the
throats of 330 healthy persons who had not, so far as known,
been in contact with diphtheria cases, and found diphtheria
bacilli in 24, or 7.3 per cent, but of these only 8, or 2.4 per cent,
of the 330 were virulent. These authors minimized the dan-
ger to be apprehended from these carriers, and although for
some time similar observations were reported, they were gen-
erally, and are even now by many, considered of little impor-
tance. It seems to me otherwise, and therefore a considerable
number of these observations are here summarized. Muller 4
examined in routine 92 children sick with diseases other than
diphtheria, in Heubner's clinic, and reported that 12, or 13
per cent, were infected with diphtheria, but only 6 of the 12
were tested on guinea pigs. Later 18 children in the same
institution, who had negative cultures on admission, were
found to harbor the bacilli. Kober 5 in Breslau found that
of 600 school children 2.5 per cent were carriers of diphtheria
bacilli, but in only 0.83 per cent of the 600 were they virulent.
At Baltimore,6 in 1899, during an exceptional prevalence of
diphtheria in a number of schools, cultures were taken from
4068 pupils apparently well, and of these 157, or 3.8 per cent,

1 Loeffler, The Bacteriology of Diphtheria, Nuttall and Graham-
Smith, Cambridge, 1908, 31.

» Roux and Yersin, Ann. de 1'Inst. Pasteur, 1890, IV, 385.

3 Park and Beebe, Med. Rec., N. Y., 1894, XLVI, 385; Sci. Bull. 1,
Dept. Health N. Y. City, 1895.

4 Muller, Jahrb. f. Kinderh., 1896, XLIII, 54.

6 Kober, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXXI,
433.

c Baltimore, Rep. Com. of Health, 1899, 96; 1900, 85.

CARRIERS AND MISSED CASES 83

showed diphtheria bacilli. The next year, of 351 examined,
23, or 6.6 per cent, were positive.

Varieties of Bacilli found. — Reports of the finding of diph-
theria bacilli in well persons would have more value if it were
made clear what was meant by the term diphtheria bacilli.
The bacillus varies greatly in form, and those who have
studied it most carefully are by no means agreed as to what
should or should not be properly so called. Fortunately
Wesbrook's classification of all these varieties, both admitted
and questioned, renders it possible to interpret properly the
reports of all who will take the trouble to use his nomencla-
ture, and it is to be regretted that more do not use it. Again,
diphtheria bacilli are quite as frequently found in the nose as
in the throat, but few observers state whether the cultures
were taken from throat or nose or both. For the purpose of
securing definite data from a large number of cases, a col-
lective investigation was undertaken by the Massachusetts
Association of Boards of Health.1 The number of persons
examined was 4250 in different parts of the United States.
Most of them had probably not been directly exposed to diph-
theria except in Minnesota, where most of the cultures were
taken from schools and institutions where there had recently
been cases of the disease. The majority of the persons ex-
amined were adults, and cultures were taken from both throat
and nose. Of all persons examined, 2.89 per cent had " granu-
lar " forms of diphtheria bacilli, i.e., those showing polar
staining with methylene blue. 'If " barred " and other aber-
rant types are to be considered as diphtheria bacilli, as they
are by some, the number of cultures to be reported positive
would have been more than doubled. Referring only to the
typical granular types, 1.46 per cent of all persons showed
throat infection (either with or without concurrent nose infec-
tion) and 2.07 per cent showed nose infection (either with or
without concurrent throat infection). In the different cities
the percentage of infected persons varied from 0.43 in Provi-

i J. Mass. Ass. Bds. Health, Boat., 1902, XII, 74.

84 THE SOURCES AND MODES OF INFECTION

dence, where particular care was taken to avoid contacts,
most of the cultures being taken from schools where there
had been no diphtheria for at least two years, to 3.66 in the
Willard State Hospital, and 6.03 in Minnesota, in both of
which localities there had recently been diphtheria in the
institutions examined. There were 47 tests for virulence, of
which 8, or 17 per cent, were positive. Recently cultures were
taken from school children in Brighton (Boston) by Slack.1
Of over 4000 children examined by throat and nose culture on
two occasions, 1.93 per cent showed diphtheria bacilli. Of
these 8 were tested for virulence, of which all proved to be non-
virulent. Most of these children were excluded from school
until free from bacilli, but all of them had, of course, been
carrying bacilli in school for a longer or shorter period. No
case of clinical diphtheria was traced to any of these carriers,
either in school or out of school. These facts have been ad-
vanced as evidence that the danger from such carriers is a
negligible factor in contrast with the danger from conva-
lescents whose bacilli are usually virulent. But in Providence
in 1908, 73 convalescents carrying diphtheria bacilli were
admitted to school, and it is not known that they were in any
instance the source of new cases in the school, and in all but
two or three instances it is certain that they were not.
Forbes 2 cites Meikle as reporting 27 carriers discharged from
the hospital without any return case. These facts merely in-
dicate that the amount of danger from a single focus of
infection is much less than is generally believed.

Value of Bacteriological Findings. — Neumann3 could not
find any diphtheria bacilli in 111 perfectly normal throats
and noses, but did find them in 8 of 95 cases suffering from
catarrhal inflammation of the mucous surfaces. He does not
think that they are found in normal throats, but certainly
all who have had any experience in taking cultures have seen

1 Slack, Arms, Wade and Blanchard, J. Am. M. Ass., 1910, LIV, 951.

2 Forbes, Brit. M. *J., 1909, II, 522.

8 Neumann, Ztschr.f.Hyg.u. Infectionskrankh., Leipz., 1902, XL, 33.

CARRIERS AND MISSED CASES 85

numberless cases where diphtheria bacilli were found in
throats which, so far as appearances go, were perfectly nor-
mal. When diphtheria bacilli are found on a reddened or
slightly inflamed mucous surface, the abnormal condition
may not in every individual case be due to their presence, for
certainly " catarrhal " inflammations of nose and throat, due
to other causes than the presence of diphtheria bacilli, are
extremely common. The evidence is that at present, in cities,
from 1 to 2 per cent of the population are infected with
diphtheria bacilli. The presence of diphtheria bacilli in an
inflamed throat may be, then, in 1 or 2 per cent of the find-
ings merely a coincidence. It is safe to assume that in all
but 1 or 2 per cent the presence of diphtheria bacilli stands
in causative relation to the inflammation of the surface on
which they are found.

Diphtheria Bacilli in Children. — In Bristol, Heaven1 ex-
amined 704 children (throat and nose) connected with schools
where there had been considerable diphtheria, and found
diphtheria bacilli in 24, or 3.4 per cent. Of 758 cultures taken
from schools in London where there had been occasional cases
of diphtheria, 58, or 7.6 per cent, showed the bacilli.2 Pen-
nington3 examined 375 well children in Philadelphia schools,
in some of which there had been more or less diphtheria, and
of 37 positive cultures from these, 24 were more or less viru-
lent. Of 125 pupils from four other schools 10.4 per cent gave
positive cultures, about half of which were virulent. But
Pennington found that of 25 cultures from convalescents 22
were virulent. In the city of Mexico, of 300 school children
10, or 3.3 per cent, showed diphtheria bacilli.4 Von Sholly 5
examined 1000 tenement-house children in New York, taking
cultures only from those whose throats appeared normal. The

Heaven, Pub. Health, Lond., 1902-03, XV, 516.

Rep. Med. Off. Health, Lond., 1904, Appendix III, 25.

Pennington, J. Infect. Dis., Chicago, 1907, IV, 36

Fabela, Rep. Am. Pub. Health Ass., 1906, XXXII, Pt. 2, 199.

Von Sholly, J, Infect, Dis,, Chicago, 1907, IV, 337.

86 THE SOURCES AND MODES OF INFECTION

children for the most part were those brought to hospitals
and dispensaries for treatment, but all cases were excluded
which had any suggestion of sore throat, nasal discharge,
etc. Of these 1000 cases 5.6 per cent showed diphtheria-
like organisms, of which, however, only 1.8 per cent (of
the 1000 cases) proved to be virulent. In 50 of the cases
nose cultures also were taken, and one of these proved to
have virulent bacilli. During a very severe outbreak of diph-
theria in Christiania, Ustvedt1 examined 4277 school children,
and found diphtheria bacilli in 191, or 4.5 per cent. Of these,
10 afterwards developed diphtheria. Of 7 cultures tested for
virulence 4 were positive. In one school in which the per-
centage of positive cultures was 9.2 during the outbreak, it
was 3.2 some months later. In a school in the country where
there had been no diphtheria for years, no bacilli were iso-
lated from 86 pupils. Ruediger2 found diphtheria bacilli in
3 of 51 normal throats. Gross3 took two throat and nose
cultures from 314 children (without diphtheria) on admission
to the Children's Hospital in Boston, and isolated diphtheria
bacilli from 26, or 7.9 per cent. Five were tested for viru-
lence, 4 of which were positive. Hewlett and Murray4
found that 58, or 15 per cent, of 385 children on admission to
the Victoria Hospital for Children in London, were diphtheria
carriers, and that in children under two years of age the per-
centage was 21. Cobbett 5 at Cambridge reported diphtheria
bacilli in 2.9 per cent of 650 well persons, many of whom were
contacts. Some had mild sore throat.

Diphtheria Bacilli in Scarlet Fever. — While the presence
of diphtheria bacilli in scarlet fever cases sheds little light
on their prevalence in normal throats, it is of much interest

1 Ustvedt, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1906, LIV,
147.

2 Ruediger, J. Am. M. Ass., Chicago, 1906, XLVII, 1173.
» Gross, Univ. M. Mag., Phila., 1896-7, IX, 45.

4 Hewlett and Murray, Brit. M. J., Lond., 1901, I, 1474.
6 Cobbett, J. Hyg., Cambridge, 1901, I, 242.

CARRIERS AND MISSED CASES 87

from an epidemiological standpoint. The exudation which
frequently covers the tonsils in scarlet fever is often mistaken
clinically for a sign of diphtheria, and such cases are, before the
appearance of the eruption, often reported as diphtheria, so
that there is a popular impression among physicians that the
two diseases are frequently combined, and that the presence
of scarlet fever predisposes to diphtheria. While bacteriolog-
ical examinations tend to show that diphtheria infection on
scarlet fever is not so common as is often assumed, and that
many cases reported as a combination of the two diseases are
not really such, yet diphtheria bacilli are frequently found in
cases of scarlet fever, more frequently than they are found
in the general public from whom the scarlet fever cases come.
In Providence during the past four years cultures have
been taken from 92 scarlet fever cases in their homes, of
which 6, or 6.5 per cent, were positive. Of 245 cultures in the
hospital, 26, or 10.6 per cent, were positive. During the pre-
ceding three years 116 cultures were taken, but it was not
recorded whether they were from hospital or home cases, and
of these, 5, or 4.3 per cent, were positive. The excessive per-
centage of positive findings in the hospital was due to infec-
tion after admission. Both throat and nose cultures were
taken in nearly every case. In the Western Fever Hospital
of London,1 of 1019 throat swabs taken from the same num-
ber of scarlet fever patients on admission, 7.86 per cent were
positive; at another hospital 5.41 per cent of 203 cases, and
at another 6.8 per cent of 87 cases. Of the 939 cases at the
Western Hospital which did not have diphtheria bacilli on
admission, 6 later developed diphtheria, ranging from the ninth
to the fiftieth day after admission, and 28 more acquired
diphtheria bacilli without being sick. Of the 80 cases that
entered infected, 4 developed clinical diphtheria on the
fourth, fifth, seventh and thirty-first days. Soerensen2 found
that of 1 547 scarlet fever cases examined on admission, 38, or

1 Cumpston, J. Hyg., Cambridge, 1907, VII, 598.

J Soerensen, Ztschr. f. Hyg. u. Infectionskrankh., 1898, XXIX, 250.

88 THE SOURCES AND MODES OF INFECTION

2.5 per cent, carried diphtheria bacilli. During a service of two
years 32 cases of diphtheria occurred in the scarlet fever wards
among about 1500 patients, and 208 others were found to be
infected but not sick. Ruediger1 found diphtheria in 4 of 75
scarlet fever throats, or 5.3 per cent. Garratt and Wash-
bourn2 found diphtheria germs in 8, or 1.2 per cent, of 666
scarlet fever cases. In Croydon,3 37, or 17.3 per cent, of 213
cases of scarlet fever were shown to be infected with diph-
theria bacilli on admission to the hospital. In Brighton,4 in
1905, 33, or 25.9 per cent, of 166 cases of scarlet fever had
diphtheria bacilli present, and all but one were infected on
admission; in 1907, 21 of 340 were infected. In Bristol,5 of
202 scarlet fever patients in the hospital in 1904, 75 per cent
had positive cultures on admission. In 1905, of 476 cases, 21 ,
or 4.4 per cent, were positive, besides 3 others which showed
symptoms of the disease. In Philadelphia, of 700 scarlet
fever admissions, 11 per cent showed diphtheria bacilli.

Percentage of Diphtheria Carriers. — The evidence thus
far adduced tends to show that diphtheria bacilli are quite
widely distributed in the urban population of Europe and
America. Perhaps on the average 1 or 2 per cent harbor the
germs of this disease in mouth or nose; but observations also
indicate that the less diphtheria there is in a community,
and the farther removed the persons examined are from cases
of the disease, the less likely they are to be infected. Thus
in Christiania during an outbreak, 9.2 per cent of the pupils
in a certain school were infected, while some months later
the ratio was only 3.2. The Massachusetts Association of
Boards of Health report an infection of 6.03 per cent of the
school children recently exposed to diphtheria in Minneapolis,
while in schools in Providence where there had been no diph-

1 Ruediger, J. Am. M. Ass., Chicago, 1906, XLVII, 1173.

2 Garratt and Washbourn, Brit. M. J., Lond., 1899, I, 893.
1 Rep. Med. Off. Health, Croydon, 1904, 68.

4 Rep. Med. Off. Health, Brighton, 1905, 42; 1907, 22.
6 Rep. Med. Off, Health, Bristol, 1904, 100; 1905, 71.

CARRIERS AND MISSED CASES

89

theria for a long time the ratio was only about one-half of
1 per cent.

Percentage among Contacts. — There are numerous obser-
vations made on persons brought more or less closely in con-
tact with the sick, which show that under such conditions

Percentage of Carriers in Diphtheria Families, Providence, 1897-1901.

Ages.

Persons ex-
amined.

Number of
Carriers.

Percentage
of Carriers.

Under 1 year ....

119

17

14.2

1 "

112

15

13.3

2 years

97

23

23.7

3 M .....

112

25

22.3

4 "

116

31

26.7

5 "

120

17

14.1

6 "

137

42

30.6

7 "

130

30

23.1

8 "

119

25

21.0

9 "

113

23

20.3

10 "

139

26

18.7

11 "

79

11

13.9

12 "

127

28

22.0

13 "

86

15

17.4

14 "

88

13

14.7

15 "

70

5

7.1

16 "

64

9

14.0

17 " '

57

9

15.7

18 "

57

6

10.5

19 " .....

45

4

8.8

20 "

34

4

11.7

Adults

2505

277

11.0

Totals . . .

4526

655

14.4

the number of carriers may be very high. For several years
in Providence l cultures were taken from all the wage earners
in the family at the time the case was reported, and from all
the members of the family for release, that is, to determine

Rep. Supt. Health, Providence, 1901, 44.

90 THE SOURCES AND MODES OF INFECTION

the end of isolation. Only throat cultures were taken. Of
4526 such contacts examined 14.4 per cent were found to be
infected. If nose as well as throat cultures had been taken,
and if cultures had been taken from the whole family about
midway in the course of the disease, I imagine the percentage
would have been several times greater. The preceding table
shows the percentage infected at different ages. The women
in the family were infected very much oftener than the men.
It was noticed that in about 10 per cent of the cases some
well member of the family remained infected after the patient
was entirely free from germs. In Glasgow,1 of 2305 contacts
in infected families 9.2 per cent were carriers.

Percentage of Carriers in Infected Schools. — In a certain
school in Minnesota,2 of 263 well children, 129, or 49 per cent,
were infected with the granular and barred types of the bacil-
lus. In Bristol3 during a school outbreak in 1907, of 190
well children, 12.5 per cent were carriers. At the Willard Hos-
pital for the Insane during an outbreak, of 1423 well persons,
189, or 13 per cent, were carriers. At the State Hospital in
California, 11.6 of 1115 persons were infected.4 In Provi-
dence during an outbreak of diphtheria in an infant asylum,
of 175 children and adults, 116 or 66.2 per cent, proved to be
infected at one time or another, but this was only after many
repeated cultures extending over a period of some months.
Many of them were persistent carriers, one of them remaining
infected for twenty-two months. Crowley 5 in a school out-
break found 42 of 93 children infected, or 45 per cent. In
Bermondsey, London,6 the children in diphtheria families
have a single throat culture taken, and of 597 children so

Rep. Med. Off. Health, Glasgow, 1908, 71.

Wesbrook, St. Paul M. J., St. Paul, Minn., 1900, II, 219 [p. 6 of
reprint].

Rep. Med. Off. Health, Bristol, 1907, 51.
Rep. Calif. St. Bd. Health, 1906-08, 201.
Crowley, J. Roy. San. Inst., 1904-05, XXV, 807.
Rep. Med. Off. Health, London, 1904, 31.

CARRIERS AND MISSED CASES 91

examined 64, or 10.8 per cent, were found to be infected. Of
115 well children in an institution in Minneapolis, Corbet 1
found 29, or 25.2 per cent, infected. In several instances the
bacilli presented atypical forms, but many of these were viru-
lent. In a school in the same city, where there had been some
diphtheria, he found that 20, or 14.3 per cent, of 140 cultures
showed the presence of typical bacilli.2 Ustvedt,3 in Chris-
tiania, found 17 per cent of contacts infected. In a school at
Oakland, California,4 25 per cent of the pupils were carriers.
In Glasgow,5 of 322 well members of diphtheria families 10.5
per cent were infected. In the Duke of York's school, Lon-
don, of 536 well pupils 117, or 20 per cent, showed diphtheria
bacilli, and of these 10 later had diphtheria, while only one
of the " free " boys developed it. Of 13 cultures from the
well boys, 7 were virulent, though mildly so, but the bacilli
from the sick also showed a low virulence. It is worthy of
note that the disease recurred in this school for two or three
years.6 von Sholly 7 found bacilli in 20 of 202 contacts, or
about 10 per cent, and of these 14 were virulent. Buchanan8
in Glasgow found that the bacilli were virulent in 66 per
cent of 21 healthy carriers, and that bacilli recovered from
56 cases of the disease proved virulent only in 74 per cent.
Aaser9 found 19 per cent of contacts in cavalry barracks in-
fected, and 20 per cent of children in a scarlet fever ward.
Hellstrom 10 in Stockholm, of 786 soldiers who had been more

1 Corbet, Rep. Com. Health, Minneapolis, 1905, 7.

2 Rep. Com. Health, Minneapolis, 1903, 16.

3 Ustvedt, Ztschr. f . Hyg. u. Infectionskrankh., Leipz., 1906, LIV, 147.

4 Bull. Calif. St. Bd. Health, November, 1906.

6 Rep. Med. Off. Health, Glasgow, 1906, 135.

8 Arkwright, J. Hyg., Cambridge, 1908, VIII, 48 ; Rep. Med. Off.
Health, Lond., 1906, 36.

7 von Sholly, J. Infect. Dis., Chicago, 1907, IV, 337.

8 Buchanan, Brit. M. J., Lond., 1909, II, 519.

8 Aaser, Deutsche med. Wchnschr., 1895, XXI, 357.

10 Hellstrom, cited by Kober, Ztschr. f. Hyg. u. Infectionskrankh.,
Leipz., 1899, XXXI, 433, and by Fibiger, Berl. klin. Wchnschr., 1897,
XXXIV, 753,

92 THE SOURCES AND MODES OF INFECTION

or less in contact with diphtheria, found 151, or 19.2 per cent,
infected. Goadby,1 in 586 pupils in a school where there had
recently been 21 diphtheria cases, found that 190, or 34.1
per cent, were carriers, and he quotes Meade Bolton as finding
45.5 per cent of 214 contacts infected. Of the children ex-
amined by Goadby 262 had enlarged tonsils and 196 abnor-
mal throats. In another school where there had been no
reported cases, 18 of 100 were infected.

Berry and Washbourn2 report 17, or 11.9 per cent, of
infected contacts in a school of 142 girls. Lister 3 found 48
per cent of 125 contacts in Shadwell Hospital infected.
Twenty-four of them had nasal discharge. Park and Beebe4
in a foundling asylum obtained virulent bacilli* and one non-
virulent form from 5 of 55 contacts. Fibiger,5 during a school
outbreak of diphtheria, recovered the bacilli in 22, or 16.4 per
cent, of 134 children. Denny6 found 11 per cent of 200 chil-
dren infected, in a school where there had recently been 4 cases
of diphtheria, and Graham-Smith,7 10.4 per cent of 519 school
children. Gabritschewsky8 in the Russian military service
found, during the prevalence of diphtheria, that 20 per cent
of the well persons examined were carriers, and in a school
in Moscow, at a time when there was much diphtheria, 21 of
66 well children showed diphtheria bacilli.

Atypical Diphtheria. — The old notion was, and I fear that
very many physicians and not a few health officers so believe
to-day, that diphtheria is a well defined disease with charac-
teristic symptoms. They cannot imagine diphtheria without

1 Goadby, Lancet, Lond., 1900, I, 236.

2 Berry and Washbourn, Brit. M. J., Lond., 1900, I, 198.
1 Lister, Erit. M. J., Lond., 1898, II, 1338.

4 Park and Beebe, Med. Rec., N. Y., 1894, XLVI, 385.

6 Fibiger, Berl. klin. Wchnschr., 1897, XXXIV, 753.
8 Denny, Bost. M. & S. J., 1900, 515.

7 Graham-Smith, J. Hyg., Cambridge, 1903, III, 216.

8 Gabritschewsky, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1901,
XXXVI, 45.

CARRIERS AND MISSED CASES 93

serious illness. There always have been, however, some acute
clinicians who thought otherwise. Jacobi in New York was
among the foremost. He recognized that the disease often
presented a mild type and that diagnosis was impossible.
He said in 1884 11

"The symptoms are often but few. A little muscular pain
and difficult deglutition are, perhaps, all that is complained
of. Women will quietly bear it; men will go about their busi-
ness. . . . There is as much diphtheria out of bed as in bed;
nearly as much out of doors as indoors. Many a mild case
is walking the streets for weeks without caring or thinking
that some of his victims have been wept over before he was
quite well himself. . . . Diphtheria is contagious. Severe
forms may beget severe or mild forms. Mild cases may beget
mild or severe cases."

Under Jacobi's teaching I early learned to look for such
mild cases, and I remember seeing an outbreak in a children's
home during the early eighties, in which there were many
very mild sore throats and walking cases. Mild diphtheria
of this type does not usually come to the notice of a physician,
and when it does is frequently not recognized. How fre-
quent such cases are it is difficult to determine, but they must
be very numerous. In a general hospital at Chelsea2 (Lon-
don), where there had been considerable trouble from out-
breaks of diphtheria, it was determined to take cultures from
all patients admitted who showed any symptoms of sore
throat. Of 815 persons so examined 65, or 7.3 per cent, gave
positive cultures. In New York City3 the school inspectors
were directed to take cultures from all children showing slight
redness of the throat or hypertrophied tonsils." -Of 11,451
cultures 757, or 6.7 per cent, were positive. Thefc,j children
were not in any sense considered as ill, yet the percentage

1 Quoted by Solis-Cohen, J. Am. M. Ass., Chicago, 1907, XLIX, 32.

2 Parkes, Pub. Health, Lond., 1902-03, XV, 538.

3 Letter from Dr. Cronin, N. Y. Bd. Health, Sept. 23, 1904.

94 THE SOURCES AND MODES OF INFECTION

infected is certainly much higher than would be looked for
among the general school population. In Hartford 1 during
the years 1900-1903 cultures were taken from 2038 mild sore
throats seen in the schools. Of these, 591 , or 29 per cent, were
positive. These children might be said to have a decided sore
throat, though they were all well enough to be in school.
They were all excluded from school. During the same period
there were reported in Hartford 1537 cases of diphtheria, so
that the cases found in the manner described equaled one-
third of the total cases. In Indianapolis,2 at a time when
there were 60 reported cases of diphtheria in the city, a dili-
gent search for, and taking of cultures from, sore throats
among the school children revealed the presence of 46 other
cases. Every one knows how extremely common sore throat
and tonsillitis are, and even if only a very small proportion,
much less than 29 per cent, of these, are true diphtheria, the
number in the aggregate must be very large. I think I am well
within bounds when I assume that for every recognized case of
diphtheria there is at least one sore throat which is also diph-
theria though unrecognized. Judging from the proportion of
mild cases in institution outbreaks, the proportion of unrec-
ognized, mild, but yet clinical cases, must be much greater
than that. Careful medical inspection often discovers that
these missed cases have been the source of reported cases.
For example, the Report on the Health of the City of Man-
chester (England), 1906, gives a list of 29 cases caused in this
way during that year.

Rhinitis and Otitis. — Diphtheria sometimes assumes a
chronic form with few symptoms. Indeed some, as Neumann,
believe that long persistence of diphtheria bacilli in throat
and nose is always accompanied by local disturbance. That
such is often the case in the nose, and that subacute
chronic rhinitis may be a form of diphtheria, and may give
rise to typical pharyngeal cases, is generally recognized.

1 Reps. Bd. Health, Hartford, 1900 to 1903.

2 Rep, Dept. Pub, Health & Charities, Indianapolis, 1908, 8.

CARRIERS AND MISSED CASES 95

Park,1 Abbott,2 Ravenel,3 Schaps,4 Treitel and Koppel,5
DeStella,6 Ballin7 and many others report cases of chronic
rhinitis due to the presence of the diphtheria bacillus. The
middle ear sometimes suppurates in diphtheria, as it does in
scarlet fever, though less frequently, and this discharge may
remain virulent for long periods.8 I happen to have such a
case under observation at the present time.

A Real Danger. — There can be no doubt, then, that mild
and unrecognized cases of diphtheria are very common, that
convalescents long harbor the bacilli, and that contacts and
other carriers are very numerous; in fine, that these sources
of infection far outnumber the cases of recognized sickness.
But are these really sources of infection, or are they, as so
many believe, an imaginary danger evolved in the brain of
the laboratory worker ? It is well again to emphasize the
fact that long before Klebs and Loeffler identified the bacillus
of this disease a few able clinical observers like Jacobi be-
lieved that evidence pointed clearly to the great danger of
these unrecognized sources of infection. The bacteriologists
have not raised the bugaboo of carrier infection, they have
simply explained the facts which observing men have long
recognized.

Carriers cause Disease. — A little perusal of medical liter-
ature will show that many cases are reported of what is
apparently the distribution of diphtheria by carriers.

Peck 9 reports an instance where a young man who was

1 Park, Med. Rec., N. Y., 1892, XLII, 121.

2 Abbott, Med. News, Phila., 1893, LXII, 505.
Ravenel, Med. News, Phila., 1895, LXVI, 537.
Schaps, Arch. f. Kinderh., Stuttg., 1905, XL, 80.

Treitel and Koppel, Arch. f. Kinderh., Stuttg., 1895-96, XIX, 107.
DeStella, Arch, internat. de laryngol. [etc.], Par., 1903, XVI, 970.
Ballin, Jahrb. f. Kinderh., 1903, LVIII, 412.
Newsholme, Rep. Med. Off. Health, Brighton, 1906, 13; Address at
Victoria Univ., Manchester, March 9, 1904, 21.
'.Peck, Brit M. J.. Lond., 1895, I, 971.

96 THE SOURCES AND MODES OF INFECTION

perfectly well but was a carrier, probably gave diphtheria to
another young man with whom he slept.

In 1897 a mother whom I saw, stayed for a few days in a
house on a neighboring street taking care of a diphtheria
case. Soon after her return home her son was taken sick with
diphtheria, and when cultures were taken from her throat she
also was found to be infected.1 In 1905 a child who was sent
to the hospital for some surgical lesion developed scarlet
fever and was removed to the scarlet-fever ward. Three days
after his return home another child in the family was taken
sick with diphtheria, and a few days later a culture taken
from the first child was positive.2 Wesbrook3 reports: " Two
always well children carrying bacilli sent home. Diphtheria
broke out shortly after their arrival. No other source of
infection discovered. One always well girl carrying diph-
theria bacilli went to her home. Stepmother and children
developed diphtheria within a few days. This family was
practically isolated in the country."

Hellstrom,4 during a diphtheria outbreak in a cavalry regi-
ment, noted two instances in which perfectly well soldiers
who had diphtheria bacilli on their mucous membrane car-
ried the disease to others. Aaser 5 reports a child with diph-
theria germs taking the disease home from a scarlet-fever
ward where there was a diphtheria outbreak.

In 1906 at Birmingham 6 there was a milk outbreak of
diphtheria of 13 cases due to infection of the milk by the
milk handlers, who had diphtheria bacilli in their throats, but
were not sick. As soon as they were isolated the outbreak
ceased. Similar milk outbreaks caused by infected well per-

1 Rep. Supt. Health, Prov., 1897, 21.

2 Rep. Supt. Health, Prov., 1905, 32.

3 Prelim. Rep. on Diphtheria in Well Persons, J. Mass. Ass. Bds.
Health, Bost., 1901, XI, 10.

4 Hellstrom, cited by Fibiger, Berl. klin. Wchnschr., 1897, XXXIV,
753.

5 Aaser, Deutsche med. Wchnschr., 1895, XXI, 357.

6 Rep. Med. Off. Health, Birmingham, 1906, 40.

CARRIERS AND MISSED CASES 97

sons have been reported from Fitchburg,1 Lowell,2 Brookline,8
Montclair, N. J.,4 Oroville, Cal.,5 Australia6 and other places.

At the Willard State Hospital 7 a watchman, who lived
with his son-in-law in the village, who had diphtheria, was
never sick but was a carrier. He played cards with two
clerks, and they all drank from the same pitcher. The clerks,
who had not been exposed to diphtheria, were a week later
taken sick with the disease. Four other instances of infection
by carriers occurred during the same outbreak, and a similar
instance some years before.

In Lagrange Co., Ind.,8 in 1902, there were three outbreaks
in a school, apparently caused by the teacher, in whose nose
diphtheria bacilli were found, and who had shown no symp-
toms but a slight cold. In Boston9 a boy who had recovered
from diphtheria, escaped from isolation and went to Brock-
ton, and within four days three clinical cases developed in the
house where he visited. Fischer 10 reports an outbreak traced
to a restaurant in which were several carriers and mild sore
throats. Solberg11 had a boy who caused several cases of the
disease. He had been kept in the hospital for several months,
but was finally discharged with the bacilli present. Cob-
bett's12 outbreak in Cambridge was due to a boy apparently

1 Prelim. Rep. on Diphtheria in Well Persons, J. Mass. Ass. Bds.
Health, Bost., 1901, XI, 9.

2 Rep. Bd. Health, Lowell, 1904, 24.

3 Prelim. Rep. on Diphtheria in Well Persons, J. Mass. Ass. Bds.
Health, Bost., 1901, XI, 9.

4 Pediatrics, N. Y., 1901, XII, 366.

5 Rep. St. Bd. Health, Cal., 1906-08, 198.

6 Armstrong, Austral. M. Gaz., Melbourne, 1908, XXVII, 350.

7 Report of Epidemic of Diphtheria at Willard State Hospital by
Russell and Salmon [Rep. State Com. on Lunacy, XVI], 35 of Reprint.

8 Letter from Dr. H. N. Hurty, Sec. Ind. St. Bd. Health, 1907.

9 Prelim. Rep. on Diphtheria in Well Persons, J., Mass. Ass. Bds.
Health, Bost., 1901, XI, 9.

10 Fischer, Miinchen med. Wchnschr., 1906, LIU, 250, 314.

11 Solberg, quoted in Pub. Health, Lond., 1902-03, XV, 515.

12 Cobbett, J, Hyg., Cambridge, 1901, I, 229,

98 THE SOURCES AND MODES OF INFECTION

well, but who for some weeks had had a slight nasal discharge
in which diphtheria germs were present. Williams 1 notes a
case in which a teacher was infected with clinical diphtheria
by the nasal discharge of a pupil containing atypical bacilli.2
Newsholme 3 saw a case of diphtheria in February, 1901,
which had ear discharge for a short time only. Suppuration
recurred April, 1904, and diphtheria germs were present.
Three cases of the disease occurred in the family soon after.
The ear again discharged in March, 1906, with bacilli pres-
ent, and a sister apparently contracted the disease from this
source. Newsholme had a similar case in 1907. Vance*
states that a nurse had diphtheria in June, and after one
negative culture from the throat, went home and infected 3
persons. In August she returned to the hospital and infected
4 other persons, at which time a culture from throat and nose
was negative. Subsequent cultures proved positive, and
continued so until late in November. Strain 5 also reports a
nurse in whose nose diphtheria bacilli were found intermit-
tently for nine months, but they were never present in the
throat. Edsall 6 notes five outbreaks due to carriers, and
Solis-Cohen7 seven such outbreaks. Cameron8 reports that
in 70 instances cases of scarlet fever discharged from the Lon-
don hospitals carried diphtheria to their homes, and that in
many of these carriers diphtheria bacilli were shown to be
present.

Similar cases in which clinical diphtheria is apparently con-
tracted from perfectly well convalescents or carriers, or from
slight unnoticed sore throat, or from rhinitis, are found scat-

1 Williams, J. Laryngol., Lond., 1905, XX, 591.
Wesbrook's W.

Newsholme, Rep. Med. Off. Health, Brighton, 1906, 13.
Vance, Intercolon. M. J. Australas., Melbourne, 1908, XIII, 152.
Strain, Lancet, Lond., 1908, II, 1143.
Edsall, J. Am. M. Ass., Chicago, 1909, LII, 125.
Solis-Cohen, J. Am. M. Ass., Chicago, 1909, LII, 111.
Cameron, Rep. to Asylums Bd. on Return Cases of Scarlet Fever
in London, 1901-02, 41.

CARRIERS AND MISSED CASES 99

tered through medical literature, and a number of such have
been collected by the Massachusetts Association of Boards
of Health,1 .Nuttall and Graham-Smith,2 Newsholme,3 Sit-
tler,4 Niven,5 Schneider,6 myself 7 and others.

Chronic Diphtheria. — Diphtheria sometimes assumes a
chronic form, with few constitutional symptoms, and little
that is noticeable, and nothing that is characteristic in the
fauces. This condition may continue for months, and is
sometimes accompanied by considerable antitoxin in the
blood. Such cases may cause typical diphtheria.8

Glanders. — According to veterinary writers 9 the infec-
tion of glanders often remains latent, and also many mild
chronic cases occur which are the source of considerable out-
breaks. The Bureau of Animal Industry10 reports an instance
where a horse was apparently infectious for eight years with-
out manifesting any symptoms which would lead even an
expert to suspect the infection. Another instance was given
of a year-long infection of an apparently well animal.

Influenza. — The rapidity with which epidemic influenza
spreads, its sudden contemporaneous appearance at many
distant points, and the difficulty of tracing the route of infec-
tion, render it almost certain that there must in this disease
be many mild atypical cases, and many persons infected, but
showing no symptoms. That the disease is not carried by

1 Prelim. Rep. on Diphtheria Bacilli in Well Persons, J. Mass. Ass.
Bds. Health, Host., 1901, XI, 9.

2 Nuttall and Graham-Smith, The Bacteriology of Diphtheria, Cam-
bridge, 1908, 311.

3 Newsholme, Address at Victoria Univ., Manchester, Mar. 9, 1904,
and Med.-Chir. Tr., Lond., 1904, LXXXVII, 549.

Sittler, Miinchen med. Wchnschr, 1906, LIII, 863.
Niven, Rep. on the Health of Manchester, 1908, 204.
Schneider, Ztschr. f. Med.-Beamte, Berl., XX, 698.
Rep. Supt. Health, Prov., 1897, 19; 1898, 23; 1905, 32.
Neisser and Kahnert, Deutsche med. Wchnschr., 1900, XXVI,
525, and Neisser, Deutsche med. Wchnschr., 1902, XXVIII, 719.

9 Law, Text-Book of Veterinary Medicine, Ithaca, 1902, IV, 235.

10 U. S. Dept. Agric., Bu. An. Ind., Circ. No. 78, 4.

100 THE SOURCES AND MODES OF INFECTION

the air over long distances, as has been urged by many, but
is invariably transmitted by persons, has been shown by Par-
sons, Schmid and others. That these persons often show no
symptoms, and are not suspected of being carriers, also seems
certain.

There is some bacteriological evidence of this, but the bac-
teriologists have not given to the study of this disease the
attention which it deserves. Finkler found the bacillus in
the sputum a year after the beginning of an attack. Lord,1
during non-epidemic periods, in examining cases of what
appeared ordinary cough, found influenza bacilli present in
60 per cent of the cases. Later,2 in 186 non-tuberculous cases,
mostly bronchitis, he found the bacillus frequently present,
often in almost pure culture. Boggs3 has also found chronic
cases somewhat simulating tuberculosis, in one instance last-
ing for a year. Holt,4 in the Babies' Hospital in New York,
obtained 112 positive cultures out of 312 taken from 198
persons. Of 48 healthy persons 16 yielded positive cultures.
I have found scarcely any other reports of the examination
of well persons for this organism, but the frequency with
which it is found in those sick with other diseases is evidence
of the widespread distribution of the bacillus. Wollstein,5 by
swabbing the throat, found the bacillus in 16 of 37 cases
of pneumonia, in 11 of 34 of bronchitis, in 8 of 18 cases of
measles, in 8 of 16 of tuberculosis, but only 5 times in 65 cases
of other diseases. Davis 6 found influenza-like bacilli in normal
throats, and in 61 of 68 cases of whooping cough and in 13
of 23 cases of measles. Inoculated on the mucous surface of
a healthy young man, fever and catarrhal symptoms devel-
oped. Both Davis and Wollstein consider that the so-called

1 Lord, Boston M. & S. J., 1902, CXLVII, 662.

2 Lord, Boston M. & S. J., 1905, CLII, 574.

3 Boggs, Am. J. M. Sc., Phila., 1905, n. s., CXXX, 902.

4 Holt, Am. J. Obst., N. Y., 1909, LX, 343.

6 Wollstein, J. Exper. M., N. Y., 1906, VIII, 681.

6 Davis, J. Am. M, Ass., Chicago, 1907, LXVIII, 1563.

CARRIERS AND MISSED C'AgES 101

pseudo-influenza bacilli in their different forms are of the
same " species " as the typical forms. Others have found
the bacillus in other infectious diseases, as Leibscher l and
Auerbach,2 the latter of whom obtained it in 5.4 per cent of
700 cases of diphtheria, scarlet fever, etc.

The Coccus of Pneumonia. — That the pneumococcus is
present in the saliva of normal mouths was early recognized
by Pasteur, Sternberg, Welch and others, and is, of course, a
now well-established fact. Recently a number of careful
observations have been made to determine the frequency of
its occurrence, its virulence, and variations in form and habits
of growth. Buerger3 found it present in 50 per cent of 78
normal mouths, and in 34.8 per cent of 204 normal throats
not so carefully examined. He also found that it could be
recovered for weeks from convalescents. Of the cultures from
normal mouths 79 per cent proved virulent. Hiss4 found it
in 14 of 22 persons, and Duval and Lewis5 in all of 24 per-
sons about the laboratory. Longcope and Fox6 obtained it in
40 per cent to 50 per cent of mouths examined. It was more
often found and more virulent in winter. Frost, Divine and
Reineking 7 found it in 36 per cent of healthy mouths, 23 per
cent in autumn, 43 per cent in winter and 50 per cent in the
spring. The organism may persist in the same individual
for months. Park and Williams 8 report the pneumococcus
as prevalent both in city and country, as do others, but note
that germs from normal mouths are less virulent for rabbits

1 Leibscher, Prag. med. Wchnschr., 1903, XXVIII, 85.
a Auerbach, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1904,
XLVII, 259.

« Buerger, J. Exper. M., N. Y., 1905, VII, 497.
< Hiss, J. Exper. M., N. Y., 1905, VII, 547.

6 Duval and Lewis, J. Exper. M., N. Y., 1905, VII, 473.

• Longcope and Fox, J. Exper. M., N. Y., 1905, VII, 430.

7 Frost, Divine and Reineking, J. Infect. Dis., Chicago, 1905 fSuppI.
No. 1], 298.

8 Park and Williams, J, Exper. M,, N, Y., 1905, VII, 403,

102 THE SOURCES' ''AWJJ 'MODES OF INFECTION

than those from cases of pneumonia. Wells 1 found pneu-
mococci in the throat or upper respiratory passages in 45 per
cent of 135 persons, and Besser 2 found them in 14 per cent
of the noses of 57 well persons, and Hasslauer 3 in 24 of 111
normal noses. Ruediger 4 got positive results in 90 per cent
of 51 normal throats, in 91.4 per cent of 71 scarlet fever
throats, in 12 of 14 cases of measles, and in 8 of 9 cases of
tonsillitis, or pharyngitis.

Gonorrhea. — That gonorrhea assumes a latent form in
both men and women, showing no signs whatever for con-
siderable periods, and relapsing into a subacute or even acute
condition after it was supposed to be cured, has long been
known. But it is only since the discovery of the gonococcus
that the latency of this infection, as well as its persistence,
has been fully appreciated, just as it has only been the rec-
ognition of this coccus which has shown the serious patho-
logical changes in important organs which often follow
gonorrhea. There has also been demonstrated recently the
frequent innocent transmission of the disease among young
children. All the text-books and monographs dealing with
gonorrhea dwell on these facts, and refer to the finding of
the germ in cases long supposed to be well, and to its per-
sistence for long periods of time.5 A physician told me of a
case lasting from the third year of life to the twenty-eighth.
Some striking instances of the latency of gonorrhea are given
by Chapman,6 and he has shown me records of many more
in which unsuspected infection had lasted for many years.
The same author states that the gonococcus may be encysted

1 Wells, J. Am. M. Ass., Chicago, 1905, XLIV, 361.

2 Besser, Beitr. z. path. Anat. u. z. allg. Path., Jena, 1889, VI, 331.
8 Hasslauer, Centralbl. f. Bakteriol. [etc.], 1st Abt. Ref., Jena, 1905,

XXXVII, 1.

* Ruediger, J. Am. M. Ass., Chicago, 1906, XLVII, 1171

6 Wertheim, Arch. f. d. Geburtsh., Jena, 1902, XLII, 192.

6 Chapman, Fiske Fund Prize Essay, Providence, 1905, The Sequel®

of Gonorrhea, etc., 31.

CARRIERS AND MISSED CASES 103

Or dormant in tubal or ovarian tissue indefinitely. Rathbun
and Dexter 1 from a clinical and careful bacteriological study
of cases show that infection often persists long after the
patient is apparently cured, and that such uncured cases are
far more common than is generally believed. Hamilton 2
refers to the difficulty of demonstrating the coccus in mild
and chronic cases, and states that such cases are very com-
mon. Nottshaft 3 followed 120 cases of gonorrhea and ob-
tained the gonococcus from 73 per cent during the second
six months. During the fourth six months the percentage
fell to 18, and in the third year the gonococci still per-
sisted in 6 per cent of the cases. This latency is the chief
cause of innocent marital infections. It is also one reason
why the medical supervision of prostitution can never
have much effect in restricting the disease. No practicable
amount of inspection would ever guarantee freedom from
infection.

Tubercle Bacilli in Mouth and Nose. — There are two
conditions under which tubercle bacilli may exist in human
beings without appreciable symptoms.

First, they may be found on the mucous surfaces of mouth,
throat, or nose. Straus 4 examined the nose of 29 well per-
sons, orderlies, nurses and physicians about tuberculosis
wards, and recovered tubercle bacilli from 9. .Cornet found
them in his own nose.5 Jones,8 by inoculating animals with
mucus from the nose of 31 persons not brought into par-
ticularly close contact with consumptives, demonstrated the
presence of tubercle bacilli in 11. Similar results have been

1 Rathbun and Dexter, N. York M. J. [etc.], 1909, XC, 241.

2 Hamilton, J. Infect. Dis., Chicago, 1908, V, 134.

3 Nottshaft, Die Chronische Gonorrhea der Mansliche Harnrohre,
Leipzig, 1905.

4 Straus, Arch, de med. exper. et d'anat. path., Par., 1894, VI, 633.

5 Cornet, Nothnagels Encyclopedia, Phila. and Lond., 1907, Tuber-
culosis, 152.

• Jones, Med. Rec., N. Y., 1900, LVIII, 285.

104 THE SOURCES AND MODES OF INFECTION

reported by Moller l and Bernheim.2 Persons in the vicinity
of consumptives must, unless great care be taken, receive a
certain number of tubercle bacilli upon their mucous sur-
faces. Whether they increase in numbers in such a location
is not shown by the observers quoted, but unless they do it
seems rather surprising that they can be demonstrated in
such a proportion of cases. On the other hand, it does not
seem likely that conditions in the mouth and nose are favor-
able for the growth of this bacillus, and the weight of evi-
dence seems to be that the tubercle bacillus does not, to any
extent at least, develop on the normal mucous membrane
of the throat or nose of well persons.

Blair 3 reports finding tubercle bacilli in the nose of various
wild animals in captivity.

Latent Tuberculosis. — That the tubercle bacilli invade
the tissues and, creating little disturbance, remain latent for
a considerable period of time, has been amply demonstrated.
While there are not many who believe that this latency is
so universal or so persistent as does von Behring, yet it is
certainly not a rare phenomenon. L. Rabinowitsch4 in a
recent article gives a short resume* of the work of various
observers, and Harbitz 5 has also discussed it in an article
on tuberculosis of children. Harbitz, in 142 autopsies of chil-
dren who were clinically free from tuberculosis, and in whose
glands no macroscopic or microscopic signs of the disease
could be seen, nevertheless was able by inoculation to dem-
onstrate the presence of virulent tubercle bacilli. Rabino-
witsch reports four cases of virulent bacilli in lymph glands
that had undergone calcareous degeneration, thus indicating
a long-standing latency. Many similar observations by other

1 Moller, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXXII,
205.

2 Bernheim, Clinique, Brux., 1905, XIX, 346.

3 Blair, J. Comp. M. & Vet. Arch., 1903, XXIV, 278.

4 Rabinowitsch, Berl. klin. Wchnschr., 1907, XLIV, 35.
6 Harbitz, J. Infect. Dis., Chicago, 1905, II, 143.

CARRIERS AND MISSED CASES 105

writers are given by those just quoted, and Gaffky * has
recently reported finding the glands of 11 per cent of 246 well
persons as containing living bacilli, demonstrated by inocu-
lation tests. Tubercle bacilli are found in the tonsils as well
as in the lymph glands. Lartigau and Nicoll 2 found infected
adenoids in 12 of 75 healthy children, and Weichselbaum and
Bartel,3 Goodale 4 and Cornet 5 report similar findings.

Leprosy. — Sticker ° says that lepra bacilli often remain
latent in the nose for long periods.

Bacteria of Suppuration, Latency. — As has been stated on
page 20, pus bacteria are normal inhabitants of the skin and
mucous surfaces, and may even invade the glands and other
deeper tissues and remain latent for long periods of time.
They may later, as the result of traumatism, be carried by
the circulation to distant organs, there causing suppuration.7
Among those who have demonstrated the latency of infection
with pus-forming bacteria may be mentioned Manfredi and
Viola,8 Kalble,9 Perez 10 and Hess.11 Conradi 12 seems to have
employed an improved technique in his work, and has demon-
strated bacteria in 72 of 162 apparently healthy organs taken

1 Gaffky, Konferenz Internat. f. Tuberk., Wien, 1907.

2 Lartigau and Nicoll, Am. J. M. Sc., Phila., 1902, n. s., CXXIII,
1031.

8 Weichselbaum and Bartel, Wien klin. Wchnschr., 1905, XVIII, 241.
4 Goodale, Boston M. & S. J., 1906, CLV, 278.
6 Comet, Nothnagels Encyclopedia, Phila. and Lond., 1907, Tuber-
culosis, 158.

6 Sticker, Arb. a. d. k. Gesndhtsamte., Berl., 1899, XVI, Anlage I.

7 Soprano, Centralbl. f. Bakteriol. [etc.], 1st Abt. Orig., Jena, 1906,
XLI, 601.

8 Manfredi and Viola, Ztschr. f. Hyg. u. Infectionskrankh., Leipz.,
1899, XXX, 64.

9 Kalble, Miinchen med. Wchnschr., 1899, XLVI, 622.

10 Perez, Jahresb. u. d. Fortschr. . . . d. path. Mik. Baumgarten,
1897, XIII, 894.

11 Hess, Centralbl. f. Bakteriol. [etc.], 1st Abt. Orig., Jena, 1907,
XLIV, 1.

12 Conradi, Miinchen med. Wchnschr., 1908, LV, 1523.

106 THE SOURCES AND MODES OF INFECTION

from 150 healthy animals. Ford1 and Dudgeon2 also report
finding bacteria, especially pus-forming types, lying latent in
healthy organs. Bardley3 found bacteria pathogenic for
rabbits in 196 of 200 atrophied tonsils, and in 101 he demon-
strated Streptococcus pyogenes.

Latent Tetanus. — Canfora 4 and Vincent 5 have both
shown that tetanus spores may be injected into the body
and remain latent for some time. Tetanus occasionally
develops in an inexplicable manner after surgical opera-
tions and under certain other conditions, as following the
administration of hypodermic injections of the salts of
quinia. Semple 6 has recently studied this subject in a
most careful manner. He finds that washed spores of
tetanus when injected into test animals do not cause the
disease, and may remain latent and virulent for as long as
7 months. Injections of the salts of quinia will in such
animals give rise to tetanus by the injurious effects of the
solution upon the tissues. It may be that surgical opera-
tions may sometimes operate in the same way to favor the
development of latent spores. It has been suggested that
the relapses which are sometimes noted in tetanus are due
to the persistence of foci of latent infection. Semple found
tetanus bacilli in the intestines of 4 of 10 healthy human
beings, and in 3 they proved virulent for guinea pigs.

Likeness of Scarlet Fever and Diphtheria. — Unfortunately
we do not know the nature of scarlet fever virus, hence many
important matters relating to the causation of this disease

1 Ford, J. Hyg., Cambridge, 1901, I, 277.

2 Dudgeon, Lancet, Lond., 1908, II, 1651.

3 Bardley, Johns Hopkins Hosp. Bull., Bait., 1909, XX, 88.

4 Canfora, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1907,
XLX, 495.

8 Vincent, J. de physiol. et de path. g6n., Par., 1908, X, 664.
8 Semple, Sc. Memoirs, Med. and San. Dept., Gov. India, 1911,
XLIII.

CARRIERS AND MISSED CASES 107

are in doubt. There are many points of resemblance between
this disease and diphtheria, such as the degree of infective-
ness, the apparent variation in the period of incubation and
duration of infection, the very brief incubation in some cases,
the persistence of infection in some instances long after recov-
ery and the contrary fact of the early loss of infectivity in
many instances, the prominence of the initial throat symp-
toms, and the occurrence of albuminuria and middle ear
inflammations. We are justified, I think, in looking for a bac-
terial infection, and the probability is that scarlet fever is
in the main, like diphtheria, a local disease of the mucous
surfaces, chiefly of the throat and nose. We certainly see
many mild atypical cases of scarlet fever just as we do of
diphtheria, probably fully as many, and we naturally expect
to find also true carriers who exhibit no symptoms at all.

Atypical Scarlet Fever. — As the bacteriologist cannot help
us, we have to rely on clinical evidence, which is much more
uncertain and difficult to secure. The layman and the inex-
perienced physician are apt to scoff at the suggestion of scarlet
fever unless the patient has a high fever and is as red as a
lobster. But all who have seen much of this disease know
that it is exceedingly common to see cases with a scarcely
discernible indefinite rash lasting for only a few hours, a rise
in temperature of only a degree or two lasting also only a
few hours, and the merest trace of sore throat. Sometimes
the rash may be entirely absent and even the fever may
escape the most careful observation.1 In institutions and
families, such cases, considered doubtful at first, or perhaps
entirely neglected, prove to be the origin of typical symptoms
in others. Every health officer will recall many such cases.
They are the missed cases which are such a factor in the
maintenance of this disease. There are many references to
them in the reports of health officers and in medical litera-
ture. Among others who report such atypical cases are News-

1 Caziot,Bull. et mem., Soc. med.d. hop. de Par., 1903, n.s., XX, 799;
Semaine med., Par., 1903, XXIII, 205.

108 THE SOURCES AND MODES OF INFECTION

holme,1 Caziot,2 Welch and Schamberg,3 Cameron,4 Butler,5
Lesarge,6 Thresh,7 Corlett,8 Ludke,9 Kerley,10 and Thorn-
ton and Mader referred to below. In most of these missed
cases there were some slight symptoms, though overlooked
or misunderstood at the time. In Manchester, in 1906,
there were discovered 229 missed cases, mostly of a mild
character. From these 139 other cases had developed.11

Number of Atypical Cases. — It is difficult to estimate the
number of atypical cases of scarlet fever. Probably it varies
according to the extent and severity of the outbreak, and
for other reasons. Newsholme12 has reported a milk out-
break in which the number of sore throats without the pres-
ence of eruption was 215, while the number of typical cases
of scarlet fever was only 38. Butler,13 at Wellesden, studied
the incidence of sore throat in families where there was re-
ported scarlet fever, and found that 31.2 per cent of 1266
persons in such families had sore throats, while only 2.8 per
cent of 1644 persons living in families where there was no
scarlet fever had sore throat. In a school with 300 children
Thornton14 found 31 typical cases, 19 cases with no rash and
slight sore throat, and 46 cases in which only desquamation

* Newsholme, Tr. Med.-Chir. Soc., Glasg., LXXXVII, 549.

2 Caziot, Bull, et mem., Soc. meU d. h6p. de Par., 1903, n. s., XX,
799; Semaine med., Par., 1903, III, 205.

8 Welch and Schamberg, Acute Infectious Diseases, Phila., 1905,
390.

4 Cameron, Rep. on Return Cases of Scarlet Fever and Diphtheria
to Asylums Bd., Lond., 1901-02, 38, 78.

6 Butler, Proc. Roy. Soc. Med., Lond., 1908, 1, Epidemiol. Sec., 225.

• Gaz. d. h6p., Par., 1909, LXXXII, 1471.

7 Med. Officer, 1910, IV, 5.

8 J. Am. M. Ass., Chicago, 1910, LV, 195.

9 Med. Klin., Berl., 1911, VII, 127.

10 Am. J. Dis. Child., 1911, I, 71.

11 Rep. on Health of Manchester, 1906, 43.

12 Newsholme, Rep. Med. Off. Health, Brighton, 1906, 48.

13 Butler, Rep. Med. Off. Health, Wellesden, 1907, 76.

14 Thornton, Brit. M. J., Lond., 1908, 1, 495,

CARRIERS AND MISSED CASES 109

was noticed. Mader l observed 9 typical and 12 very atypi-
cal cases in an outbreak in a home for epileptics. Ewart,2
from an examination of 8000 school children at Middle-
borough, estimates that there must be at least 3 missed cases
annually for each 1000 children. True " carriers," that is,
perfectly well persons, are sometimes reported. Thus I
have noted an instance where a woman apparently in this
way carried scarlet fever to her child. She had been taking
care of another child, and after an entire change of clothing,
bath and shampoo, visited the first-named child, who was
taken sick two days later. Newsholme reports what he
thinks are possibly, or even probably, similar cases. New-
man,3 of Finsbury (London), noted 5 carriers among school
children, three of whom, though they had never had the
disease, transmitted it to others. Kerr 4 also reports 5 such
cases. When a diphtheria patient discharged from a hospi-
tal carries scarlet fever home to his family, he must in most
instances be considered a true carrier, for it is unlikely that
symptoms of scarlet fever would often escape notice in the
hospital. Ten of this sort of return carriers are reported by
Simpson,5 30 by Cameron 6 and 44 by Turner.7

Smallpox. — There is still doubt as to the specific cause
of smallpox. The claims that have been advanced in favor
of this or that micro-organism have not as yet been substan-
tiated. Hence there is no " laboratory " evidence that car-
riers of this disease exist, nor is there much clinical evidence

1 Mader, Cor.-Bl. f. schweiz. Aerzte, 1908, XXXVIII, 169.

2 Ewart, Pub. Health, Lond., 1910-11, XXIV, 275.

3 Newman, Rep. Med. Off. Health, Lond., 1904, 27.

4 Kerr, Rep. Med. Off. Education, Lond., 1907, 43.

5 Simpson, Rep. on Return Cases of Scarlet Fever and Diphtheria
to Asylums Bd., Lond., 1898-99, 8.

6 Cameron, Report on Return Cases of Scarlet Fever and Diphtheria
to Asylums Bd., Lond., 1901-02, 43.

7 Turner, Rep. on Return Cases of Scarlet Fever and Diphtheria to
Asylums Bd., Lond., 1902-04, 3.

110 THE SOURCES AND MODES OF INFECTION

that perfectly well persons transmit the disease. There are,
it is true, scattered through medical literature numerous
reports of instances of such transmission. These reports, as
often in other diseases, are, as regards the evidence, rather
unsatisfactory. There is usually a possibility of such trans-
mission, no actual demonstration. Nevertheless it is quite
possible that there may be true carriers of smallpox, and
there may be a considerable number of them. Mild cases
are often very numerous. This was well illustrated in the
recent epidemic in the United States and England. In the
United States in the year ending June 30, 1901, the fatality
in 38,506 cases was only 1.79 per cent. At such a time great
numbers of cases escape recognition. The patients often
have no idea that they are sick with a dangerous disease.
They may be at their work even during the prodromal stage.
There may be only half a dozen, or even fewer, atypical
pustules. I have in rather a limited experience seen several
such cases. They are also reported by Welch and Scham-
berg.1 These authors also state that cases occur where there
is no eruption at all. Davies 2 reports a nurse who was
exposed to smallpox early in February and who had some
fever and headache without any eruption or other symp-
toms. Twelve days after she returned to her ward work,
which was on February 28th, one of her patients in the
ward developed frank smallpox. Davies from epidemio-
logical evidence thinks that chronic carriers do not occur
in this disease. The step from half a dozen points of
eruption to none at all is so slight, and the extremely mild
cases at times are so numerous, that a considerable number
sine eruptione may reasonably be expected. Armstrong 3 re-
cently reports three such cases with some fever and subjec-
tive symptoms but no eruption. They occurred in infected
families and subsequently proved refractory to vaccination.

1 Welch and Schamberg, Acute Infectious Diseases, Phila., 1905, 207.

2 Davies, Rep. Med. Off. Health, Bristol, 1909, 24.
» Armstrong, Arch. f. Diagnosis, 1909, II, 126.

CARRIERS AND MISSED CASES 111

Measles. — There is little clinical evidence that "carriers"
of measles are common. Most health officers consider that
measles is rarely carried by a " third person." It usually
has a quite definite clinical picture. In Aberdeen, so Dr.
Matthew Hay writes to me, judging from a census taken in
certain schools, it was estimated that from 90 to 93 per cent
of the children over ten years of age had had a recognized
attack of measles. That such a high percentage of children
had recognized attacks indicates that atypical cases cannot
be very common. Mild, atypical and unrecognized cases of
this disease must be far less numerous than are such cases
in scarlet fever, diphtheria and typhoid fever. Levy at
Richmond l investigated 2331 cases during an outbreak.
They were all, without exception, traced to previous frank
cases of the disease. Levy could find no evidence of in-
fection from either carriers or fomites. His careful work
in the study of this outbreak renders his conclusions of the
highest value.

Protozoan Diseases. — Although the marks of distinction
between animal and vegetable life as seen among the lower
forms are ill defined and uncertain, yet it is generally agreed
that though the group of organisms known as bacteria have
characters belonging to both the vegetable and the animal
kingdom, they are more nearly allied to the former, while an-
other group, known as protozoa, are allied to the lower forms
of animal life. Although one of the latter class was discovered
to be the cause of malaria in 1880, at a time when most of the
disease-producing bacteria were unknown, the bacteria have
received far more study, and indeed it is only recently that
the protozoa as the cause of disease have received much at-
tention. It is known that a number of diseases both of men
and of animals are caused by protozoa of various types.
As has recently been emphasized by Daniels,2 latency is a

1 Richmond Rep. Health Dept,, 1910, 38.

2 Daniels, Brit. M. J., Lond., 1909, II, 767.

112 THE SOURCES AND MODES OF INFECTION

common phenomenon of protozoan infection, and is of the
utmost importance from an epidemiological standpoint.
Not only do the blood parasites maintain long continued in-
fection with few symptoms, but such forms as Ameba, Balan-
tidium and Lamblia live in the intestines indefinitely, repro-
ducing themselves asexually. A number of the diseases to
be referred to do not affect man, but they serve to illustrate
the prevalence of latency in protozoan infections.

Texas Cattle Fever. — It has long been recognized that
the Texas fever of cattle could be transmitted by apparently
healthy animals. The explanation of this fact was not, how-
ever, forthcoming until Smith and Kilborne's 1 classical re-
searches in 1893 demonstrated that the disease was due to a
blood parasite, a protozoan (Piroplasma bigeminum), not a
bacterium, and also demonstrated that it is only transmitted
from animal to animal by means of a species of tick (Boophi-
lus annulatus) , in which the parasite passes through a cycle
of changes necessary for the maintenance of the species. This
work of Smith and Kilborne's was one of the most important
steps in the development of our knowledge of the insect
carriage of disease, a knowledge which has been of such inesti-
mable value in connection with malaria, sleeping sickness,
yellow fever and other blood diseases. It was soon deter-
mined that animals which had recovered from Texas fever
and were immune to it, carried the piroplasma in the blood
for an indefinite time. In one instance it was known to
have remained for thirteen years.2 Reports from the Philip-
pines 3 show that many of the cattle in those islands are in-
fected with the parasites of Texas fever though showing no
symptoms. Slightly different types of this disease occur
among cattle in Europe and in Africa caused by different
species of piroplasma and spread by carriers.

1 Smith and Kilborne, U. S. Dept. Agric., Bu. An. Ind., Bull. No. 1,
1893, 57.

* U. S. Dept. Agric., Rep. Bu. An. Ind., 1904, XXI, 26.

* Bull. 14, Bu. Gov. Lab., Manila, P. I., 1904, 11.

CARRIERS AND MISSED CASES 113

Nagana. — Nagana 1 is an African cattle disease which,
like the sleeping sickness, is caused by a trypanosome, T.
brucei, and is also transmitted by a tsetse fly, G. morsitans.
Wild herbivora are very generally infected, but because of
immunity, probably acquired, they show few or no symp-
toms. These carriers are the real source of the disease.

Dourine. — This is a contagious disease of horses spread
almost exclusively by the sexual act. It is caused by a
trypanosome, T. equiperdum, found in the secretions,2
blood and tissues. Although recovery may occur, the
trypanosomes remain for months in the sexual organs of
apparently cured animals and thus spread the disease.

Malaria. — The most important of the protozoan diseases
is malaria. It had always been believed that this disease
might remain latent for months and years, but what latency
really meant could not be determined until after the dis-
covery of the specific cause of the disease. Many of the
protozoa pass through various metamorphoses, or fixed cycles
of development, and it was found that in malaria the Plas-
modium, which is its cause, may in certain stages persist in
the blood or organs of the body without causing any charac-
teristic symptoms, or indeed any symptoms at all. Then at
any time, from one cause or another, its reproduction may
again become active and more or less marked symptoms
appear. Thus relapses may occur after a period of several
years, when the bodily resistance is from any cause impaired.
Thus fever after surgical operation is not rarely due to a
latent malaria becoming active, the parasite being found in
the blood and the symptoms yielding to quinia. As malaria
is a strictly transmissible disease, the plasmodium which
causes it being borne from one person to another by mosqui-
toes, a latent case of the kind described may be, and doubt-
less often is, the means of introducing the disease into hitherto

1 Minchin, Gray and Tullock, Proc. Roy. Soc., Lond., 1906; Nature,
Lond., 1906, LXXVII, 57.

1 U. S. Dept. Ag., Bu. An. Ind., Bull. No. 142, 1911.

114 THE SOURCES AND MODES OF INFECTION

uninfected localities. Unless such cases drift into hospitals,
they are almost certain to be unrecognized. Craig1 made a
careful study of 424 such latent cases found among 1653 sol-
diers examined in the Philippines. Since I first wrote this
chapter Craig's book has appeared in which he discusses in
much detail latency and recurrences.2 He has noted an asex-
ual conjugation of the parasites in the blood cells, which is
followed by a resting stage, and which he believes has some
relation to the latency of the infection. Where malaria pre-
vails extensively, as in the tropics, it has long been noticed
that a large part of the adult population is immune. It is
now known that this immunity is to a large extent acquired,
and is due to the invasion of the body in infancy by the
parasites. This invasion, while sometimes causing symptoms
and death, frequently gives rise to few or no symptoms, or
if some reaction appears at first, it soon disappears, and the
children may seem perfectly well though the parasites are
constantly found in the blood. They disappear year by year
and infection is rarely found after adolescence. Koch3 in
Africa found large numbers of children infected, even as high
as 100 per cent. Plehn4 found many adults infected though
not sick. Christophers and Stephens 5 found the parasites in
the blood of 90 per cent of infants examined in one locality
on the Gold Coast in Africa, and the Thompson Yates ex-
pedition to Nigeria6 reports finding them in 63 per cent of
children under 3 years of age. Ziemann7 found that in the
Cameroon country 37 per cent of children under 5 years of

1 Craig, J. Infect. Dis., Chicago, 1907, IV, 108.

2 The Malarial Fevers, N. Y., 1909, 228.

8 Koch, cited in Thompson Yates' Lab. Rep., 1900, No. 4.
4 Plehn, cited by Marchiafava and Bignani, Twentieth Cent. Prac-
tice, XXI, 807.

6 Christophers and Stephens, Reports of the Malarial Commission of
the Roy. Soc. (Eng.), 2nd Ser., 1900-03, 15.

0 'Thompson Yates' Lab. Rep., 1900, III, Pt. 2, 201.

7 Deutsche med. Wchnschr., 1900, XXVI, 399, 642, 753, 769.

CARRIERS AND MISSED CASES 115

age were infected. In Panama Kendall1 found 57 per cent
infected of the natives of all ages examined in the village
of Bahio, and 73 per cent of foreigners. While many of
the latter were more or less sick, many were entirely well.
Darling2 in villages in the Panama Canal zone where there
were no Anopheles, nevertheless found that 10 per cent of
the laborers at work were infected though they were not at
all sick. Among the families of the Spanish and the West
Indians, the latent infection reached 30 per cent. It is this
latent infection in the blood of the native population which
is the cause of the malaria which so certainly attacks arrivals
from non-malarial regions. The greater the distance that can
be placed between the natives and the strangers the less
the danger of the latter contracting the disease.

Sleeping Sickness. — African sleeping sickness has been
shown to be due to a protozoan, Trypanosoma gambiense.
This disease has been much studied of late, and it seems cer-
tain that it is transmitted by means of the tsetse fly (Glossina
palpalis) , though whether it is a purely accidental mechanical
transference on the proboscis of the fly, or whether it passes
through a part of its life history in the body of the fly, as
the Plasmodium of malaria does in the mosquito, is still uncer-
tain. In any event the parasite is frequently found in the
blood of apparently healthy subjects, just as is the malarial
parasite. According to Todd 3 it may remain in the blood
for 15 years, causing no symptoms, and frequently remains
for many months. The expedition from the Liverpool School
of Tropical Medicine4 found many natives infected, but who
exhibited no symptoms, or only slight symptoms. The greater
the prevalence of the disease the more common are these
latent cases. In Gambia, where the disease is rare, not more

1 Kendall, J. Am. M. Ass., Chicago, 1906, XLVI, 1151, 1266.

2 Darling, J. Am. M. Ass., Chicago, 1909, LIII, 2051.
8 Todd, Tr. Epidemiol. Soc., Lond., 1905-06, XXV, 1.

4 Liverpool School Trop. Med. Memoirs, 1904, XIII; Med. News,
N. Y., 1904, LXXXV, 526, 615,

116 THE SOURCES AND MODES OF INFECTION

than one native in 1000 examined showed the parasites,
while in the Congo 46 in 100 were infected, and in Uganda
the percentage was still higher. Whether the disease is
always transferred from man to man, or whether some of
the lower animals also harbor the parasites and thus serve
as a " reservoir " from which the human disease is derived,
is as yet uncertain.

Syphilis. — The spirochete of syphilis has in one instance
been reported as remaining latent in a healed lesion of that
disease.1 According to Bosquenet 2 the spirochetes are
commonly found in gumma, where they may apparently
remain latent for a long time. The fact that gumma have
not been considered infectious has been urged as an~argu-
ment against the pathogenicity of the spirochetes. It is
now, however, generally thought that the spirochetes in
these tumors are infective.

Relapsing Fever. — There are several types of relapsing
fever having more or less well marked geographical limita-
tions. The disease is characterized by well denned febrile
" relapses " with equally well denned afebrile intervals.
The different forms of the disease are caused by slightly dif-
fering species of spirochetes, and, as will be referred to in the
last chapter, these parasites are undoubtedly transmitted
by insect carriers. According to Craig,3 it has been demon-
strated by Breinl and Kinghorn and by Button and Todd
that though Spirocheta duttoni is not found in the blood
during the afebrile period, the blood nevertheless is infec-
tious. Mackie has shown the same for S. carteri, and
Darling 4 has shown that the blood from the afebrile stage
and from entirely recovered animals is still infectious.
Darling's work was done in Panama.

1 Pasini, cited by Rosenberger, New York M. J. [etc.J, 1908,
LXXXVII, 394.

2 Bosquenet, Spirochsetes, Phila., 1911, 51.

1 Craig, The Malarial Fevers, New York, 1909, 447.
« Darling, Arch. Int. Med., Chicago, 1909, 150.

CARRIERS AND MISSED CASES 117

Vincent's Angina. — It seems highly probable that this
disease is caused by the spirochetes which are found con-
stantly in the lesions. That many mild missed cases, and
perhaps true carriers occur is probable. Farley,1 in report-
ing an institutional outbreak of this disease, says that a
number of children were found without any constitutional
symptoms, but with a slight exudation on the tonsils or with
spongy gums, and that in these the spirochetes were found
to be present. These cases were not discovered until a
systematic examination was made of all the children and
the outbreak ceased on their removal from the institution.
Kerr 2 reports an outbreak in the Linden Lodge School in
London, consisting of about 20 cases, and due apparently
to unnoticed chronic and subacute cases which attracted
no attention until they were sought for.

Yellow Fever. — The parasite which is the cause of yellow
fever is still unknown, although fortunately for preventive
medicine we have very accurate knowledge of the manner in
which the disease is transmitted. As in scarlet fever, so in
yellow fever lack of knowledge of the parasite renders diffi-
cult the recognition of carrier cases if they exist. But there
is abundant clinical evidence that many very mild and atyp-
ical cases occur which it is impossible to recognize. It is in
young children chiefly that this slight disturbance is pro-
duced by the infection. A similar phenomenon is noted in
malarial disease, and young children are the chief source of
infection in both yellow fever and malaria. The fact of the
mildness of these cases, their frequency, and the impossibility
of making a diagnosis, has been insisted upon by Finlay,
Gorgas, Guiteras, Carter, Agramonte, Marchoux and others.
Even in adults, walking cases, which it is impossible to dis-
cover by an ordinary examination, are not rare. Thus it was
claimed that during an outbreak in Louisiana a single walk-
ing case carried the disease to three different communities.

1 Farley, J. Am. M. Ass., Chicago, 1910, LIV, 1516.

2 Rep. Med. Off. of Education, Lond., 1909, 63.

118 THE SOURCES AND MODES OF INFECTION

Amebic Dysentery. — A number of observers have re-
ported finding Entameba histolytica, which is the cause of
this disease, in the intestines of healthy persons. It is
now believed that most of these reports are based on error
due to confusing the pathogenic species named above with
harmless saprophytes. This is the view of Vedder J and
Craig.2 Although the laboratory evidence of the existence
of carriers of dysentery amebae is uncertain or lacking, there
is epidemiological evidence that these parasites are some-
times found in persons for a long time after recovery, and
may also be found in the feces before thq disease develops.
Martini 3 reports a case in which the sickness lasted from
the 15th of September to the 1st of December, 1907, but in
which the amebae persisted until the last of January, 1908.
Vincent 4 reports several instances in which persons were
carriers for five months after their return to France from
Tonkin. Lemoine 5 had under observation a man who con-
tracted the disease in China in 1897 and returned to France
and transmitted the disease to another in 1908. Cameron 6
reports a soldier returned to Scotland from the Boer War,
and apparently well for 6 years, but who then developed a
liver abscess in which amebae were found.

Poliomyelitis. — The disease known as acute anterior
poliomyelitis, or infantile paralysis, has been, during the
past fifteen years, occurring with increasing frequency.
It usually appears in well-defined local outbreaks lasting
from a few weeks to a few months. Except in .the largest
cities the outbreaks are not often of long duration. " Spo-
radic " or somewhat isolated cases also doubtless occur,

Vedder, J. Am. M. Ass., Chicago, 1906, XLVI, 870.

Craig, J. Infect. Dis., Chicago, 1908, V, 324; The Parasitic
Amoebae of Man, Phila., 1911.

Martini, Arch. f. Schiffs- u. Tropen-Hyg., Cassel, 1908, XII, 588.

Vincent, Bull. Soc. path, exot., Par., 1909, II, 78.

Lemoine, Bull, et m6m. Soc. med. d. h6p. de Par., 1908, 3d Ser.,
XXV, 640.

6 Cameron, Brit. M. J., Lond., 1911, I, 973.

CARRIERS AND MISSED CASES 119

though it is not unlikely that some which are reported under
this name may be due to a totally different cause from that
which gives rise to the " epidemic " form, just as cases of
cerebro-spinal meningitis, closely resembling the disease
caused by the Weichselbaum diplococcus, may be produced
by several other infective agents. For many years epi-
demiologists have considered poliomyelitis as an infectious
and contagious disease, though many have thought the
evidence of communicability not to be entirely satisfac-
tory. In fact, the epidemiological evidence of the conta-
giousness of this disease, and of cerebro-spinal meningitis,
is almost on a par. The contagiousness of both diseases
is certainly not very marked ; that is, the chance of a frank
case giving rise to another frank case, either in homes or in
institutions, is very small. When an apparently contagious
disease, yet one only slightly so, appears in well-marked
outbreaks, it is pretty safe to assume the existence of large
numbers of carriers or of mild atypical and unrecognized
cases. The epidemiology of cerebro-spinal meningitis was
inexplicable until the carriers of that disease were discov-
ered. The explanation of its spread now involves no more
difficulties than does that of scarlet fever or diphtheria.
The epidemiological, and indeed clinical resemblance of
cerebro-spinal meningitis and poliomyelitis led some, even
before the work of Flexner and Clark, and Lucas and
Osgood, to suspect the existence of carriers. The exist-
ence of numerous " abortive " cases, as they have been
called, had long been considered probable.

Laboratory Studies. — To determine definitely the ex-
istence of either carriers or atypical cases of any infectious
disease by means of epidemiological studies is extremely
difficult, and to determine their number is entirely impos-
sible. Fortunately experimental laboratory work, culmi-
nating in the investigations of Flexner and Lewis,1 have

1 Flexner and Lewis, J. Am. Med. Ass., Chicago, 1909, LIII, 1639,
1913.

120 THE SOURCES AND MODES OF INFECTION

shown, by repeated transfers of the virus through monkeys,
that poliomyelitis is in all probability due to a living con-
tagium. Subsequent studies have shown that the virus is
filterable; that it is found in various fluids and organs of
monkeys including the nasal and pharyngeal mucous mem-
brane, and presumably also in the mucus ; and that animals
may be infected not only by injection into the central
nervous system, but also by injection into the subcuta-
neous tissue, and by application of the virus to the scari-
fied and healthy mucous surfaces. The spinal fluid very
early shows quite characteristic changes, and " immunity
principles " (capable of neutralizing the virus in the test
tube) are found in the blood of animals, and human beings
who have had the disease.1

Poliomyelitis Carriers. — By means of the experimental
methods rendered possible by this work, it has been shown
that in monkeys, at least, the virus of this disease remains
in the nose and throat after the clinical symptoms have
subsided. Thus Osgood and Lucas 2 found the virus in
the naso-pharyngeal mucous membrane of two monkeys
6 weeks and 5J months respectively after the acute symp-
toms of the disease had disappeared, and Flexner and
Clark found the virulence to persist 4 weeks in a monkey
studied by them. Thus the existence of chronic conva-
lescent carriers in inoculated monkeys is definitely proved.
As the virus has been shown by Flexner and Clark 3 to
exist in the tonsillar tissue of human beings during the dis-
ease, just as it is found in the mucous membrane of monkeys,
it seems reasonable to expect that it only needs further
investigation to demonstrate the existence of human car-
riers of this disease. Occasionally observers have, for epi-
demiological reasons, suspected certain persons of being

1 Flexner and Lewis, J. Am. Med. Ass., Chicago, 1910, LIV,
1780.

2 Osgood and Lucas, J. Am. Med. Ass., Chicago, 1911, LV, 495.
» Flexner and Clark, J. Am. Med. Ass., Chicago, 1911, 1685.

CARRIERS AND MISSED CASES 121

carriers. Thus Lovett 1 states that he has a number of
such cases in his records, and in 11 instances the disease
followed intimate contact with persons who had previously
suffered from it. Such cases have also been reported by
Krause.2 Anderson 3 in Nebraska notes a number of in-
stances in which the disease seems to have been carried by
well persons. Thus a girl who had visited in Stromberg,
where the disease prevailed, returned to a farm many
miles distant and a few days later a case of poliomyelitis
developed, followed by others on that and a near-by farm.
In another instance a thresher apparently carried it from
his own family to the place where he worked. In two other
instances peddlers seemed to be carriers. Similar observa-
tions have, in England, been made by Reece and Farrar.4

Poliomyelitis : Atypical Cases. — Ever since Wickman's
studies in Sweden it has been believed that so-called " abor-
tive " cases of this disease are quite common. These cases
may show some slight symptoms referable to the nervous
system, such as headache, pain, tenderness and rigidity of
neck, hypersthesia and neuralgic pains; and some think
that tonsillitis and gastro-intestinal irritation without ner-
vous symptoms may be due to the same pathogenic agent.
According to Frost,5 Wickman found such abortive cases
to equal in number 15 per cent of the frank cases. In
Massachusetts in 1909 there were 49 possibly abortive
cases observed in connection with 150 acknowledged cases.
Frost himself in an outbreak in Hancock County, la.,
found 25 abortive cases and 5 frank cases. That some at
least, and perhaps a large proportion, of these abortive
cases are really poliomyelitis, is shown by the finding of

1 Lovett, Rep. State Bd. of Health, Mass., 1909.

2 Krause, Therapie der Gegenwart, 1911, LII, 145.

3 Anderson, Pediatrics, N. Y. & Lond., 1910, XXII, 543.

4 Reece and Farrar, Rep. to Local Gov. Bd. Lond., 1912, No. 61.

6 Frost, U. S. Pub. Health and Mar. Hosp. Serv., Pub. Health Bull.
No. 44, 1911.

122 THE SOURCES AND MODES OF INFECTION

11 immunizing principles " in their blood. Anderson and
Frost 1 state that Netter and Levaditi demonstrated a
case in this way, and the authors themselves showed the
presence of " immunizing principles " in the blood of 6 of
9 suspected abortive cases.

In view of the epidemiological facts, and the experi-
mental data, it seems not unsafe to surmise that, with the
perfection and extended application of laboratory methods
of diagnosis, carriers and mild atypical cases of poliomye-
litis will be shown to be common, and it is not unlikely that
they may prove, as in cerebro-spinal meningitis, many times
more numerous than are frank cases of the disease.

Latency a Common Phenomenon. — The laity and not a
few physicians are still incredulous that there can be diph-
theria infection unless the patient is sick in bed and the
throat choked with exudation, or that there can be scarlet
fever without high fever and an extensive eruption. Still
less are they willing to admit that perfectly well persons
can carry and reproduce in their bodies the virus of the
infectious diseases. It is difficult for many to realize that
the virus of disease may remain latent in the body, for
long periods, without causing symptoms. The facts pre-
sented in the preceding pages have been gathered to show
that latency is a common phenomenon, and for this pur-
pose, besides many common human infections, a number of
diseases have been referred to, not transmissible to human
beings and of little interest to health officials, but which
seem to illustrate this phenomenon. Infection without
symptoms is no cause for surprise, but may be expected in
any parasitic disease. There may be exceptions, as is said
to be the case in relapsing fever and African East Coast
cattle fever (caused by Th. parva). But these exceptions,
rather than the existence of latency and atypical types,
should be cause for surprise.

1 Frost, J. Am. Med. Ass., Chicago, 1911, LVI, 663.

CARRIERS AND MISSED CASES 123

Laboratory Evidence of Carriers. — Definite knowledge
as to the existence and number of carriers must rest almost
entirely upon evidence furnished by the finding in the body
of the causative agents of disease, bacteria or protozoa, or
the demonstration in the blood or tissues of an unknown
infective principle, as in the case of poliomyelitis. Actu-
ally, almost all such evidence depends upon the finding of
disease-producing bacteria or protozoa in the bodies of
healthy persons or animals. Those persons who do not
believe that the causative relation between these minute
forms of life and disease has been established with a rea-
sonable degree of certainty will of course attach little weight
to the bacteriological evidence of the existence of carriers.
With such I cannot agree. On the contrary, the causation
of certain diseases by bacteria seems to me to be one of the
best-established truths of medical science, and it is conse-
quently entirely legitimate to make use of bacteriological
evidence in developing the theory and practice of preven-
tive medicine. Many, while admitting that we have satis-
factory evidence that some diseases are caused by specific
bacteria or protozoa, are unwilling to admit that the evi-
dence is conclusive as to the causative relation of certain
other alleged pathogens and the diseases said to be due to
them. Indeed, almost everyone takes this attitude towards
some disease. It happens, however, that in the opinion of
the majority of the most competent observers the evidence
is especially strong as regards the causative agent of chol-
era, typhoid fever, diphtheria, cerebro-spinal meningitis,
sleeping sickness and malaria. And it is particularly for
these diseases that we have abundant evidence as regards
the numbers and importance of carriers.

Laboratory Evidence of Atypical Cases. — While acute
clinical observers have at all times recognized many atypical
cases which would have been overlooked by the average
man, it was from the nature of things impossible to know
how far or how often a disease could deviate from the nor-

124 THE SOURCES AND MODES OF INFECTION

mal when the only criterion for the determination of the
disease was the symptom complex of the normal type. It
is only the discovery of the germ, and the ability to recog-
nize it, which enables us to take a wider view of an infec-
tious disease, and to see that it includes far more types and
cases than was at first suspected. It is only the work of
the laboratory that makes it possible to recognize the mild
atypical cases of diphtheria, typhoid fever and malaria.
The clinicians have been willing to accept, though, it must
be confessed, with a little hesitancy, the teachings of the
laboratory men in regard to prevalence of atypical cases.
Some -are even emboldened to proceed along lines where
the laboratory men cannot as yet go, and to recognize
atypical scarlet fever which they would formerly have
passed by, and to postulate the " abortive " type of polio-
myelitis. Yet when the laboratory man points out the
true carrier and suggests danger from him as well as from
the atypical case, many a clinician, and not a few epidemi-
ologists as well, hesitate to follow.

Carriers Vary in Numbers. — One of the important facts
noticeable in the study of this subject is the great varia-
tion in the number of carriers in different diseases. It
ought to be determined by a sufficient number of observa-
tions just how many carriers there are in different diseases
among the general public and among contacts at different
ages, at different seasons, and in different places. Un-
fortunately the observations have not as yet been exten-
sive enough to warrant any definite conclusions, though
some quite striking differences are apparent. Probably the
number of carriers is on the whole greater in pneumonia
than in any other disease, including at times half the total
population. In certain protozoan diseases the percentage
of carriers may be even higher within limited groups, as
among children in intensely malarious districts. So, too,
carriers of the trypanosomes of sleeping sickness are at
times very numerous. In the case of animal diseases, as

CARRIERS AND MISSED CASES 125

Texas cattle fever, the proportion of infected animals is
very large. Among human diseases it would appear that
influenza carriers are very numerous, though the identity
of the causative agent has not been so well determined, and
the number of observations is not as large as for many other
diseases. It is certain that cerebro-spinal meningitis car-
riers are many times more numerous than the cases — in the
opinion of many, from ten times to twenty times as numer-
ous. Under ordinary conditions, with a moderate preva-
lence of the disease, diphtheria carriers equal about one per
cent of the population, and during outbreaks the percent-
age, particularly among children, is many times as great.
Probably typhoid carriers are not so numerous as are
diphtheria carriers, but there are not many data as to the
numbers among either the general public or among con-
tacts. In neither human nor rat plague do carriers appear
to be either numerous or important. There is even less
evidence of the existence of carriers, certainly of chronic
carriers, of smallpox; and for measles the clinical evidence
that there are no carriers is very strong.

Relation between Number of Carriers and Infectivity. —
Here again data are too few to warrant more than a sugges-
tion, but one cannot but be struck by the fact that the most
infectious diseases show the fewest carriers, while some of
the diseases which are only slightly contagious show very
many more carriers than cases. Thus smallpox and mea-
sles, considered perhaps the most contagious of diseases, are
not certainly known to give rise to the carrier condition.
Diphtheria is not so infectious as measles; that is, a case,
brought in contact with susceptibles, does not so surely
give rise to other cases ; and carriers of diphtheria are quite
numerous. Cerebro-spinal meningitis exhibits many car-
riers, and its ineffectivity, as measured by contagion in the
family and institution, is small. Pneumonia has the most
carriers of all and is the least contagious.

126 THE SOURCES AND MODES OF INFECTION

Virulence of Germs in Carriers. — The question of the
infecting power of carriers is the crucial one of the whole
subject. One way of testing this is by testing by animal
experiment the virulence of the germs which the carrier is
producing. Unfortunately, owing to insusceptibility of the
lower animals, or the difficulties of the technique, or the
cost of susceptible animals, like the apes, renders numerous
observations out of the question. In one disease, diphthe-
ria, in which experimentation is comparatively easy, it has
been shown that in a varying percentage of carriers the
bacilli are virulent, often exceedingly virulent, though there
is a large number of carriers in which they are not virulent.
The actual number of carriers of virulent germs is never-
theless shown to be very large. The pneumococcus found
in well persons is virulent for rabbits, though less so than
are cocci from cases of pneumonia. Tests on the lower
animals for the virulence of typhoid bacilli are of little
value, but it is worthy of note that a case of typhoid fever
in a human being has recently resulted from drawing into
the mouth a culture of a typhoid bacillus derived from a
carrier.

In some of the diseases of the lower animals direct experi-
ment has demonstrated the virulence of disease germs from
healthy individuals. This has been abundantly proved in
Texas cattle fever and in certain trypanosome infections.

Carriers as a Source of Protozoan Disease. — There
seems to be little disposition in any quarter to question the
importance of carriers of the malarial plasmodium in the
extension of malarial disease. Yet so far as I know there
are no experiments which demonstrate such transmission,
and few if any clinical observations which would indicate
such transmission. It is simply assumed that the numer-
ous well persons carrying plasmodium must be a source
from which oftentimes the mosquitoes get the infection
which they transmit to others. Probably the reason why
this view is so readily accepted without any demand for

CARRIERS AND MISSED CASES 127

rigid experimental or epidemiological proof is that the mode
of transmission of this disease is established on such con-
vincing evidence. It seems certain that it is only trans-
mitted by mosquitoes, and that mosquitoes obtain the
parasites only by previously biting infected human beings.
So, too, because the mode of transmission of sleeping sick-
ness is definitely determined, no one looks for the origin of
this disease outside the bodies of living beings, and as many
human beings are shown to be " carriers " of the trypano-
some, such persons are believed to be a principal source of
the disease.

Carriers as a Cause of Bacterial Diseases. — In at least
one bacterial disease, namely " white diarrhea of chicks,"
due to B. pullorum, it has been definitely shown that adult
carrier hens infect their eggs and are thus the principal
factor in the spread of the disease. In no other animal
disease due to bacteria does the carrier question appear to
have been so well worked out.

While the existence of numerous carriers has been amply
demonstrated in such diseases as cholera, typhoid fever,
dysentery, diphtheria and cerebro-spinal meningitis, their
importance in the spread of these and similar diseases has
been questioned by many who never question the impor-
tance of carriers in the spread of malarial disease. The
reason for this, doubtless, is that the mode of transmission
of these diseases is not so well understood because they do
not readily permit the use of experimental methods. Every-
one feels sure how every case of malaria is caused. On the
other hand, we are often, perhaps usually, in doubt in cases
of typhoid fever, cholera and diphtheria as to the mode of
transmission of the infection and the source from which it
comes. When there are many possible sources and modes
of infection it is not easy to determine the right one.

Evidence of Contagion the Same as for Cases. — It is
fair to claim that the evidence of the infectivity of carriers
is the same as the evidence of the infectivity of cases. We

128 THE SOURCES AND MODES OF INFECTION

believe that frank cases of diphtheria are contagious be-
cause it very generally happens that persons exposed to
them develop the disease, and, conversely, because a certain
proportion of the recognized cases have been in relation to
other cases. We believe that typhoid fever is contagious
because a certain number of persons exposed to cases of
the disease contract it, even though the proportion is
smaller than in diphtheria. It is considered a further
proof that many cases of typhoid can be shown to have
had some connection with previous cases. When a con-
siderable number of cases of typhoid fever are caused by
the consumption of milk, and the milk is known to have
been handled by a typhoid patient, or even to have come
from premises occupied by such a patient, it is generally
assumed that the contagion was derived from the patient.

In precisely the same way it has been shown in the pre-
ceding pages, and much more additional proof of the same
kind is obtainable, that persons in contact with diphtheria
carriers, and indeed the carriers themselves, not rarely
develop the disease. It has also been shown that persons
with diphtheria have often been in relation to carriers.
The number, it is true, is not very great, and it cannot be
so great as for cases; for while it is comparatively easy to
recognize any frank case in the environment of any given
patient, it is not possible to recognize any carrier there
may be, except by cultures from every one with whom the
patient has come in contact.

In typhoid fever it is easier to find carriers because ty-
phoid fever, more than any other disease, is spread through
milk, water or food; and outbreaks so caused are often
traced to some definite locality, often to a single house, so
that the search for carriers becomes easier. Hence we
have more definite proof that typhoid fever is caused by
carriers than we have that any other human bacterial dis-
ease is so caused. The outbreaks of typhoid fever which
have been shown to have a definite relation to carriers

CARRIERS AND MISSED CASES 129

afford as good evidence that carriers are the cause of this
disease as there is that cases give rise to other cases.

Lack of Statistical Evidence Alleged. — Hamer l has
urged that the mere association of a carrier and a case is no
proof of a causative relation, and he says that the finding
of carriers in connection with certain outbreaks of typhoid
fever is of little importance unless we have some idea of
the number of carriers in the general population. He does
well to call attention to the lack of satisfactory data, but
when only 1 carrier is found among 250 persons and 3
among nearly 1000, it is fair to assume that carriers are not
very common.

Carriers Often Appear not Dangerous. — It is certainly a
fact that carriers often appear to be non-infectious. Many
carriers of typhoid bacilli and of diphtheria bacilli have
been known to remain such for long periods of time with-
out apparently infecting members of their families or others
brought in close contact with them. As shown on page 84,
diphtheria carriers have been followed in schools in Boston
and Providence and no infection from them has been dis-
covered. Diphtheria carriers have been discharged from
hospitals and no cases have developed in their homes.
These and similar facts certainly demonstrate that all car-
riers do not at all times cause disease in those with whom
they are brought in contact. Why this is so may be due
to a variety of causes. Thus the excretion of bacteria is in
many cases notoriously intermittent. There is also good
reason for believing that the bacteria have in many cases
lost their virulence. Perhaps sometimes they are produced
in relatively small numbers. Again, many of the persons
exposed are doubtless immune. Furthermore, we must bear
in mind, as will be referred to in the following chapter, that
the infecting power of even frank cases of disease is very
much less than has generally been supposed. Such cases
may often remain in close association with susceptible per-

1 Hamer, Proc. Roy. Soc. Med., 1911, IV, Epidemiol. Sect., 105.

130 THE SOURCES AND MODES OF INFECTION

sons without the conditions being present for a transfer
of an effective amount of infective material. One would
expect this to happen still more often with carriers.

On the other hand, there is much evidence which shows
that carriers at times do cause disease. To most persons
the evidence of this is conclusive. The question at issue
merely is, How often does this occur? To what extent are
carriers a factor in the maintenance of the infectious dis-
eases? To the writer it appears that many things strongly
point to carriers as a factor of great moment. The prob-
ability of this is, I believe, sufficient to warrant our modi-
fying our restrictive measures accordingly. I nevertheless
freely admit that we are greatly in need of more statistical
evidence.

There are still many problems concerning the relation of
carriers to disease which need further careful study, and it
is not for a moment claimed that the " carrier theory," so
called, satisfactorily explains every epidemiological phe-
nomenon. There are many which it does not explain at
all. On the other hand, some of the most important phe-
nomena of the extension of the contagious diseases are far
better explained by the newly discovered facts concerning
carriers than by any of the theories of former years. We
still have much to learn, but we are not on that account
justified in neglecting the facts which have already been
learned and in basing our practice on disproved theories
of the sources of infection.

No Sharp Separation between Varieties of Carriers. —
That typical cases of disease are the source of similar cases
follows necessarily from the very definition of contagious-
ness. That mild cases, even very mild and atypical cases,
may give rise to typical as well as to other mild cases is
recognized by everyone. Both clinicians and epidemiolo-
gists have always believed that perfectly well persons may
be the bearers of infection from the sick to others. It
was believed by most, and is now by many, that such per-

CARRIERS AND MISSED CASES 131

sons carry the infection in the hair or clothing or on the
hands. The discovery that the germs of disease may grow
in the body without causing symptoms has forced most
rational persons to the belief that when well persons carry
infection it is because they are " carriers," that is, are
growers, of germs.

It is probably true that all carriers are not dangerous,
certainly not at all times, even when they are excreting
bacilli. We know that in some diphtheria carriers the
bacilli are not virulent for test animals and probably not
for human beings. Certain observers of high standing have
assumed that it is only the convalescent carriers and the
carriers who are in immediate contact with the sick who
are dangerous; that is, it is these only who carry virulent
germs. If these can be controlled the carrier problem is
solved, they think. I am willing to admit that the severe
case is potentially more dangerous than the mild case,
that the mild case may be more dangerous than the car-
rier, and that the convalescent carrier may be more dan-
gerous than the chance carrier found among the public at
large. There is some evidence of this, and it is not im-
probable that bacteria may tend to lose their virulence in
passing through a succession of immune persons (as it is
not improbable that they may increase in virulence by
passing through susceptible persons) ; but I can see no
ground at present for the assumption that virulent bacilli
derived from a sick person may be carried by one well person
but that when they pass to another well person they cease to
be dangerous. There seems to be no ground for assuming
that a virulent germ cannot pass from carrier to carrier.

Conclusions. — We are justified from the- evidence pre-
sented in coming to the following conclusions:

1. Mild atypical and unrecognized cases of the infectious
diseases are often extremely common. In many diseases
they may be more numerous than the recognized cases.

2. Disease-producing micro-organisms, whether bacteria

132 THE SOURCES AND MODES OF INFECTION

or protozoa, frequently persist in the body without causing
symptoms.

3. Sometimes the germs remain only a few weeks or
months after convalescence, and sometimes they may persist
for years, perhaps for life. Sometimes these carriers give
no history of ever having been sick.

4. While the bacteria found in carriers are sometimes
lacking in virulence, many times they show the highest
degree of virulence.

5. There is ample epidemiological evidence that healthy
carriers as well as mild unrecognized cases are the source of
well-marked outbreaks.

6. The number of carriers varies greatly in different dis-
eases. From 20 to 50 per cent of the population are carriers
of pneumococci. It seems probable that the influenza bacil-
lus is as widely distributed. During outbreaks of cerebro-
spinal meningitis the number of carriers may be from 10 to
30 times as numerous as the number of cases. Even when
diphtheria is not prevalent 1 per cent of the population may
be carrying the bacilli, and during outbreaks the number
may be several times greater. Probably 25 per cent of all
typhoid fever cases excrete bacilli for some weeks after con-
valescence, and it is estimated that from 1 in 500 to 1 in 250
of the population are chronic carriers. What little evidence
there is indicates that carriers are as numerous in dysentery
and cholera as they are in typhoid fever. In yellow fever,
sleeping sickness, and particularly in malaria, carriers are
very numerous. There is evidence that there are not many
carriers of measles or smallpox.

7. Any scheme of prevention which fails to take into ac-
count carriers and missed cases is doomed to partial and
perhaps complete failure,

CHAPTER III.

LIMITATIONS TO THE VALUE OF ISOLATION.

Number of Mild Cases and Carriers. — In the first chapter
the attempt was made to show that pathogenic organisms do
not usually develop outside of the body. Except for a few
diseases, or under unusual circumstances, the saprophytic
existence of disease germs is not to be looked for. Such
sources of infection are much rarer than is generally assumed,
and for most diseases may be entirely neglected. In the
second chapter, evidence was presented that certain other
sources of infection are very much more numerous than is
generally believed, and it is here contended that no scheme
of sanitation can have a scientific basis, or can have any
possibility of success, which does not take full cognizance of
them.

It must be admitted by all that mild atypical cases of con-
tagious diseases are very numerous. Every one who has had
any experience with the last epidemic of smallpox in the
United States and England must have had many unpleasant
reminders of this. Health officers' reports are full of instances
of the introduction of the disease into a community by per-
sons unsuspected by any one of having the disease, and who
often give rise to a whole series of cases. Similar experiences
with scarlet fever are often reported. The most critical inves-
tigation, such as that of our surgeons in the Spanish War,
indicates that mild unrecognized cases of typhoid fever fully
equal, if they do not exceed, the number of cases which
are recognized and reported. Even with every facility for
diagnosis, the amount of sore throat due to the diphtheria
bacillus, but not so suspected, is fully equal to the amount

of recognized diphtheria; and in many other infectious

133

134 THE SOURCES AND MODES OF INFECTION

diseases these mild cases occur with varying degrees of
frequency.

Usually not Recognized. — The extent to which these mild
atypical cases escape recognition varies with the disease, the
social condition of the people affected, the intelligence and
conscientiousness of the physician, and the attitude of the
health officer. That the majority of people will not consult
a physician unless they are decidedly sick, is certain. That
they will refrain from doing so if they expect to be reported
to the health officer and to be placed under various restric-
tions, is but in accord with human nature. A slight sore
throat, or a fleeting rash, little suggestive of danger, will be
lightly passed over, no physician will be called and no pre-
cautions taken, and often there will be no thought of danger
to others. It has always been known that a certain number
of mild cases, difficult to recognize, could be expected in
almost all infectious diseases, but it remained for the labora-
tory worker to show how numerous they are in such diseases
as typhoid fever, diphtheria, plague and malaria. The micro-
scopic demonstration of the frequency with which clinically
unrecognizable attacks of the above named and other
diseases occur, had called the attention of clinicians and epi-
demiologists to their probable occurrence in such other dis-
eases as scarlet fever, smallpox, and yellow fever, the specific
organisms of which have not as yet been discovered. So that
at the present time the most careful epidemiologists, clinicians
and laboratory workers begin to realize that very large num-
bers of mild atypical and unrecognizable cases are bound to
occur in most infectious diseases. But as yet few text-books
on sanitation, clinical medicine, or even on bacteriology, lay
sufficient emphasis on this fact. Nothing is more common
than to find the young man just from the medical school, as
well as the old practitioner, quick to deny the presence of
scarlet fever, diphtheria, or typhoid jever because the symp-
toms are not severe enough or because they deviate too
much from the text-book description. But the large number

LIMITATIONS TO THE VALUE OF ISOLATION 135

of the mild and aberrant cases, which usually remain " missed
cases," and their importance in the extension of the infectious
diseases, must now be admitted.

Carriers Exceedingly Numerous. — Still more numerous
are the pure carriers, those persons in whose bodies the
pathogenic bacteria and protozoa develop without causing
symptoms. The recognition of this element of danger is due
entirely to laboratory investigation, but, strange to say, most
workers on bacteriology lay no more stress on this epidemio-
logical factor than do the writers of treatises on hygiene or
of text-books of medicine. In the preceding chapter sufficient
evidence was presented to demonstrate the very great fre-
quency with which these carrier cases occur. Their exist-
ence and the virulence of the germs which they carry are
now established facts. Numerous instances were given where
such carriers appeared to have transmitted the disease to
others. Indeed it is almost inconceivable that it should be
otherwise. It is hardly possible that virulent typhoid bacilli
or diphtheria bacilli produced in large numbers, as they fre-
quently are in carriers, should not be equally as dangerous as
those which develop in the bodies of the sick. That is, they
are equally dangerous potentially; actually the well person
moving freely about may be more dangerous to the com-
munity than the sick person who is confined to the house.

Approaching the subject from another standpoint, it is
interesting to see how the discovery of these missed and car-
rier cases has explained so much which we formerly did not
understand.

Effort to " Stamp Out " Disease. — Twenty-five or thirty
years ago we heard a great deal about "stamping out" the
contagious diseases. That was the era of the building of hos-
pitals for these diseases, of the organization of the sanitary
service, of the discovery of pathogenic bacteria. The wonder-
ful decrease in smallpox, the successful fight against cholera,
the almost total disappearance of typhus fever, and the com-
plete disappearance of plague, only foreshadowed, it was said,

136 THE SOURCES AND MODES OF INFECTION

the extermination of typhoid fever, diphtheria, scarlet fever
and measles. It was claimed that in those diseases which
are exclusively contagious, if every case can be isolated until
it is free from infection, the disease will be exterminated. It
was believed that if people, and especially physicians, would
take only a little more care, practically all cases of these
diseases could be recognized and isolated. It was also thought
to be not very difficult to control them until infection had
disappeared. This confidence in the efficacy of isolation was
in the then existing state of knowledge not unreasonable.

Isolation and its Results in Providence. — Previous to
1884 there had been in Providence no isolation to speak of
in any of the contagious diseases except smallpox. In fact
very many physicians did not consider that scarlet fever and
diphtheria were very contagious, if contagious at all, but
were inclined to look upon them as filth diseases. Restric-
tive measures, including isolation at home and fumigation,
began to be applied in 1884 and were quite steadily strength-
ened during the next sixteen or seventeen years. I hoped,
as did most health officers, that if scarlet fever and diph-
theria could not be stamped out, they could be reduced to
an insignificant remnant. But they were not stamped out
in Providence, as they have not been in other cities. On the
contrary, we had twice as many deaths in 1887 from scarlet
fever as we had had during any year for seven years. Diph-
theria from 1886 to 1890 also caused nearly double the num-
ber of deaths that it had in the preceding four years. Of
course we talked about epidemic waves, and noted that the
mortality from the last wave was very much lower than from
many that had preceded it, and congratulated ourselves that
the outbreak was not so severe as in former years. But I
began to ask myself what there was about epidemic waves
that made restrictive measures of little use, and also to inquire
if there was anything wrong about the restrictive measures.
If we were limiting these diseases at all, it was certainly in a
very moderate way.

LIMITATIONS TO THE VALUE OF ISOLATION 137

Infection by Air and Fomites Thought Most Important. —

It was fully appreciated that in cities at least most cases
of contagious disease cannot be traced to their source. Two
theories have from antiquity been advanced in explanation.
One is that most contagious diseases are easily carried
by the atmosphere. Thus a person going by a house where
there is scarlet fever, or passing an infected person in the
street, might contract the disease. Such unconscious exposure
might be quite common. But the principal source of the
untraced cases of contagious disease was believed to be fo-
mites. Walls and furniture were thought to become infected
with the virus, and for weeks and months persons entering
the room might contract disease through the breath. Books,
toys, clothing and, in fact, every material thing, might readily
become a source of infection and retain its virulence for
months and years. These were perhaps not unreasonable
a priori hypotheses, and they had some apparent backing of
facts. At any rate they were the best theories we had. So
health officers everywhere, including Providence, set about
improving methods of disinfection. Sulphur fumigation was
abandoned and the use of formaldehyde gas adopted in
its place. Many cities set up a steam disinfecting plant, in
Providence as early as 1887, and carpets, bedding and cloth-
ing were disinfected by steam. Some cities, particularly on
the continent of Europe, sent a band of uniformed disinfec-
tors to wash and scrub everything in the infected house.
Scarlet fever and diphtheria refused to be exterminated,
though in Providence we did have rather less during the early
nineties than we had had before. But I was not satisfied.
It seemed to me that we were having too much of these
diseases, that there must be a leak somewhere.

Cultures Expected to Discover Much Diphtheria. — Then
for one disease a new weapon was put into our hands. Many
had long recognized that the diagnosis of diphtheria was diffi-
cult. It was suspected that many cases, because of this diffi-
culty, escaped isolation entirely. When the culture method

138 THE SOURCES AND MODES OF INFECTION

of diagnosis was devised I became enthusiastic and hopeful.
We adopted it in Providence in January, 1895, and soon after
required a negative culture before the patient was released
from isolation. Hill has shown that without cultures the
chance of error in the diagnosis of diphtheria is 50 per cent,
which corresponds entirely with my frequently expressed
opinion before the advent of the culture method. It is evi-
dent, then, that the general use of cultures ought to bring to
light great numbers of cases of diphtheria which were formerly
unrecognized, and this it certainly does. If such an im-
provement in diagnosis, and consequently in isolation, is
brought about by the use of cultures, and if by the same means
isolation can be maintained until the patient is certainly free
from infection, there ought to follow a marked reduction
in this disease. But it was quite otherwise. The deaths in
Providence, which in 1894 had numbered 45, rose to 79 in
1895 and 125 in 1896, nearly twice as many in proportion to
the population as there were in 1883, when there was no isola-
tion, no disinfection and no antitoxin. The cases rose from
166 in 1894 to 386 in 1895 and 890 in 1896. The apparent
reduction in the fatality rate from 27.71 to 14.07 indicates
very plainly that the culture method of diagnosis had dis-
covered a very large number of mild cases that would have
previously been unrecognized, for antitoxin was only a minor
factor in reducing the fatality, as it had been used in only a
little over one-third of the cases. Isolation, disinfection, the
use of cultures, and the opening of the contagious hospital
had been accompanied by the greatest prevalence of the dis-
ease for ten years. I do not mean to say that the adoption
of the measures described had no effect upon the amount
of diphtheria in Providence. I am sure that they had, and
that this disease on the whole has been lessened, cases pre-
vented and lives saved. But better results were expected.
I was disappointed, and I think other health officers have
been disappointed also. It seemed that the measures, car-
ried out as they were, ought to have given better results.

LIMITATIONS TO THE VALUE OF ISOLATION 139

It seemed that there was something which we did not
understand.

Failure of Hospital Isolation. — One of the most effectual
means of isolating cases of contagious diseases is by removal
to the hospital. Certainly while in the hospital they can do
no harm, and with reasonable care there is not much danger
of their carrying infection back to their homes. Return cases
do not occur in scarlet fever and diphtheria in more than
about one to three per cent of discharges, and are not a factor
of moment in the extension of these diseases. The idea that
such hospitals would be a powerful factor in the extermina-
tion of these diseases was not unreasonable. Smallpox hos-
pitals have been in general use for a very long time, but they
are not here under consideration, though it is questionable
whether the hospital isolation of smallpox can ever accom-
plish much alone and unaided by vaccination. But it is the
hospitalization of scarlet fever and diphtheria that is par-
ticularly instructive.

English Hospitals. — The use of hospitals for contagious
diseases has been carried farther in England than elsewhere.
Fifteen or twenty years ago the larger municipalities began
building them on a considerable scale, and at present most
of the English towns are provided with large hospitals for
scarlet fever, and to a less extent for diphtheria. At the time
when their construction was first strongly urged it was be-
lieved that their use would result in the eradication of the
diseases for which they were provided. The result has cer-
tainly been disappointing, and there has recently been an
active discussion as to whether they do an amount of good in
restricting disease at all commensurate with their cost.1 There
is no doubt that the mortality from scarlet fever, both in

1 O'Connor, Geo. Wilson, Waddy and others, Brit.M. J., Lond., 1905,
II, 630; Millard, Biss, Fraser, etc., Med. Press & Circ., Lond., 1904,
LXXVIII, 215, 218, 241, 327, 377; Newsholme, Tr. Epidem. Soc.
Lond., n. s., 1900-01, XX, 48; J. Hyg., 1901, I, 145; Millard, Pub.
Health, 1901, XIII; J. T. Wilson, Pub. Health, 189&-97, IX, Sup., p. 21.

140 THE SOURCES AND MODES OF INFECTION

England and the United States, has greatly diminished during
recent years, but whether this has been due to restrictive
measures or to lowered virulence has been disputed. The
small death rate would indicate that the disease is really
milder. This is also indicated by the fact that plural deaths,
that is, more than one death in a family, are less frequent
now than formerly. Again, local outbreaks of the old-time
severe type occasionally appear. There was such an out-
break in Providence in 1906-07, during which the case fatal-
ity ran up to 12.85 per cent. Similar outbreaks have been
noted in Keene, N. H., Haverhill, Worcester and Mont-
real. The relatively lowered fatality in recent years has
rendered it difficult to determine from the number of deaths
just what influence restrictive measures, like hospital isola-
tion, have had on the prevalence of scarlet fever, and increas-
ing accuracy and care in reporting cases render it difficult
to draw conclusions from the number of cases. But after
all has been said it is clear that hospital isolation in scarlet
fever has checked the disease very much less than was ex-
pected, and sometimes appears to have had little effect. In
Huddersfield, a city of nearly one hundred thousand people,
from 1890 to 1899 the percentage of removals to the hospital
was 90, yet the mean attack rate for the period was 4.3 per
thousand. From 1900 to 1908 the removals to the hospital
were 92.4 per cent and the attack rate 2.96. This is certainly
a surprisingly high morbidity rate for a city where practically
all reported cases have for twenty years been subjected to
most excellent isolation. A similar state of things is noted
in other cities. Some cities with a high per cent of removals
to the hospital have more of the disease than do cities with
no hospitals. The same is noted in rural communities.
O'Connor,1 medical officer of health of Leicestershire and Rut-
land combined sanitary districts, reported -that in five par-
ishes where the percentage of hospital isolation had for ten
years reached 66 per cent, the attack rate was 6.2 per thou-
1 O'Connor, Brit. M. J., Lond., 1905, II, 630.

LIMITATIONS TO THE VALUE OF ISOLATION 141

sand, while in five other parishes where hospital isolation was
applied to only 14 per cent of the cases, the attack rate was
2.8. In another district, one parish which sent 60 per cent of
its scarlet fever to the hospital, had three times as many
cases as contiguous parishes which had little hospital isola-
tion. Neech1 says that in Cornwall from 1898 to 1907 there
was practically no difference in the incidence of scarlet fever,
in both urban and rural districts, connected with the use or
failure to use isolation hospitals. Again, a city after it builds a
hospital may have more of the disease than before. The aban-
donment of hospital isolation may do no harm. In Leices-
ter2 the hospital was closed temporarily, and all the scarlet
fever cases sent to their homes, with an actual decrease in
number of reported cases. The hospital in other cities also
has been closed without harm. I have thought it possible
that at times hospital isolation might actually favor the
spread of disease. When a case is removed to the hospital,
comparatively little restraint is placed upon other members
of the family, and if some of them are carriers, as they cer-
tainly usually are in diphtheria, much harm may result. If,
however, the case is kept at home, the rest of the family,
particularly the children, are under considerable restraint for
several weeks. Observations in Providence demonstrate that
the retention of the case at home very rarely leads to the
extension of the disease to other families in the house, and
presumably to still less extension outside of the house.

Why the Failure ? — It may be admitted, and is doubtless
true, that hospitals have prevented very many cases of dis-
ease, and they may have been somewhat of a factor in its
lessened prevalence. Nevertheless it must also be admitted
that, notwithstanding the complete and excellent isolation
secured in some cities like Huddersfield, scarlet fever has
still prevailed to an alarming extent. When eighty to ninety
per cent of the cases are removed to the hospital it is certain

1 Neech, Pub. Health, Lond., 1908-09, XXII, 296.
* Rep. on Health of Leicester, 1902, 36.

142 THE SOURCES AND MODES OF INFECTION

that the remainder will be so situated that home isolation will
quite effectually prevent extension from them. In such cities
almost all of the reported cases are thus effectively isolated,
either in hospital or home. Yet the disease continues to pre-
vail. There is evidently some source which escapes control.
Newsholme has very clearly shown this. While admitting
that isolation does prevent much sickness, he says that the
disease still prevails and outbreaks still occur owing to some
"epidemic influence." Now our purpose is to determine, if
possible, what this "epidemic influence" is. What is the
factor which is so powerful and continuous in its action that
90 to 95 per cent of perfection in the isolation of a contagious
disease is unable to prevent a continued high prevalence and
repeated epidemic waves ?

American Hospital Experience. — It is not merely English
hospital experience which has shown the inefficiency of isola-
tion. We have numerous instances in our own country. I
have already referred to the experience of Providence, where
increasing stringency in isolation was not followed by a
decrease in contagious diseases. There has been much dif-
ference between American cities as regards the strictness of
enforced isolation. Some have been notoriously lax, while
others have for years endeavored to secure a complete regis-
tration of scarlet fever and diphtheria, and have adopted
rigorous measures of isolation and disinfection, though in no
American city has hospitalization been carried so far as it has
in many English towns. But no one would be able, by study-
ing mortality rates, to pick out the cities which pursue a
rigorous policy of isolation. I collected data relating to the
prevalence of scarlet fever and diphtheria in American cities
for the decade 1890-1 899,1 and it is surprising to note that
the cities with the best sanitary administration frequently
have a comparatively large amount of scarlet fever and diph-
theria. During the period mentioned, Boston, in my opinion,
had the best sanitary administration of any of the large cities,

1 Chapin, Municipal Sanitation, Providence, 1901, Table op. 480.

LIMITATIONS TO THE VALUE OF ISOLATION 143

though New York stood high. The death rate from diph-
theria in both cities was 84 per 100,000 living, from scarlet
fever 25 for Boston, and 33 for New York. Certainly neither
Chicago nor Cincinnati enforced such rigorous measures, yet
the rates in these two cities were 72 and 71 for diphtheria
and 17 and 7 for scarlet fever. Among the smaller Mas-
sachusetts cities Fall River has usually had a rather ineffi-
cient health service and little hospitalization, yet the death
rate from diphtheria was 21 and from scarlet fever 15 per
100,000 living, while in Worcester the figures were 48 and 8,
and this notwithstanding the fact that in Fall River the
proportion of children is much greater than in most Amer-
ican cities, and that the population is exceptionally ignorant
as measured by illiteracy. Worcester has had a contagious-
disease hospital since 1897, and has removed to it in some
years as high as 63 per cent of its diphtheria cases. In general,
Worcester secures an excellent registration of cases, and con-
sequent isolation. Nevertheless Worcester has recently had,
notwithstanding its increasing hospitalization and good home
isolation, a severe outbreak of the disease. It seems a fair
assumption that some factor much more important than the
recognized cases of the disease has been at work in Worcester.
If it were not so, the reported cases of the disease should not
have risen from 132 in 1905 to 1178 in 1907.

Diphtheria Isolation at Home. — In Providence for some
years previous to March, 1902, isolation in diphtheria was
terminated only when a negative culture had been obtained
from the patient. At that date a change was made, terminat-
ing isolation ten days after the disappearance of exudation.
This certainly liberated many cases still infectious, but this
factor was of so little moment that the disease continued to
decrease until on a certain day in August the city was entirely
free from reported cases, a condition again nearly reached in
1907.

Isolation in Institutions. — It is by no means uncommon
to see outbreaks in institutions lasting many months, al-

144 THE SOURCES AND MODES OF INFECTION

though every case is promptly isolated and all persons in the
institution are repeatedly examined by cultures from throat
and nose. I have records of several such in Providence, and
most health officers and managers of institutions have been
through such unpleasant experiences. Here again we are
forced to look for sources other than the recognizable cases.

At a school in Owatonna, Minnesota,1 there had been more
or less diphtheria for years until a sharp outbreak in 1896
caused a very energetic effort to be made to stamp it out.
Cultures were taken from everybody, and all persons in whom
diphtheria bacilli were found were isolated, at first in a com-
mon ward, but later each person in a separate room. Isola-
tion was maintained until three successive negative cultures
were obtained from throat and nose, and efficient disinfection
was practiced. Still the disease persisted month after month,
and some of the carriers, after release, were found to be still
harboring the bacilli. One boy carried the bacilli for nineteen
months. The attempt at the isolation of carriers was aban-
doned, and several years later there was still more or less
diphtheria in the institution.

A still more notable failure of the isolation of diphtheria
was at the Willard State Hospital for the insane in New York.2
In this institution practically the same methods were adopted
as at Owatonna, but nevertheless the disease persisted sev-
eral years. Patients and carriers released from isolation after
several negative cultures were still found to be carriers. If
in an institution of this kind, where the strictest discipline
is maintained, and where carriers can be isolated indefinitely,
the most stringent measures of isolation and disinfection fail
to stamp out the disease, what is to be expected of any such
measures which can be adopted in an ordinary community?

An outbreak in a hospital in California was managed in

1 St. Paul M. J., 1900, II, 223; Brit. M. J., Lond., 1898, I, 1008; Rep.
St. Bd. Health, Minnesota, 1897-98, 465; Rep. Am. Pub. Health Ass.,
1899, XXV, 546.

8 Rep. State Commission in Lunacy, N. Y., 1904, XVI.

LIMITATIONS TO THE VALUE OF ISOLATION 145

the same way.1 At first no effect seemed to be produced,
and the outbreak lasted from April to August, but was as-
sumed to have been finally checked by the same measures
which failed at Owatonna and Willard. But as many out-
breaks last even a shorter time, though no attention is paid
to carriers, it is uncertain whether the outbreak was stamped
out or died out. Sidney Davies2 reports that at the Bostall-
lane School in Woolwich diphtheria prevailed for three years
notwithstanding repeated closures and bacterial examinations
of the pupils. There is little doubt that at times a careful
search for carriers and missed cases in schools and institu-
tions and their isolation until two successive negative cul-
tures from throat and nose are obtained, prove successful in
checking outbreaks; there is no doubt, too, that outbreaks
frequently die out of themselves; and it is shown by the exam-
ples given above that it is at times impossible by any degree
of isolation to stamp out an outbreak. How to manage such
institutional epidemics cannot be determined until the con-
trol of carriers is more carefully studied than it has been.

Isolation a Failure in Measles. — Measles is a disease
which in cities it seems to be impossible to check to any
appreciable extent by isolation. In Aberdeen 3 this was faith-
fully tried for twenty years, 1883 to 1902, but no apparent
effect was produced on the prevalence of the disease. Similar
failures have been noted elsewhere. During the last half of
this period in Aberdeen, when there was far more accurate
registration and better control than before, the number of
cases rose to 24,254, about fifty per cent more than in the first
half of the period. A census of the children in certain schools
indicated that from ninety to ninety-three per cent of children '
over ten years of age had had the disease. Restrictive meas-
ures which protected only seven to ten per cent of the popu-
lation from attack were then wisely abandoned. The very

1 Rep. State Bd. of Health, California, 1906-08, 201.

2 Rep. Med. Off. Health, Lond., 1908, 36.

8 Aberdeen, Report of Med. Off. of Health, 1904, 41.

146 THE SOURCES AND MODES OF INFECTION

excellent report of the medical officer of health of Aberdeen
discusses the subject very fully. In New York measles was
first isolated in 1896, but not until 1902 were the regulations
very rigorously enforced. The average death rate from 1895
to 1904 was 2.40, and the highest death rate since 1896 was
reached in 1906, when it was 2.69. There is no evidence that
the measures adopted in New York have had any more
influence on the prevalence of the disease than did isolation
and disinfection in Aberdeen. It seems in the highest degree
probable that the disease prevails because of the unrecognized
but infectious prodromal stage. No amount of isolation after
the disease is recognized can atone for the harm done before
the diagnosis is made.

Isolation a Failure in Meningitis. — Within a recent period
the city of New York suffered from an exceptionally long and
severe outbreak of cerebro-spinal meningitis. The attack
rate was higher than in any of the other large American cities,
and the outbreak lasted longer; and it lasted longer than it
has in most of the German cities. On April 19, 1905, after
the epidemic had continued for about two years, very strin-
gent restrictive measures of isolation and disinfection were
adopted. The outbreak was then declining, and that it would
afterwards decline still more was to have been expected.
But it was not "stamped out," for in 1907 there were reported
642 deaths and in 1908, 351. In Leith l great efforts
were made to isolate all suspects and carriers, and the
outbreak appeared to be checked, but in Edinburgh also
it died out, though without such energetic measures. The
experience in Germany seems to be that if in the beginning
diligent search is made for carriers, and strict isolation is
maintained, the outbreak may sometimes be checked, but
after the disease has become established, and carriers are
numerous, isolation is of little avail. These appear to be
the views of Lingelsheim, Ostermann, Selter, Flatten and
others.

1 Ker. Practitioner, Lond., 1908, LXXX, 66.

LIMITATIONS TO THE VALUE OF ISOLATION 147

Isolation often a Failure in Smallpox. — Smallpox is less
likely to escape detection than is any other disease. Yet
epidemics grow in the face of the most rigid isolation. Unrec-
ognized cases, even in this disease, are so numerous that the
isolation of the recognized cases often seems to be a complete
failure. The State Board of Health of Minnesota, realiz-
ing this,1 has had the boldness to advise that no attempt be
made to isolate, and that entire reliance be placed on vacci-
nation. It was hoped that this would lead to more complete
vaccination. There has been no alarming increase in small-
pox in Minnesota, and Montana 2 is now following the lead
of the former state.

Why Does Isolation Fail ? — The epidemiological evidence
is conclusive that the isolation of recognized cases of con-
tagious diseases often fails to check outbreaks which grow
in spite of it; that it does not stamp out disease, and that it
only reduces in a moderate degree the prevalence of the
disease. We are forced to conclude that there is some defect
in our procedures, or some other source of infection more
important than the recognized cases. The error cannot be
that isolation is too brief, for return cases and recurrences
are not important factors. It is not that isolation is imper-
fect, for isolation in the hospital is well-nigh complete, and
careful observations in Providence indicate that the danger
of extension outside the family from cases at home is very
slight.

Importance of Carriers and Missed Cases. — As has been
previously stated, two theories have been advanced to account
for the appearance of untraced cases of contagious diseases:
that of distant aerial infection, and that of the persistence of
infection on things, i.e., fomites. In other chapters of this
book it is shown that both of these modes of infection are
of little moment, and in the first chapter it was shown to be

1 Resolution adopted July, 1906, to go into effect January, 1908,
Am. J. Pub. Hyg., 1907, III, 227.

2 Am. J. Pub. Hyg., 1909, V, 815.

148 THE SOURCES AND MODES OF INFECTION

unlikely that the specific organisms of our common diseases
grow outside of the body. In the second chapter it was
shown that they may grow in the body and yet produce few
or no symptoms. It appears, then, highly probable that by
far the most important factor in the causation of the con-
tagious diseases are the "carrier" and " missed" cases.

It is not for a moment suggested that the existence of
carriers explains all epidemiological problems. There are, for
instance, quite a number of contagious diseases which exhibit
a varying degree of periodicity. We have as yet only the
vaguest notions as to the causes of the variations in the
prevalence of disease, its seasonal irregularities, and the rise
and fall of epidemics. There is no reason to believe that
these phenomena depend on the extra-corporal growth of
pathogenic organisms, and it is probable that the seasonal
distribution of such diseases as smallpox and scarlet fever is
only in the most indirect manner dependent upon tempera-
ture, rainfall, etc. It is not unlikely that the factors affecting
the extension of these diseases are so numerous that their
prevalence is really to a large extent a matter of chance.
But whatever the factors may be, we must believe that they
affect carriers as well as clinical cases. Why we have more
carriers and more cases at one time than another, we cannot
at present say. But it appears to be in the highest degree
probable that in times of outbreaks, as well as in inter-
epidemic periods, the chief factor in the extension of the
disease is the existence of unrecognized infection in human
beings or, in some diseases, in the lower animals.

Need of Further Study. — I would be the last person to
assert that the views here set forth are unassailable and unal-
terable. They doubtless will be modified; it may be that
they are entirely erroneous. What is needed is further inves-
tigation along the lines indicated, and a great deal of it.
Unfortunately sanitary science is far from exact. We have
few established truths, but many theories of greater or less
probability, on which to base our practice. The probability

LIMITATIONS TO THE VALUE OF ISOLATION 149

of each theory must be carefully weighed. Which is the most
probable source of infection, the cesspool in the yard, the
fomites that escaped disinfection, the patient who was re-
leased too soon, or the unknown carrier ? The chances are
greatly in favor of the last, and yet to-day sanitary practice
almost completely ignores the carrier. It is not surprising
that many health officers fail to appreciate the importance
of the question which is here discussed. It is remarkable,
however, that bacteriologists as a rule minimize the danger
to be apprehended from carriers.

Carriers not to be Ignored. — It may be admitted that
only 1 per cent of the population are diphtheria carriers,
though as a matter of fact it must often be more than that;
also that only 15 per cent of these carry virulent germs, though
it is really often much greater. It may also be admitted that
these bacilli are not usually so numerous as in the sick, though
it is known* that they sometimes appear in pure culture; yet
even admitting all this, there will be in inter-epidemic periods
in a city of 100,000 people at least 150 well persons carrying
virulent diphtheria bacilli. Why should the bacteriologist
ignore these 150 sources of infection and insist on two throat
and nose negative cultures from every reported patient ? Or
why should he, as he occasionally does, recommend isolating
the carriers in the family but ignore all other carriers ? Why
neglect this whole question of carriers, and spend endless time
in devising methods of liberating formaldehyde to kill the
few bacilli that may remain about the house ? Why worry
about the saliva that may have gotten onto a book, a coat,
or the wall of a room, and neglect the reader of the book,
the wearer of the coat and the dweller in the room, who is
probably growing the germs in his throat and nose ? It is
not easy to answer these questions, and so the bacteriologist
withdraws to his laboratory and ingores them. But the
health officer must answer them in one way or another.

Principles of Diphtheria Isolation. — To confine the dis-
cussion for the moment to diphtheria, concerning which we

150 THE SOURCES AND MODES OF INFECTION

have more accurate knowledge than we have for any other
disease, there appear to be only two logical positions which
the health officer can take in regard to its management.
There is the possible policy of non-regulation. This the public
would certainly not permit, and I think with very good
reason. Or we can go to the other extreme and attempt to
isolate every person carrying diphtheria bacilli, until they
disappear from throat and nose, or until they are shown to be
non- virulent. That is, of course, entirely impracticable except
in small isolated communities or institutions, and it is often
very difficult, and sometimes impossible then. In every large
city there are hundreds of carriers who can never be detected,
and who could not be controlled if they were. Virulent bacilli
may remain for months, and the attempt to isolate for a pro-
longed period prominent lawyers, business men, or physi-
cians, would result in a breakdown of the whole system.
As a matter of fact, such carriers escape from isolation while
still infected. If cultures are taken daily, or every few days,
it is usually not very long before the two negatives required
by rule are secured and the patient is released, though sub-
sequent cultures not infrequently reveal the bacillus. The
ideal of health officers has been to keep up isolation until
every spark of infection has died out, — a very reasonable
ideal, until it was learned that there are many hidden sparks
scattered about the community, some of which are sure sooner
or later to burst into flame. As it is impossible to attain the
ideal of stamping out all infection, and as it is certain that
many infected persons cannot be restrained at all, it is unrea-
sonable to require restraint to the uttermost limit, of the
recognized cases. It is often argued that the fact that all
infected persons cannot be isolated, is no reason for not iso-
lating all that can be found, anymore than the fact that many
thieves escape is no reason for not imprisoning those thieves
who are caught. But there is no true parallel here. The
thief is a criminal, and his imprisonment is a punishment.
The bacillus carrier is not a criminal, and he is isolated, not

LIMITATIONS TO THE VALUE OF ISOLATION 151

as a punishment, but to protect the community. If the
protection secured is not commensurate with the hardship
inflicted, the procedure is unjust, and unwise from a socio-
logical standpoint.

A Compromise Necessary. — But just as soon as we depart
from the orthodox ideal, absolute isolation, we have to adopt
some sort of compromise, a compromise which has no strictly
logical defense. We are thus forced to follow one of three
policies, — either do nothing, which is unwise and would not
be permitted, or attempt absolute isolation of all sources
of infection, which is manifestly impossible, or employ a
moderate degree of restriction, which, though not strictly
logical, is nevertheless practicable, reasonable and to a con-
siderable degree effective.

Isolation of Real Value. — While the most rigorous isola-
tion does not stamp out diphtheria, we know that restriction
does some good. The removal of a case to the hospital is
shown by statistical evidence to prevent to some extent the
development of secondary cases in the family. So too does
good isolation at home. The warning sign on the house
keeps out many people, some of whom would otherwise cer-
tainly contract the disease. The regulation of school attend-
ance prevents some school outbreaks, and others are checked
by the prompt removal of the infecting child. The evidence
points to an appreciable reduction in diphtheria prevalence
in recent times, which I believe it is fair to consider as due,
in part at least, to deliberate separation of the sick from the
well under the direction of the sanitary authority or other-
wise. Why more has not been accomplished by isolation is
explained by the facts set forth in the preceding chapter.
If because of the existence of so many unrecognized sources
of infection, isolation can effect only a limited degree of pre-
vention, could not substantially as much be accomplished
by more moderate methods than are usually employed ?
Why keep the patient indoors for six weeks when other mem-
bers of the family with infected throats are going about ?

152 THE SOURCES AND MODES OF INFECTION

Why keep from work the wage earners in an infected family
when scores of other carriers are attending to their business
without restraint ?

Typhoid Carriers cannot be Isolated. — The isolation of
typhoid carriers is no more possible, just, or effectual in check-
ing disease, than is the isolation of diphtheria carriers. As
was shown in the preceding chapter, the number of carriers
and convalescents excreting bacilli is probably much larger
than the number of cases confined indoors. It seems to me
useless to attempt to confine convalescents two or three
months after their recovery. There certainly would be most
energetic opposition on the part of the public, which probably
would ultimately be sustained by the courts. The health
officer who attempted to isolate convalescents until bacilli
were no longer to be found in their urine, would be in an
awkward position if he allowed chronic carriers to go at large,
and he would be in a still more awkward position if he at-
tempted to isolate all chronic carriers indefinitely. There
are probably 200,000 cases of typhoid fever in the United
States each year, and 3 per cent of these would be 6,000.
To attempt to isolate 6,000 carriers would of course be futile.
Not one-tenth of them could even be discovered. To isolate
the small fraction of carriers who can be discovered is practi-
cally useless, and therefore unjust. It may be, and probably
is, wise to regulate the life of such carriers as may be dis-
covered, and at times to forbid their engaging in certain
occupations, such as those of cook, waitress and milk dealer,
but to attempt their isolation under present conditions seems
to me most unwise.

Isolation too Rigorous. — I believe that, on the whole, iso-
lation in our prevailing contagious diseases is carried farther
than is necessary; that less rigorous measures would accom-
plish practically as much good, and that there would be less
temptation to conceal cases and to interpret doubtful symp-
toms in line with the patient's desires. It is impossible here
to lay down in detail a scheme for the proper isolation of

LIMITATIONS TO THE VALUE OF ISOLATION 153

contagious diseases. What are the best methods of dealing
with these diseases is a matter for free discussion, and also
for experiment. It is to call attention to the new facts, and
to elicit discussion from the new viewpoints, that these pages
are written.

Degree of Contagiousness. — Before considering some of
the principles which should guide the application of meas-
ures of isolation, we should consider what are, under the
ordinary conditions of life, the chances of infection from a
single individual. Until recently it has been believed to be
very great. That smallpox hospitals are placed a mile or
more from inhabited districts, that physicians clothe them-
selves with gowns and caps and rubber boots on their visits
to infectious cases, and that it is insisted that the ceiling of
a room occupied by such a case be disinfected, indicate a
belief that the virus of the contagious diseases is exceedingly
diffusive and exceedingly virulent. Why such a belief is not
well founded will be shown in succeeding chapters, and it
will suffice here to call attention to the fact that the chance
of an infected person transmitting the infection to another
is not nearly so great as is generally supposed. This is a
mathematical necessity. We now know that the number of
infected persons is very much greater than was formerly
believed, and that they often remain infected for much longer
periods than was suspected. It necessarily follows that the
danger to be apprehended from any one person at any one
time is much less than was once thought.

Factors Involved. — The chance of an infected person giv-
ing rise to the disease, or of transmitting the infection to
others, varies greatly with the intensity of the infection, or
the number of disease germs in the secretions or excre-
tions, their virulence, the volume of the excretions, the care
exercised, the occupation, and the surroundings generally.
Conditions in the home and in institutions are usually much
more favorable than elsewhere for the transmission of infec-
tion. Yet in the home the chance of such extension of disease

154 THE SOURCES AND MODES OF INFECTION

is not as great as has been supposed. In Providence,1 the
chance of persons contracting diphtheria from another mem-
ber of the family who has the disease is only about 1 in 15.
The chance of a child between two and six years contract-
ing it is 1 in 5; of an adult, 1 in 40. The chance of a woman
contracting the disease in the family is about three times as
great as that of a man. For scarlet fever the figures are
not very different. Every one must have noted repeated
instances where only one of a family of children is attacked
by scarlet fever or diphtheria, the rest of the family remaining
well, though isolation may have been far from satisfactory.
Danger Less outside Family. — Outside of the family, in
school, in factory and in ordinary social relations, except per-
haps in the play of young children, the chance of transmitting
the disease must be very much less. Instances are not rare
where children, presumably in the infectious stage of scarlet
fever, have mingled freely with others for many days, or
perhaps weeks, with little or no extension of the disease. I
have notes of an instance where a boy with scarlet fever in
the sore-throat stage attended a Sunday-school festival, and
no other case developed among the large number of children
present. At an infant asylum a child was sick with mild
scarlet fever for 17 days, mingling freely with about 75 chil-
dren, mostly under 5 years of age, and only 3 other cases re-
sulted. At a large school a girl returned at the end of the
first week of an attack of scarlet fever, and continued her
attendance for 20 days. Only 3 or 4 cases developed in that
school. Similar and even more marked instances of appar-
ently feeble infectivity are reported by others. I have made
no attempt to collect such cases, but I happen to have before
me three reported by Butler.2 One child, taken sick Novem-
ber 5, attended school from November 11 to November 19.
Two other children, attacked December 26 and 27 respec-
tively, attended different schools up to January 16. In only

1 Rep. Supt. of Health, Providence, 1909.

8 Butler, Proc. Roy. Soc. Med., Lond., 1908, I, Epidemiol. Sec., 225.

LIMITATIONS TO THE VALUE OF ISOLATION 155

one of the three schools did even a single case occur. I have
known of a teacher with virulent diphtheria bacilli in her
throat from the first of January to the middle of April, who
taught in a kindergarten all that time, but who did not trans-
mit the disease to any one. Many of the instances of typhoid
carriers which have been referred to, show that such persons
may for long periods of time fail to infect any one, even though
employed as cooks or handlers of milk. I have known of
cases of smallpox remaining for several days in lodging
houses or hospital wards, or traveling on railroads, without
any one contracting the disease. Failure to infect may be
due sometimes, perhaps, to intermittency in the excretion of
germs, sometimes to lack of contact, and sometimes to lack
of susceptibility on the part of the receiver of the infection.
Whatever the cause may be, we must recognize that an
infected person often mingles freely with the public without
transmitting the infection to another.

Carriers less Infective than the Sick. — As has been
stated, it is not improbable that the infectivity of a well car-
rier may for several reasons be less than that of a person sick
with the disease. It may then be argued that the danger
from carriers has in the preceding pages been very much
exaggerated. It is not to be denied that the probable danger
from a single carrier is small. Their number, however, is
large, so that the danger from all carriers is large, and is,
I believe, a very considerable factor in the maintenance of
the contagious diseases.

Isolation should Vary. — In applying isolation to the
prevention of disease it must be kept in mind that different
conditions require different procedures. Thus when an ordi-
nary contagious disease first appears, after a considerable
absence, in a small community, or in an institution, very
rigorous measures of isolation are usually desirable, as expe-
rience has shown that very often an outbreak is thus effec-
tively checked in its beginning.

Isolation in Villages. — The efficacy of isolation under such

156 THE SOURCES AND MODES OF INFECTION

circumstances is well illustrated by the history of outbreaks
of the common contagious diseases in the smaller cities, town-
ships and villages of Michigan. The data given in the annual
reports of the board of health of that state are of great epi-
demiological interest, and my discussion of the subject on
another occasion is here given.1

" Only those places are considered which have remained
free from the disease for at least sixty days, and this unfor-
tunately is never true of a city of any considerable size. The
outbreaks reported are arranged in groups, one in which iso-
lation and disinfection were both enforced, one in which they
were both neglected, and one in which the reports did not
state with sufficient exactness what restrictive measures were
carried out. The following is a summary of some of the
tables in the report :

Number of Cases
per Outbreak.

Number of Cases
per Outbreak.

Restric-
tive Meas-
ures not
Enforced.

Restric-
tive
Meas-
ures
En-
forced .

Restric-
tive Meas-
ure? not
Enforced.

Restric-
tive
Measures
Enforced.

2.22
1.71
2.53
4.67
3.80

Typhoid fever, 10 years . .
Diphtheria, 14 years
Scarlet fever, 14 years ....
Measles, 1 1 years

5.82
11.12
11.95
48.30

3.13
2.11
2.32
3.03

1900

a

14

n

6.72

4.85
10.43
27.60
32.00

Smallpox

Isolation Effective. - " Several things are to be noted in
connection with these figures. In the first place, isolation
and disinfection accomplish very much in preventing the
extension of all these diseases. The number of facts is so
great, the outbreaks of each disease running into the hun-
dreds, and the difference between good and bad sanitation is
so manifest in each one of the years for each one of the

1 J. Mass. Ass. Bds. Health, Bost., 1904, XIV, 226.

LIMITATIONS TO THE VALUE OF ISOLATION 157

diseases, that the success achieved must be a very real one.
It appears certain that isolation and disinfection as practiced
in the smaller communities of Michigan reduce the cases of
contagious disease in round numbers from forty-five to ninety-
five per cent. In scarlet fever, diphtheria, measles and small-
pox, isolation appears from the reports, as one would expect,
to have very much more restrictive effect than disinfection.

"As will be shown later, disinfection probably has little
influence in restricting contagious diseases, but in view of
existing ideas and practices it is probable that isolation and
disinfection were not so distinctly separable as is indicated by
the reports of the local health officers. One of the things
which appeared most remarkable to the writer in these re-
ports is the apparently great restriction of measles. It has
certainly been the experience in all our larger cities that
restrictive measures, no matter how energetic, have had very
little effect in reducing the mortality from this disease. But
in the smaller communities in Michigan it appears that where
isolation and disinfection are well carried out there is nearly
ninety-four per cent less cases per outbreak than where pre-
cautions are neglected.

Cause of Success. — " But if one examines the original
returns of the health officers it appears plain why such good
results are obtained. If the first case of the disease coming
to a community is early recognized and isolated the chances
are good that the outbreak will be at once checked. The
chances that such a case will be so recognized in a village are
very much greater than in a city. If, however, the outbreak
is not checked at its very outset, the chances are, even in the
country, that its extension will be very considerable. As a
matter of fact, a great many of the first cases coming to small
communities are promptly recognized and isolated, and as a
result an outbreak is prevented. It is because outbreaks are
in rural communities so often nipped in the bud that the
application of restrictive measures in such communities makes
such a good showing. What is true of measles is true also

158 THE SOURCES AND MODES OF INFECTION

of other diseases, particularly of scarlet fever and diphtheria.
It is success in applying restrictive measures to the first case
that is the principal cause of the apparent efficiency of these
methods. If every appearance of contagious disease in these
rural communities had gained some headway before restric-
tive measures were applied, the showing would not be nearly
so favorable. In fact, in the large cities in Michigan, where
these diseases are always epidemic, restrictive measures, even
of a very rigorous type, have not enabled the health officers
to * stamp them out.' Thus in Detroit it was the custom
for many years to l quarantine' absolutely every house
where there was scarlet fever or diphtheria. No one was
allowed to go out, and the inspectors visited the house twice
a day and furnished provisions for the poor at an expense
of thousands of dollars annually. Yet these diseases were no
more 'stamped out' in Detroit than they have been in
Chicago or other cities where milder methods have prevailed."

Isolation in Common Diseases. — When measles, diph-
theria or scarlet fever appears in a town or institution which
has for some time been free from the disease, the patient ought
to be isolated until it is as certain as medical science can
determine that he is free from infection. At least this should
be attempted. If it should happen, as it sometimes does,
that the diphtheria patient retains virulent bacilli for many
months, or that the scarlet-fever patient has a discharging
ear for a similar period, isolation will probably have to be
abandoned. But isolation is worth trying, for in the majority
of instances safety is secured in a few weeks. Contacts also
should be carefully examined and isolated, or otherwise re-
stricted as to their relations with the community. If after
a reasonable time the disease is " stamped out " the health
officer is to be congratulated. If it is not stamped out he
may well temper the rigor of his restrictive measures.

When a rare disease, as plague, leprosy, or cholera, appears
in Europe or North America, equally stringent measures
should be employed. At present smallpox belongs rather

LIMITATIONS TO THE VALUE OF ISOLATION 159

to this class of rare diseases, and strict isolation of the first
case and careful examination and supervision of contacts is
desirable and useful.

Many Carriers make Isolation Useless. — The effective-
ness of isolation, and the consequent reason for its practice,
varies inversely as the number of carriers and missed cases.
With the enormous number of carriers of pneumococci, it is
entirely useless to enforce isolation of cases of pneumonia.
Influenza belongs to the same class as pneumonia, in which
compulsory isolation is useless. It is probable also that noth-
ing which the health officer can do in the way of isolation
will have any effect on the extent of outbreaks of cerebro-
spinal meningitis. The enforced isolation of typhoid fever,
owing to the number of carriers, will usually prove of little
value. While it is probable that there are few carriers of
measles, yet the long prodromal but extremely infectious
stage renders ineffectual measures of restriction. If isolation
of measles is attempted, little can be hoped for other than a
slight postponement of the age of attack, and no measures
should be adopted which inflict any great hardship. No one
advocates the isolation, in the ordinary meaning of the term,
of pulmonary tuberculosis. It is not attempted in this
disease simply because infected persons are so numerous that
it is impossible. If tuberculosis were as rare as leprosy, strict
isolation would be, and should be, demanded. On the other
hand, there are so few carriers of smallpox that, even with
the mild type of the disease prevailing, strict isolation is often
advisable.

To discover the proportion of carriers to recognized cases,
for each disease, is a matter of the greatest practical impor-
tance. We must have a fairly clear idea of how many un-
recognized human foci of infection there are before we can
determine upon what methods of isolation, if any, are likely
to prove effective. Yet the investigation of carriers has
received comparatively little attention at the hands of
bacteriologists.

160 THE SOURCES AND MODES OF INFECTION

Value of Hospitals. — Hospitals are useful for protecting
the family, for checking outbreaks in institutions, for receiv-
ing cases from lodging houses and hotels, for furnishing better
medical service, and for relieving the overworked housewife
in the families of the poor. It is an unnecessary expense
to provide hospital accommodations for all cases of scarlet
fever and diphtheria, or for ninety per cent or even eighty
per cent. That half or two-thirds of the cases of these
diseases can, for all practical purposes, be equally well cared
for at home, is not unlikely.

Home Isolation. — In home isolation of scarlet fever, diph-
theria and measles, the patient should, for the benefit of the
public, be kept in the house. It does not seem reasonable to
prolong isolation until all possible chance of infection has
ceased. Exactly what the period should be in each disease
should now be a topic for renewed discussion.

Isolation in Providence. — In Providence at present the
period of isolation for scarlet fever is four weeks from the
beginning of the case. Up to 1902 the period was five weeks.
Since the reduction in the period of isolation, the attack rate
has been about 33 per 100,000 living, although one of our
largest epidemic waves occurred during this period, and regis-
tration is certainly far better than formerly. From 1884 to
1901 the attack rate was about 34 per 100,000. In diph-
theria, isolation is maintained for ten days after the disap-
pearance of the membrane. The decrease in the prevalence
of the disease which followed the adoption of less rigorous
isolation has been previously alluded to.

Family. — When the attempt is made to isolate the patient
in the family from the family, in order to protect other mem-
bers, the duration of isolation may well be left to the dis-
cretion of the family. It is the duty of the health officer
to explain that the longer the separation of sick from well
is maintained, the more likely it is to be effective. In scarlet
fever there is no means of determining when the patient is
free from infection. I am in the habit of recommending

LIMITATIONS TO THE VALUE OF ISOLATION 161

separation from the family for six weeks, perhaps a somewhat
shorter time if the case is a mild one, and longer if there is
aural or nasal discharge. In Providence, scarlet-fever cases
are usually sent home from the hospital in four or five weeks,
if free from nose and ear discharge. No regard need be paid
to desquamation, as the experience of English hospitals has
shown that no danger is to be feared from that source.1 It
is absurd to isolate with strictness a diphtheria patient from
the rest of the family unless cultures have shown that the
rest of the family are free from the bacilli. If such home iso-
lation is to be undertaken at all, it should be continued, if
possible, until two or three successive negative cultures have
been obtained.

Hospital. — Diphtheria cases in Providence are usually in
the hospital until two, three or four successive negative cul-
tures from the throat have been secured, the greater precau-
tion being taken when the child is to return to an institution.
Sometimes the patient is sent out while still harboring bacilli.
The duration of isolation of the living cases is 19.65 days,
and the percentage of return cases has been 1.9. In London 2
the period of detention of 6866 cases was 57.5 days, and the
percentage of return cases was 1.2, about 0.5 per cent less
than in Providence, although the period of detention was
almost three times as long. The duration of stay in the Provi-
dence hospital for scarlet fever, surviving cases, is 45.67 days,
and the percentage of return cases is 3.4. In the London
hospitals the period of detention is 64.9 days,3 and the per-
centage of return cases, 1902-04, was 3.22.

1 Metropolitan Asylums Board Report on Return Cases of Scarlet
Fever and Diphtheria, 1902-04, 6. See also my discussion of this subject
in FiskeFund Essay LII, published by R. I. Medical Society, Providence,
1909.

* Metropolitan Asylums Board Report on Return Cases of Scarlet
Fever and Diphtheria, 1901-02, 59, 62.

3 Metropolitan Asylums Board Report on Return Cases of Scarlet
Fever and Diphtheria, 1902-04, 5, 23.

162 THE SOURCES AND MODES OF INFECTION

School. — School children in infected families should be
excluded from school for a liberal period, for this works little
hardship, and the state should make special effort to keep
its schools free from disease. .Yet school exclusion may be,
and often is, carried to excess. In Providence children living
in a family where there is scarlet fever are excluded from
school for four weeks from the beginning of the last case.
In all except the poorer class of houses children living in
other families in the house are permitted to attend school.
In diphtheria, children in the infected family are not allowed
in school .for four weeks, provided that if all the school chil-
dren in the family yield two throat and nose negative cultures,
they may attend school after the warning sign has been
removed (which is ten days after the disappearance of the
exudation). Children in other families in the better class of
houses are admitted to school if one negative culture is ob-
tained from throat and nose. After the expiration of a month,
all children in families where there has been diphtheria are
usually admitted to school whether or not they are carrying
morphologically typical bacilli, and of course quite a number
do go to school while infected. Doubtless these carriers may
at times infect others in school, but even if two negative cul-
tures were required before readmission, some children would
probably still prove infectious. Absolute security is impos-
sible, and the rule should be so framed as to accomplish a
maximum amount of good with a minimum amount of an-
noyance.

Wage Earners. — There is little reason for excluding wage
earners from their work except in a few occupations. Most
of the carriers we cannot restrain, and therefore why penalize
those who have the additional misfortune of sickness in their
families ?

The most dangerous carriers are those who handle milk;
hence milk producers and dealers living in infected families
should be excluded from work. Judging from the number
of reported outbreaks, the danger is probably greater for

LIMITATIONS TO THE VALUE OF ISOLATION 163

typhoid fever than for any other disease. It is probably wise
to regulate the occupation of all typhoid contacts who
handle any kind of food that is eaten raw. Perhaps the same
should be done with diphtheria contacts. Teachers and
nurses may very properly be prevented from following their
usual vocations if they live in infected families. It has been
my custom also to exclude from work car conductors, post-
men, barbers and department-store clerks. I doubt, however,
whether this is always advisable. I am very sure that
laborers, mill operatives and office clerks need not, under
ordinary circumstances, be kept from their business.

Summary. — As regards the employment of isolation for
the prevention of the spread of infectious disease we may
fairly conclude:

1. The danger to be apprehended from a single infected
person is much less than has been supposed.

2. Isolation is of far less value than was believed a few
years ago.

3. The fewer the infected persons in any community or
institution the more likely is isolation to be successful. Iso-
lation in an extensive outbreak rarely accomplishes much.

4. The effectiveness of isolation varies inversely as the
number of missed cases and carriers.

5. Hospitalization in such diseases as scarlet fever and
diphtheria cannot be expected to exterminate them, and the
majority of patients can be as well cared for in their homes.

6. In diphtheria, scarlet fever and measles there is rarely,
and only in certain occupations, any necessity for interfering
with the freedom of the wage earners of the family.

7. The isolation of school children should be more strict
than that of adults, for less hardship results, and there is
more danger in the mingling of children than in the inter-
course of adults.

CHAPTER IV.

INFECTION BY CONTACT.

Most Obvious Mode. — Contact infection is the most
obvious mode of transmission of the infectious diseases. For
the sick to touch the well, and thus infect them, seems to be
the most natural way of accounting for the spread of these
diseases. If contact infection can explain epidemiological
phenomena, there is no occasion for assuming the growth of
pathogenic germs outside of the body, or of infection by fo-
mites or infection by air, or any other similar theory, and
no such theory should be adopted as a working hypothesis
unless pretty strong evidence can be brought to its support.

Venereal Diseases. — Gonorrhea and S3^philis are univer-
sally believed to be transmitted exclusively by contact, and
almost invariably by a special kind of contact. This idea is
so firmly fixed in the minds, of medical men and the laity,
that no matter how many cases occur which it is impossible
to trace to their source, no one ever suggests that these
diseases are air-borne, or that their germs maintain a sapro-
phytic life. No matter how much the patient may protest,
it will still be held that the infection is due to contact, and
in the vast majority of cases to contact involved in the sexual
act. Gonorrhea, particularly, is believed to be almost never
transmitted except by the most direct contact; yet there is
at times as much reason for assuming that the gonococcus
is air-borne, or clings to the walls of rooms and thence infects
their occupants, as there is to assume the same for scarlet
fever. Yet so firmly are we held by tradition that if any of
us should suggest such an origin for gonorrhea it would pro-
voke only a smile, while such sources of scarlet fever are
accepted as well established.

164

INFECTION BY CONTACT 165

Gonorrhea in Babies' Hospital. — The spread of gonococ-
cus infection in institutions for children is very suggestive of
the ways in which other infections are transmitted.

A very interesting account of institutional infection is given
by Holt.1 At the Babies' Hospital in New York, from 1894
to 1898 inclusive, 64 cases of gonococcus vaginitis were admit-
ted, and 16 cases developed in the hospital. In the summer
of 1899, three children suffering from gonococcus vaginitis
were inadvertently admitted to the country branch of the
hospital, and though the danger was realized, and every effort
was made to prevent the extension of the disease, by pro-
viding separate nurses for the infected cases, by washing the
napkins separately, and boiling and disinfecting them, never-
theless 15 girls contracted vaginitis. In 1901 similar trouble
was experienced, and notwithstanding the most vigorous
measures qf isolation and napkin disinfection, 22 cases devel-
oped in the one cottage to which the three original cases
were admitted. The physicians were inclined to look upon
general house infection as the only explanation of the origin
of the cases. In November, 1902, a new hospital building
was occupied for the first time, and it was hoped that it could
be kept free from the disease, but 5 cases were unwittingly
admitted during the first six months, and 29 cases of vagi-
nitis and 8 of gonococcus arthritis developed in the institu-
tion. During the year, 13 cases were admitted, and 66 were
contracted in the hospital. Although the infected cases were
all strictly isolated, on two occasions a child, even in another
part of the hospital, developed the disease. For a time
napkins were discarded and pads used, which could be
burned; separate thermometers, baths and supplies were
required for each child; wash cloths were burned, and tub
baths forbidden. Bed clothing was disinfected with the
greatest care. There was thought to be absolutely no
opportunity for direct contact between child and child.
When diphtheria or scarlet fever persists in an institution in

1 Holt, N. York M. J, [etc.], 1905, LXXXJ, 521.

166 THE SOURCES AND MODES OF INFECTION

such a manner, it is at once attributed to persistent infection
of the building itself, and it is suggested by Holt as a pos-
sibility in this instance also.

Carried by Nurses. — But the gonococcus is an excep-
tionally frail organism, and it is impossible to believe that
persistent infection of a building or its contents can occur.
Actually, in the Babies' Hospital, it was found that the nurse
was the carrier of the germs from child to child, and the two
cases which developed in distant wards, and which were sup-
posed to be perfectly isolated and under the care of different
nurses, were attended by the same night nurse who looked
after the infected cases. Finally, when the strictest disin-
fection had failed to check the disease, it was at last con-
trolled by requiring that the nurses should practice a strict
medical asepsis, and disinfect the hands in every instance
immediately after bathing or changing the napkins of each
child. Here was a disease which continued to spread after
the erratic and persistent manner of scarlet fever and diph-
theria, and which was shown to depend exclusively upon con-
tact infection. Air-borne infection and fomites infection can
have no part in institutional gonococcus infections, for the
gonococcus dies so quickly that such modes of transfer-
ence are impossible. Yet this infection is most persistent
and troublesome in many institutions for the care of young
children.

Contact not always Direct. — Gonorrheal infection is not
only quite common in institutions, but it is often found in
infants and young children in their homes. It is believed
to result usually from the child sleeping with its parents, or
to direct contact with the hands of the mother while washing
or dressing the child. I have also known of cases of the direct
transfer of the disease on instruments in a physician's office.

It has thus been necessary to modify our conception of
the mode of transmission of gonorrhea, and although it is
still conceived of as due exclusively to contact infection, it
is now recognized that the contact need not always be direct

INFECTION BY CONTACT 167

between the sick and the well, but some infected person or
thing may act as intermediary. Yet from what is known of
the weak resistance of the gonococcus, the interval of time
must be brief. The term contact infection as now employed
means just that kind of transference of rather fresh infecting
material from one to another. It does not necessarily imply
actual contact between sick and well, but it does imply that
there are no long intervals of time in which the infective
materials may become dry and inert. The transfer of gonor-
rheal pus from child to child on a syringe, or on the fingers
of the nurse, are examples of contact infection. If the pillow
used by an infected child were put away for a week or two,
and when brought out caused ophthalmia in the infant who
used it, it would be an example of fomites infection. The
distinction between the two types of disease transference,
though not sharply or accurately defined, is a reasonable and
practical one. Thus every one admits that gonorrhea is fre-
quently transferred by indirect contact infection, as it may
be called, but it is never suggested that this disease is spread
by fomites.

Syphilis spread solely by Contact. — While it appears to
be pretty well determined that Treponema pallida is the
cause of syphilis, we have as yet no data as to the cultivation
of this organism or its persistence outside of the body. Its
cultivation outside the body certainly is not easy, and it may
be suspected that its life is short; but it is not impossible
that the virus may retain its virulence for some days or weeks,
as does that of rabies, variola and vaccinia. At present we
have to rely solely on clinical evidence as to these points, and
the universal opinion of this much studied disease is that
the virus develops only in human beings (and some of the
apes), that it is not very persistent, and that it is trans-
mitted solely by contact. No one has ever suggested that
syphilis is an air-borne disease in the ordinary sense, though
one or two cases of droplet infection have been reported.1
1 Buckley, Syphilis in the Innocent, N. Y.; 1894, 176.

168 THE SOURCES AND MODES OF INFECTION

No matter how obscure may be the origin of cases, no one
would attribute them to aerial infection, or to dwelling in an
infected house. Infection during the sexual act is undoubt-
edly the cause of by far the larger proportion of the cases of
this disease, yet in the aggregate a great number of cases
are caused in other ways.

Non-sexual Contact. — Buckley in the work just quoted
shows that probably ten per cent of the initial lesions are
extragenital. Exceedingly infectious lesions at times are
found in the mouth so that kissing is a not unusual mode
of infection. At a party in Philadelphia, where there
were kissing games, 8 persons were infected by a young
man. One of these was a man who received his infection
indirectly on the lips of a girl just previously kissed by the
syphilitic.1 Syphilis is often transmitted by nursing and
the sucking of wounds. Mediate contact by all sorts of
infected articles is common and Buckley has collected many
hundreds of instances. Among the articles named are
cups, glasses, spoons and other eating-utensils, pipes, toilet
articles, underclothing, bathing suits, handkerchiefs, bed-
ding, pins, string, wind instruments of all kinds, glass
blowers' tubes, pencils, coins, nursing-bottles, sponges,
syringes, surgeons' instruments, dentists' tools and barbers'
utensils. Buckley's book is mainly taken up with an
enumeration of these modes of infection. A perusal of
these reports is extremely interesting, for though it is not
claimed that every case narrated was caused in the man-
ner assumed, yet it is evident that all such modes of trans-
mission are possible, and one is greatly impressed by the vast
number of ways in which fresh secretions may be passed from
one to another. While the time during which the various
articles retained their infection is unknown, or at least is not
given in the reports of cases, it is apparent that almost none
of the instances of mediate contact were the result of long
persistent infection. Fomites infection is not an important
1 Shamberg, J. Am. M. Ass., Chicago, 1911, LVII, 783.

INFECTION BY CONTACT 169

factor in the spread of syphilis. Buckley1 says that the
danger from soiled clothing, rags, or second-hand garments
" is infinitely less than is commonly supposed, and relatively
few instances have been recorded, and none of these are very
clear or satisfactory." In the few instances which he gives,
the infective material might well have been very fresh.

Typhoid Fever by Contact. — Of late years a great deal
has been written about the transmission of typhoid fever by
contact from case to case, and this mode of infection appears
to have attained greater prominence in connection with this
than with any other of the common infectious diseases. My
attention was first drawn to the importance of contact infec-
tion in typhoid fever by the vivid description given by Sedg-
wick of an outbreak which he investigated in 1892 in
Bondville, Massachusetts. He says:2 "Children abound; and,
as there are no fences, and because it is the custom, they mingle
freely, playing together and passing from house to house.
The families are of that grade in which food always stands
upon the table ; meals are irregular except for those who must
obey the factory bell. The children play awhile, then visit
the privies, and with unwashed hands finger the food upon
the table. Then they eat awhile and return to play. Or,
changing the order of things, they play in the dirt and eat
and run to the privy, then eat, play, and eat again, and this
in various houses and in various privies. For them, so long
as they are friendly, all things are common, — dirt, dinners
and privies; and, to illustrate exactly how secondary infection
may go on, I may describe in detail one case which I personally
witnessed. A whole family (of six or more) was in one room.
Four of them had the ' fever.' Two of these were children
in the prodromal stage. A table stood by the window covered
with food, prominent among which was a big piece of cake.
It was early September, and a very warm day; but every
window was shut and the odor sickening. Flies innumerable

1 Buckley, Syphilis in the Innocent, New York, 1894, 156.

2 Sedgwick, Rep. Bd. Health Mass., Bost., 1892, 736.

170 THE SOURCES AND MODES OF INFECTION

buzzed about, resting, now on the sick people, now on the
food. A kind-hearted neighbor was tending the baby. By
and by one of the children having the fever withdrew to the
privy, probably suffering with diarrhea, but soon returning,
slouched over to the food, drove away some of the flies, and
fingered the cake listlessly, finally breaking off a piece, but
not eating it. Stirred by this example, another child slid
from his seat in a half-stupid way, moved to the table, and,
taking the same cake in both hands, bit off a piece and swal-
lowed it. The first boy had not washed his hands, and if
the second boy suffered from secondary infection, I could not
wonder at it.

"This was one case; but I have seen so often the table of
food standing hours long in the kitchen, and serving as one
station in the dirty round of lives like these, that it is easy
for me to understand how dirt, diarrhea and dinner too often
get sadly confused. The privies had been obviously in bad
condition, and, from some, filthy streams ran down between
them and the houses. In and around these streams the chil-
dren played. Given any original imported case, the infection
might easily have reached these trickling streams. Children's
fingers might thence carry the germs to the food, and thus
the journey of the germs from one living intestine to another
be completed. Or, again, given in such a community an
imported case and no disinfection, as was the condition here
at first. The importer, while in the early stages, handles with
unclean hands food for others; or the clothing of such a per-
son gets infected and is handled; there need be, then, no diffi-
culty in completing the history. It follows as a matter of
course."

Contact Typhoid in Spanish War. — Probably the report
of Reed, Vaughan and Shakespeare1 did more than anything
else to call attention to the importance of contact infection
in the epidemiology of this disease. This commission found

1 Abst. of Rep. on the Origin and Spread of Typhoid Fever in U. S.
Military Camps during the Spanish War of 1898, Wash,, 1900.

INFECTION BY CONTACT 171

that infected water played little part in the development of
typhoid fever in the camps. They also found that probably
every regiment brought into camp one or more mild unrec-
ognized cases or carriers, and that these were the starting
points of outbreaks. It was shown that the fever was not
evenly distributed through the regiments, but was more or
less localized in companies or squads (p. Ill et seq.).

While they attributed a certain amount of the disease to
carriage by flies and by dust, they considered contact infec-
tion from man to man the most important cause. Of 1608
cases especially studied, and which were accurately located
as to place and time, 35.01 per cent were directly connectible
and 27.79 per cent indirectly connectible attacks; total con-
nectible attacks, 62.80 per cent (p. 184). Owing to the unsat-
isfactory methods, or lack of method, of excreta disposal, the
shoes, clothing and hands of the men, as also the blankets
and tentage, became more or less soiled with excreta, and
infection of the men became easy, and in fact unavoidable.
Men detailed as hospital orderlies were, after they had per-
formed the duty of emptying bedpans, seen to go directly
to their meals without washing their hands, and even to dis-
tribute food to their comrades.

In South Africa. — Similar conclusions were arrived at by
the surgeons who studied typhoid fever, which proved equally
disastrous to the English, in the Boer War, but on the
whole, however, the English, while recognizing the impor-
tance of contact infection, did nol place so much stress
upon it as did the Americans.1

Contact Typhoid in Civil Life. — Outbreaks in civil life
have of late frequently been attributed largely to contact
infection. Winslow2 reported an outbreak in Newport, R. I.,
which probably had its origin in an infected well, but which

1 Col. Lane-Notter, Tr. Epidemiol. Soc., Lond., 1904, XXIII, 149;
J. Roy. Army Med. Corps, Lond., 1905, IV, 587, 693; Tooth, Brit. M.
J., Lond., 1901, I, 642.

* Winslow, Technology Quarterly, 1901, XIV.

172 THE SOURCES AND MODES OF INFECTION

was continued by contact infection. He coined the word
" prosodemic " to describe this extension of the disease from
case to case. Water-borne and milk-borne outbreaks of
typhoid fever usually develop suddenly, a large number of
cases being reported within a short time. That they do not
end so suddenly but, even after the infected water and milk
are eliminated, continue, decreasing gradually, is a phenome-
non which becomes apparent from the examination of the
charts of such outbreaks. This prolongation of the outbreak
is due to the prosedemic infection of Winslow, and has been
noted by Whipple l for water outbreaks, and by Trask2 for
milk outbreaks. During a water-borne outbreak at Lincoln,
England, from 50 to 60 per cent of the cases were due to
personal contact.3

Municipal Outbreaks. — Many local outbreaks have been
believed to be due almost, if not quite, exclusively to contact
infection. Such an outbreak was reported by Weston and
Tarbett in Knoxville,4 Jordan in Winnipeg,5 Magrath in
Springfield,6 Noetel in Beuthen,7 and Weil in Rathsweiler.8
Freeman9 states that the majority of the outbreaks in the
smaller towns of Virginia appear to be due to contact infec-
tion. The authority of Koch has done much to direct atten-
tion to the importance of contact infection in the spread of
typhoid fever in civil life. In a report on a village outbreak
in Trier 10 he says that small country epidemics will usually

1 Whipple, Typhoid Fever, 1908, 209.

1 Trask, U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab. Bull.
No. 41, 38.

3 Pub. Health, Lond., 1905, XVIII, 129.

4 Weston and Tarbett, Am. Pub. Health Ass. Rep., 1907, XXXIII,
Pt. 1, 63.

5 Jordan, Abst. in J. Am. M. Ass., Chicago, 1905, XLIV, 563.
8 Magrath, Am. J. Pub. Hyg., Bost., 1905, I, 467.

7 Noetel, Ztschr. f . Hyg. u. Infectionskrankh., Leipz., 1904, XLVII,211.

8 Weil, Med. News, N. Y., 1904, LXXXIV, 467.

8 Freeman, J. Am. M. Ass., Chicago, 1909, LIII, 1263.
10 Koch, Die Bekampfung des Typhus, Berlin, 1903.

INFECTION BY CONTACT 173

be found to be due to this mode of infection. In the particu-
lar outbreak investigated all the cases were found to be due
to contact, that is, to the direct transfer from person to per-
son, and the source was in most instances unrecognized mild
cases, often in children. He then describes conditions very
much like those reported by Sedgwick. Fecal matter was
observed in the dooryards where it could readily infect the
children playing about, and would certainly be carried indoors
on their shoes.

• Fulton,1 Egbert 2 and others have shown that typhoid fever
prevails far more extensively in the country than in the city.
Formerly, when less was known about the etiology of the
disease, polluted well water was believed to be the chief factor
in its causation, but it now appears that it is far more likely
to depend on contact infection.

In Hospitals. — Typhoid fever frequently develops in hos-
pitals and other institutions, presumably by contact infec-
tion. Occasionally explosive outbreaks due to infected milk
or food are observed, but many institutional outbreaks are
characterized by a slow and irregular development of cases.
The disease chiefly attacks those who are brought in close
contact with the sick, physicians and especially nurses. Fifty
or sixty years ago few cases of hospital infection were re-
corded. It is suggested that this may have been due to the
fact that nurses in those days were usually women well
along in years, and therefore not so susceptible to the
disease, and usually immune. During recent years typhoid
fever among nurses and ward tenders has been quite com-
mon. It is stated that in 1902-6, of an annual aver-
age of 322 nurses in six hospitals, 26 contracted typhoid
fever and 12 of an average of 94 ward tenders contracted
typhoid fever while on duty. It is said that in the Lon-
don hospitals typhoid fever is twenty times as common

1 Fulton, J. Am. M. Ass., Chicago, 1904, XLII, 73.

2 Egbert, Am. Med., Phila., 1905, IX, 649.

3 Joslin and Overlander, Boston M. & S. J., 1907, CLVII, 428.

174 THE SOURCES AND MODES OF INFECTION

among nurses as among women of the same ages in other
occupations.1 Goodall 2 says that during the years 1892-
1899 there were treated in the London hospitals 5913 typhoid
patients, and 100 attendants contracted the disease. In four
hospitals not admitting typhoid fever no cases developed.
Schuder,3 Neufeld,4 McCrae,5 Talayrach,6 Edsall7 and others
have reported an excessive prevalence of typhoid fever among
the attendants on cases of this disease in hospitals, and they
have also noticed numerous instances in which it has devel-
oped among other patients. Nurses and others infect their
hands while caring for typhoid patients, and then without
washing the hands, or after careless washing, infect their
own mouths directly, or by handling their food and drink.
Through the same carelessness they also infect other patients.
Neufeld refers to the transfer of the germs on a thermometer
and in a bath. Edsall has seen a nurse given the double
duty of emptying the bedpans of typhoid cases and pre-
paring special diet, and a patient was observed to empty a
bedpan and then proceed to his dinner without washing his
hands. Nurses also wipe out the mouths of patients with a
bit of gauze on the finger, a procedure which was believed by
Holt to be one of the means of spreading gonorrheal infec-
tion in the Babies' Hospital. By introducing strict cleanli-
ness Edsall was able to stop this transference of the disease.
In Other Institutions. — Sometimes quite marked out-
breaks occur in institutions, due presumably to contact in-
fection. Usually water, milk and other food as sources of
infection can be excluded, and the irregular and slow devel-
opment of the cases, and perhaps the discovery of carriers or

Pub. Health, Lond., 1905, XVIII, 142.
Goodall, Trans. Epidem. Soc., Lond., 1900.

Schuder, Ztschr. f . Hyg. u. Infectionskrankh., Leipz., XXXVIII, 251.
Neufeld, Kolle u. Wassermann, Handbuch [etc.], Jena, 1903, II, 296.
McCrae, Mod. Med. [Osier], Phila. & N. Y., 1907, II, 82.
Talayrach, Arch. d. med. et de pharm. milit., Par., 1903, XLII, 393.
1 Edsall, Am. J. M. Sc., Phila., 1908, n. s., CXXXV, 469.

INFECTION BY CONTACT 175

missed cases, renders the extension of the disease by per-
sonal contact highly probable. In an almshouse in New
Haven,1 37 cases occurred during a period of some months.
Mild cases were concealed and worked in the kitchen, and
it was impossible to teach the inmates cleanly habits. At a
soldiers' home at Lafayette, Indiana, there were 65 cases
due to contact infection, probably chiefly from the physician.2
At the State Hospital for the Insane at Trenton there
occurred between April 8 and August 13, 1907, 80 cases
with 16 deaths, due in all probability to contact infection,
largely in the kitchen and pantry.3 Ravenel4 reports an
outbreak in a boarding house at the University of Wisconsin,
where 41 cases developed, probably as the result of a patient
working in the pantry during the prodromal period of his
illness. Sedgwick5 reports the case of a chambermaid, and
also of a laundress, who contracted typhoid fever from
handling soiled linen. A waitress, also, who ate bread cut
by the other two, developed the disease. Wolcott tells of
the matron of a hospital who caught the disease from han-
dling infected bedding and neglecting to wash the hands.
McCrae 6 reports an instance of infection from careless han-
dling of typhoid cultures in the laboratory, and a similar
case occurred in an insane asylum at Northampton, Mass.7
Dr. Mann tells me that a number of nurses eating at the
same table in the Homeopathic Hospital in Boston con-
tracted typhoid fever, probably from a waitress whose sister
was sick with the disease, and who herself was perhaps a
carrier.

1 Rep. Bd. Health, New Haven, 1905, 6.

2 Hurty, J. Am. M. Ass., Chicago, 1909, LIII, 1263.
1 Rep. St. Bd. Health, New Jersey, 1907, 149.

< Ravenel and Smith, K. W., J. Am. M. Ass., Chicago, 1909, LII,
1635.

6 Sedgwick, J. Mass. Ass. Bds. Health, Bost., 1900, X, 148.

6 McCrae, Mod. Med. [Osier], Phila. & N. Y., 1907, II, 82.

7 Rep. Bd. Health, Mass., Bost., 1899, 762.

176 THE SOURCES AND MODES OF INFECTION

Contact on Shipboard. — Many of the instances of infec-
tion by carriers referred to in Chapter II were almost cer-
tainly the result of contact infection, and many more could
be easily collected. Occasionally small outbreaks occur on
shipboard, where it is possible to exclude all modes of infec-
tion but contact. Thus, 9 cases on an Atlantic liner were
traced to a saloon steward.1 On the United States ship Con-
necticut there were three outbreaks in 1907, almost certainly
due to contact infection.2

In the Family. — During 1908 I noted three small out-
breaks of typhoid fever in Providence, due apparently to con-
tact infection. In one instance there were 10 cases in three
closely related families, in another 8 cases in three such
families, and in the third outbreak there were 7 cases in
two families. Such family outbreaks are noticed in Provi-
dence nearly every year, and they are referred to so often
in current medical literature that it does not seem necessary
to give additional references. Contact outbreaks in hotels
and boarding houses, as well as in various public institutions,
are also frequently reported. Hill in Birmingham, in 1898,
traced 10 per cent of the typhoid-fever cases to contact with
other cases. In Manchester in the same year about 13 per
cent were traced in the same way, and in 1906, 36.8 per cent.
Edsall attributed 27 per cent of 250 cases to contact, and
Forster 117 of 386 cases, and McCrae 68 of 500 cases. It
would be easy to prolong the discussion and show that med-
ical men are everywhere attributing more and more impor-
tance to contact infection in this disease. It is to be noted
that in few of the reported instances of contact infection is
there any direct and conclusive proof that the disease was
caused in the manner alleged. The evidence is by no means
so convincing as to mode of infection as it is in so many
water-borne and milk-borne outbreaks, and from the nature
of things it rarely can be so.

1 Rep. Med. Off. Health, Glasgow, 1907, 147.

2 McDonnold, Mil. Surgeon, Carlisle, Pa., 1908, XXIII, 29.

INFECTION BY CONTACT 177

Evidence of Contact Infection. — But when water and food
as vehicles of infection are excluded, when the disease devel-
ops in persons who are in contact with the sick or with car-
riers, and when the fingers, possibly or perhaps evidently
soiled with excreta, are seen to touch food, eating-utensils,
and sometimes the lips, the probability of the development
of the disease in this manner is so evident that observers
are constantly seeing in contact infection the cause of family,
institutional and community outbreaks of this disease. The
importance now attributed to contact infection is the result
of a careful estimate of probabilities.

Amount of Contact Infection. — In 1908 there were 11,375
deaths from typhoid fever in the registration area of the
United States, which doubtless means considerably over
100,000 cases. The registration area includes about one-half
the population of the country. While a certain amount of
typhoid fever is caused by milk, it must be a very small
fraction of the great total. As will be shown, milk outbreaks
are usually well marked, and probably not a very great deal
of milk-borne infection escapes notice. Again, while in some
cities the larger part of the typhoid fever is due to infected
water, it is not so in most cities, and certainly only a small
part of the total typhoid fever in the country can be traced
to such a source. It is not likely in my own city for instance
that more than 20 per cent of the 4300 cases of typhoid fever
during the past twenty-five years have been due to infected
water or milk; and infected oysters, celery, etc., certainly
play a very unimportant part in the causation of the disease.
As we shall see, it is improbable that air-borne infection
or dust infection is of any moment in civil life. The only
other important source of typhoid fever which has been sug-
gested is fly-borne infection. It will be shown also that
infection by means of flies probably accounts for only a
small part of our typhoid fever. By exclusion, then, we
are led to the conclusion that contact infection is the chief
source of our typhoid fever.

178 THE SOURCES AND MODES OF INFECTION

Why should we not, in the absence of direct evidence as
to other modes of infection, assume at once that contact,
that is, the more or less direct transfer of infective material
from person to person, is the principal source of contagious
disease ? Why neglect the most obvious and direct mode
of transfer in favor of more circuitous paths ? There are
several reasons why contact infection in this disease has been
neglected.

Why Contact Infection has been Neglected. — One reason
has been the undue emphasis laid upon other modes of trans-
mission. Formerly air infection was considered of great
moment, and as this fell into disrepute, undue weight was
attached to water and milk infection, partly, no doubt, because
of the dramatic character of so many outbreaks. Another
reason for doubting the importance of contact infection is
the prevalent notion that typhoid fever is essentially an intes-
tinal disease, and that it must be caused by infected food.
It now appears that typhoid fever is not an intestinal disease,
and in some cases the intestines are not involved and the
feces are free from bacilli during the whole sickness. But
the bacilli are found in the blood stream in the earliest stages,
and it seems to be more likely that they enter the circulation
through the upper part of the alimentary tract than through
the intestines after running the gauntlet of the gastric secre-
tions. Even if it be denied that the tonsils and stomach are
the portals of entry of the typhoid bacilli, it is highly probable
that bacilli taken into the mouth in small masses find their
way to the lower part of the alimentary tract by means of
the small quantities of saliva which are constantly swallowed.
It has certainly been demonstrated that tubercle bacilli are
swallowed and thus reach the intestines.

People are not Cleanly. — Then, too, contact infection
in typhoid fever supposes an intimacy with the most offen-
sive substances, which most persons would vehemently deny.
Nevertheless it appears that the fingers of human beings,
and secondarily everything that the fingers touch, are fre-

INFECTION BY CONTACT 179

quently contaminated with excremental matter. The colon
bacillus is for all practical purposes a good test for the pres-
ence of excrement, and it is somewhat of a shock to learn that
it is found on the hands of five to ten per cent of ordinarily
cleanly people. Winslow * and Hall,2 while studying the dis-
semination of typhoid fever by carriers, recovered fecal
bacteria from the finger nails and hands after the usual pro-
cedures following the use of the closet and toilet. An inspec-
tion of the privies or water-closets in railway stations,
factories, shops and tenement houses shows that they usually
present evidence of contamination with feces and urine, and
in many instances are constantly in a horribly filthy condition.
It is only in the better class of hotels and residences that these
apartments are kept in even an apparently cleanly condition,
and this is only by dint of constant vigilance and frequent
cleansing. There can be no doubt that even very careful
people frequently infect the seat, their fingers, the pull, the
door, etc., and that in a large proportion of privies and water-
closets the users almost certainly infect their fingers with at
least traces of their own or others' excremental matter. Yet
how many persons are there who invariably wash the hands
after the use of a closet ? How many make it a rule never
to put the fingers in the mouth ? Yesterday I saw a workman
carrying a can of beer to his friends. His thumb was im-
mersed a couple of inches in the beverage. Had he washed
his hands after leaving the barroom water-closet ? At a
recent sanitary convention I noticed the colored waiter stick
his finger into a glass which he, however, did not remove, and
which the speaker soon drank from. What was the recent
history of that finger ? Does the fruit peddler wash his hands
after using the tenement privy before he ventures to sort his
fruit ? Do the waitress, the milk peddler, the candy seller,
the Pullman porter, the soda-water clerk, the baker's boy,
the delicatessen man always wash the hands before taking

.. l Winslow, J. Mass. Ass. Bds. Health, Bost., 1903, XIII, 144.
2 Hall, Rep. Med. Off. Health, Bristol, Eng.. 1908, 27.

180 THE SOURCES AND MODES OF INFECTION

up their work? Are the toilets in their places of business
so cleanly that such a precaution is not necessary ? How-
ever shocking it may seem, it is certain that it requires only
a little observation to demonstrate that the path from intes-
tines to mouth is not always a circuitous one.

Contact with Carriers. — Until recently it would have
been argued that contact infection is not an important factor
in typhoid fever because only a small proportion of the cases
can be shown to have been in contact with this disease, and
because the large proportion of cases of this fever are more
or less isolated in their homes because of the sickness. But
we now know that there are vast numbers of mild unrecog-
nized cases, and most important of all, that the number of
carriers is very great. There are doubtless 200,000 cases of
this disease in the United States each year. If only 3 per
cent of these become chronic carriers, and if a carrier remains
such only three years, we should have a carrier population
of 18,000 persons, practically all unknown and taking no pre-
cautions against infecting others. If we add to these the
25 per cent of convalescents, who for some weeks are excret-
ing the bacilli in their urine, it appears that there is a very
respectable army of unrecognized sources of typhoid infection.

At present we have no definite knowledge of the origin of
the larger number of our cases of typhoid fever. In view
of the almost universal careless habits of the people, and
the great number of carriers, why not adopt as a working
hypothesis the most obvious source of infection, infection by
contact ?

Danger from Privies. — This view that contact infection
is the most important factor in the causation of this disease,
is borne out by the observation that the more promptly and
effectively human excrement is disposed of, the less chance
there is for contact infection and the less the disease prevails.
When the disposal is exceptionally bad, as in army camps,
lumber, mining and railroad camps, then this disease is al-
most always very common, In thoroughly sewered and clean

INFECTION BY CONTACT 181

cities, provided the water and milk are not contaminated,
typhoid fever is comparatively rare. The privy vault stores
up fecal matter on the premises and is rarely kept as clean
as a water-closet, and the area around privies is often filthy
also. It has often been noted that the removal of privy
vaults has been followed by a decrease in typhoid fever. In
Providence the disease fell off forty per cent after most of the
privy vaults were abolished. Many other American cities in
which typhoid fever is not maintained by a polluted water
supply, show a steadily decreasing prevalence of this disease
as privy vaults are gotten rid of. In England Pringle1 has
shown that in fourteen towns with middens the typhoid rate
was 0.25 per 1000, while in fourteen water-closet towns it
was 0.19. In Ipswich there was a marked decrease in typhoid
fever following the removal of the middens.2 A similar de-
crease was noted in Oldham, Leicester3 and other cities. On
the other hand more " pail closets " remain in Nottingham
than in most English cities, and to this Boobbyer4 attributes
the fact that typhoid fever has decreased less in that town
than in the other great towns of England. Like reports of
a decrease in the disease following removal of privies come
from such widely separated places as Winnipeg 5 and Mel-
bourne.8

There are also numerous reports to the effect that in the
same town typhoid fever is much more common in houses
without water-closets than in those with them. This may
be due in part to the storing of fecal matter on the premises
of the former, and it may be due also to the fact that houses
with privies are usually of a poorer class, and the inhabitants
less cleanly in their personal habits, or to other reasons; but

1 Pringle, Pub. Health, Lond., 1902-03, XV, 630.

2 Pub. Health, Lond., 1908-09, XXII, 414.

3 Rep. Med. Off. Health, Leicester, Eng., 1908, 29.

4 Boobbyer, Rep. Health of Nottingham, 1908, 53.
6 Rep. Dept. Pub. Health, Winnipeg, 1908, 4.

6 Jamieson, J,, Australas. M; Gaz., Sydney, 1903, XXII, 56.

182 THE SOURCES AND MODES OF INFECTION

these facts of typhoid distribution accord with the view that
the disease is largely spread by contact from person to
person. In Birmingham,1 the incidence of typhoid fever in
" pail " and water-closet houses was as 65 to 43, and there
has been a steady decrease in the disease as the " pail
closets " have been abolished. In Nottingham there was (in
ten years) 1 case in each 37 houses with privies, and 1 in
each 558 water-closet houses; in Salford it was 1 to 20 and
1 to 42 respectively,2 and in Gorton the cases were two
or three times as numerous in privy houses as in water-
closet houses.3 At Leigh,4 typhoid fever was four times
as prevalent among colliery workers as among the rest of the
population, owing, in the opinion of the inspector, to the filthy
mode of excreta disposal in the mines. In Glasgow5 second-
ary cases of the disease developed in 23 per cent of the cases
in houses with privies and in 6 per cent of the cases in houses
with water-closets.

Every one who has had practical experience in sanitary
inspection work knows that privy vaults not only serve as
storehouses for excrement, but their presence encourages its
careless disposal in the yards and on ash heaps. The filthy
condition of the ground about the houses, leading to contami-
nation of feet and hands of children, as reported by Sedgwick
and by Koch, is rarely noticed on premises provided with
water-closets. It may be urged that the excess of typhoid
fever in privy towns is due to infection by flies rather than
infection by contact, and this may be true to some extent;
but if flies were the chief factor we should scarcely expect
a great difference between water-closet and privy houses in
the same town. So also the infection of miners, as noticed

1 Rep. Health of Birmingham, 1906, 49, and 1908, 49.
a Pringle, Pub. Health, Lond., 1902-03, XV, 630.
» Martin, Pub. Health, Lond., 1904-05, XVII, 709.
4 Sweeting, Rep. Med. Off. Local Gov. Bd., Lond., 1907-08, XXXVII,
57.

6 Rep. Med. Off. Health, Glasgow, 1902, 97.

INFECTION BY CONTACT 183

by Sweeting, cannot be due to flies. In Providence after the
removal of privies the decrease in typhoid fever was as great
in winter, when there were no flies, as during the fly season.

Amebic Dysentery. — There is no reason why amebic
dysentery may not be transmitted by personal contact, but
if, as is1 generally believed, the ameba is quite widely dis-
tributed in the soil, it is likely that in countries where the
disease is endemic, man is the least common source of infec-
tion. Nevertheless cases do develop from contact with t)ther
cases. Lemoine1 has reported such infections in France, in
one instance at least, probably due to the use of the same
bedpan, douche, etc. The infecting case was a carrier of
ten years' standing. Allan2 writes that in Charlotte, N. C.,
he observed four cases of amebic dysentery which he believed
were due to contact with chronic carriers.

Bacillary Dysentery. — Bacillary dysentery, like typhoid
fever, is a great scourge of military life, and outbreaks of a
serious character are not rare in institutions such as hospitals
for the insane. The bacillus is found in the feces of the
patients and also in convalescents. There is every reason
for believing that this disease, like typhoid fever, is frequently
spread by more or less direct contact.3 Conradi 4 described
a village outbreak near Metz where 70 cases occurred during
a period of three months. Several carriers were found in
infected families, and conditions favoring contact infection
were noted, very similar to those observed in the outbreaks
of typhoid fever studied by Koch in Trier. Dodge5 reports
the case of a laboratory worker who got some of a culture
of the dysentery bacillus hi his eye; the tears ran profusely
and were swallowed, and in twenty-four hours an attack of

1 Lemoine, Bull, et mem. Soc. med. de hop. de Par., 1908, 3 s., XXV,
640.

2 Allan, Med. Rec., N. Y., 1910, LXXVII, 63.

3 Shiga, Mod. Med. [Osier], Phila. & N. Y., 1907, II, 781; Scheube,
Diseases of Warm Climates, 2d Ed., Phila., 466.

4 Festschrift von Robert Koch, 1903, 555.

5 Dodge, Am, Pub, Health Ass. Rep., 1905, XXX, 310.

184 THE SOURCES AND MODES OF INFECTION

dysentery developed. An outbreak of 49 cases, developing in
the characteristic slow irregular manner of contact outbreaks,
was reported in the Connecticut Hospital for the Insane.1
Kruse2 reports a similar outbreak in Germany, and two in
Holland. Epidemic dysentery is quite common in hospitals
for the insane, and its prevalence is explained as due chiefly
to contact infection, for the carelessness of many patients
and the impossibility of controlling their habits offer every
opportunity for this sort of diffusion.3

Cholera and Contact. — If typhoid fever and dysentery
are spread by means of contact infection, we should expect
that cholera would be. That less is written about it is due
probably to the fact that in recent years less attention has
been, paid to cholera than to the more common diseases.
The literature relating to typhoid fever has of late been many
times more voluminous than that relating to cholera. Never-
theless most writers attribute some importance to contact
infection in this disease, and some consider it an etiological
factor of very great importance. The filthy conditions of vil-
lage life described by Sedgwick in America and Koch in Ger-
many as giving rise to typhoid fever, are far surpassed in
danger by the habits of vast numbers of the poorer people
who dwell in cholera-infested countries. The opportunities
for the direct transference of fecal matter from person to
person are far greater in Asiatic countries than they are with
us, and a number of writers have emphasized the part played
by personal contagion in this disease. Gotschlich 4 refers to

1 Rep. St. Bd. Health, Connect., 1903, 234.

2 Kruse, Deutsche med. Wchnschr., 1901, XXVII, 370, 386.

8 See Reports of Commissioners in Lunacy (Eng.) since 1903; also
Heuser, Deutsche med. Wchnschr., 1909, XXXV, 1694; Ryder, Boston
M. & S. J., 1909, CLXI, 681; Haenisch, Ztschr. f. Hyg. u. Infections-
krankh., Leipz., 1908, LX, 245; Mott, Tr. Epidem. Soc., Lond.,
1901-02, and Arch. Neurol. Path. Lab., Lond. Co. Asyl., Lond., 1903,.
II, 735; Prior, Australas. M. Cong. Tr., Victoria, 1909, III, 383.

4 Gotschlich, Kolle and Wassermann, Handbuch [etc.], Jena, 1904,
IV, 108.

INFECTION BY CONTACT 185

this factor, and a number of our officers in the Philippines
have dwelt upon its importance. Woodruff l says that while
infected water played some part in the great epidemic in
Manila, the filthy habits of the people were the chief cause
of the extension of the disease. He speaks of an outbreak
of eighty cases in a provincial town due to food prepared in
Manila by a caterer who soon died of cholera. Heiser 2 con-
siders that the handling of foodstuffs and of the leaves in
which the betel nut is wrapped, by dealers and prospective
purchasers, is one of the chief ways in which cholera is
spread, and McLaughlin 3 considers the " carrier " with
filthy habits the greatest source of danger. He says that
contact infection of visitors in the houses of the sick is a
common means of disseminating the disease. Schumburg 4
reported several small outbreaks of cholera near Hamburg
caused by contact infection. Shakespeare 5 reports several
instances in which cholera was apparently caused by han-
dling soiled linen, some of which might perhaps be considered
rather as examples of fomites infection. Macrae6 reports
the infection of hospital nurses with cholera in a manner
similar to the infection of nurses with typhoid fever.
Heiser 7 says that in the Philippines, physicians and nurses
who had been trained to aseptic methods did not contract
cholera but untrained attendants frequently did through
contact with patients.

Diarrhea. — Newsholme 8 believes that much infantile
diarrhea is due to direct contact infection in the home, and
he suggests that it is brought about by " sucking infective

Woodruff, J. Am. M. Ass., Chicago, 1905, XLV, 1160.
Heiser, J. Am. M. Ass., Chicago, 1907, XLVIII, 856.
McLaughlin, J. Am. M. Ass., Chicago, 1909, LII, 1153.
Schumburg, Ztschr. f. arztl. Fortbild., Jena, 1905, II, 567.
Shakespeare, Rep. on Cholera in Europe and India, U. S. Gov.
Printing Office, 1890, 81.

Macrae, Indian M. Gaz., 1909, XLIV, 361.
Heiser, Bull. State Bd. Health, N. Y., Sept., 1911.
8 Newsholme, J. Hyg., Cambridge, 1906, VI, 139.

186 THE SOURCES AND MODES OF INFECTION

matter from dirty fingers, from dummy teats and in other
ways." Naish ! and Niven 2 are largely in accord with this
view. Sandilands,3 in an extended article, cites others
as reporting hospital and laboratory infection as well as
evidence from house and neighborhood outbreaks. From
his own study of multiple cases in houses, and of local dis-
tribution, he is convinced that contact plays an important
part in the causation of this disease. Doubtless the most
complete epidemiological study of summer diarrhea has
been made by Peters,4 and it should be read by every one
interested in the subject. Flies are considered as possible
carriers of infection, and while no evidence against this
theory is found, much positive evidence is presented to show
that contact infection is an important factor in the causa-
tion of this disease.

Hook Worm Disease. — The European type of the disease
is caused by Anchylostoma duodenalis, while the American
type is due to a slightly different species, Uncinaria ameri-
cana.5 The symptoms are caused by the growth of the
worms, which are about half an inch long, in the intestine,
where they fasten themselves to the intestinal wall, and
not only suck blood themselves, but also cause considerable
free bleeding from the wounds which they make. The
eggs, which are laid in the intestine, do not there develop,
owing to lack of oxygen, but hatch soon after the feces are
voided, and the young worms then pass through several
stages of growth in moist earth or mud.

Until recently it was believed that human infection resulted
from drinking water containing the young worms, or by get-

1 Naish, Pub. Health, Lond., 1909-10, XXIII, 168.

2 Niven, Proc. Roy. Soc. Med., Lond., 1909-10, III, Epidemiol.
Sect., 131.

3 Sandilands, Proc. Roy. Soc. Med., Lond., 1909-10, III, Epidemiol.
Sect., 95.

4 Peters, J. Hyg., Cambridge, 1910, X, 602.

6 Stiles, U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab. Bull.
No. 10, 1903.

INFECTION BY CONTACT 187

ting them onto fingers or into food, and so into the mouth.
It was first suggested by Looss in 1898 that infection might
take place through the skin, and he believed that he had
himself become infected in that way. In 1901 he proved the
correctness of his surmises by experiments. Later Grassi,
Fieri and Noe placed a few drops of water containing worms
upon the skin, but only one of the three became infected.
In 1902 Looss successfully repeated his experiments. It had
meantime been noticed by Bentley and Boycott and Haldane
that dermatitis was apt to be found in regions where the
disease prevailed, and it was suspected that it might be due
to the passage of the worms through the skin. Smith1 showed
that by placing infected earth on the arm, not only was infec-
tion caused, but there resulted a dermatitis at the site of the
application. Ashford, whose careful study of the disease
in Porto Rico, and whose brilliant success in curing its vic-
tims give his opinion great weight, believes that the skin
is by far the most important avenue by which the worms
infect the body.2 Uncinariasis is, then, par excellence, a dis-
ease due to contact infection.

Fortunately the disease is usually easily curable under
medical treatment, and the freeing of individuals from worms
by this treatment is necessarily an important part of the
prevention of the disease. It is evident from the mode of
infection that the pollution of the soil with human feces is
the principal factor in the spread of the disease. Properly
constructed privies or water-closets, with the ultimate dis-
posal of the fecal matter by deep burial, or some other means
to prevent the pollution of the upper layers of the soil, are
the essentials of prophylaxis.

Contact Infection less Easy in Some Diseases than in
Others. — Gonorrhea is a disease in which the infecting secre-
tion is not likely to be much handled, and when it occurs

1 Smith, C. A., J. Am. M. Ass., Chicago, 1905, XLV, 1142.

2 Ashford, Rep. of Commission on Study and Treatment of
"Anemia" in Porto Rico, 1904, 37.

188 _ THE SOURCES AND MODES OF INFECTION

in young children the secretion is often received on a diaper.
Yet it has been shown in the preceding pages that infantile
gonorrhea is spread exclusively by contact infection. Al-
though the children themselves take little part in the spread
of the disease in hospitals where they are isolated from one
another, and although nurses and physicians have been shown
the danger of carrying the disease, and have been instructed
to take the greatest precautions, yet this disease may be
maintained for months in an institution solely by means of
infection borne on thermometers, syringes, etc., but particu-
larly on the hands of attendants.

In typhoid fever the bacilli are contained in the feces and
urine, which even careless people are supposed to avoid touch-
ing. Nevertheless it appears that the fingers of careful people,
and even of trained nurses, are infected in this manner, and
that transfer to the mouth with the subsequent development
of typhoid fever results. There is much evidence that this
mode of transference is an important, if not the most
important, factor in the spread of this disease.

Contact with Saliva. — If contact infection is important
in such diseases as gonorrhea, typhoid fever, dysentery and
cholera, in which the infecting material is not constantly at
hand, and is usually strenuously avoided, how much more
important must this mode of transference be in diseases in
which the specific germs are found in the secretions of the
nose and mouth or in the sputum.

Danger from Fingers. — Probably the chief vehicle for
the conveyance of nasal and oral secretion from one to another
is the fingers. If one takes the trouble to watch for a short
time his neighbors, or even himself, unless he has been par-
ticularly trained in such matters, he will be surprised to note
the number of times that the fingers go to the mouth and
the nose. Not only is the saliva made use of for a great
variety of purposes, and numberless articles are for one reason
or another placed in the mouth, but for no reason whatever,
and all unconsciously, the fingers are with great frequency

INFECTION BY CONTACT 189

raised to the lips or the nose. Who can doubt that if the
salivary glands secreted indigo the fingers would contin-
ually be stained a deep blue, and who can doubt that if
the nasal and oral secretions contain the germs of disease
these germs will be almost as constantly found upon the
fingers? All successful commerce is reciprocal, and in this
universal trade in human saliva the fingers not only bring
foreign secretions to the mouth of their owner, but there ex-
changing them for his own, distribute the latter to everything
that the hand touches. This happens not once but scores
and hundreds of times during the day's round of the indi-
vidual. The cook spreads his saliva on the muffins and rolls,
the waitress infects the glasses and spoons, the moistened
fingers of the peddler arrange his fruit, the thumb of the milk-
man is in his measure, the reader moistens the pages of his
book, the conductor his transfer tickets, the " lady " the
fingers of her glove. Every one is busily engaged in this dis-
tribution of saliva, so that the end of each day finds this
secretion freely distributed on the doors, window sills, fur-
niture and playthings in the home, the straps of trolley cars,
the rails and counter and desks of shops and public buildings,
and indeed upon everything that the hands of man touch.
What avails it if the pathogens do die quickly? A fresh
supply is furnished each day.

Drinking Cups. — Another important vehicle of transfer
must be the common drinking cup. Davison1 estimated that
there were as many as 20,000 epithelial cells on a drinking
glass that had been in use in a school for nine days, which
well illustrates the amount of infection which may be thus
carried. Cars, steamboats, hotels, schools, offices, factories,
theaters, churches, all provide a common vessel from which
large numbers of persons drink, thus furnishing an almost
ideal method by which perfectly fresh saliva may be trans-
ferred from one to another. Hundreds of thousands of
persons must be each day in this manner exchanging the

1 Davieon, Tech. World Mag., Chicago, 1908, IX, 623.

190 THE SOURCES AND MODES OF INFECTION

secretions of the mouth. When traveling in the steam cars I
have noted the shocked expression on the face of passengers
as a fashionably dressed woman was seen to allow her pug
dog to drink from the common glass, — not a pleasant thing,
of course, but infinitely less dangerous than for the woman to
drink from it. She might have tuberculosis, or carry diphthe-
ria bacilli, or perhaps even have mucous patches on her lips.

Other Kinds of Contact. — Besides the moistening of the
fingers with saliva and the use of the common drinking cup,
the mouth is put to numberless improper uses which may
result in the spread of infection. It is used to hold pins,
string, pencils, paper and money. The lips are used to
moisten the pencil, to point the thread for the needle, to wet
postage stamps and envelopes. Children "swap" apples,
cake and lollipops, while men exchange their pipes and women
hatpins. Sometimes the mother is seen "cleansing" the
face of her child with her saliva-moistened handkerchief,
and perhaps the visitor is shortly after invited to kiss the
little one.

Children have no instinct of cleanliness, and their faces,
hands, toys, clothing and everything that they touch must
of necessity be continually daubed with the secretions of the
nose and mouth. It is well known that between the ages of
two and eight years' children are more susceptible to scarlet
fever, diphtheria, measles and whooping cough than at other
ages, and it may be that one reason for this is the great
opportunity that is afforded by their habits at these ages
for the transfer of the secretions. Infants do not of course
mingle freely with one another, and older children do not
come in such close contact in their play, and they also begin
to have a little idea of cleanliness.

Contact Dangerous because of Missed Cases. — A little
observation and reflection will show that the ways are num-
berless in which fresh secretion of nose and mouth is passed
from person to person. Enough has been written to demon-
strate that the opportunities for contact infection are sufli-

INFECTION BY CONTACT 191

ciently numerous to account for the spread of the contagious
diseases without invoking any other mechanism. The
chief objection to this view is that while it is true that there
is much interchange of secretions, it is between well per-
sons, and not between the sick and the well. That this
objection is untenable is amply demonstrated by the evi-
dence presented in Chapter II. There can no longer be the
slightest doubt that there are large numbers of mild and un-
recognized cases of infectious disease mingling freely with
the public, and that in some diseases, and perhaps in most,
there are also larger numbers of perfectly well carriers who
also are unknown. These unrecognized foci are clearly
numerous enough to cause, by their contact with others, the
recognized cases. The transfer of the disease by fairly
direct means is so obvious and easy that there is no neces-
sity for invoking the agency of other and more circuitous
modes of dissemination. Indeed it is sometimes said that
the arguments here presented prove too much, and that if
carriers were as numerous, and contact with them as fre-
quent, as is here alleged, none could escape. But, as was
stated in Chapter II and on page 153, infection does not
take place so readily as is generally believed. This is
demonstrated clinically, and the reasons are apparent.
There are often long intervals in which carriers are not
eliminating the pathogenic organism, and the saliva may
not contain the germs, even when they are in the throat
and nose. Germs are not evenly distributed through
saliva, sputum or feces, and the particle transferred may
be free from them. The .infective material is often small in
amount and spread in a thin layer, and the contained organ-
isms very speedily die. Lastly, small numbers of pathogens
are often, perhaps usually, destroyed by the body. We
must believe that usually continued or somewhat massive
infection is necessary to cause disease, but that neverthe-
less sometimes a single infection with a very small number
of germs suffices. The number of unrecognized foci of

192 THE SOURCES AND MODES OF INFECTION

infection in human beings, the opportunities for contact
infection, and the natural obstacles to successful infection,
appear admirably to explain many otherwise inexplicable
phenomena of epidemiology.

Bacteria on the Hands. — Reference has already been
made to this on page 179. Neumann,1 by finding fecal bacilli
on water-closet fixtures, showed how the hands are infected,
and he also showed 2 how by another transfer on the hands
they may infect bread, rolls, fruit, butter and milk. Colon
bacilli have also been found on 3 of 12 roller towels.3 Pus-
forming organisms have their natural habitat on the skin,
and the work of the surgeons has shown the tremendous
importance of contact infection, which they have now
learned most successfully to avoid.

Danger from the Shuttle. — As weavers habitually thread
the shuttle by sucking the yarn through the eye, mouth
bacteria may in this way be transferred from one to another.
This has been alleged as a mode of infection in tuberculosis,
and indeed Brown4 claims to have traced three cases to this.

Bacteria on Cups, Pencils. — Vincenzi found diphtheria
bacilli in the holy water in a church font. Kinyoun6 found
them in 2 of 85 swabbings from the woodwork and drink-
ing glass of a railway car. They have also been found on
glasses and cups by Albert and Boyd,7 Perrow,8 and in
Chicago.9 Forbes, in Rochester, found them upon a drinking
glass which was believed to have been the cause of an out-
break. This has been frequently referred to, and though I

1 Arch. f. Hyg., Miinchen. u. Leipz., LIX, 174.

2 Neumann, Deutsche med. Wchnschr., Leipz. u.Berl., 1910, XXXVI,
2046.

Bull. Mass. State Bd. Health, Nov., 1911.

Brown, Med. Officer, 1911, IV, 27.

Vincenzi, Semaine me"d., 1898.

Kinyoun, Med. News, N. Y., 1905, LXXXVII, 193.

Albert and Boyd, Bull. State Bd. Health, la., Oct.-Dec., 1911, 37.

8 Perrow, Rep. Health Dept., Lynchburg, Va., Sept., 1911.

9 Bull. Dept. Health, Chicago, April 22, 1911.

INFECTION BY CONTACT 193

i

have not been able to find Forbes' original article, Dr. Goler,
the present health officer of Rochester, informs me that the
facts are as reported. Williams1 recovered diphtheria bacilli
from pencils moistened by the lips of children sick with the
disease. The observations referred to in the chapter on
fomites show that the germs of this disease are rarely found
on fomites and then only on objects that have been recently
and grossly infected.

Tubercle Bacilli in Mouth. — Not only are tubercle bacilli
found in enormous masses in the true sputum, but they
are often present in the saliva as well. Neild and Dunkley2
found them in saliva from the tip of the tongue in 29 of 50
cases of pulmonary tuberculosis. Park found them in the
saliva of 10 of 15 cases,3 and refers to Moller recovering them
in 3 of 20 cases. Le Noir and Camus4 found virulent
tubercle bacilli in the nose as well as from the mouth of
tuberculous cases.

Pathogenic Bacteria on the Hands. — With the present
habits of human beings these germs must be constantly trans-
ferred to the fingers, and to a lesser degree to everything that
the fingers touch. Graziani5 found tubercle bacilli on the
hands of 4 of 8 tuberculous patients, and on 3 of them 3
hours after washing with soap and water. He also obtained
the bacilli from the hands of 4 out of 6 non-tuberculous
patients. After shaking hands with tuberculous patients he
was able several times to recover the bacilli from his own
hands. Baldwin6 found bacilli on the hands of patients in
the Adirondack Sanatorium. They did not use handker-
chiefs. Of 10 patients seen in private practice 8 had tubercle

Williams, Scientific Bull. 2, 1895, Health Dept., N. Y. City, 14.
Neild and Dunkley, Lancet, Lond., 1909, I, 1096.
Park, Sixth Internat. Cong, on Tuberc., Wash., 1908, I, 157.
Le Noir and Camus, Comp. rend. Soc. de biol., Par., 1908, LXV, 464.
Graziani, Ann d' Ig. Sper., XV, 709, referred to by Rosenau, Sixth
Internat. Cong, on Tuberc., Wash., 1908, I, 28.
6 Baldwin, Tr. Am. Climat. Ass., 1898, XIV, 202.

UNIVERSITY OF CAUFORNIA

194 THE SOURCES AND MODES OF INFECTION

bacilli on their hands. The other two were exceptionally
careful. Preisich and Schiitz J found tubercle bacilli on the
hands of children in a children's hospital at Budapest. Of
66 examinations of the dirt from under the finger nails 14
were positive under the microscope, but owing to the death
of the animals from sepsis their virulence was not demon-
strated. Of the 14 positive cases, 11 had tuberculosis, or
were associated with it, while of the 52 negative cases only
5 were associated with tuberculosis. Dieudonne*2 by inocu-
lation demonstrated tubercle bacilli on the hands of 2 of
15 children. In this connection may be mentioned the
experiment of Schumburg,3 in which he rubbed an ose of a
culture of bacteria on 'his hands, and recovered the germs
after 15 and 16 handshakes. Ostermann,4 on the other hand,
does not consider contact infection of much importance in
tuberculosis. While he recovered tubercle bacilli from the
hands of 7 of 14 phthisical patients and from 1 attendant,
he obtained them only 4 times from 42 children living in
tuberculous families and 2 times from the floors occupied by
these families. He does not find that bacteria are transferred
from hand to hand as readily as have other observers. He
also made a few cage experiments with guinea pigs to show
that infection by contact is less effective than infection by
air.

It scarcely needed the experiments of Annett at Liverpool 5
and Higgins6 at Birmingham to show that virulent tubercle
bacilli may be found in the sputum on sidewalks, or those of
Dixon7 to demonstrate that they may be swept up on the
skirts of ladies' dresses.

1 Preisich and Schiitz, Berl. klin. Wchnschr., 1902, XXXIX, 466.

2 Dieudonne, Munch, med. Wchnschr., 1901, XLVIII, 1439.
1 Schumburg, Ztschr. f. arztl. Fortbild., Jena, 1905, II, 567.

4 Ostermann, Ztschr. f . Hyg. u. Infectionskrankh., Leipz., 1908, LX, 375.

6 Annett, Thompson Yates Laboratory Rep., 1901-02, IV, Pt. 2,
359.

« Higgins, Pub. Health, Lond., 1909-1910, XXIII, 100.

7 Letter from Dr. Samuel A. Dixon to author.

INFECTION BY CONTACT 195

In Communion Cups. — Anders1 found tubercle bacilli in
the dregs from communion cups in a Philadelphia church,
and Moller2 from the communion cup of a sanatorium
chapel. Davison3 found them on a glass used for some
weeks in a high school. He also demonstrated pneumococci.
Klein4 obtained a positive tuberculous reaction once after
inoculation of the swabbings of six telephones, although on
twelve telephones on another occasion5 he was unable to
find either diphtheria or tubercle bacilli, and Rickards6 was
unable to find either bacillus on twenty-four mouthpieces
of lung-testing machines in Boston. Hubs7 found tubercle
bacilli on napkin rings in the sanatorium at Stadtwald, but
did not find them on the spirometer which was in daily use.
Price8 used some water in which sanatorium dishes had been
washed to inoculate eight guinea pigs, all of which died of
tuberculosis. Washings from dishes which had been first
washed in the ordinary way showed no bacilli.

Contact Chief Mode of Infection. — Since it is true that
pathogenic organisms begin to die or lose their virulence
when thrown off from the body, we are forced to conclude
that the closer the relationship in time and space with the
bearers of the germs, the greater the chance of infection.
Now that the number of unknown foci of infection and the
opportunities for direct transfer of secretions have been
demonstrated, the deduction is certainly permissible that
contact infection is more important than the more indirect
infection by fomites or by air.

We are also compelled by inductive methods to place the
greatest emphasis upon contact infection. In the chapter on

Anders, J. Am. M. Ass., Chicago, 1897, XXIX, 789.

Moller, Deutsche med. Wchnschr., 1905, XXXI, 548.

Davison, Tech. World Mag., Chicago, 1908, IX, 623.

Klein, Lancet, Lond., 1908, I, 1862.

Ref. J. Am. M. Ass., 1905, XLIV, 1866.

Rickards, Rep. Bd. Health, Boston, 1906, 91.

Hubs, Ztschr. f. Tuberk. u. Heilstattenw., Leipz., 1906, IX, 396.

Price, Sixth Internat. Cong, on Tuberc., Wash., 1908, I, 167.

196 THE SOURCES AND MODES OF INFECTION

infection by fomites, and particularly in that on infection by
air, much evidence is presented to show that infection usually
does not take place unless contact is fairly close. The views
here presented as to the great importance of contact infec-
tion, and the comparatively slight importance of air infection
and fomites infection, gradually developed after considera-
tion of much experimental and epidemiological evidence.
The two facts which more than all others have contributed
to these views have been the restriction of scarlet fever and
diphtheria to single families in the same house, and the suc-
cess of certain hospitals in preventing cross infection, when
contact infection is strictly guarded against.

Disease Spreads in Dwellings only by Contact. — The
rarity with which scarlet fever and diphtheria pass from
one family to another in the same house has already been
alluded to, but it is worth considering again. Of 4306 other
families in Providence living in the same house with a case
of scarlet fever, only 295, or 6.8 per cent, were later invaded
by the disease. This includes all families, whether careful
or careless, and whether or not the initial case was removed
to the hospital; but as less than 10 per cent of the cases
have been removed to the hospital during the 23 years
covered by the figures, hospital isolation can have been of
little moment. This amount of infection is surprisingly small,
and of itself indicates very strongly that close and intimate
contact is usually necessary for the extension of this disease.
A further study of the facts indicates this still more clearly.
During the last 5 years, scarlet fever has extended to 118
of 1888 other families, or 6.3 per cent. In 54, or 22 per cent,
of the cases infection occurred during the first two weeks, of
which 37 were during the first week. Most of these cases
during the first two weeks probably derived their infection
from the first family before the disease was recognized. Only
24 second families developed their infection after the end of
the second week and while the warning sign was on the house.
This is the time when the patients are desquamating and

INFECTION BY CONTACT^ 197

when the disease is popularly believed to be most infectious.
These 24 cases are 1.3 per cent of the 1888 other families in
the infected houses. As in a number of these cases there was
known to be considerable communication between the fami-
lies, it seems to be certain that in an ordinary tenement
house scarlet fever is not readily carried from one family to
another, and that effective isolation is comparatively easy.
In fact, everything goes to show that two or more families
may live in the same house, using hallways, doors, and even
water-closets in common, without scarlet fever extending
from one to the other. All that is necessary is that there
shall be absolutely no visiting between the families, and that
the children shall never meet in play. I do not know that
similar data have been collected elsewhere, but Cameron1
has shown by his study of " return" cases of this disease
that in many instances the infecting case may remain in the
family for some time, providing there is no direct contact
with the well persons.

Diphtheria is as little likely to extend from family to family
without direct contact as is scarlet fever. During the last
twenty-one years, in Providence, of 3667 other families living
in a house with diphtheria, only 263, or 7.2 per cent, were
invaded by the disease. During the five years 1904 to 1908,
of 1648 other families, 114, or 6.9 per cent, have acquired the
disease. Of these, only 38, or 2.3 per cent, were attacked after
the first week of the initial sickness and before the warning
sign was removed. As was shown above, many of these infec-
tions are due to the fact that the first case is not recognized
and hence the different families in the house continue to visit
one another; or in some instances to the fact that persons in
the second family carry the bacilli for a time without being
sick. Probably a large part of the infections during the first
week take place before the disease is recognized, so that it
seems very probable that less than 2 and perhaps less than

1 Cameron, Rep. on Return Cases of Scarlet Fever and Diphtheria,
Lond., 1901-02, 98.

198 THE SOURCES AND MODES OF INFECTION

1 per cent of families living in a house with a case of diph-
theria contract the disease while the warning sign is on the
house. I have given particular attention to the cases of diph-
theria arising in other families in the house while the warning
sign remained on the apartment where the disease first ap-
peared, and almost always there is known to be direct inter-
course between the families. Contact infection, then, seems
to be necessary for the transfer of diphtheria from one family
to another living in different apartments of the same house.

Disease spreads in Hospitals only by Contact. — Another
strong reason for this view that contact infection is the chief
mode of extension for the common contagious diseases is the
success of certain hospitals in preventing cross infection by
minimizing in every way the opportunities for infection by
contact. The Parisian hospitals, particularly the Pasteur
Hospital, have been leaders in demonstrating the success of
this method. In most hospitals for contagious diseases great
effort is made to prevent cross infection by separating the
different diseases to a considerable distance, since the theory
prevails that air-borne infection is the chief mode of spread.
This view so dominates hospital practice that numberless
ways escape notice by which disease is spread by very direct
contact, with the result that in most hospitals cases of cross
infection are continually developing.

French Hospitals. — The French seem to have been the
first to appreciate the importance of guarding against contact
infection, and an interesting account of the efforts made is
given by Grancher,1 who was one of the pioneers. Strict
methods of "antisepsie medicale," as Grancher calls it, were
introduced, and it was found possible to care for patients
with different diseases even in a common ward with far less
danger of transfer than before. He employed wire screens
around the beds to impress upon the nurses the necessity for
guarding against infection by contact. During the years

1 Grancher, Cong. Internal, de med., XIIL, C.-r. Par., 1900, Sec.
de meU de 1'enf ance, 478.

INFECTION BY CONTACT 199

1890 to 1900, 6541 patients were treated in Grancher's wards,
and diphtheria was introduced 43 times, but only once did
a case develop in the wards. Scarlet fever was introduced
19 times, and 7 cases developed in the wards. Less suc-
cess was obtained in isolating measles, but infections were
reduced two-thirds. Grancher was satisfied that even this
disease was spread by contact. He believed that in clean
wards infection is rarely air-borne; it is " objective," not
" atmospheric." In several other French hospitals the impor-
tance of preventing contact infection has been realized. In
some of them isolation is made easier by placing the patients
in separate rooms or cubicles; in some, partitions are placed
between the beds. These partitions may be full height, or
may stop short of the floor, and reach only just above the
head. Sometimes screens only are used. In 1897 the Min-
turn Hospital in New York, a small private hospital, was
built to provide isolation in separate rooms for cases of scarlet
fever and diphtheria. But the most notable example of the
new method is the Pasteur Hospital in Paris, opened in 1900.
The success of these various hospitals in preventing cross
infection is correlated, not with the amount of isolation as
ordinarily understood, but with the care with which aseptic
measures are carried out by the attendants. The cubicles,
partitions and screens certainly cannot prevent infection if
the nurses without taking proper precautions pass from one
case to another. This is shown by the constant development
of cross infection in the ordinary hospital. It was in the
Pasteur Hospital that the principles of medical asepsis were
first fully appreciated and carried out in a practical manner.
The hospital consists of two pavilions with about fifty beds
each, designed ostensibly, one pavilion for scarlet fever and
one for diphtheria; but, as is shown below, all sorts of cases
are admitted. For the care of all these " other diseases "
and mixed and doubtful cases, twelve single rooms are pro-
vided in each ward, all opening into a common corridor, the
doors of which are usually left open. The same nurses look

200 THE SOURCES AND MODES OF INFECTION

after different diseases, often in adjoining rooms. The disci-
pline of the hospital attempts to secure an entire avoidance
of contact infection, and is remarkably successful. Yet the
procedures are all very simple. The nurse always sterilizes
her hands after waiting on the patient; she wears a gown,
which is kept in the room, when anything is done which
would be likely to infect her clothing. Nothing goes into the
room except what is sterile, and nothing comes out without
being at once sterilized. To train nurses sufficiently to take
charge of a ward, two years are necessary, but the nurses
whom I saw there were carrying out these details with the
same precision and unconsciousness which are shown by the
bacteriologist in his laboratory.

English Hospitals. — The principles of aseptic nursing as
applied to contagious diseases have been employed in a num-
ber of places in England, as, for instance, in the North
Eastern, South Western and London Fever hospitals in
London, at Walthamstow and Manchester, and doubtless in
other places. The methods adopted in the Monsall Hospital
at Manchester, and described by Gordon,1 are as follows:

"The patient's bed in the general ward is surrounded with
a screen covered with sheets, which are kept constantly wet
with a weak solution of some disinfectant. The main pur-
pose of this screen is to serve as a label, and to remind the
nurses that certain precautions must be taken for the patient
behind it. At the same time, I think the wet sheets may pos-
sibly arrest infective particles that are projected against them
in the acts of coughing and sneezing. The only other requi-
sites are two glass shelves fixed on the wall behind the bed,
and a locker or portable cupboard made of metal, with an
enameled surface, which can easily be disinfected and kept
clean.

" The precautions to be taken by the nurses in attending
patients behind this screen or ' barrier ' are printed on a
card fixed to the screen, and are as follows:

1 Gordon, Rep. on Health of Manchester, 1908, 154

INFECTION BY CONTACT 201

Rules in Manchester. — " Precautions to be Observed in
the Nursing of Barriered Cases.

"(l) Rubber gloves are to be worn by the Nurse for all
manipulations connected with the case, including the han-
dling of clothes. The gloves are to stand in a bowl of 1 in 400
Izal solution.

" (2) The following utensils are to be marked and kept on
the glass shelves or in the locker provided:

Spatula
Nozzles
Clinical Thermometer

To be kept completely
immersed in a 1 to
400 Izal solution.

At least two bowls.

All feeding utensils (plates, spoons, forks, etc.).

"(3) A plentiful supply of wet swabs, with a bowl contain-
ing Izal solution to receive these when used, is to be kept on
the locker. Handkerchiefs or muslin squares are not to be
employed.

"(4) No toys or books that have once been used inside
the barrier are to be taken outside it except to be destroyed.

" (5) In every case a square of jaconette is to be placed on
the pillow slip, and over this a piece of muslin; the latter is
to be renewed whenever soiled.

"(6) An overall is to be worn by the Nurse whenever
either the patient or the clothes are handled. This is to be
kept inside the barrier."

The success of the methods described in the preceding
pages is well shown by the published figures. At the
Pasteur Hospital from October 1, 1900, to April 19, 1903,
the following cases were received:

Diphtheria 443 Scarlet fever 92

Sore throat 166 Erysipelas 163

Smallpox 524 Phlegmon of tonsil 20

Chicken pox 55 Other diseases 219

Measles.. 126 Mothers with infants. . 192

202 THE SOURCES AND MODES OF INFECTION

Such a combination furnishes a remarkably fertile field for
cross infections. During the next year about 750 cases were
admitted.1

During this whole period the only cases which developed
in the hospital were 5 of smallpox, 2 of erysipelas and 1 of
diphtheria. Dr. Loiseau writes me that since 1904 the per-
centage of cross infections has been less than 0.1 per cent.
At the Minturn Hospital in New York there has been no in-
stance of infection in the hospital. At the Monsall Hospital
in Manchester in 1908 a large number of persons admitted
with mistaken diagnosis were cared for in the wards by the
methods described without contracting the disease, and a
number of cases of mixed infection were cared for in the open
wards without infecting others, except in one instance when,
owing to a mistake in orders, " barrier " isolation was not
promptly instituted. A letter just received from Dr. Arnold,
who has succeeded Dr. Gordon at Monsall, states that in
order that the patient may see and be seen, and to avoid
the dampness of the wet sheet, he has substituted a tape
which as effectually prevents the mythical aerial flight of the
germs, or, to speak literally, equally well reminds the nurse
that she must be clean.

Is Tuberculosis Air-borne ? — It is assumed that tuber-
culosis, as it occurs in human beings, is usually an air-borne
disease, and as will be shown on another page, there is more
reason for such an assumption concerning this than concern-
ing most diseases. Yet there is in certain quarters a growing
tendency to attribute to contact infection more importance
than formerly. Yet it must be confessed that most writers
on this disease lay no emphasis on contact infection, and
some scarcely mention it. Cornet,2 in speaking of kissing,
the most direct means of contact infection, says that it is
incredible that tuberculosis should be transmitted in this
way, for the saliva is ordinarily germ free, and the germs if

1 Martin, Bull, med., Par., 1904, XVIII, 251.

2 Cornet, NothnagePs Encyclopedia, Tuberculosis, 187.

INFECTION BY CONTACT 203

present would not be carried to the lungs. Cornet's first
contention is certainly not correct, for Neild and Dunkley, as
before mentioned, found tubercle bacilli on the tip of the
tongue of phthisical patients examined during intervals of
freedom from cough. They also report cases of lupus, due in
all probability to inoculation with saliva, and refer to others
reported by Wild.1

There has been little experimental work to determine the
part played by contact infection in tuberculosis. The work
of Bartel and Spieler2 indicates that guinea pigs exposed
under natural conditions to contact with the members of a
tuberculous family more often develop the disease than do
animals exposed merely to the air of the room, while the
experiments of Packard, though inconclusive, indicate little
difference.

Infection by Alimentary Tract. — There seems to be evi-
dence that the tonsils may be the seat of infection in many
cases of tuberculosis. Harbitz 3 has very forcibly called atten-
tion to the probability that tubercle bacilli frequently gain
access to the body through the tonsils. He examined the
tonsils and lymphatic glands of a large number of children,
and found latent tuberculosis in many of the tonsils, and
latent bacilli in the cervical glands in 17 instances, much more
often than in the mesenteric glands. He thinks that a more
careful study of the tonsils and lymph nodes would reveal
much latent tuberculous infection, and he thinks that a con-
siderable amount of infection takes place through the tonsils.
Harbitz mentions Grawitz,4 Aufrecht5 and Beckmann as sup-
porting this view. Ravenel and Reichel 6 review much litera-
,ture, and refer to Wood's experimental infection of swine

1 Wild, Brit. M. J., 1899, II, 1353.

2 Referred to in the chapter on infection by air.

3 Harbitz, J. Infect. Dis., Chicago, 1905, II, 143, and especially 198.

4 Grawitz, Deutsche med. Wchnschr., 1901, XXVII, 711.

6 Aufrecht, Verhandl. d. deutsch. path. Gesellsch., Berl., 1901-02,
IV, 65.

0 Ravenel and Reichel, J. Med. Research, Bost., 1908, XVIII, 1.

204 THE SOURCES AND MODES OF INFECTION

through the tonsils, the course of the bacilli being apparently
through the submaxillary and cervical glands. Benome l
caused infection in animals through the mouth and pharynx.
Bandelier2 finds primary tuberculosis of the tonsils not so rare
as is generally believed, but he does not consider the tonsils
as a frequent starting point of phthisis. Mohler and Ravenel
from experiments and observations consider the mouth as
a frequent site of infection in the tuberculosis of hogs.

Contact Infection in Tuberculosis. — The nose also may
be the seat of infection. Cornet,3 by applying infective ma-
terial by means of a feather to the nasal mucous membrane
of guinea pigs, was able to produce disease of the nose and
submaxillary glands. Renshaw4 was able in the same way
to infect seven of eight animals. As tubercle bacilli are nu-
merous upon the hands of consumptives and upon various
articles used by them, it is evident that fresh bacilli must be
frequently carried to the mouth and nose of persons near by,
and may either infect directly through the mouth, nose and
pharynx, or may be swallowed and enter the circulation
through the lower part of the alimentary canal. The only
question is, How frequently does this happen? As was stated
above, it is the opinion of many that it is a very common
mode of infection. Moore5 is of the opinion that in cat-
tle tuberculosis is spread chiefly by the animals licking
one another, and by their eating and drinking from the
same vessels. Bartel6 believes that infection by pharynx,
stomach and intestines is more common than has been sup-
posed, in which view he supports Weichselbaum,7 Volland,8

1 Benome, Ref. J. Am. M. Ass., Chicago, 1907, XLIX, 888.

2 Bandelier, Beit. z. Klin. d. Tuberk., Wiirzb., 1906, VI, 1.

3 Cornet, Nothnagel's Encyclopedia, Tuberculosis, 154.

4 Renshaw, J. Path, and Bacteriol., Lond., 1901, VII, 142.
6 Moore, Conference of Sanitary Officials, N. Y., 1907, 37.

6 Bartel, Sixth Internat. Cong, on Tuberc., Wash., 1908, I, 95.

7 Weichselbaum, Festsch. VI Konf. Internat. Tuberk., Wien u.
Leipz., 1907.

8 Volland, Berl. klin. Wchnschr., 1899, XXXVI, 1031.

INFECTION BY CONTACT 205

Kavacs1 and Preisich and Schu'tz.2 Among others who at-
tribute much importance to contact infection may be men-
tioned Wassermann,3 Calmette and Landouzy.4

Certainly the opportunities for the direct transfer of fresh
moist infective material in the home of a phthisical patient
must be very great, while the chance of the infective material
becoming dried, pulverized and, while still virulent, being car-
ried to the pulmonary alveoli, must be comparatively small.
Unless there is some good reason to think otherwise, one
would naturally attribute to contact infection the chief role
in the extension of this disease, at least in the family. The
only objection is offered by the pathologists, many if not
most of whom affirm that the evidence points to direct infec-
tion of the lungs by the inspired air. There are, however,
many able experimenters who think otherwise, and who main-
tain that tubercle bacilli may enter the body at various points
and reach the lungs through the lymph channels. It is impos-
sible for the writer properly to weigh pathological evidence,
but that the question is still sub judice must be admitted.
Under the circumstances it seems to be wise to assume
as a working hypothesis that contact infection is a factor of
great importance in the causation of human tuberculosis. It
is certainly essential to guard against such infection in every
way, and from a person who does thus conduct himself in a
cleanly manner at all times, diffusion of the disease through
the air would be impossible. Undue emphasis laid upon the
invisible and therefore terrifying infection in the air has done
more than anything else to develop the unfortunate phthisi-
phobia which so often renders miserable the life of the tuber-
culous, and seriously interferes with rational measures for the
restriction of the disease.

1 Kavacs, Zeiglers Beitrage zur. Path. Anat., 1906, XL.

2 Preisich and Schiitz, Berl. klin. Wschnschr., 1902, XXXIX, 466.

3 Wassermann, Berl. klin. Wchnschr., 1908, Nr. 48.

, * Calmette and Landouzy, Sixth Internat. Cong, on Tuberc., Wash.,
1908, I, 110.

206 THE SOURCES AND MODES OF INFECTION

Importance of Contact Infection. — I have sometimes been
told that I lay too much emphasis on contact infection, but
if it is the principal way in which disease spreads, too much
emphasis cannot be placed upon it, and it seems to me that
the evidence is that it is the chief mode of infection. Even
if it is not so important as is here alleged, every one must
admit that it is of considerable importance, yet until recently
very little attention has been paid to it. If contact infection
is the chief mode of extension of the contagious diseases, then
defense against them becomes more largely a personal affair
than we have been taught. We do not have to rely exclu-
sively on the municipality for our protection, awaiting forced
isolation, hospital construction, disinfection and the like, but
can largely protect ourselves by keeping our fingers out of
our mouths, and also everything else except what belongs
there. It may not be possible to prevent all contagious sick-
ness in this way, but some can be avoided, and perhaps most
of it. Contact infection is avoided by personal cleanliness,
and personal cleanliness is demanded by decency, good man-
ners and refinement, as well as by hygiene. It is not much
trouble to be decently clean, and it is not very expensive.
It is a serious mistake to build an expensive isolation hos-
pital that does not check disease, or to construct a filter when
it is not needed; but no harm would be done if the views here
presented should be found to be erroneous and people should
learn to wash their hands and keep their fingers out of their
mouths to no purpose so far as disease prevention is concerned.
It seems certain that much can be done to prevent the spread
of disease in the family and in hospitals, schools and institu-
tions, if only personal cleanliness be insisted upon. Yet such
cleanliness or medical asepsis is sadly neglected by physicians
as well as by nurses. Rather strong preaching is needed
when, as was referred to on another page, a typhoid nurse is
detailed for the double duty of washing bedpans and preparing
food. It is certainly necessary to insist somewhat strenuously
on reform when such occurrences as the following take place.

INFECTION BY CONTACT 207

Disregard of Contact Infection. — Thus, at one of the finest
hospitals in this country, with separate wards for scarlet fever
and diphtheria, a considerable number of cases have arisen
in the general wards. The germs were supposed to be air-
borne, as it was said there was no other possible avenue of
infection. When I saw the head nurse lick her finger to
facilitate turning the bedside charts of diphtheria patients, I
suspected that the principles of medical asepsis had not been
entirely mastered. Called to see a case of scarlet fever in a
well-to-do family, I found the door of the sick-room carefully
hung with a sheet to keep the infection from the other chil-
dren. After examining the throat with a spatula I handed
the latter to the mother. She took it into the hall and put
it on an upholstered sofa, and with her saliva-infected hands
opened the door of an adjoining room. The attending physi-
cian meanwhile sat on the bed and handled the patient, an
entirely unnecessary proceeding at that time, and except for
the example set him, would have forgotten to wash his hands
before leaving. A certain hospital determined to copy in one
of its wards the cubicle system of the French, but had so
missed its essential features that I found doctors and nurses
going from cubicle to cubicle feeling the pulse, smoothing the
bedclothes, and handling dishes without even stopping to
wash their hands. Meanwhile the screen was supposed to
prevent the microbes from passing from bed to bed, and we
all carefully wore gowns and caps so that the wicked little
germ might not jump into our hair and then jump off again
onto the next patient. In another fine hospital for conta-
gious diseases, where great stress is laid upon ample space
between different diseases so as to prevent cross infection,
the superintendent was observed freely to touch articles about
the ward, and handle the patients, and then go to the public
office without even washing his hands. Such incidents could
be multiplied indefinitely. The superintendent of another hos-
pital invited another visitor and myself to eat ice cream from
the same spoon as himself, which spoon was then replaced

208 THE SOURCES AND MODES OF INFECTION

in the freezer which was to supply the wards. I was most
of all impressed with the fact that at the International Con-
gress on Tuberculosis in 1908 a large number of the readers
of papers moistened their fingers with their tongue when turn-
ing the pages, and in each of the sections only one drinking
glass was provided for all the speakers; and this continued
without protest for a day or two. If the most distinguished
investigators and health officials of the world, gathered to
study the most important contagious disease, show no appre-
ciation whatever of the importance of contact infection^ it
is certainly tune for some one to be emphatic.

Personal Prophylaxis. — The discovery that disease germs
are rarely able to maintain themselves outside of the body
clouded the hopes of those who expected by municipal house
cleaning to " stamp out the zymotic diseases," and the later
discovery of numerous missed cases and carriers has shown
that isolation of the sick controls infection far less than was
believed. These somewhat discouraging facts are to a large
degree offset by what has been learned of the modes of
infection. Formerly air infection was chiefly feared; now it
appears that contact infection is of prime importance. For-
merly dependence was placed upon the state to isolate and
disinfect; now it appears that the individual can protect him-
self, and as easily protect others if he chance to be infected.
It is usually comparatively simple so to live as not to allow
the secretions of others to come in contact with one's own
mucous surfaces, and it is easy, and should be considered
immoral, to allow one's own secretions to be so placed that
they may infect another. Personal cleanliness is less expen-
sive than municipal cleanliness, and is within the reach of all.

Need of Education. — When one notes the utter disregard
of medical asepsis shown in our hospitals and medical con-
gresses, one is apt to feel that the education of the public in
habits of personal cleanliness will take a very long time; but
some encouragement is felt when it is remembered that it is
not so very long ago that excrement was commonly thrown

INFECTION BY CONTACT 209

into the street and garbage was tossed under the tables of
the great. It is only a few years since our sidewalks were
flecked with saliva, but now nineteen persons out of twenty
are ashamed to be caught spitting on the pavement. Perhaps
we may all soon learn to stop distributing our secretions so
freely among our friends.

Education in School. — As the avoidance of contact infec-
tion is chiefly a personal matter, the present need is for educa-
tion. First of all, the teachers in the medical schools and the
staffs of hospitals must learn to appreciate the importance of
this mode of infection. It is not unreasonable to hope that
in a few years the schools and the hospitals will place as
much emphasis on medical asepsis as they now do on surgical
asepsis, and it is to these centers that we must look for the
education of physicians, health officers and nurses. To edu-
cate the general public is a more difficult matter. Some years
ago Dr. Theobald Smith called my attention to the desirabil-
ity of teaching school children something about the require-
ments of personal cleanliness, and since then I have each year
distributed to each school child the following " dont's "

REMEMBER THESE THINGS.

Do not spit if you can help it. Never spit on a slate, floor,
or sidewalk.

Do not put the fingers into the mouth.

Do not pick the nose or wipe the nose on the hand or sleeve.

Do not wet the finger in the mouth when turning the leaves
of books.

Do not put pencils into the mouth or wet them with the
lips.

Do not put money into the mouth.

Do not put pins into the mouth.

Do not put anything into the mouth except food and drink.

Do not swap apple cores, candy, chewing gum, half-eaten
food, whistles or bean blowers, or anything that is put into
the mouth.

210 THE SOURCES AND MODES OF INFECTION

Never cough or sneeze in a person's face. Turn your face
to one side.

Keep your face and hands clean; wash the hands with
soap and water before each meal.

PROVIDENCE, May, 1901.

An explanatory circular is sent to the teachers, and of late
a short account of the sanitary reasons for personal clean-
liness has been distributed to children above the primary
grade. Large printed copies of the " don'ts " have been
framed and hung in the schoolhouses.

Much kindergarten work is of such a nature as to inculcate
rather than discourage cleanly habits. Children work in
common in moist clay and sand, use the same " gifts " and
toys and are brought into close contact in the games.
Miss Bessie M. Scholfield, late supervisor of these schools
in Providence, succeeded, without any undue expense or
violent change of methods, in employing kindergarten work
as a means of teaching some of the principles of personal
hygiene.

Municipality should encourage Cleanliness. — Besides
efforts that are directly educational the municipality can do
much indirectly to encourage habits of personal cleanliness
and to prevent the distribution of the secretions and excre-
tions of the body. The common drinking cup is a most
efficient means of such distribution, and it should be abol-
ished in all schools and other public institutions. This has
been done in many places, and individual cups or specially
designed drinking fountains substituted. The example is
now being followed by some railways, factories and shops.
The states of Kansas, Michigan and Mississippi were the
first to forbid the use of a common drinking glass on rail-
ways, and the Lackawanna and some other roads, and the
Pullman Company advertise as an attraction free cups of
paraffin paper for each passenger.1 Since the above was

1 J. of Outdoor Life, 1909, VI, 371.

INFECTION BY CONTACT 211

written a dozen other states have forbidden the use of the
common glass. Hundreds of churches have adopted the
individual communion cup. In the first edition, I said
that " the roller towel should go the way of the common
drinking cup." Since then Massachusetts and Connecticut
and the cities of New York and Chicago have enacted
statutes or adopted rules forbidding the use of a common
towel in public places.

People are more likely to keep clean if it is easy to do
so. Hence the establishment of public baths may be con-
sidered a real sanitary measure. While compulsion can
have little share in the campaign for cleanliness, certain
prohibitions are entirely reasonable and feasible. Thus
ordinances against spitting on the sidewalks and the floors
of public places have done much to teach people to take
proper care of their secretions. Reference has already
been made to evidence that the abolition of privy vaults
results in a decrease in typhoid fever. Privy vaults cer-
tainly encourage the improper disposal of excreta and
general uncleanly habits. A good ^wftge system and the
removal of vaults and cesspools do much to prevent con-
tact infection, at least in the fecal-borne diseases.

CHAPTER V.

INFECTION BY FOMITES.

Definition of Term. — As was shown in the preceding
chapter, it seems very probable that contagious disease may
often be caused by the quite direct transference of the germs
from one person to another on such objects as cups, pencils,
pipes, the fingers, etc. This mode of transference should
properly be considered a form of contact infection. The term
contact infection, as commonly used at the present time, does
not necessarily imply the immediate touching of two persons,
but it does imply the comparatively direct transference of
quite fresh material from one to another. Although almost
any object may in this manner be the bearer of infection, it
would not ordinarily be considered as fomites. By fomites
are usually meant infected objects which retain the infection
for some time. A toy used by a diphtheria patient and sent
to a distant town and there giving rise to the disease, the
dress of a scarlet-fever patient put away for weeks or months
and brought out only to cause another case, a library book
carrying the infection of smallpox from one household to
another, blankets loaded with typhoid bacilli in South Africa
transferring infection to England, infected hides from Asia
causing anthrax in Philadelphia, blank cartridges as the bear-
ers of tetanus germs, and the various objects in a room lately
occupied by a case of any contagious disease giving rise to
the same affection in newcomers, would all be recognized as
fomites. The cup which carries the moist saliva from one
school child to another, the borrowed pencil which transfers the
fresh syphilitic virus from lip to lip, and the urine-moistened
closet seat which infects the fingers and then the mouth of
the next user, are not thought of as fomites but as the neces-

212

INFECTION BY FOMITES 213

sary media for that intimate mode of disease transference
which is coming to be called contact infection. This dis-
tinction between the two classes of bearers of infection is
somewhat arbitrary, and not very definite, but is eminently
practical. In this book, by infection by fomites is meant a
transference of infecting material on objects under such con-
ditions that considerable time elapses, days at least, usually
weeks, sometimes months.

Yellow Fever and Fomites. — If one takes up the older
text-books on yellow fever it will be found that fomites were
considered the most important means in the extension of this
disease. The invasion of cities and countries was usually
attributed to this mode of carriage. This was the general
view up to, and indeed after, the discovery of the role played
by the mosquito, and numerous instances of such transference
are given. Thus the federal inspectors1 attributed the out-
break at Brunswick, Ga., to ballast brought from Cuba. The
disease was supposed to have been carried from New Orleans
to Havana by means of second-hand oyster buckets.2 Lice-
aga3 gives instances of the transmission of yellow fever by
a shipload of grain from New Orleans, by cloth spread out
on the grass to dry, by general merchandise, by bagging, by
clothing, and by ballast. Horlbeck4 says that two persons
at Key West contracted yellow fever from sleeping on a mat-
tress that was brought from Cuba.

Never so Transmitted. — In all these instances the evi-
dence is the same : a locality has long been free from yellow
fever, something is imported from an infected place and the
disease develops. What could be clearer? The proofs that
it is a fomites-borne disease were far more numerous and
stronger for yellow fever than for almost any other disease.

1 Rep. Surg. Gen. U. S. Mar. Hosp. Serv., 1893, II, 33.
a Report on Shipment of Merchandise, U. S. Mar. Hosp. Serv.,
Special Report, 1899, 9.

1 Liceaga, Am. Pub. Health Ass. Rep., 1898, XXIV, 122.
4 Horlbeck, Am. Pub. Health Ass. Rep., 1897, XXIII, 436.

214 THE SOURCES AND MODES OF INFECTION

Yet we now know that yellow fever never was, nor could be,
transmitted in any such way. Such a mistake, a mistake which
cost millions upon millions because of the needless interrup-
tion of commerce, and disinfection, should make us careful
how on similar, but weaker, evidence we attribute importance
to fomites as a means of infection in other diseases, and should
lead us to inquire what proof there is that the long persist-
ence of infection on things is a weighty factor in the trans-
mission of disease.

Smallpox. — It does not require much search in medical
literature to find numerous instances of the alleged trans-
mission of disease by fomites. Recent text-books and jour-
nals are full of them. Welch and Schamberg1 state that
smallpox was brought to Philadelphia on cotton from the
South, but the only reasons for thinking so were that there
was much smallpox in the cotton region and none in Phila-
delphia, and that the patient handled cotton. Not long since
the health officer of a western city reported that the principal
source of smallpox in that city was lumber, his assumption
being based on the facts that there was much of the disease
in the lumber camps, that the rough lumber was well fitted
to carry contagion, and that in most families the first person
attacked was engaged in some sort of woodworking. The
above are fair samples of the kind of evidence on which the
theory of fomites infection rests. Smith2 reports that a man
from Paris died in London of what was probably malignant
smallpox. Two people who afterwards slept in the same bed,
on different days, developed the disease, as did the girl who
sorted at the laundry the soiled linen from this hotel.

Scarlet Fever. — A recent writer in Public Health3 reports
six instances of house infection giving rise to scarlet fever,
in one case nine months after the first patient was sick. The

1 Welch and Schamberg, The Acute Infectious Diseases, Phila., 1905,
160.

' Smith, Pub. Health, Lond., 1901-02, XIV, 211.
1 Trotter, Pub. Health, Lond., 1906-07, XIX, 745.

INFECTION BY FOM1TES 215

only evidence was the recurrence of the disease in the house.
Welch and Schamberg1 quote from others reports of fomites
infection in this disease. Boeck states that the hair of a
scarlet-fever patient caused the disease twenty years after.
Another physician caught the disease from a coat which he
wore while attending a case a year and a half before. The
health officer of Detroit2 reports two cases due to infection
from books which had been used by a patient some months
before. Wende3 states that quilts used by scarlet-fever
patients in August, and put away without disinfection,
caused the disease in November. The reports of the state
board of health of Michigan4 give instances of the per-
sistence of the scarlet-fever virus for years in houses, letters,
books, etc.

Diphtheria. — Buckley5 quoted from the Newton, Vic-
toria, Health Report an instance where a cornet used by a
diphtheria patient was put away for four years, and was then
found by some children, who contracted the disease from it.
He gives another instance where the disease recurred in the
house after a period of two years. At a time when Manila
was absolutely free from diphtheria, an American child who
had been there over a year received some Christmas presents
from St. Louis, and was taken sick a few days later. Chris-
tian6 writes of the transmission of diphtheria on carpenters'
tools which were sent from one shop to another.

Cholera. — A number of instances of the transmission of
cholera by soiled clothing are given in the Report of the
Marine Hospital Service for 1893. 7 In nearly every instance
the clothing was brought from foreign countries. Because

1 Welch and Schamberg, The Acute Infectious Diseases, Phila., 1905,
344.

2 Rep. Bd. of Health, Detroit, for year ending June 30, 1903, 11.
8 Buffalo San. Bull, Nov. 30, 1908.

« Rep. St. Bd. Health, Mich., 1906, 134; 1907, 133.

• Pub. Health, Lond., 1906-07, XIX, 296.

« Bull. N. Y. St. Board of Health, June, 1907, 5.

' Rep. Surg. Gen. U, S. Mar. Hosp. Serv., 1893, Vol. II, 353.

216 THE SOURCES AND MODES OF INFECTION

a company of soldiers had suffered from typhoid fever for
two years, and the disease ceased on disinfecting the bar-
racks, the outbreak was believed to have been due to room
infection.1

Before weighing the value of this evidence it is perhaps
worth while to consider some of the things which are most
often alleged to serve as fomites.

Infected Clothing. — From the time when the priest was
directed how to detect leprosy in woolen and linen2 to the
present, clothing has been considered an important vehicle
of infection. It is not to be doubted that disease germs may
be carried on clothing. If clothing is soiled with a consider-
able quantity of infected saliva, feces, urine or pus, and if
while fresh, say within a few hours or days, it is brought in
contact with susceptible persons, disease may result. If it is
folded and put away in the dark, especially if it be in a damp
place, it may remain infectious perhaps for months. But as
bacteria as a rule die rapidly, and as there must be a sequence
of gross infection, favorable conditions for survival, and con-
tact with susceptible persons, it does not seem likely that
disease is often caused in this way.

Many instances are recorded, usually only of possible, rarely
of probable, transference of disease by clothing. Of yellow
fever more than of any other disease has this been alleged, yet
we now know that such transmission of this disease is impos-
sible. Many writers also report the spread of bubonic plague
by means of clothing, but, as will be seen, the evidence is that
plague is only under very exceptional circumstances carried
by fomites.

Typhus not carried in Clothing. — When typhus fever
appeared in New York in 1892, from sixty to seventy-five
officers of the health department were, according to Doty,3

1 Cited by Germane, Ztschr. f. Hyg. u. Infectionskrankh., Leipz.,
1897, XXIV, 404.

2 Leviticus, Chapter xiii, verses 47-59.

3 Doty, Med. News, N. Y,, 1905, LXXXVI, 730.

INFECTION BY FOMITES 217

more or less in contact with the cases, often in intimate
contact; gowns were not used, and the officials went freely
between the patients and their own homes, and about their
other business, yet no case of this disease could be traced to
fomites infection.

Butler l speaks of a parlor maid in a contagious-disease
hospital who for six months had been in daily contact with
nurses coming directly from scarlet-fever cases without
change of clothes, but she did not contract the disease until
exposed to an incipient case in the person of a nurse, when
she promptly developed scarlet fever. This is only one
illustration of many of the failure of supposedly infected
clothing to infect.

Physicians rarely carry Disease. — In scarlet fever and
diphtheria physicians are constantly passing from the sick
to the well. Some of them take great precautions to avoid
carrying the disease in their clothes, but the majority take
few or none, — or at least did not until very recently.
Yet records of their transmitting disease are extremely
rare. For many years I was on the lookout for this mode
of transference, and only once or twice found any evidence
that the physician was at fault. In such cases how much
greater is the chance that the physician carries infection on
his hands than in his clothes. Barlow 2 found that of 500
cases of scarlet fever in only 4 had the physician previously
been treating the disease and these 4 cases were apparently
traced to other sources.

Infection by Clothing is Rare. — If the gross infection of
clothing is only in rare instances the cause of disease, how
little must be attributed to slight infection ! A few droplets
of tuberculous saliva, a slight smear of moisture from the
lips of a diphtheria patient, will soon lose their virulence after
exposure to light and air. But usually the visitor in a sick-
room will escape all infection except the hypothetical floating

1 Butler, Proc. Roy. Soc. Med., Lond., 1908, 1, Epidemiol. Sect., 225.

2 Barlow, Med. Officer, 1910, III, 319.

218 THE SOURCES AND MODES OF INFECTION

particles. The danger from this floating matter will be shown
in another chapter to be a negligible quantity. It is the com-
mon practice for physicians, nurses and visitors in a room
occupied by a contagious case to wear a gown and cap. For
physicians and nurses who are to move or lift the patient, or
otherwise come into intimate contact with him, the gown is
a reasonable protection against possible gross contamination.
For the careless visitor too it may be useful. The cap is a
useless frivolity. It is amusing to see how religiously it is
worn, while no protection is given the feet, though Denny
and Nyhen1 have shown that infection by means of the shoes
is bacteriologically possible, while the aerial infection of the
hair has never been demonstrated and is highly improbable.
More than once have I seen a physician don his cap to keep
the germs from flying into his hair, and pass freely from one
diphtheria patient to another, inspecting the throat, feeling
the pulse or smoothing the bedclothes, and occasionally
stroking his own chin or scratching his head, all without
washing his hands. And physicians and nurses will continue
to do such things as long as they fancy that air infection is
all they have to fear, and forget that there is such a thing as
contact infection. It has for some years been my custom
not to wear a gown when called in consultation to cases of
contagious disease, or when visiting hospital wards, unless I
am to do much work about the patient. Under the condi-
tions of a casual visit, even in smallpox, I do not wear a
gown. It is not necessary to touch anything except with the
hands, and these should be carefully washed before leaving.
I have never carried infection to my home, or elsewhere so
far as known, and I believe that there is no chance of my
doing so.

Laundries and Disease. — If clothing can carry infection,

sickness ought to make its appearance in laundries, to which

enormous quantities of clothing go even while the infecting

material is still fresh. Attempts have been made to show

1 J. Maes. Ass. Bds. of Health, Boat,, 1904, XIV, 109.

INFECTION BY FOMITES 219

that disease is carried to laundry workers in this way, and
it is probable that some of the alleged instances are true.
Thus Sedgwick, and also Walcott,1 report cases of typhoid
fever where the evidence was very strong that the disease
was contracted by handling soiled linen. Thorne-Thorne 2
also reports an interesting case of probable transmission of
typhoid fever by clothing. That such accidents are common
has never been shown. Certainly in Providence there is no
excess of scarlet fever, diphtheria or typhoid fever among
laundry workers. This cannot be due to the disinfection of
the clothing, for much disinfection is not effective, and for
six years there has been no disinfection after diphtheria,
and moreover the linen from carriers and missed cases is
nowhere subjected to disinfection.

It is true that Landouzy,3 after examining 1590 laundry
workers in Paris, states that tuberculosis is twice as prevalent
among them as among other workers, and he attributes it to
infection from soiled clothing. But certainly other occu-
pations show an even greater excess of this disease, and
there is no suggestion that it is due to infection from the
materials handled. There may be many other reasons why
laundry workers should show an excess of tuberculosis, such
as age distribution, poverty and overwork.

Infection of Rooms. — It is generally believed that the
room, or rooms, which have been occupied by a case of con-
tagious disease are a fertile source of danger to others. The
germs of diphtheria, scarlet fever and smallpox are sup-
posed to become attached, not only to books, playthings,
bedding and furniture, but also to the walls and ceilings. It
is true that anything which can be reached by the patient
or attendant may possibly receive infection, the chance in-
creasing according to the frequency with which the thing is

1 Sedgwick and Walcott, J. Mass. Ass. Bds. Health, Boston, 1900,
IX, 145.

2 Thorne-Thorne, Clin. Soc. Trans., Lond., 1892, XXV, Suppl., 67.
» Landouzy, Presse med., Par., 1905, XIII, 633.

220 THE SOURCES AND MODES OF INFECTION

handled. That parts of the room or its contents not touched
should be infected by floating particles, is highly improbable.
The secretions and excretions which in various ways become
attached to the contents of the room are usually small in
amount and thinly smeared on the surface. Such material
usually rapidly loses its virulence by drying, so that, as is
shown in these pages, virulent germs are recovered from the
sick-room in only a small proportion of tests. Nevertheless
if people should, after the termination of the sickness, crowd
into the room, rub their moistened fingers over the various
objects and put the fingers in the mouth, infection might
sometimes result. But as the germs die rapidly, as probably
not many persons enter the room, as even visitors would
usually run little chance of taking up whatever pathogenic
organisms might be there, we are, I think, justified in assum-
ing that infection from the room or its contents is not very
likely to take place.

" Lung Blocks." — The celebrated " lung block " in New
York, bounded by Cherry, Catherine, Market and Hamilton
streets, has had enormous influence on modern views con-
cerning the transmission of tuberculosis. The great excess
of tuberculosis in certain tenement houses has given rise to
a strong belief in its causation by infection which remains
attached to the interior of the dwelling. We even hear much
loose talk about the germs of the disease developing in the
filth and dampness of these dark houses. What an important
factor house infection is believed to be in the causation of
this disease is well illustrated by the exceedingly prominent
place which is given to room disinfection. To judge from the
attention, time and money bestowed on room disinfection
after the removal or death of a consumptive, this practice
is considered of equal importance with hospitals, sanatoriums,
dispensaries or district nursing as a preventive measure. Let
us see what is the evidence on which house infection is sup-
posed to rest. It is apparently that cases continue to occur
in the same house during successive years, This judgment

INFECTION BY FOMITES 221

is based on the spot map. Thus in New York1 during five
years, 42 per cent of the deaths from tuberculosis occurred in
23 of the total houses infected, or in 5.25 per cent of all the
houses in the city. In Ward IV, 55.8 per cent of the cases
occurred in 10.5 per cent of the houses in the ward and in 28
per cent of the infected houses. In Ward VI, 44.7 per cent of
the cases were in 7 per cent of the total houses and in 19
per cent of the infected houses. Many in other cities have
noted the same phenomenon, but it has been especially well
recorded by the New York Department of Health. But to
say that these facts are to be largely explained by the per-
sistence of the tubercle bacilli in the houses is no more rea-
sonable than it would be to explain the recurrence of an excess
of murders in certain areas to the persistence of a hypothet-
ical microbe of homicide. It is true that the advocates of
the importance of house infection point to numerous in-
stances in which an apparently healthy family moving into
a house recently occupied by a consumptive later develop
the disease. That such should quite often happen merely as
a coincidence is necessitated by the great prevalence of the
disease. That in " lung blocks " exposure to living cases in
the other tenements is a much more likely source of the
disease than exposure to bacilli on the walls, is probable. To
demonstrate the relation of the disease to house infection it
would be necessary to go fully into the history of at least a
large proportion of the cases, and that, particularly as we
really know nothing about the latent period of the disease,
is at present impossible. The excessive incidence of tubercu-
losis on certain houses is no proof of house infection, and we
are obliged to appeal to the facts of bacteriology and the
general principles of infection to estimate the probable dan-
ger from this source. When we consider the number of per-
sons who are continually throwing off great numbers of
tubercle bacilli, and the numberless chances there are, particu-
larly in the crowded tenement districts, of coming in contact
1 Rep. Dept. Health, City of New York, 1896, 244.

222 THE SOURCES AND MODES OF INFECTION

with fresh infective material, there seems to be no necessity
of assuming that infection must be traced to the more or less
feeble and scattered germs that may be clinging to the walls
of a vacated apartment. What necessity or excuse is there
for assuming that such infection plays more than an inap-
preciably minute part in the causation of this disease ?

Rags and Disease. — Rags have often been considered an
important vehicle of disease. At one time much attention
was given to this subject in the United States, and great
danger was apprehended from the importation of foreign rags,
and stringent measures were taken to secure their disinfec-
tion. Lengthy discussions of the matter, and references to a
great volume of literature, may be found in a report for the
New York City Board of Health by Smith in 1886, and in the
Marine Hospital Report for 1893.1 Numerous references
are given of the alleged transmission of various diseases by
means of rags. Among the diseases mentioned are smallpox
(126 outbreaks), influenza, scarlet fever, erysipelas, typhoid
fever, septicemia, cholera, and a disease peculiar to rag dust,
called " flock cough."

Rags and Smallpox. — It has been believed that small-
pox is frequently introduced among the workers in paper
mills by the rags which they handle. Numerous instances
have been reported from Maine, Massachusetts and Wiscon-
sin, as well as from foreign countries. Most of the evidence
is very inconclusive, as no effort is made to exclude other
sources of infection, and the disease is usually prevailing
generally at the time. The most suggestive outbreaks are a
series reported by the Massachusetts State Board of Health.2
Here from one to three cases occurred in six mills in different
localities at a time when there was no smallpox in the town.
Dr. Abbott, who personally studied these outbreaks, was con-
vinced that they were due to handling rags. Yet it is curious
that almost nothing has been heard of this sort of infection

1 Rep. Surg. Gen. U. S. Mar. Hosp. Serv., 1893, II, 330.
8 Rep. St. Bd. Mass., 1888, xvi.

INFECTION BY FOMITES 223

during the last twenty years, and this at a time when small-
pox was of a remarkably mild type and great quantities of
clothing worn by patients must have escaped disinfection.
Abbott was also convinced that, owing to the length of time
between the collection of the rags and the opening of the
bales, no danger was to be apprehended from foreign rags.
Doty1 says " that the most careful investigation has failed
to present satisfactory evidence that either foreign or domes-
tic rags act as a medium of infection." He has personally
carefully studied the question in Egypt, where many rags are
collected for the American market, and he says that there is
no evidence of the infection of the handlers of even the fresh
rags.

Rugs and Plague. — Remlinger2 has recently called atten-
tion to the supposed danger to be apprehended from draper-
ies, and particularly also rugs, from the Orient. He rightly
says that many of the rugs are very filthy, and must have
been infected during their use or manufacture. Yet during
all the prevalence of cholera and bubonic plague in Asia dur-
ing recent years not a single case of these diseases has been
brought into Europe or America in this way, though great
quantities of rugs, draperies and rags have been imported
without disinfection or with very imperfect disinfection.

Money and Disease. — Money is popularly believed to be
a common means of spreading disease. Indeed there are few
things which at first sight seem more likely to do so than paper
money. Germs readily become attached to its surface, it
passes rapidly from one person to another, it is kept in inti-
mate contact with the person, held closely in the hand and
often put to the lips. Nevertheless there is no good evidence
that money has ever actually been the means of spreading
contagious disease. If money is frequently a carrier of infec-
tion, persons who handle a great deal of money ought to be
particularly subject to infectious disease. This does not seem

1 Doty, Med. Rec., N. Y., 1900, LVIII, 681.

2 Remlinger, Hyg. ge"n. et appliq., Par., 1907, II, 257.

224 THE SOURCES AND MODES OF INFECTION

to be the case, though there are unfortunately no good pub-
lished statistics bearing on the subject. However, it is the
business of the supporters rather than the opponents of the
theory to produce the figures. I have been on the lookout
for contagious disease among bank clerks, but the very few
cases that have come to my knowledge during the past
twenty-five years have evidently been contracted in other
ways. According to Hilditch,1 the " United States treasurer,
who has given the subject long and careful consideration, is
emphatic in his statement that ' there is not the slightest evi-
dence to show that the employees in his department contract
infectious diseases any oftener than others who are not in
this line of work.' ' It may be argued, and there is some
truth in this, that tellers are accustomed to take considerable
precaution, such as keeping the fingers away from the lips
and washing the hands before eating. Tram-car conductors
are, however, I know from observation, particularly prone to
hold bills and coins between the lips, and are in other ways
extremely careless, yet they certainly show no excess of scarlet
fever, diphtheria or smallpox. A bacteriological study of
paper money has been made by Hilditch above referred to.
He examined twenty-four bills and found the number of bac-
teria varied from 14,000 to 586,000 per bill. Pus bacteria
were found, as was to have been expected, but no other patho-
genic forms. Hilditch could find accounts of only four other
similar investigations, none of which, however, were as
thorough as his. Bacteria are not found in any large num-
bers on coins, chiefly because of the germicidal action of the
metal, as shown by the researches of Park,2 Vincent 3 and
Bolton.

Much Evidence Unsatisfactory. — It would be easy to find
hundreds of alleged instances of fomites infection, in some
of which the infection was supposed to have persisted for

1 Hilditch, Pop. Sc. Month, N. Y., 1908, LXXIII, 157.

2 Cited by Hilditch.

3 Vincent, Abst. Med. News, N. Y., 1892, LXXX, 275

INFECTION BY FOMITES 225

years. Those mentioned in the preceding pages are only a
few which I happened to have at hand. In most of them
there is no real evidence that the disease was produced in the
manner claimed. The error made in claiming so much for
fomites infection in yellow fever shows how great is the lia-
bility of error for other diseases.

Persons, not Things, are Dangerous. — It must also be
borne in mind that in very many of the reported cases the
supposed infected articles were carried by some person. That
the person may be the " carrier" of living germs on his own
mucous surfaces, though showing no symptoms, we now know
full well. Until recently this was not known, hence it was
universal to consider things, not persons, as the bearers of
infection. We can now see that persons in whom the germs
are growing are much more likely to be the agents of infec-
tion than are things on which the germs are dying. Some of
the instances of alleged fomites infection, such as the room
infection in scarlet fever referred to, are doubtless really
instances of carrier infection.

Fomites and Tetanus. — It is not for a moment to be
assumed that there are no instances of fomites infection. It
is not impossible, or at all improbable, that occasionally
typhoid fever, smallpox, diphtheria and other diseases are
caused by material things holding the living bacteria for some
weeks or even for months. In some instances the clinical
evidence of fomites infection is very strong, though perhaps
it can rarely if ever be in any individual instance entirely
conclusive. If it can be substantiated by bacteriological evi-
dence, it becomes so much the stronger. The strongest evi-
dence we have of fomites infection is concerning anthrax and
tetanus. This is not surprising when it is recalled that the
bacilli of both of these diseases are spore-forming and capable
in that state of resisting unfavorable conditions of life for
years. Thus Smith1 finds that tetanus bacilli will survive
boiling for sixty minutes at a time, or twenty minutes on
1 Smith, Theobold, J. Am. M. Ass., Chicago, 1908, L, 929.

226 THE SOURCES AND MODES OF INFECTION

each of three successive days. This explains why they have
been found alive and virulent in commercial gelatine and in
that situation have been known to give rise to the disease in
human beings.1 Of six samples of cotton lamp wick pur-
chased in various shops in Havana, five were shown by the
inoculation of white mice to be infected with tetanus. This
material was used by midwives for tying the umbilical cord,
and after sterile material was furnished them by the depart-
ment of health it is said that almost no deaths from infantile
tetanus occurred in Havana.2 The number of deaths from
tetanus in children under one year of age in Havana decreased
from 128 in 1901 to 18 in 1908. Some of the Fourth of July
tetanus is believed to be due to the presence of the spores
in the wads of blank cartridges, and they were demonstrated
in them by Dolley,3 and are said to have been found in car-
tridges in Germany byMusehold of Strassburgand others, but
several other American observers failed to find tetanus germs
in a total of 759 cartridges examined.

Fomites and Anthrax. — The spores of anthrax are so
resistant that they have been kept for ten or twelve years,
but the bacilli themselves do not survive any longer than
typhoid bacilli. Anthrax, though rather rare in the United
States and England, is very common in some parts of the
world, particularly in Asia, and the spores are frequently
imported in dry animal products from Asiatic countries.
Legge4 cites from several observers who recovered the spores
from hair and hides imported from China and Siberia. More
recently Eurich5 has examined nearly 750 specimens of wool,
hair and dust from these materials. In over 600 specimens

1 Tuck, Jour. Path. & Bacteriol., Edinb. & Lond., 1904, IX, 38.

2 Junta Sup. de San. de la Isla de Cuba, Supplement*) y Note Adi-
cional, 1902-03, 4.

3 Dolley, J. Am. M. Ass., Chicago, 1905, XLIV, 466.

4 Legge, Lancet, Lond., 1905, 1, 694, and Rep. Insp. of Fact., Lancet,
Lond., 1904, I, 1206.

6 Rep. Anthrax Investigation Bd., Bradford, Eng., No. 3, 1908, 8.

INFECTION BY FOMITES 227

free from blood he found no anthrax germs, but he did find
them in 20, or 14.4, per cent, of 139 bloodstained specimens.
These findings are substantiated in his last report (1909), and
he speaks of a case of anthrax in a man who handled wool
in which anthrax spores were actually found. In 64 samples
of dust, anthrax germs were demonstrated only once.1 Page2
also gives references to other similar findings. But such
observations are not necessary to show that the disease is
transported in this way, for there is ample clinical evidence
that such goods are the direct cause of anthrax in men and
animals. In fact, most industrial anthrax in Western Europe
and North America is caused by handling wool, hair and
hides imported from anthrax-infested countries.

Fomites and Typhoid Fever. — One of the most remarkable
of the authenticated instances of fomites infection is the
transmission of typhoid fever by means of army blankets
from South Africa.3 These blankets came from Africa in
October, 1902, and were then sold. They went to 290
different parties. One lot used on the transport Cornwall
apparently gave rise to the disease in May, 1903; the use
of another lot in England was also followed by typhoid
fever. Some of the blankets were considerably soiled, and
living bacilli were found on several that were examined in
London.

Fomites and Diphtheria. — A young man working in a
laboratory in an American city spilled some bouillon culture
of diphtheria bacilli on his coat. This coat, without disin-
fection, he wore when calling on the young woman to whom
he was engaged, and she developed diphtheria two days later.
The culture which was spilled contained only the branching
forms of the diphtheria bacillus, and the culture from the
patient's throat showed the same forms.

But very few instances other than the above are on record

1 Rep. of Chief Inspector of Factories and Workshops, 1907, 57.

2 Page, J. Hyg., Cambridge, 1909, IX, 357.

3 Parkes, Practitioner, Lond., 1903, LXXI, 297.

228 THE SOURCES AND MODES OF INFECTION

where pathogenic bacteria have actually been found on mate-
rials which presumably have carried infection. According to
Simpson,1 plague bacilli were found by Kitasato on cotton
goods imported into Japan, and this was thought by Kitasato
to be the way in which the disease was introduced into that
country, but from what is now known about the mode of
extension of the disease this seems highly improbable.

Few Instances of Fomites Infection. — But while we may
admit that occasionally the virus of even many of the com-
moner diseases may be retained on fomites for a considerable
length of time, and ultimately give rise to new cases, there
is no clinical evidence to show that such instances are at all
common. Even when carefully sought for, fomites infection
is not very often found. In my early work as health officer
I firmly believed in the importance of this factor, and dili-
gently sought for evidence. The fact that I found very little
was one thing which led to a more careful consideration of the
subject. Of 13,970 cases of scarlet fever reported in Michi-
gan,2 only 335 were attributed to fomites infection. To an
even less degree are diphtheria and measles attributed in
this report to fomites infection. Of 221 cases of bubonic
plague in Natal,3 only 8 were by Hill attributed to fomites
infection, and Mitchell in Port Elizabeth attributed only
6 of 337 cases to the same source. When we consider that
most of the evidence is extremely flimsy, and that much of
the alleged fomites infection is probably carrier infection, and
remembering also how the history of yellow fever has taught
us to be wary of such proofs, we are forced to the conclu-
sion that there is little in the history of the more prevalent
infectious diseases to indicate that fomites infection is at all
common.

Reasons for Belief in Fomites. — One reason why fomites
infection looms so large in the minds of health officers, as

1 Simpson, Treatise on Plague, Cambridge, 1905, 204. '
* Rep. St. Bd. Health, Mich., 1906, 134.
» J. Hyg., Cambridge, 1907, VII, 712.

INFECTION BY FOMITES 229

well as of the laity, is that the striking character and air
of mystery about the alleged incidents are so impressive.
" Death in a Toy," or " A Child Succumbs to the Dread
Disease from Infection Lurking in its Mother's Shawl,"
appeal to the imagination. That the invisible emanations of
disease should cling to a garment for years is too near akin
to the stories of the Arabian Nights not to impress the average
mind. Hence it is that the comparatively few instances of
real fomites infection have far more than their due weight
in our estimation of the relative importance of different modes
of infection.

Even if all the alleged instances of fomites infection were
true, the amount of disease apparently caused in this way
is relatively very small. The frequency with which conta-
gious disease can be traced to fomites is not the reason for
the general belief in the importance of this mode of infection.
The real reason, I have no doubt, is, that until very recently
there seemed to be no other way of explaining isolated cases
of disease. As a rule it is impossible, even with modern aids
to the diagnosis of obscure cases, to trace the source of infec-
tion of most cases of contagious disease, particularly in cities.
The theory of long persistent fomites infection seemed to
offer a reasonable solution, and hence met with universal
acceptance. The theory was almost a necessity to explain
the facts as they were formerly understood. Now we have
no need- for such a theory, and a much more satisfactory
explanation is at hand.

Evidence against Theory. — It is only within a very few
years that the frequency with which mild atypical cases of
disease occur has been recognized, and the existence of
numerous entirely healthy carriers is a modern discovery,
which is even now denied by some. The more carefully
individual cases and outbreaks of disease are studied, the
more often are they traced to missed cases and carriers.
It is not probable that we shall ever be able to discover the
origin of all our contagious disease. We can only infer its

230 THE SOURCES AND MODES OF INFECTION

source from the data we have. As was shown in the first
chapter, there is every reason for thinking that disease germs
rarely grow outside of a living body. Two other theories
are open to us. Disease may be due to the persistence of
infection on things, or it may be due to exposure to mild
cases or carriers. There should be no hesitancy in choosing
between the danger from rapidly dying germs on books,
money, furniture or clothes, and rapidly growing germs in the
mouth, nose and intestines of persons. Moreover things
must be carried, people move freely at will.

If the danger from fomites infection were as great as is
generally believed, the contagious diseases would be much
more common than they are. The advocates of this theory
are constantly telling us how easily everything near the sick
becomes infected and how long the infection lasts. Every
one knows that at the best disinfection is imperfect, and that
much that passes for disinfection is no disinfection at all.
Then the missed cases, which all admit occur in considerable
numbers, to say nothing of the carriers, are constantly infect-
ing large numbers of things which are not subjected to any
disinfection. Yet our scarlet fever and diphtheria are not
increasing, which means that one case of the disease gives rise
to no more than another case. If fomites infection occurred
as easily as is alleged, each case would ramify through
fomites into a dozen more cases. One reason for doubt about
the prevailing ideas of fomites infection is this: if fomites
infection were as common and as easy as is alleged, few
could escape it, and the infectious diseases would be much
more prevalent than they are. As was shown in Chapter IV,
the chances for the transfer of fresh infective material are so
extremely numerous that there is no necessity for assuming
the far more difficult and uncertain modes of aerial convection
and transmission by fomites, and indeed there seems to be
little opportunity for their action.

Bacteriological Evidence. — Having considered some of
the clinical evidence of the part played by fomites in the

INFECTION BY FOMITES 231

transmission of infection, it is desirable to inquire what light
the laboratory study of disease has thrown on the problem.
One of the first labors of the discoverer of a pathogenic organ-
ism is to determine its resistance to various hostile influences,
such as heat, cold, drying, light and disinfectants.

The Effect of Drying upon Bacteria. — Drying, exposure
to light and lack of nourishment are the principal factors
which determine the life of micro-organisms on fomites.
Besides the study of the germs of special disease by those
particularly interested, Germano, whose work is mentioned
in the chapter on aerial infection, Ficker,1 Zonchello,2 Heim,3
and Buckley,4 among others, have given careful and system-
atic attention to the effect of drying on the vitality of bac-
teria. Exceedingly divergent results have been reported by
these different observers. This, however, is not surprising if
the number of factors involved is taken into consideration.
Among the most important of these factors is the amount
of light. Germs that are killed in a few minutes in direct
sunlight may live for weeks in a dark place or even in diffused
light. The thicker the layer of infectious material, the longer
is its virulence likely to be maintained. This thickness
depends largely upon the nature of the medium. In a dried
watery medium, bacteria may die quickly, while they may
survive long in sputum or feces. The more complete the
drying, the shorter the life, and alternate drying and damp-
ening is unfavorable. The higher the temperature, the
sooner the germs perish. Their vitality also varies with
the rapidity of the drying process and the material on
which they happen to be. Old cultures die sooner than
fresh ones, and different strains have different powers of
resistance. The chemical composition of the medium and

1 Ficker, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1908, LIX, 367.

3 Zonchello, Giornale della Real Soc. Ital. d' Igiene, 1905, XXVII,
489, 537.

1 Heim, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1905, L, 122.

4 Buckley, Pub. Health, Lond., 1906-07, XIX, 290.

232 THE SOURCES AND MODES OF INFECTION

the presence or absence of other organisms may have an
influence.

The spores of bacteria are so resistant that we should natu-
rally expect the diseases caused by the spore-forming kinds
to be readily carried on fomites. As has been shown, this is
true of anthrax and tetanus, diseases which belong to this
class.

Typhoid Bacilli. — Among the more hardy non-spore-
forming bacteria is the bacillus of typhoid fever. In the first
chapter it was shown that in the presence of moisture, as in
privy vaults, the soil, milk, water, etc., this bacillus some-
times lives for some months, though it often dies out in a
much shorter time. It remains to consider the duration of
its life when in a more or less dry condition.

Firth and Horrocks1 found that typhoid bacilli would live
on khaki for 78 days, in feces dried on serge for 9 to 17 days,
on serge for 10 days after it had been exposed to direct sun-
light for 50 hours. Pfuhl2 says that dried on linen they lived
97 days. Germano3 cites Gaffky as reporting that the
typhoid bacillus would live for 3 months when in a dry con-
dition, and that Uffelmann recovered it from various dry
materials after a period of from 21 to 80 days. But Germano
suspects that the substances were not perfectly dry. Ger-
mano himself was able to preserve typhoid bacilli dried on
wood or linen for 90 days, but when he inoculated sterile
dust with a bouillon culture, the bacilli did not survive over
4 days and sometimes perished in 1 day. Most of them
died off very rapidly. Buckley 4 found they would live for
from 5 days when dried on paper in a room to 119 days
when kept on wood in a moist chamber. The consensus of
opinion seems to be that while under unfavorable conditions,

1 Firth and Horrocks, Brit. M. J., Lond., 1902, II, 936, 1094.

2 Pfuhl, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1902, XL, 555.
* Germano, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1897,

XXIV, 403.

4 Buckley, Pub. Health, Lond., 1906-07, XIX, 290.

INFECTION BY FOMITES 233

as when quite dry and exposed to light, the typhoid bacillus
may die in a few days, yet under conditions which must fre-
quently prevail it may remain alive on such things as cloth-
ing and bedding for some months.

Mediterranean Fever. — The micrococcus of Mediterra-
nean fever has about the same resistance as that of typhoid
fever. Like the typhoid bacillus, it is killed in an hour or
two in direct sunlight. Dried on glass it survives 16 days, in
moist soil 72 days and on a blanket 80 days.1

Diphtheria. — Loeffler kept dry diphtheria bacilli alive for
from 9 to 16 weeks, Roux and Yersin for 5 months, D'Espine
and Morignac for between 3 and 4 months, and Park for 4
months.2 Germano3 found that they would retain their
virulence after remaining in dry earth or dust for 20 to 40
days, and Reyes4 found them virulent in sand and on cloth
after 14 days. Buckley5 recovered living bacilli, when dried
in the air on paper, after 6 days, on wood after 8 days, on
cotton and on glass after 24 days, and on plaster after 37
days. Hill 6 exposed to ordinary room conditions, glass rods
which had been rubbed on a culture of diphtheria bacilli.
Of these 28 per cent survived 14 days and 9 per cent 20
days. Houston 7 found that they died very quickly in earth.
Leighton8 recovered them from warm moist modeling clay
up to 18 days. Williams 9 could not recover the germs
after 24 hours from pencils moistened by the lips of patients
who had the bacilli in the throat,

1 Horrocks, Rep. Commission Roy. Soc., Pt. I, 1901.

2 Cited by Germano.

3 Germano, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1897,
XXV, 439.

4 Reyes, Jahrb. u. d. Fortschr. . . . d. path, mik., Baumgarten,
1895, XI, 203.

6 Buckley, Pub. Health, Lond., 1906-07, 290.

6 Hill, Am. Pub. Health Ass. Rep., 1902, XXVIII, 209.

7 Houston, Rep. Med. Off. Loc. Gov. Bd., Lond., 1898-99, XXXIII, 413.

8 Leighton, Pediatrics, 1901, XII, 360.

9 Williams, N. Y. Health Dept., Sci. Bull. 2, 1895, 16.

234 THE SOURCES AND MODES OF INFECTION

Dysentery Bacilli. — According to Pf uhl 1 the bacillus of
dysentery may remain alive for 17 days when dried on cloth,
or 10 days when in dry sand. In direct sunlight it dies in 30
minutes.2 Kruse3 claims that when dry it will retain its vital-
ity for months.

Tubercle Bacilli. — More attention has been paid to the
vitality of the tubercle bacillus than to that of other bacteria.
It is generally believed to be one of the most resistant, but
Hill 4 has shown that under the same natural conditions of
dryness, light, etc., the diphtheria bacillus will outlive it.
Many of the earlier writers claimed a very considerable lon-
gevity for the tubercle bacilli in dried sputum. Villemin,
Schill, Fischer, Koch, De Thoma, Sormani, Maffuci and
Cadeac and Malet claimed a life of from 1 to 9 months.5
Ransome and Delepine 6 found that the bacilli if exposed to air
and light would not survive 45 days, but if kept in dim light
they did survive. Twichell 7 placed sputum in a folded hand-
kerchief, in a folded carpet, and spread on wood, and ex-
posed it to the air at ordinary temperatures and in diffused
light. The bacilli survived for 39 to 70 days. In sunlight
they died in a few hours. Migneco 8 found that when dried
on cloth in the sun they lived from 20 to 30 hours.

Not so Resistant as Believed. — Many recent observers
do not find this bacillus so resistant as has been supposed.
That it perishes in direct sunlight in less than an hour seems
certain. Weinzirl,9 using improved methods, finds that it
will not survive 10 minutes, and frequently dies in 2 minutes.

1 Pfuhl, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1902, XL, 555.

2 Shiga, Deutsche med. Wchnschr., 1901, XXVII, 765, 783.
Kruse, Deutsche med. Wchnschr., 1901, XXVII, 370, 386.
Hill, Am. Pub. Health Ass. Rep., 1902, XXVII, 209.

Kolle and Wasserman's Handbuch [etc.], Jena, 1903, II, 108.
Ransome and Dele*pine, Proceedings Royal Society, No. 336.
Twichell, Med. News, N. Y., 1905, LXXXVII, 642.

8 Migneco, Arch, of Hyg., Miinchen u. Leipz., 1895, XXV, 361.

9 Weinzirl, J. Infect. Dis., Chicago, 1907 [Suppl. No. 3], 128.

INFECTION BY FOMITES 235

Cade*ac * spread sputum on marble and could find no living
germs after the fourteenth day. On a porous plaster plate
they died within 2 days. Hill 2 dried sputum on glass rods
in the air under ordinary room conditions, and found no liv-
ing bacilli at the time of his first test, which was made after
16 days. Rickards, Slack and Arms3 have made very careful
tests by exposing sputum on wood and cloth in the rooms of
ordinary tenements. They find that when dry and kept in
diffused light the bacilli will live about 1 month, in dark and
dry rooms up to 85 days; another strain survived only 45
days under the latter conditions. Rosenau4 says that further
work upon the viability of the dried tubercle bacillus may
change our views as to its hardiness, and failure to recognize
lesions produced by the dead bacillus is responsible for some
of the false conclusions reached by certain experimenters.

Plague Bacilli. — Simpson 5 states that the German Plague
Commission found that in a large number of experiments with
sputum, blood, etc., dried on all sorts of materials, under
natural conditions, the bacilli of bubonic plague do not sur-
vive over 8 days. Of many specimens of the organism dried
on cover glasses and sent to England, none survived the
journey. Kitasato8 found that plague pus dried on cover
glasses lost its virulence, when exposed to the sun, in from
3 to 4 hours, and this has been substantiated by others. As
was referred to in the first chapter, the work of the last Eng-
lish Plague Commission shows that virulent plague bacilli
cannot be found in the dirt floors of native houses after 48
hours. According to the careful experiments of Buckley,7

1 CadSac, Lyon meU, 1905, CV, 865, Abst. Brit. M. J., Lond.,
1906, I.

Hill, Am. Pub. Health Ass. Rep., 1902, XXVIII, 209.

Rickards, Slack and Arms, Am. J. Pub. Hyg., Bost., 1909, V, 586.

Rosenau, U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab. Bull.

No

57, 1909.

Simpson, Treatise on Plague, Cambridge, 1905, 96.

Kitasato, Lancet, Lond., 1894, II, 428.

Buckley, Pub. Health, Lond., 1906-07, XIX, 290.

236 THE SOURCES AND MODES OF INFECTION

plague bacilli remain alive after drying in the air, for 11 hours
when dried on cotton, 2 hours on wood, 5 hours on plaster,
2 hours on glass, and 3 hours on paper. When kept in a
desiccator they survived on cotton for 22 hours, and in a
moist chamber for only 36 hours. Gotschlich,1 by folding
material containing the germs in cloth, could preserve them
alive for from 3 to 4 weeks. Simpson2 reports that infected
cloth may retain its virulence for 80 days. According to
Verjbitski,3 the crushed viscera of experimental animals and
the crushed bodies of fleas when smeared on cloth and dried
will preserve the bacilli alive for 130 days at a temperature
of 4-5° C., and for 35 days at room temperature. Bandi and
Stagnitta-Balistreri state that these bacilli may survive in the
bodies of dead rats for 2 months. The vitality of the plague
bacillus has been carefully investigated by Rosenau,4 who
does not consider it a frail organism. Temperature is the
most important factor in its life. It may lose its virulence
before it loses its vegetability. It dies in a few days on the
dry surface of hard objects and on paper. Rosenau says that
bedding may harbor the infection for a long time. Tidswell,5
experimenting with a large number of materials, found that
plague bacilli dried under natural conditions lived only
from 3 to 4 days, but when dried slowly on muslin they
might live for 21 days. The colder the climate the greater
is the chance of the persistence of infection. In this all are
agreed.

Pus-forming Bacteria. — The pus-forming bacteria are quite
resistant to drying. According to Germano,8 streptococcus

1 Gotschlich, cited by Kolle and Wasserman's Handbuch [etc.],
Jena, 1903, II, 496.

2 Simpson, Treatise on Plague, Cambridge, 1905, 93.

3 Verjbitski, J. Hyg., Cambridge, 1908, 203.

4 Rosenau, U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab. Bull.
No. 4, 1901.

6 Tidswell, Report on Plague in Sydney, J. A. Thompson, 1902, 67.
8 Germano, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1897, XXVI,
66.

INFECTION BY FOMITES 237

withstands drying for a month, but different strains have
varying degrees of resistance. See also Heim1 and Neisser,2
who found that these organisms would withstand drying for
a long time. Buckley3 could keep staphylococcus alive for
only 7 days when on paper in the air, and on other sub-
stances and under different conditions for varying times, up
to 130 days on cotton kept in a desiccator.

Cholera Spirilla. — Germano4 in a number of experiments
found that the cholera spirillum in dried feces lived only 3
days and in other experiments only 1 day. He also cites Zon-
chello as reporting that it is among the least resistant bac-
teria. Kitasato5 states that it may retain its virulence up
to 8 days, but that it may die sooner, especially when dried on
glass. Usually it lives from a few hours up to 4 days. Koch
and Gaffky6 state that when dried on glass it survives only
a few hours, but when dried on fabrics it may retain its viru-
lence up to 4 days. Buckley3 found that cholera germs would
survive when dried in the air, 9 hours on cotton, 8 hours on
wood, 1} hours on glass, and 5 hours on paper. They did
not survive nearly so long when dried in a desiccator. This
is contrary to the experience of others, for as a rule bacteria
live much longer when dried in a desiccator than when dried
in the open air under natural conditions. Gotschlich7 says
that cholera germs will live in dejecta dried in the air on
clothing for 36 days, and when damp, according to Karl-
niski's observations, for 7 months. He considers that such a
long life is exceptional, and that generally the spirillum dies
in a few days. A duration of only a few days, or even hours,

Heim, Ztsch. f. Hyg. u. Infectionskrankh., Leipz., 1905, L, 122.

Neisser, Ueber Luftstaub-Infection, Inaug. Dis., Breslau, 1898.

Buckley, Pub. Health, Lond., 1896-97, XIX, 290.

Germano, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1897,
XXIV, 403.

Kitasato, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1885, V, 134.

Koch, Arb. a. d. k. Gesundsamte, Berl., 1886, I, 199.

Gotschlich, Kolle and Wasserman's Handbuch [etc.], Jena, 1902,
I, 211.

238 THE SOURCES AND MODES OF INFECTION

was also found by Gamaleia,1 Hesse2 and Koch and
Gaffky.

Pneumococci. — The pneumqcoccus is widely distributed
in healthy human mouths, and the opportunities for the
direct transference of fresh secretion are so numerous that
it probably is of no importance whether this organism lives
long or not. Wood3 found that while pulverized sputum lost
its virulence in a few hours when dried in mass, it might,
under favorable conditions, retain it 35 days. Buerger4 recov-
ered the pneumococci from a handkerchief 7 days after it had
been in use. Germano5 and some others claim a consider-
ably greater resistance. Germano kept it alive in dust for
140 days, but the tendency of later observers is to consider
it a much feebler organism.

Influenza Bacilli. — According to Pfeiffer,8 the influenza
bacillus retains its vitality when dried in sputum for 36 to 40
hours. When dried on a cover glass and kept at 37° C., it
survives for only 2 hours, and when kept at room temperature
for from 8 to 20 hours.

Meningococci. — The evidence in regard to the germ of
cerebro-spinal meningitis appears to be somewhat conflicting.
Germano and Neisser claim considerable resistance for it, as
also does Jaeger.7 Germano said it would live for 80 to 90
days, but it is said that he did not work with the true menin-
gococcus.8 More careful and recent observers do not find it
so resistant. Councilman9 found that it would live when dry

Gamaleia, Deutsche med. Wchnschr., 1893, XIX, 1350.

Hesse, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1893, XIV, 30.

Wood, J. Exper. M., N. Y., 1905, VII, 592.

Buerger, J. Exper. M., N. Y., 1905, VII, 518.

Germano, Ztschr. f. IJyg. u. Infectionskrankh., Leipz., 1897, XXVI, 66.

Pfeiffer, Nothnagel's Encycl. Pract. Med., Phila. and Lond., Influ-
enza, 1895, 584.

Jaeger, Med. Klin., Berl., 1905, I, 990, 1011.

Arkwright, J. Hyg., Cambridge, 1907, VII, 193.

Special Report on Cerebro-spinal Meningitis, Mass. St. Bd.
Health, 1898, 78.

INFECTION BY FOMITES 239

in a dark room less than 72 hours. Albrecht and Ghon 1
could keep it only 24 hours in the dark, and Bettencourt and
Franca 2 less than 9 hours. Kache,3 Kutscher 4 and Fliigge 5
had similar results, and Arkwright 6 succeeded in keeping
the organism alive only from 20 to 48 hours. Lingelsheim 7
says that in culture media it dies in a few hours, but in
sputum it may be kept alive for 5 days.

Gonococci. — According to Schaffer and Steinschneider,8
and Ullmann 9 the gonococcus lives only a few hours on
textiles, at most 36 hours, and often dies as soon as thor-
oughly dry.

Hertmanni 10 from his own observations and those of
others, whom he cites, concludes that the Tryponema pal-
lida of syphilis may retain its motility for some months if
left moist and in the dark. Drying quickly kills it.

Bacteria on Fomites. — Pathogenic bacteria have fre-
quently been sought for on various articles believed to be
likely to be the means of transporting disease, but with the
exception of the spores of anthrax and tetanus they have
rarely been found. The finding of anthrax and tetanus spores
has already been referred to. Besides the places mentioned,
tetanus germs have been found in the soil in various places,
in the dirt filling between the floors of houses u and in one

Albrecht and Ghon, Wien. klin. Wchnschr., 1901, XIV, 984.

Bettencourt and Franca, Ztschr. f. Hyg. u. Infectionskrankh.,
Leipz., 1904, XLVI, 463.

Cited by Fliigge.

Kutscher, Deutsche med. Wchnschr., 1906, XXXII, 1071.

Fliigge, Klin. Jahrb., Jena, 1905, XV, 373.

Arkwright, J. Hyg., Cambridge, 1907, VII, 193.

Lingelsheim, Klin. Jahrb., Jena, 1905, XV, 373; Ztschr. f. Hyg. u.
Infectionskrankh., Leipz., 1908, LIX,. 457.

8 Verhandl. d. IV Kong. d. deutsch. dermatol. Gesellsch., Breslau,
1904.

9 Ullmann, Wien. med. Blatter, 1897, XX, 703 et seq.

10 Hertmanni Dermat. Ztschr., Berl., 1909, XVI, 633.

11 Heinzelmann, Munchen med. Wchnschr., 1891, XXXVIII, 185,
200.

240 THE SOURCES AND MODES OF INFECTION

instance in a house where there had been a death from
tetanus.1

Distribution of Germs of Suppuration. — The pus organ-
isms are quite resistant to drying, and if they were not, they
are so widely distributed, being found constantly on the skin
and mucous surfaces of human beings, that their presence
might be expected wherever human beings are found. They
have as a matter of fact been found almost wherever sought,
as on clothing, books, money, instruments, floors and wood-
work, and indeed on anything that is touched by the hand
of man.

Diphtheria Bacilli on Fomites. — Diphtheria bacilli have
been frequently searched for on all kinds of objects and fre-
quently found. Abel2 and Wesbrook3 found them on toys,
and in Abel's case it was 86 days after infection. Trevelyan4
recovered them from a handkerchief 11 weeks after it had been
used by a diphtheria patient. Park5 took cultures which
proved positive in almost every instance, from dried stains on
bedclothing soiled by children sick with diphtheria. He also
found the bacilli alive in a piece of membrane after 4 months.
Wright and Emerson 6 made 20 cultures from various articles
in the Boston City Hospital, and found 5 positive. Of these
3 were from the shoes, 1 from the hair of an attendant, and

1 from a floor brush. Schumburg7 in 40 cultures from a
room occupied by a diphtheria patient recovered virulent
bacilli from a drinking glass and the handle of a mirror. In

2 of the 5 cultures the virulence of the organism was low.

1 Gotschlich, Kolle u. Wasserman's Handbuch [etc.], Jena, 1902, 1, 210.

2 Abel, Centralbl. f. Bakteriol. [etc.], I Abt. Orig., Jena, 1892, XIV,
756.

3 Wesbrook, Wilson and McDaniel, Am. Pub. Health Ass. Rep.,
1899, XXV, 546.

4 Trevelyan, Lancet, Lond., 1900, I, 1443.

6 Park, Med. Rec., N. Y., 1892, XLII, 116.

6 Wright and Emerson, Centralbl. f. Bakteriol. [etc.], I Abt. Orig.,
Jena, 1894, XVI, 412.

7 Schumburg, Ztschr. f. arztl. Fortbild., Jena, 1905, II, 567.

INFECTION BY FOMITES 241

Welch ardt1 took 300 swabbings from various things in a
sick-room and 250 from other parts of the house, and found
diphtheria germs 3 times on objects which had been in contact
with the patient's mouth. Hill2 took 532 swabbings from a
room occupied by a diphtheria patient, and obtained 4 posi-
tive results, all of which were from objects handled by the
patient. In Providence about 200 swabbings taken under
similar circumstances showed no diphtheria bacilli. These
last three observations indicate that diphtheria bacilli are
not very numerous, even on objects brought into close con-
tact with the patient. Kober3 could find no bacilli on the
floor, bed linen, etc., of 10 houses in which there had been
diphtheria, and he states that Heymann did not find them
in the Hygienic Institute at Breslau. Klein could not find
them on telephones in London,4 and Hill in Boston5 could
not find them on 24 mouthpieces of lung-testing machines.

Tubercle Bacilli on Fomites. — While the tubercle bacillus
is not so resistant to drying as was formerly thought, it is
discharged in such numbers in the sputum that it has been
found outside of the body more often than have any other
organisms except the pus-forming bacteria. Reference to
finding it in dust will be given in the next chapter. When
in considerable masses of sputum, and kept damp, the bacillus
will survive longer than when mixed with dust. Besides on the
floors and various articles in rooms, the bacillus has been
found in books which were in use for some years in a circu-
lating library.6 Petersson7 examined the history charts kept
by the bedside of tuberculous patients and put away for

1 Weichardt, Jahresb. u. d. Fortschr. . . . d. path. Mik., Baum-
garten, 1900, XVI, 197.

Hill, Am Pub. Health Ass. Rep., 1902, XXVIII, 209.
Kober, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXXI, 449.
Klein, Abst. J. Am. M. Ass., Chicago, 1905, XLIV, 1866.
Hill, Rep. Bd. Health, Boston, 1906, 91.

Mitulescu, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1903,
XLIV, 397.

7 Petersson, Ztschr. f. klin. Med., Berl., 1907, LXIII, 346.

242 THE SOURCES AND MODES OF INFECTION

periods varying from months to years. He found by micro-
scopic examination tubercle bacilli on two books that were
kept for six years. Bissell 1 washed the pockets of some uni-
forms that had been used by soldiers sick with consumption,
and obtained two positive results by guinea-pig inoculation.
Friberger 2 used a vacuum cleaner to remove the dirt from
clothing fresh from use by tuberculous patients, and found
virulent bacilli in 3 of 12 tests.

Cholera Spirilla. — Although the life of the cholera spiril-
lum outside of the body appears to be short, usually only a
few days and often less, some early observers, as Babes,3 claim
to have found it on the personal effects of cholera patients.

Resistance of Protozoa. — There is no theoretical reason
why the protozoan blood parasites might not, in a spore-like,
resistant stage, withstand drying and remain alive for some
time outside of the body. But there is no experimental evi-
dence to show that in any of the well-known protozoan dis-
eases such spores are formed. There is certainly no clinical
evidence to show that such diseases are ever carried on
fomites, and for malaria, yellow fever, sleeping sickness and
Texas cattle fever there is convincing evidence that they
are not so carried.

Resistance of Vaccine Virus. — With two exceptions, prac-
tically no experimental work has been done with the virus
of any of the infectious diseases, a specific organism for which
has not been determined. Vaccine virus has an extensive
use, and it is desirable to store it and transport it long dis-
tances, so that its keeping qualities have received considerable
attention. When the infectious material dries naturally in
the crust which forms from the vesicle, it retains its virulence
for a considerable time. These crusts were largely used in

1 Bissell, Med. News, N. Y., 1899, LXXIV, 156.

2 Friberger, Ztschr. f . Tuberk. u. Heilstaltenw., Leipz., 1908-09, XIII,
37.

8 Shakespeare, Report on Cholera in Europe and India, U. S. Gov.
Print. Off., 1890, 606,

INFECTION BY FOMITES 243

Providence for maintaining the Jennerian strain of vaccine,
which was used in the health department for nearly fifty
years. The crusts, when taken from the arm and kept
wrapped in paper in a dark place, could be relied upon to
retain their virulence for a month, and often did retain it
longer. When kept in a tightly corked bottle in a refrigerator,
they will generally remain virulent over 6 months. If exposed
to light and air and varying temperature, the virulence may
be lost in less than a month. A thin layer of the same lymph
on a quill does not remain active when exposed to the air
for more than a week or ten days. The ivory points covered
with vaccine matter, which were so much used a few years
ago, were usually guaranteed to keep 3 weeks, and often did
remain virulent a month or more. But there was usually
more than one layer, and the thickness of the material was
further increased by the presence of blood and leucocytes.
According to Seaton,1 dried vaccine matter on points may
keep for from 6 or 9 months or more, and he quotes Husband
as securing successful vaccinations from 7 of 93 dried points
kept for periods varying from 6 to 30 months. Vaccine matter
in powder also has been kept by Warlomont and others for
several months.2 Recently Green3 has kept dried pulverized
vaccine matter in sealed tubes for periods varying from 20
days at 37° C. to 252 days at 10-15° C. and 301 days at
4° C. Although under exceptional circumstances dried vac-
cine matter may be kept a considerable time, it requires great
care and often results in failure. Hence early in the nine-
teenth century, in order to introduce vaccination into her
American possessions, the Spanish Government4 sent out a
ship with a number of children on board, so that by successive
arm-to-arm vaccinations fresh lymph might be carried across
the seas. At the end of the nineteenth century, when Porto

1 Seaton, Handbook of Vaccination, 1868, 172.

1 Warlomont, Manual of Animal Vaccination, Phila., 1886, 139.

» Green, J. Hyg., Cambridge, 1908, VIII, 528.

4 Life of Jenner, Baron, I, 606; II, 78.

244 THE SOURCES AND MODES OF INFECTION

Rico came under the control of the United States, the diffi-
culty of transporting vaccine virus to the island became so
great, so little active virus surviving even this short voyage,
that it was necessary to establish a vaccine farm on the island.
In the Philippine Islands it was found to be necessary to
transport the virus packed in ice to inland villages.1

Resistance of -Smallpox Virus. — It is a common belief that
the crusts in variola are infectious, and they are supposed to
have been used at times in the practice of inoculation. I do
not remember to have seen any evidence of this till I came
across the work of Brinckerhoff.2 He refers to a rather limited
literature, and details his own experiments, which determined
a persistence of virulence in the crusts for periods of from 22
to 52 days, and in one instance to 88 days.

Comparative Resistance. — Experiment shows that some
pathogenic organisms, like those of gonorrhea and cerebro-
spinal meningitis, are of such feeble resistance that infection
by fomites, as ordinarily understood, is in these diseases highly
improbable. It is also shown that certain other bacteria, as
those of diphtheria, typhoid fever and tuberculosis, are resist-
ant enough to make infection by fomites very possible. But
it appears that all known pathogens, except those having
spores, tend to die off quite rapidly outside of the body, and
that under natural conditions it is rather the exception for
them to persist for any great length of time. It is worth
noting that the typhoid bacillus is more resistant than any
other pathogenic organism referred to. Yet we hear almost
nothing about fomites infection in that disease. Is it not
because attention has been directed to other sources which
have been believed to be sufficient, so that there has been
little temptation to assume infection by fomites ? Infection
by fomites has, on the other hand, been considered of much
importance in cholera, the spirillum of which is far less resist-
ant than is the bacillus of typhoid fever, probably because

1 Rep. Bd. Health, Philippine Isl., 1904-05, 23.

2 Brinckerhoff, J. Med, Research, Bost., 1904, XI, 284.

INFECTION BY FOMITES 245

in the absence of the knowledge of carriers which we now
possess the world-wide diffusion of this disease seemed inex-
plicable except by means of fomites.

Experiments with Yellow Fever. — With a few notable
exceptions, little experimental work has been done to deter-
mine the part played by fomites in the spread of disease.
When Reed, Carroll Lazear and Agramonte proved that yel-
low fever is transmitted by the mosquito, it still remained
uncertain whether it might not also be carried in other ways,
and sanitary officials were generally so convinced of the over-
whelming evidence of the great part played by fomites, that
definite proof or disproof of the fact was earnestly desired.
The complete failure of the commission to produce the disease
in human beings by fomites is well known, and their experi-
mental work has now been corroborated by abundant clinical
evidence from all the great endemic foci of this disease. In
these experiments, non-immunes were confined in a small
room for a number of days in close contact with bedding and
clothing fresh from yellow-fever cases, most of which was
soiled with excreta and vomitus.1 The experiments were
repeated a number of times, but in no instance did the disease
develop. When not long after I stood in one of those dirty
little rooms in company with Dr. Finlay and Dr. Gorgas, and
saw some of the old bedding lying in the corner, I realized
as never before how very much greater are the difficulties in
observation than in experiment, and it seemed that the time
had come when the prevailing views as to the importance of
fomites infection in other diseases than yellow fever should
be seriously reviewed.

Observations on Plague. — In another disease, bubonic
plague, exceptionally brilliant work, from the purely epidemi-
ological standpoint, has been done by Thompson in Australia.
It is true enough that conditions are much more favorable for
observation when a disease first invades a country than when
it has become as widespread as plague is in India, but in no
1 Phila. M. J., 1900, VI, 790.

246 THE SOURCES AND MODES OF INFECTION

place have they been taken advantage of as they were in
Sydney. Thompson 1 showed conclusively that rat plague
preceded human plague, that the flea was the intermediary
between rat and man, and that in New South Wales, at least,
fomites had no part in the diffusion of the disease. These
observations have been confirmed in other parts of Australia.2
Experiments with Plague. — The present English Plague
Commission have, by their ingenious and painstaking ex-
periments and observations in India, confirmed Thompson's
work. Reference will be made in another chapter to the
experiments on the transmission of the disease by the air, in
which they definitely proved that aerial infection does not
take place. They also studied experimentally the conditions
of house infection, and their work is among the best planned
and executed and the most convincing of any that has ever
been attempted for any disease. Besides much other work,
they placed guinea pigs in the houses of an Indian village
just vacated by the inhabitants because of plague, thus sub-
stituting an experimental population under control for the
normal human population.3 The results were in every way
in accord with the view that plague is primarily a rat disease,
and is transmitted from rats to man, or to other animals, by
means of fleas. In these and in similar less extensive experi-
ments 4 there was evidence to show that the animals could
not contract the disease from the dwellings themselves. To
test more definitely the possibility of transmission by fomites
other experiments were made. The floors of houses were
soaked with cultures of the bacilli, and guinea pigs placed
upon them. Of 24 exposed during the first hour after in-
fection 4 died; of 12 exposed up to six hours 4 died; but
during all this interval pools of the culture were upon the
floor. Of 24 animals exposed after twelve hours only 1

1 Thompson, J. Hyg., Cambridge, 1906, VI, 537.

2 Ham, Rep. on Plague in Queensland, 1900-07, Brisbane, 1907.
» J. Hyg., Cambridge, 1907, VII, 799-875.

4 J. Hyg., Cambridge, 1906, VI, 450-482.

INFECTION BY FOMITES 247

succumbed. Guinea pigs were also allowed to run freely
about hospital wards in which cases of bubonic plague were
being treated, and 15 were confined in a room with freshly
soiled bedding, which was renewed every day for six weeks.
In none of the animals did the disease develop. Again (p.
887), clothing from infected houses, removed without pre-
cautions to prevent the transport of fleas, was placed in
cages containing 26 guinea pigs, 1 of which died of the
plague. A few fleas which had been brought in the clothing
were found in the cages. It seems to be quite clear from
these experiments that while infection by fomites is pos-
sible, it is probable only when the infection is exceedingly
gross and the contact is intimate, as in the case of the in-
fected floors referred to above; or it may also be possible
when infected fleas are transported in the goods. How
common this is has not yet been determined.

Tuberculosis. — I know of no other diseases in which seri-
ous experimental work has been done to determine the part
played by fomites in their diffusion. It is particularly re-
markable that tuberculosis has not been thus critically
studied. Every one has been calling for room disinfec-
tion as a powerful means of combating this disease, yet it
does not seem to have occurred to anyone to place suscep-
tible animals in a series of supposedly infected rooms (as
was done in the case of plague in India) to determine if pos-
sible how great the danger from room infection really is.

Disinfection Unnecessary after Diphtheria. — By a study
of the facts presented in this chapter the writer was con-
vinced some time since that the danger from fomites in-
fection is for most diseases very much less than is generally
believed. In diphtheria particularly it appeared that little
if any sickness could be due to infection remaining about
the house or its contents after the termination of isolation.
As isolation is almost invariably terminated without any
reference to the presence of carriers in the family, it ap-
peared to be absurd to disinfect the material things in the

248 THE SOURCES AND MODES OF INFECTION

house when in all probability some carrier was still growing
the bacilli. Hence in March, 1905, terminal disinfection
was abandoned in Providence, except in those very few
instances in which the family was willing to wait for two
successive negative throat and nose cultures from each of
its members. The attempt was thus made to test the im-
portance of fomites infection by abandoning disinfection
and noting any change that might occur in the prevalence
or distribution of the disease. One obvious way of testing
the influence of disinfection is to compare the recurrence of
the disease after disinfection with the amount which occurred
after the warning placard was removed without disinfection.
The following shows the number of recurrences within 60
days after disinfection, the number of infected families and
the ratio between the two, during the years 1902 and 1905.

YEAR.

Infected
Families.

Recurrences.

Ratio.

1902..
1903.

358
453

6

7

1.67
1 54

1904

559

10

1 78

1905

87

2

2.30

Total

'1457

25

1.71

The following shows the same facts where there was no
disinfection.

YEAR.

Infected.
Families.

Recurrences.

Ratio.

1905..

258

4

1 55

1906

259

4

1 55

1907

343

7

2 04

1908

687

17

2 34

1909

472

10

2 12

1910

431

4

93

1911

550

8

1 49

Total....

3000

54

1 80

INFECTION BY FOMITES

249

The period from 1902 only is taken, because previous to
that time the duration of isolation was somewhatjpnger, but
there has been no change since. Very few health officers
keep any record of the recurrences in diphtheria and other
infectious diseases, but such records are kept in Baltimore,
where terminal disinfection is practiced in an unusually
thorough manner, and is. in a large majority of cases tested
for its efficiency. In order to make the Providence figures
comparable to those of Baltimore, it was necessary to in-
clude recurrences in other families in the house as well as
in the family first invaded, and to calculate the percentage
on total cases, rather than on invaded households. The
following are the figures and there is nothing in them to
indicate that terminal disinfection is of any value in pre-
venting recurrences.

BALTIMORE. DISINFECTION.

YBAKS.

Cases.

Recurrences.

Ratio.

1903-1909

6931

122

1.76

PROVIDENCE. NO DISINFECTION.

YEARS.

Cases.

Recurrences.

Ratio.

1905-1911

4189

69

1.64

Again, it may be objected that recurrence in the family is
no criterion of the infection of the house, for it may be that
the family is largely immune. This would certainly be
much less true of those members of the family who went
away during the sickness. Of 634 minors who thus re-
turned to the non-disinfected house only 2 were taken sick.
So, too, if terminal disinfection is such an important matter,

250 THE SOURCES AND MODES OF INFECTION

its neglect should be followed by an increase in the disease.
As a matter of fact, after disinfection was abolished diphthe-
ria diminished, until at one time in the following August
there was not a single reported case in the city. Again in
August, 1908, the disease was reduced to a single recog-
nized case. There has been a marked increase since, common
to many parts of New England, and the excess in Boston,
where disinfection is practiced, has been greater than
in Providence. Still another test of the danger from the
omission of disinfection is the extension of the disease to
other families in the same house. Of 2592 such families
with susceptible children only 19 or 0.73 per cent were in-
vaded by the disease within two months after the termi-
nation of isolation in the non-disinfected apartment. It
may be suggested that, though no official disinfection was
practiced, the houses were perhaps thoroughly cleansed by
the occupants. I must confess that I hoped that the aban-
donment of official disinfection would cause people to do
more cleansing for themselves, but there has thus far been
little improvement, and as a large part of our diphtheria
occurs among poor and ignorant people, many of whom are
recent immigrants, house cleaning by the family can scarcely
be expected to be very efficient at present. It appears,
then, that this experiment shows that house infection in
diphtheria is in Providence a negligible factor in the dis-
semination of the disease.

Disinfection Unnecessary after Scarlet Fever. — So sim-
ilar, from an epidemiological standpoint, are scarlet fever
and diphtheria that I have been gradually abandoning
disinfection after the former disease also. The following
shows the recurrences where there was and where there was
not official disinfection.

The first table shows the number of recurrences within
60 days after disinfection for scarlet fever, the number of
infected families, and the rate of recurrence during the
years 1904-09.

INFECTION BY FOMITES

251

YBAB.

Infected
Families.

Recurrences.

Ratio.

1904..
1905

868
298

12
2

1.38

67

1906

398

9

2 26

1907

540

8

1 48

1908

273

3

1 09

1909

52

3

5 77

Total

2429

37

1 52

During the last four years the recurrences where there
was no disinfection were as follows:

YEAR.

Infected
Families.

Recurrences.

Ratio.

1908..

40

1

2 50

1909

377

10

2 65

1910

389

5

1 29

1911

434

4

92

Totals .

1240

20

1 61

As in diphtheria so in scarlet fever the infection of well
persons who return to the house after the termination of
isolation may perhaps be a measure of the value of terminal
disinfection. In Providence, during the years 1887-89, of
1816 persons, mostly under 14 years of age, who had never
had scarlet fever, and who returned home after disinfection,
33, or 1.8 per cent, developed the disease. Of 207 persons
who, since 1908, have returned to non-disinfected houses,
3, or 1.4 per cent have contracted scarlet fever. During
the years 1908-10, of 138 persons so returning none were
attacked. The only apparent recurrences were in 1911.

Again one may examine the recurrences in other families in
the same house. Of 1329 such additional families in 1904-
08 there was recurrence within two months after disinfection

252 THE SOURCES AND MODES OF INFECTION

in 21, or 1.6 per cent. Of 1005 families in 1909-11, living
in houses where there had been no disinfection, there were
recurrences in 15, or 1.5 per cent.

A further comparison showing the results with and with-
out disinfection can be made with the data from Baltimore.
To make these comparable to the Providence figures there
have been included, in the latter, recurrences in other fam-
ilies in the same dwelling. The ratio is calculated on the
number of cases instead of on the number of families as in
the table next preceding.

BALTIMORE. DISINFECTION.

YEABS.

Cases.

Recurrences.

Ratio.

1903-09

5792

84

1 44

PROVIDENCE. No DISINFECTION.

YEABS.

Cases.

Recurrences.

Ratio.

1908-11

1801

31

1 72

The health commissioner of Buffalo * reports, as illustrat-
ing the efficacy of disinfection, that after 3029 cases of
scarlet fever where terminal disinfection was practiced and
tested, there were 117 recurrences, or 3.8 per cent.

Disinfection Unnecessary in Other Diseases. — In any
disease which is widely distributed, and in which there is
ample opportunity for extension from mild cases, from car-
riers, from cases in the early stages, and from convalescents,
the absolute disinfection of all possible fomites at the as-
sumed termination of the sickness will probably have no
influence in checking the disease. Thus I can see no use in
disinfecting after measles, whooping cough, influenza, pneu-
monia or cerebro-spinal meningitis, and I think that this
1 Rep. Dept. of Health, 1909, 20, 44.

INFECTION BY FOMITES 253

view is held by the majority of our health officers and epi-
demiologists. In fact it is the expressed opinion of most
health officers that disinfection after measles is unneces-
sary, as there is no evidence that fomites are a factor in the
spread of the disease. Disinfection after measles, which
was practiced in Aberdeen for twenty years, had no influ-
ence on the prevalence of the disease. When official dis-
infection after measles was temporarily suspended in New
York from January 10, 1908, to March 1, neither the sus-
pension nor the resumption of the practice appeared to
have any effect on the epidemic curve. Disinfection after
cerebro-spinal meningitis is also absurd, as the meningo-
coccus lives only a few hours, and carriers are almost the
sole means by which this disease is spread. This is coming
to be the accepted view and in a recent publication of the
U. S. Public Health Service 1 Frost makes no mention of
terminal disinfection among prophylactic measures. Much
disinfection after tuberculosis is also without reason. If a
patient has been living with his family and taking no pre-
cautions, certainly no amount of terminal disinfection will
atone for long-continued sanitary sin. If precautions are
taken, the phthisiologists tell us that there is no danger in
living with the patient, and if so, certainly there can be no
possible danger in living in his house after he is dead.
The only occasion for disinfection after consumption, then,
is when the apartments of a careless patient are to be occu-
pied by another family, and, after all, we have no evidence
to show how much danger there is even then. Yet many
health officers make disinfection after every death from
consumption' the first and often the only measure taken to
restrict tuberculosis.

Dr. William C. Woodward of Washington in the discus-
sion of a paper read before the American Public Health
Association presented some figures which he believes indi-

1 U. S. Pub. Health and Mar. Hosp. Serv., Pub. Health' Rep., 1912,
XXVII, 97.

254 THE SOURCES AND MODES OF INFECTION

cate that terminal disinfection after tuberculosis is a valu-
able prophylactic measure. He says that during a period
of about two years ending in May, 1910, 1650 houses in
Washington were disinfected because of tuberculosis. Sub-
sequent to the disinfection there were reported from the
disinfected houses cases of tuberculosis sufficient to give
an annual attack rate of 223 while the attack rate through-
out the remainder of the population was 414.

Even in smallpox there are bits of evidence to show that
fomites play an unimportant part. Jordan l refers to two
unvaccinated men who fumigated and handled infected
smallpox clothing for two months without contracting the
disease but who succumbed to it as soon as they came in
contact with patients.

School Disinfection. — When a number of cases of scar-
let fever or diphtheria develop in a school, the public is in-
clined to consider room infection the source and to demand
" disinfection." This subject has lately, perhaps owing in
part to the efforts of sellers of disinfectants, been much dis-
cussed in England, but fortunately the leading officials
refuse to be influenced by public clamor. Among others,
Kerr 2 has clearly set forth the reasons why the room can
rarely be at fault in school outbreaks of the contagious
diseases. In another place Kerr 3 reports 26 instances in
which scarlet fever patients in the desquamating stage were
found in school. In 2 instances the room was disinfected
after their removal and in 1 of these a secondary case de-
veloped 9 days later. In the other 24 instances there were 2
secondary cases. Hope4 says, "All evidence points strongly
to the fact that when children contract infectious disease in
school the channel of infection is not by means of school
desk or floor but by the personal infection of another child."

1 Jordan, Am. J. Pub. Hyg., 1910, VI, 755.

2 Kerr, Pub. Health, Lond., 1909, XXIII, 49.

* Kerr, Rep. Med. Off. Education, Lond., 1909, 59.

4 Hope, Rep. to Education Authority, Liverpool, 1909, 9.

INFECTION BY FOMITES 255

French Views on Disinfection. — Recently,1 at a discus-
sion in Paris, Comby, Courmont and Lemoine took the
ground that the active disinfection which has been insisted
on in that city in recent years has not reduced the mortality
from scarlet fever, diphtheria or measles, for which dis-
eases it is required. Comby 2 is emphatic in his contention
that it is persons, not things, that are the bearers of con-
tagion. The same position was taken by Semaine medicate
for February 14, 1906. Lemoine 3 has found disinfection
not so essential as has been claimed. At the hospital at
Val de Grace certain rooms were used for isolating single
cases of contagious disease. There was often such demand
for their use that cases of different diseases quickly suc-
ceeded one another, sometimes without any disinfection
and often with slight disinfection by washing the walls as
high as could be reached. Sixty-five cases of scarlet fever,
.41 of measles, 25 of mumps, 31 of diphtheria, 4 of smallpox,
1 of chicken pox and 136 of other diseases are reported as
rapidly succeeding one another in these rooms. One case
of measles and 4 of scarlet fever developed in the rooms.
Often it was necessary to shift entire wards of twenty to
thirty beds, with only a slight attempt at disinfection.
This was done a good many times during nine years with-
out bad results. We have had a similar experience at the
Providence City Hospital where no attempt is made to
disinfect the walls and ceilings of the cubicles in which
different diseases rapidly succeed one another, and there
has never been an instance in which infection could pos-
sibly be attributed to neglect of this rite.

English Views. — There is a growing tendency on the
part of many English health officials to attach less impor-
tance to the part played by fomites in the spread of disease
and consequently less value to terminal disinfection as

1 Bull, et m&n. Soc. meU d. h6p. de Par., 1909, n. s., XXVII, 585.

2 Comby, Presse med., Paris, 1909, XXVII, 249.

3 Lemoine, Rev. d'hyg., 1907, XXIX, 1057.

256 THE SOURCES AND MODES OF INFECTION

ordinarily practiced. Richards l does not consider disin-
fection of the room necessary, provided the floors are
washed and things which came in contact with the patient
are cleansed and disinfected. Butler, Barlow, Hogarth
and others have expressed a growing belief that fomites
play a comparatively unimportant part in the spread of
disease. The views of Hope and Kerr are referred to
above. The writer has received a number of letters from
English and Colonial medical officers of health indicating
agreement with the main proposition of this chapter.

American Views. — It is not, of course, to be expected
that the majority of health officers would be willing to
abandon terminal disinfection as now practiced. There are,
however, very many of the most thoughtful, and at the
same time most conservative, who agree that the present
methods of disinfection should either be made really effi-
cient or else given up. They are ready to admit that
fomites are a far less important factor in the spread of dis-
ease than was formerly thought, but they think that even
minor factors should if possible be efficiently controlled.
A few health officers are willing to test the matter for
themselves and have given up or are gradually omitting
gaseous disinfection after scarlet fever and especially after
diphtheria. The health officer of one important American
city was trying gradually to abandon terminal disinfection
until his progressive attitude was frowned upon by the
state health officials. In Newton disinfection is not com-
pulsory.

When Terminal Disinfection is Desirable. — When a new
or rare disease invades a locality, it may at times be desir-
able to take extraordinary precautions to prevent its exten-
sion, which would be entirely useless if the disease were
established. This, it is true, is not the popular or legal way
of regarding preventive measures, but it is the scientific
one. If a case of smallpox should occur in Providence,
1 Richards, Pub. Health, Lond., 1909, XXIII, 42.

INFECTION BY FOMITES 257

which has been free from it for several years, it would be
worth while, perhaps, to expend considerable time and
money in disinfection, even though the chance of infection
from the room or goods might not be one in a thousand.
But if there were hundreds of cases of measles in the city,
it would be folly to go to the same trouble and expense for
each case, even if the chance of infection were ten times as
great. A spark in the dry grass should be stamped out at
any cost, but it is useless to waste time in extinguishing the
smoldering flames left here and there as the line of fire is
sweeping across the prairie.

Objections to Disinfection. — There are several objec-
tions to the present practice of terminal disinfection. One
is that it only partially disinfects. If disinfection is to be
honestly and efficiently applied, methods must be changed.
But even as now practiced, disinfection is expensive. Many
cities employ disinfectors, with horses and apparatus, while
their laboratory languishes, their medical inspection is poor
and their diphtheria patients must secure antitoxin as best
they can. Another serious objection to routine terminal
disinfection is that it misleads the public. They are given
a false sense of security, and they are encouraged in the old
belief that it is things, not persons, which are dangerous.
We can never successfully preach the truth about carriers,
or teach the necessity for stricter personal cleanliness, so
long as we continue to make so much of a fetish of the
practice of disinfection.

Cleanliness versus Disinfection. — I am here referring to
terminal disinfection, which is often only a fumigation, or
smell-producing process. The continued practice of clean-
liness all through the sickness, and indeed at all times, by
which the transference of fresh infective material may be
prevented, is another matter, the value of which was con-
sidered in the preceding chapter.

Conclusions. — It seems to me, in view of the considera-
tions here presented, that we are justified in concluding that,

258 THE SOURCES AND MODES OF INFECTION

1. There is no good epidemiological evidence that any
diseases except those due to spore-forming bacteria are to
any great extent transmitted by fomites.

2. Judging from our experience with yellow fever most of the
alleged evidence of infection by fomites is no t to be relied upon.

3. Even if all the alleged fomites infection is real, only a
very small part of contagious disease is traceable to this
source.

4. The theory of fomites infection was an a priori not an
a posteriori theory, and is no longer demanded.

5. Other modes of transmission so much more satisfac-
torily account for the spread of disease, that there seems
to be really little opportunity for infection by fomites.

6. Laboratory investigation shows that fomites infection
with spore-forming bacteria is common; that such infection
in typhoid fever, tuberculosis, diphtheria and with other
resistant organisms doubtless sometimes takes place; that
it is possible in cholera and plague, while such infection in
gonorrhea, influenza, cerebro-spinal meningitis and pneu-
monia must be practically impossible.

7. Experiment and epidemiological observation have
demonstrated that fomites infection is practically unknown
in yellow fever and is probably so in the other diseases car-
ried by flying insects.

8. Experiment and observation show that fomites are of
little moment in the diffusion of bubonic plague, and of no
moment in the extension of Mediterranean fever, both
formerly believed to be spread in this way.

9. Observations made in Providence indicate that fo-
mites infection is of no practical importance in the diffusion
of diphtheria.

10. Observations in Providence indicate that fomites are
of no practical importance in the diffusion of scarlet fever.

Finally, it may be affirmed that the evidence has been
rapidly accumulating that fomites infection is of very much
less importance than was formerly believed.

CHAPTER VI.

INFECTION BY AIR.

Reasons for Former Belief. — From time immemorial the
air has been considered the chief vehicle of infection. This
was but natural, for until recently the virus of the infectious
diseases was believed to be gaseous, or at least readily dif-
fusible, and readily borne by air currents. The infective
material was supposed to be given off in the expired breath,
and to emanate from the surface of the body and from moist
soil and decomposing matter of all kinds. Contagious dis-
eases were known to arise without any apparent connection
with other cases, and what could be more natural than to
assume that the invisible, imponderable materies morbi is
mixed with and carried by the air? Moreover one of the
most widespread and best known diseases, malaria, was
shown by a great mass of clinical evidence to be an air-borne
disease, and yellow fever, another infectious disease of great
importance, was also on good grounds believed to be trans-
mitted in the same manner. What seemed to be well estab-
lished for these two diseases was assumed on much more
slender evidence to be true of most, if not all, the infectious
diseases. It is true in a certain sense that the two diseases
just mentioned are air-borne, that is, they are transmitted by
small insects, which " diffuse," as it were, from their breeding
places and are readily wafted by air currents. It is not in
this sense that the term air-borne is used in this chapter, but
the inquiry here made is whether the virus of the infectious
diseases is borne by the air, either free or attached to small
particles of inanimate matter.

Then, again, the first symptoms of measles, and often of
influenza, are connected with the nose> diphtheria appears to
be a throat disease, while consumption and pneumonia infect

259

260 THE SOURCES AND MODES OF INFECTION

the lungs. What is more natural than to assume that the air
which bathes these parts is the vehicle of infection ? But since
it has been shown that the pneumococcus is constantly found
in the blood in pneumonia, and has been demonstrated before
the initial chill, and since tubercle bacilli readily reach the lungs
through the circulation, the force of this argument is lost.

Omitting the insect-borne diseases, let us see first what
epidemiological evidence there is that the contagious diseases
are air-borne.

Smallpox Air-borne from Hospitals. — Smallpox is be-
lieved to be more widely air-borne than is any other disease.
The modern doctrine of the aerial transmission of smallpox
received its greatest support from the investigations of Power1
in 1881 concerning the conditions about the Fulham Hospital
in London. That smallpox could be carried by the air long
distances had been claimed by many before that time, for
Dr. Waterhouse of Cambridge, more than one hundred years
ago, believed the disease had been wafted across from Boston
to Charlestown, though later he was inclined to deny this
mode of transmission. Power showed that smallpox had not
prevailed to any extent in the vicinity of the Fulham Hospital
before the hospital was opened, and that on a number of occa-
sions soon after it was occupied by a considerable number of
patients the disease began to develop in the neighborhood.
Furthermore he showed that there was a progressive decrease
in the amount of smallpox as the distance from the hospital
increased, the alleged influence of the hospital extending to
at least one mile. It was furthermore shown by the doctor
that this distribution of the disease was uniform in every
direction from the hospital, that is, in every quadrant of the
circle surrounding it. Later investigations at this hospital
yielded similar data.2 Smallpox almost always developed
in the vicinity whenever the hospital was occupied by from

1 Rep. Med. Off. Loc. Gov. Bd., Lond., 1880-81, X, Supl. 302.
* Rep. Med. Off. Loc. Gov. Bd., Lond., 1884-85, XIV, 55, 1885-86,
XV, 111.

INFECTION BY AIR 261

eighty to one hundred acute cases. But on one occasion, at
least, there was no outbreak even when the hospital was fully
occupied. The conclusion was that when a considerable
number, eighty to one hundred, acute cases of smallpox are
gathered in a hospital, there is great danger that the virus
of the disease will be carried by the air a mile or more from
the hospital in quantity sufficient to infect persons at that
distance. Power considered that the hospital was thus the
focus of infection on one occasion when there were only
twenty patients, and at another time when there were only
nine, of which five were acute. It was thought, from a study
of the conditions at the time of the outbreak, that the dis-
semination was favored by still, damp weather. A somewhat
similar though not generally so well defined or nicely graded
distribution of smallpox around hospitals was believed by
many to have been demonstrated in the cases of theHomerton,
Deptford, and Hempstead hospitals. It was natural that
twenty-five years ago, when knowledge of modes of infection
was far more vague than at the present time, and with such
evidence at hand, the officers of the Local Government Board
should have been firmly convinced of the importance of aerial
transmission in the spread of smallpox, and that this opinion
should have been shared by many medical officers of health.
Subsequent to the investigations referred to above, outbreaks
due to the spread of this disease from hospitals are said to
have occurred, among other places, at West Ham, 1884-85,1
Nottingham, 1887-88,2 Oldham, 1888 and 1892,3 Warrington,
1892-93,4 Bradford, 1893,5 Liverpool, 1902-03,6 and in Gates-

1 Rep. Med. Off. Loc. Gov. Bd., Lond., 1886-87, XVI, 97.

2 Whitelegge, Practitioner, Lond., 1888, XLI, 65.

1 Report on the Health of Oldham, 1892, by Niven; also Niven,Pub.
Health, 1892-93, V, 324, 366.

4 Gornall, Rep. on the Epidemic of Smallpox in the years 1892-93
In Warrington, 1885, 111.

6 Evans, Brit. M. J., Lond., 1894, II, 356.

6 Reece, Special Rep. Loc. Gov. Bd., Lond., No. 208, 1905, Smallpox
in Liverpool.

262 THE SOURCES AND MODES OF INFECTION

head and Felling, 1903-04.1 In Glasgow2 Chalmers states
that smallpox seemed to develop around the hospital when
it contained many patients, but that this did not invariably
occur. When the hospital was removed to another location,
it again appeared to be a focus of disease. Much has been
made of the alleged aerial transmission of smallpox from the
ships lying in the Thames below London and used for the
reception of cases of that disease from the metropolis. Bu-
chanan3 and Thresh4 attempt to show that after the ships
were brought into use the disease was carried by the air to
the Essex shore at Purfleet and West Thurrocks in the Orsett
Union. It is true enough that the incidence of the disease
was very great in these districts, but it is difficult to under-
stand why it is not as well explained by contact infection, as
were hundreds of similar outbreaks in England and the United
States. The chief evidence on which the theory of aerial
infection is based is the existence around the hospital, in
every quadrant, of a graduated incidence of the disease. No
such evidence is presented in this instance, and the area of
infection attributed to the ships lies only in one direction
from them. Smallpox appeared on the shore nearest the
ships, and then gradually extended to a distance of two or
three miles. This sort of extension is just what would be
expected in contact outbreaks. If air-borne, the near and
distant communities should have been affected at the same
time. It was claimed by Dr. Thresh that the influence of
the ships could be noted at a distance of four or five miles.
It was also claimed that several vessels anchored near the
hospital ships developed smallpox twelve days later. That
ships leaving London during the period of the extensive out-

1 Buchanan, Special Rep. Loc. Gov. Bd., Lond., Smallpox in Gates-
head and Felling, 1904.

8 J. Royal San. Inst., 1905, XXVI, 212, and Tr. Epidemiol. Soc.,
Lond., 1904-05, n. s., XXIV, 151, 244.

3 Rep. Med. Off. Loc. Gov. Bd., Lond., 1902-03, XXXII, 81.

4 Thresh, Tr. Epidemiol. Soc., Lond., 1902, n. a., XXI, 101.

INFECTION BY AIR 263

break in that city should occasionally carry smallpox with them
is not remarkable. Finally it was admitted that surrepti-
tious communication with the ships occasionally occurred.

Aerial Convection Denied by Some. — Many sanitary offi-
cials did not, and do not, accept these conclusions, and nu-
merous instances are given where the disease has not extended
from hospitals. Thus Renney1 says that in 1883-84 he saw
300 cases of smallpox treated in wards which were situated
between twenty and two hundred and twenty-four feet of
other hospital wards, schools and houses, without any exten-
sion, though only the school was protected by vaccination.
At another time he saw a considerable number of cases cared
for without harm in a ward from forty to one hundred and
thirty-eight feet from other occupied buildings. So Wilson at
Rugby2 had a hospital within a few yards of a much fre-
quented road with no untoward results. At many other
times he has seen smallpox hospitals maintained in close
proximity to other occupied buildings without aerial trans-
mission resulting. Boobbyer3 treated 20 cases near a high-
way where a thousand workmen passed daily, and not a case
was contracted from them. Ker4 at Edinburgh had a small-
pox hospital in connection with a general hospital, and close to
other institutions, and with a population of 3000 persons liv-
ing within a mile circle. There were only 4 cases within this
circle, of which 2 were known to be contracted elsewhere.
Dr. Thorne Thorne5 stated that in two instances only had
he seen evidence of the aerial extension of smallpox from a
hospital, namely, at Maidstone and at Stockton, while he had
seen numerous instances where there was no extension,
notably at Leeds and Nottingham.

Renney, Jour. Roy. San. Inst., 1905, XXVI, 210.
Wilson, Brit. M. J., Lond., 1905, II, 630.
Boobbyer, Tr. Epidemiol. Soc., Lond., 1905, n. s., XXIV, 219.
Ker, Tr. Epidemiol, Soc., Lond., 1905, n. s., XXIV, 174.
Thorne Thorne, Rep Med. Off. Loc. Gov. Bd., Lond., 1880-81,
X, Supl. 40.

264 THE SOURCES AND MODES OF INFECTION

At Manchester Niven1 reported only 13 smallpox cases
out of a population of over 40,000 living within a half mile
to a mile area around the hospital and none among the 606
persons living within the half-mile circle. Other instances
can be given where smallpox hospitals have not infected
their neighborhood, and these facts should have some weight,
though according to the advocates of the theory aerial trans-
mission is to be expected only under certain conditions of
the atmosphere. There is not much evidence bearing upon
this subject to be obtained in the United States, partly
because smallpox hospitals have been much more rarely
situated in thickly populated districts and partly because
less attention has been given to the subject here than in
England.

American Evidence against Theory. — In Philadelphia it
is claimed2 that the hospital has been the source of smallpox
in its neighborhood. Thus in one outbreak in the municipal
ward in which the hospital was situated the case rate was
61 per 10,000, nearly twice that of any other ward, and it
decreased as the distances from the hospital increased. The
same conditions were noted in another outbreak.

In Boston in 1902-03 there was some discussion as to
whether the disease spread from the hospital, which was
on a busy street and near many occupied buildings. The
evidence was that there was not much smallpox in the
neighborhood, and also that contact infection from the hos-
pital could not be excluded.

In New Orleans3 a large number of cases of smallpox were
in 1900 treated in a hospital in close proximity to a dense
population, but without evidence of extension. Dr. Theard
writes me that his observations, extending over nine years
since that time, have only strengthened his views then ex-
pressed, namely, that smallpox virus is not carried from hos-

1 Niven, Tr. Epidemiol. Soc., Lond., 1905, n. s., XXIV, 157.

8 Rep. Bu. of Health, Phila., 1903, 29.

1 Rep. Bd. Health of the City of New- Orleans, 1900-01, 33.

INFECTION BY AIR 265

pitals by the air. In Brooklyn smallpox is cared for at the
contagious-disease hospital, in pavilions about twenty feet
from those occupied by measles and scarlet-fever patients.
There has been no extension of the disease, though this may
be largely due to the effort to keep the other patients well
protected by vaccination. But there is also a considerable
population within a mile of the hospital which has never been
injuriously affected by it. In Providence the smallpox hos-
pital is distant only four hundred or five hundred feet from a
number of cottages and an excursion ground frequented by
hundreds of persons daily. It is true that only about a dozen
patients have ever been there at one time, but it is hard to
understand why ten patients should not be more dangerous
at five hundred feet than one hundred patients a mile dis-
tant. In Detroit, Dr. Kiefer writes me, there has been no
extension from the hospital, and in Chicago * there was less
smallpox around the hospital than in other parts of the city.
Theoretical Objections. — I have been led to question
this theory of the aerial transmission of smallpox for various
reasons. From what is known of the nature of the virus of
so many other diseases it seems highly improbable that they
are carried any great distance by the air, and in fact it is
only for smallpox that this mode of transmission is claimed.
But smallpox virus is certainly solid matter, and it certainly
after a time loses its vitality, and in all respects other than
the one under consideration it behaves not unlike the mate-
ries morbi that we are better acquainted with. Again, it
would be most remarkable, if the disease extends from, say,
one hundred cases to the distance of a mile with sufficient
intensity to infect many persons, that it should not extend
one hundred feet from ten cases or even from one case. Why
should we not expect aerial infection frequently to operate
at short distances from single cases ? Yet such transmission
does not occur unless it be with great rarity. How rare it ia
for any claim to be made that this disease has been carried
1 Rep. Health Dept., Chicago, 1907-10, 60.

266 THE SOURCES AND MODES OF INFECTION

across the street from house to house, and how unique a rigid
demonstration of such an occurrence would be! How often
a single case in a crowded lodging house, ship's steerage, or
hospital ward, fails to infect others! Yet we are asked to
believe that one hundred cases can give rise to a whole circle
of cases a half mile away. Either the amount of virus must
depend upon the number of patients, or it must under hospi-
tal conditions develop in some marvelous way outside of
the body.

Contradictions in Claims. — If the evidence adduced in
favor of this theory is examined, several suspicious circum-
stances are noticed. It is very curious that a material sub-
stance should be borne by the air without reference to air
currents ; yet in the earlier reports by Powers all evidence of
such currents was lacking. It is true that in later reports
the wind has been claimed as a factor, as at Gateshead and
Felling, but the evidence in this case has been made valueless
by more detailed search for the origin of the cases. At Liver-
pool the disease was distributed in different directions around
the three hospitals, so that if air-borne it must have been
independent of air currents. Savill at Warrington even
claimed that the virus diffused against the wind. The reports
of the Fulham Hospital give one the impression that the
virus of smallpox must diffuse like a gas, which certainly is not
thinkable. If air-borne at all, it must be carried as is dust or
as are liquid particles. If the virus does diffuse in all direc-
tions like a gas, the intensity of the infection should diminish
according to the square of the distance, which it did not do
at Fulham and Liverpool. Whether it diffuses or is carried
like solid particles, the houses nearest the hospital should be
by far the most intensely infected. Solid particles are speed-
ily, under ordinary conditions, precipitated to the ground, as
one may easily note by observing a cloud of dust of any kind.
The particles of smallpox virus ought, then, to work their
chief havoc close to the hospital. Yet this incidence was not
always the case, even at Fulham, and similar absence of near-by

INFECTION BY AIR 267

infection was noted at Stockwell1 and Liverpool.2 Again,
quite a number of instances are given where large institu-
tions, like schools, workhouses and general hospitals, have
been located within the area alleged to have been severely
infected from the hospital, and yet have nearly or entirely
escaped. It is curious, too, that in the only instance I have
noticed in which the sex of the patients in the infected area
was given, namely Fulham,3 twenty-four were male and
seventeen female. As so many more men are away from their
homes at work, a much larger female population must be
exposed to the hospital influence, and the female patients
ought to be more numerous than the male. In most out-
breaks of smallpox the male patients are more numerous,
because the men move about more, and are thus more likely
to be exposed to cases of disease. Again, it is remarkable
that extension should be more likely to take place from
acute than from chronic cases. In the former it is probable
that the moist mucous membrane is the only source of infec-
tion, while in the latter the dried crusts are known to be
infectious.

Distribution of Cases Opposed to Theory. — It is thus seen
that there are a number of facts and a number of theoretical
considerations opposed to the theory of aerial transmis-
sion. That smallpox is distributed with decreasing inten-
sity around smallpox hospitals is not a demonstration that
the hospital is the cause, for as even the advocates of the
theory admit, sucti circles of infection can be drawn around
other points in a city during epidemic times, as was indeed
shown by Hope in Liverpool in 1902-03 and Clayton at Gates-
head in 1903-04. In fact most outbreaks, not only of small-
pox but also of scarlet fever and diphtheria, are in a general way
arranged around a center, with more cases toward the center

1 Rep. Roy. Com., Smallpox and Fever Hospitals, Lond., 1882, 92.

2 Hope, Observations by the Med. Off. Health on the Report of
Dr. Reece on Smallpox at Liverpool. C. Tinling & Co., 1905, 11.

• Rep. Med. Off. Loc. Gov. Bd., Lond., 1880-81, X, Supl.

268 THE SOURCES AND MODES OF INFECTION

and fewer toward the periphery. It is not remarkable that
occasionally a smallpox hospital is found near the center of
such a localization of disease. As favoring this chance, it must
be remembered that the population near a smallpox hospital
is likely to be of the poorer classes, upon whom the weight of
this disease most often falls. That surrounding outbreaks
occur chiefly after the hospital has been occupied, is only to
be expected, as the hospital is occupied by a number of cases
only in epidemic times. Much stronger evidence would be
offered by the advocates of this theory if they could show
that no other explanation of the origin of the cases could be
found than hospital infection. This they are not able to do,
and in the absence of such evidence, and in the face of the
evidence against the theory, the theory must be considered
not demonstrated.

Cases often traced to Other Sources. — A good deal of
evidence in regard to the influence of hospitals in the spread
of smallpox was collected in England at a time when there
was no registration of the disease, when the frequency and
importance of mild cases were not recognized as at present,
and when its administrative control was not so complete.
Thus it was stated that many cases of smallpox walked to
the Homerton Hospital to apply for admission, ambulance
drivers stopped at public houses, children of the neigh-
borhood rode on the steps of the ambulance and the patient's
friends inside. It would not be surprising if, under such con-
ditions, smallpox spread by contact — and indeed it was admit-
ted that this was a factor. It is interesting to note that even
Power found personal exposure the cause of nine out of thirty-
two cases near Fulham. In many outbreaks, where there is
no question at all of hospital infection, to trace the source of
such a proportion of cases is all that can be expected. Again,
in the block of houses nearest the same Fulham Hospital
Dudfield l showed that twenty of forty-one cases were due to

1 Dudfield, Rep. Roy. Com., Smallpox and Fever Hospitals, Lond.,
1882, 101.

INFECTION BY AIR 269

contact infection. According to Clayton,1 the medical officer
of health of Gateshead, of the fifty-six cases of smallpox
within one-half mile of the hospital, on which Buchanan bases
his conclusion that the disease was carried by the air, fifty-
two were traced to contact infection. Clayton in his report
on this outbreak very clearly shows the fallacy of most of
the arguments presented by Buchanan. In Liverpool it was
shown by the advocates of aerial transmission that within
one-quarter of a mile of the Parkhill Hospital the rate of
incidence of smallpox was five hundred and twenty-six per ten
thousand houses, while in the city outside of hospital areas
it was only eighty-five. A detailed study shows that this
apparent high rate depended on only nine patients in one
hundred and seventy-one houses, and Hope shows that of these
nine, four were known to be due to direct exposure to other
cases. A careful study of the report by Reece2 of this Liver-
pool outbreak is well worth while by all interested in this
subject. A most ingenious use has been made of the facts,
but an impartial critic must see that the conclusions arrived
at are entirely unwarranted. If one is still in doubt, he should
read the report of Hope,3 the medical officer of health of
Liverpool, which clearly and briefly refutes all the arguments
of the government inspector.

Conclusions concerning Smallpox. — It appears that the
evidence for the aerial transmission of smallpox from hospi-
tals consists solely of the alleged distribution of the disease,
at a gradually decreasing rate, around the hospital, the exist-
ence of the cases being assumed to be otherwise unexplained.
It will be noted:

1. That there are comparatively few instances of such dis-
tribution recorded.

1 Clayton, J. Roy. San. Ins., 1905, XXVI, 199.

2 Reece, Rep. Local Gov. Bd., Lond., No. 208, Smallpox in Liverpool,
1905.

3 Hope, Observations by the Med. Off. Health on the Report of
Dr. Reece on Smallpox at Liverpool. C. Tinling & Co., Liverpool, 1905

270 THE SOURCES AND MODES OF INFECTION

2. That many instances are noted where there was no such
diffusion.

3. That in some of the alleged instances, as at Fulham,
Gateshead and Liverpool, a large number of the surrounding
cases have been shown to be due to contact infection.

4. That in the long run the amount of infection around
the hospital should diminish according to the square of the
distance. This it does not do, but it diminishes irregularly,
just as it does in most outbreaks of this and other diseases
due to contact infection.

5. There may sometimes be contact infection from the hos-
pital. The surrounding population is often of the poorer sort,
and is consequently particularly subject to the 'disease.
Contact infection and chance may be sufficient to account
for those instances where a smallpox hospital is the center
of a local outbreak.

The evidence in favor of the aerial transmission of small-
pox from hospitals is so slight that it should never influence
a municipality in its selection of a hospital site.

Chicken Pox. — Caiger,1 while able successfully to isolate
scarlet fever, diphtheria and whooping cough in his hospital
wards by the cubicle system, had several transfers of chicken
pox, and concludes from his experience that this disease is
frequently air-borne. Others in England and in the United
States have had difficulties in isolating chicken pox. Good-
all,2 however, has isolated this disease in cubicles with only
1 cross infection which he does not think was air-borne.

Scarlet Fever believed to be Air-borne. — Scarlet fever also

is generally believed to be an air-borne disease. One reason

for this is doubtless because until recently the desquamating

epidermis was considered to be the chief vehicle of infection.

t As the epidermis comes off to a large extent as very fine light

particles, it was but natural to assume that these would be

1 Rep. Metropol. Asylums Bd., 1907, 1908.

2 Pub. Health, Lond., 1911-12, XXV, 17.

INFECTION BY AIR 271

readily carried by the air. Recently much clinical evidence
has accumulated which indicates that the epidermal scales
are not infectious, and this has in turn developed doubts as
to the disease being commonly air-borne. Whether or not
the epidermis is infectious, there seems to be no really good
evidence that the disease is caused by air-borne infection. On
the contrary, there is considerable evidence that it is not air-
borne.

Scarlet Fever in Hospitals. — The writer, like every health
officer, has frequently noted that a case of this disease may
remain in school or hospital ward for days, or sometimes for
weeks, without another case developing, or at most only one
or two cases. Such facts indicate that the disease is not
easily air-borne. Visitors to fever hospitals do not contract
scarlet fever. Thus, of three hundred to four hundred non-
immune students who visited the scarlet-fever wards of the
Philadelphia hospital, remaining in the ward from twenty
minutes to an hour, not one contracted the disease.1 Often-
times scarlet fever does attack other patients in hospitals,
but it is in a manner to indicate contact rather than air-borne
infection. When contact infection is rigidly guarded against,
as in the Pasteur Hospital in Paris and in many English
hospitals, scarlet fever may be, and is, treated in the same
ward with other diseases without cross infection. The failure
of contagious disease to spread in hospitals when cpntact
infection is guarded against was referred to in some detail
in the chapter on contact infection, and is a striking demon-
stration of the small part played by aerial infection in the
transmission of the common contagious diseases.

Scarlet Fever in Dwellings. — I have been much impressed
by the fact that scarlet fever and likewise diphtheria do not
extend from one family to another in the same house. Most
people in Providence live in houses of two or three stories,
rarely more, with one or two families on each floor. Of 4306

1 Welch and Schamberg, Acute Infectious Diseases, Phila., 1905,
346.

272 THE SOURCES AND MODES OF INFECTION

" other families " living in the same house with scarlet-fever
families, only 6.8 per cent were invaded. Investigation has
shown that with very few exceptions the infection takes place
through close intercourse before the disease is recognized or,
more rarely, after the isolation has been terminated. Most
of the disease in the "other families " develops within a few
days after the report of the primary case, and is doubtless
due to contact infection before the disease is recognized.
Between the end of the second week and the termination of
isolation, the disease extends to other families in the house
in only 0.6 per cent of the cases, and in most of these it is
known that isolation is not carried out, and that there is free
intercourse between the families. If the disease were air-
borne, it would certainly pass from one family to another in
the house, which it does not do.

Scarlet Fever and Outdoor Air. — If scarlet fever is not
air-borne from family to family in the house, one would not
expect it to be borne from house to house by the air. Yet
such a claim is sometimes made, and even that the virus of
the disease may thus be transmitted a considerable distance.
A number of the reports of the health department of Phila-
delphia contain shaded maps purporting to show an excess
of this disease, as well as of smallpox, in those parts of the
city near the hospital. I do not think that much value at-
taches to such maps, for there are too many factors involved,
and very rarely is the intensity of the disease as great close
to the hospital as the theory demands. Moreover, around
very many hospitals no such distribution of the disease can
be shown. Thus Tarnissier,1 in Paris, found that the En-
fants Malades and Trousseau hospitals could not be con-
sidered foci of infection. The same is true of the scarlet-fever
wards in Providence, in Detroit and in Boston. -In the latter
city,2 for the period studied, there were no cases of the disease
within one-eighth of a mile of the hospital, while in the next

1 Tarnissier, Semaine med., 1903, 267.

2 Med. and Surg. Rep., Bost. City Hosp., 1897.

INFECTION BY AIR 273

eighth of a mile circle there were sixty-eight cases, in the next
seventy-one, in the next seventy-five and in the next seventy-
two.

Where various contagious diseases are treated in different
wards of the same hospital there is sometimes cross infection.
But this occurs so irregularly as to time and place, and is so
limited in amount, that it can scarcely be attributed to any-
thing but contact infection. As most of the physicians and
nurses in our contagious hospitals have no appreciation of
what true medical asepsis really means, it is surprising that
we see as little cross infection as we do. If scarlet fever does
not spread within the walls of the Pasteur Hospital, it would
indeed be marvelous if it should extend to the neighboring
houses. If it does not pass from family to family in the same
house, it would be most surprising if it could be wafted by the
air over large areas around the Philadelphia Hospital.

Diphtheria and Sewer Air. — Diphtheria was formerly be-
lieved to be a filth disease, and it was also believed that air,
especially sewer air, was frequently the vehicle of infection.
Graham-Smith refers to this,1 and shows that there is no
foundation for this belief, and that diphtheria bacilli have
never been found in sewer air. He says that Shattock culti-
vated bacilli of low virulence in sewer air for two months, but
could not thereby increase their virulence. As I was, years
ago, prejudiced in favor of the filth origin of this disease, I
gave the matter careful consideration in my investigation of
cases, but was never able to find any evidence that sewer gas
was an etiological factor. Indeed my observation of diph-
theria and typhoid fever had as much to do with my dis-
carding the filth and sewer-gas theories as had the slowly
accumulating mass of bacteriological evidence.

Diphtheria in Hospitals and Dwellings. — From an epi-
demiological standpoint diphtheria and scarlet fever are much
alike. As the latter disease has been supposed to be air-

1 Nuttall and Graham-Smith, The Bacteriology of Diphtheria,
Cambridge, 1908, 321.

274 THE SOURCES AND MODES OF INFECTION

borne from person to person, so has the former, and there is
the same lack of positive evidence for both diseases ; and the
evidence against the theory is much the same for diphtheria
as for scarlet fever. As is stated by Graham-Smith, bacterio-
logical evidence is all against diphtheria being an air-borne
disease except in rare instances, yet probably most medical
men and most health officers consider that the disease is
commonly spread in this way. But visiting students in hos-
pital wards do not contract it, and it does not spread when
cases of this disease are treated in pavilions together with
other diseases. In Providence I am certain, from a careful
study of about eleven thousand cases, that it practically never
extends from one family to another in a house except by per-
sonal contact ; and it does not extend from one hospital ward
to another through the air. At North Brother Island in New
York there is a diphtheria ward only a few feet from a tuber-
culosis ward; and Dr. S. A. Knopf tells me that there is no
cross infection. Similar conditions are noted in many other
hospitals. Yet Coutts1 recently suggested, without any evi-
dence, that certain cases of diphtheria were due to street dust,
and Cornell 2 attempted to show that the development of the
disease in a certain locality in Philadelphia was due to air-
borne infection. At best there was in the instances reported
by the latter only a possibility of aerial infection, with the
probabilities very much against it, while the facts as stated
did point very strongly to the existence of unrecognized cases
probably spreading the disease by contact.

Typhoid Fever and Sewer Air. — During the heyday of
the sewer-gas theory of disease, numerous outbreaks of
typhoid fever were supposed to have been traced to infection
by means of air from sewers and drains. A number of typical
reports are given by Roechling,3 and others may be found
scattered through medical literature. In none of these is real

1 Coutts, Pub. Health, Lond., 1906-07, XIX, 297.

2 Cornell, N. York M. J. [etc.], 1905, LXXXII, 1318.

» Roechling, Sewer Gas and Health, Lond. & N.Y., 1898, 30.

INFECTION BY AIR 275

proof given that the disease was thus caused; it was merely
a plausible hypothesis. Now in the light of present-day
knowledge of bacteria and sewer air it is no longer a plausible
hypothesis. For years past we have been able to trace most
of our outbreaks of this disease to water, milk, oysters or
other food, or to contact infection. In most instances they
could not be due to sewer air. Usually investigation shows
that house or institution outbreaks cannot possibly be due
to sewer air, and where such an hypothesis is permissible, it
usually appears highly improbable. I see almost every year
small house outbreaks of typhoid fever. There is rarely any
evidence of the escape of drain air into the house, and in
almost all instances such escape is impossible. Most of these
house outbreaks indicate contact infection, and in none can
contact infection be excluded. I have never seen the slightest
evidence that typhoid fever is ever due to sewer air, though I
began my public-health work with a fairly strong belief in
the danger from this source and sought diligently for evi-
dence of it.

Typhoid Fever and Dust. — It is also claimed that infected
dust may be the cause of outbreaks of this disease. This
mode of infection was considered to be of some moment in
the Spanish- American and Boer1 wars, and certainly bacterio-
logical evidence points to its possibility. Many outbreaks sup-
posed to be due to dust infection have been reported. Some
of these are referred to by Germano2 and Visbecq3 and in the
Report on Typhoid Fever in the war with Spain,4 but if the
original reports of these outbreaks are examined it will be seen
that the evidence is very weak indeed. Because the houses

1 Tooth, Brit. M. J., Lond., 1900, II, 1368; Tr. Clin. Soc., XXXIV,
1213.

2 Germano, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1897, XXIV,
403.

3 Visbecq, Arch, de med. et pharm. milit., Par., 1903, XLI, 536.

4 Abst. of Rep. on the Origin and Spread of Typhoid Fever in U. S.
Military Camps during the Spanish War of 1898, Wash., 1900, 215.

276 THE SOURCES AND MODES OF INFECTION

or apartments or barracks where an excess of typhoid fever
appears are situated not far from a place where possibly
infected feces are deposited, it is argued that the disease is
caused by the wind blowing over the spot alleged to be in-
fected. Quill 1 reports that typhoid fever was brought to a
certain garrison in India by a company of five thousand Boer
prisoners, many of whom were infected. The disease con-
tinued to spread among the prisoners for three months, until
there were from six to eight hundred cases. Then it slowly
appeared among the garrison, until there were twenty-four
cases. It was supposed to be caused by dust blown from the
latrines, though the possibility of fly-borne infection is men-
tioned; but an extraneous source, or unsuspected contact
infection, is highly probable, though neither is excluded or
even mentioned. Mewius 2 gives an excellent report of what
he considers an air-borne outbreak, but it appears rather to
have been due to contact infection, a typical outbreak of
what Winslow calls prosedemic infection.

The fact that typhoid fever, dysentery and cholera3 can
be treated in a well-managed hospital without spreading to
other patients is good evidence that these diseases are not
air-borne under such circumstances.

Infantile Diarrhea and Dust. — Newsholme,4 judging
largely from the fact that the summer diarrhea of infants
occurs with greater frequency during dry seasons, and in
towns with poor scavenging, infers that it is due, to some
extent at least, to the infection of milk and other foods by
dust. This also is the view of Hope,5 who states that in
Liverpool in six Septembers with an average rainfall of 13.8
inches there were 373 deaths from diarrhea, while in four-
Quill, Brit. M. J., Lond., 1902, I, 383.

Mewius, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1896, XXIII,
497.

Woodruff, J. Am. M. Ass., Chicago, 1905, XLV, 1160.
Newsholme, Pub. Health, Lond., 1899-1900, XII, 139.
Hope, Pub. Health, Lond., 1898-99, XI, 435.

INFECTION BY AIR 277

teen Septembers with 10.9 inches of rain the deaths num-
bered 573.

Influenza. — During the present pandemic of influenza,
especially at its commencement, it was frequently stated that
the disease was chiefly air-borne, and air-borne to great dis-
tances, even across the Atlantic. This view prevailed be-
cause the disease spread with such great rapidity, appearing
on this side of the Atlantic at about the same time that it
did in England, and developing almost simultaneously in a
large number of localities. The subject has been thoroughly
studied by Leichtenstern,1 Parsons 2 and Schmid. The lat-
ter's observations were in Switzerland chiefly, where he had
ample opportunity to study the outbreak in small isolated
communities. His work is reviewed by Leichtenstern, who
states that the disease never developed except after the ad-
vent of some stranger. Parsons studied the incidence of the
disease on several thpusand deep-sea fishermen and on four
hundred offshore lighthouse keepers, and in no instance did
the disease develop except as the result of contact with the
sick or within two or three days after leaving shore. He
says there is no evidence whatever that the disease is air-
borne. Leichtenstern studied the extension of the disease to
distant countries, and found that all the evidence pointed to
personal contact as the only factor in its spread. Thus care-
ful epidemiological investigation is entirely in accord with
the findings of bacteriology, for the weak resistance of the
bacillus makes it difficult to understand how the disease can
be carried by the air as readily as is often alleged. This
feebleness of the germ renders it unlikely that influenza is
a dust-borne disease. Of course it may spread by droplet
infection, but the increased volume of the secretions, the per-
sistence of the bacillus after recovery and the great number
of carriers give such opportunities for contact infection that

1 Leichtenstern, Nothnagel's Encycl. Pract. Med., Influenza, Phila.
and Lond., 1905, 523.
^ * Parsons, Brit. M. J., Lond., 1891, II, 303.

278 THE SOURCES AND MODES OF INFECTION

it is hardly necessary to attribute much importance to
aerial infection, as the term is generally used, or to fomites
infection.

Poliomyelitis. — Hill,1 from a careful study of this disease
in Minnesota, was led to suggest the theory that this might
be a dust-borne disease. He finds that usually outbreaks occur
in hot, dry weather, and many cases had long exposure to dust
infected with animal feces, especially those of the horse.
Hill says that at Winona and some other places the disease
ceased soon after the watering of the roads was begun.
The fact that poliomyelitis occurs chiefly in the driest
season of the year has led some others to accept Hill's
proposition, and some have made similar observations as to
the checking of outbreaks. Hillier2 reports an outbreak
at Stowmarket, Eng., brought to a sudden stop by the
watering of streets. Experiments indicate that the virus
has considerable resistance and may possibly be transported
by dust. Landsteiner, Levaditi and Pastia3 found that it
was virulent after being dried thoroughly for 24 days.
Neustadter and Thro 4 in at least one instance succeeded in
causing the disease in a monkey by inoculating dust from a
room occupied by a case of the disease. Fifty grams of
dust were shaken up with 30 c.c. of water and 5 c.c. of the
filtrate were used for the experiment.

Measles. — Measles is considered a typical air-borne
disease, at least within doors, but the experience of the
Parisian hospitals5 shows that the danger of infection within
wards decreases as the opportunity for contact infection is

1 Hill, Northwestern Lancet, Sept. 1, 1909 [reprint].

2 Hillier, Med. Off., 1912, VII, 78.

3 Landsteiner, Levaditi and Pastia, Ann. de 1'Inst. Pasteur, Par.,
1911, XXV, 805.

4 Neustadter and Thro, N. York M. J., 1911, XCIV, 614.

5 Grancher, Cong. Internat. de me"d., 1900, XIII, C.-r., Par., Sect,
de med. de Tenfance, 478. Moizard, Bull, et mem. Soc. med. d. hop. '
de Par., 1900, 3 s., XVII, 683. Martin, Rev. d'hyg., Par., 1903, XXV,
256; Bull, et mem Soc. med. d. hop. de Par., 1904, 3 s., XXI, 297.

INFECTION BY AIR 279

lessened. Grancher was the first to avowedly disregard
infection by air and to attempt to control the spread of
disease by strict attention to medical asepsis. At Tho-
pital des Enfants-Malades he did not at first use cubicles
but merely wire screens or low partitions between the beds.
During the 10 years when this experiment was going on
measles was introduced 139 times and from these cases 115
cross infections resulted, less than one-third as many as
occurred when attention was not focused on the prevention
of contact infection. A suggestive instance is reported of a
child with measles who remained for 24 hours in an open
ward with many susceptibles, and it is stated that in the
3 cases which later developed there had been contact with
the first patient, either direct, or by freshly infected fomi-
tes. At the Pasteur Hospital 126 cases of measles were
cared for without a single cross infection. At the Provi-
dence City Hospital which was opened in March, 1910,
many different diseases are cared for in rooms open most
of the time into a common corridor as in the Pasteur Hos-
pital, while other mixed cases are " barriered " in other
wards, as described on pages 200 to 202. Up to September,
1911, 56 cases of measles had been admitted, with extension
in only one instance to one child.1 Since then there have
been two small outbreaks which Richardson thinks were
most likely not due to aerial infection but to some failure
in technique. One of these outbreaks arose from a single
case, though at other times several cases had been cared for
at one time without extension. In this outbreak one of
the cases was on a different floor from the infecting case.
Rohmer2 from his hospital experience at Cologne is satisfied
that measles is not air-borne further than between adjoin-
ing beds.

English Hospital Experience. — Most English hospital
superintendents who have tried cubicles and barriers do

1 Richardson, Internat. Hosp. Rec., 1911, XV, 18.

2 Rohmer, Jahrb. f. Kinderheilkunde, 1912, LXXV, 7&

280 THE SOURCES AND MODES OF INFECTION

not think that measles can be safely treated in this way
because they consider it likely to be air-borne, or that is
the explanation offered by some at least. Caiger,1 Thom-
son,2 Goodall,3 Gordon4 and Biernacki5 may be mentioned
as holding this view in regard to measles and chicken pox.
All these and Crookshank6 do not find that diphtheria is
air-borne under the hospital conditions mentioned and
uncomplicated scarlet fever usually is not. Caiger does
not think that rubella and whooping cough are readily
air-borne, while Thomson is inclined to think that they are.
Biernacki agrees with Caiger as to rubella and also as re-
gards whooping cough, if the beds are at least 12 feet apart
so as to prevent droplet infection in coughing. Even then
he usually employs a canopy as an additional precaution.
Biernacki thinks that ring-worm is readily air-borne.
This also is the view held by Gates from a study of this
disease in schools.7

Just as this is going to press an article by Rundle and
Burton 8 has appeared reporting two years* experience with
a hospital ward at Liverpool in which a variety of infectious
and non-infectious cases were cared for without any at-
tempt at air isolation. In all, 668 persons passed through
the ward, of whom 69 had scarlet fever, 40 diphtheria, 37
measles, 38 varicella, 9 whooping cough and 215 erysipelas.
There were only 2 cross infections, 1 of diphtheria and 1
of scarlet fever. These authors believe that in hospital
wards the " danger of aerial infection is to be disregarded
for practical purposes. "

1 Caiger, Rep. Metropol. Asylums Bd., 1907, 1908, 258; Med.
Officer, 1910, V., 76.

2 Thomson, Rep. Metropol. Asylums Bd., 1908, 261; Med. Officer,
1910, V, 197.

» Goodall, Rep. Metropol. Asylums Bd., 1910, 257.
4 Gordon, Rep. on Health of Manchester, 1908, 154.
6 Biernacki, The Nursing Times, 1908.

6 Crookshank, Essays and Clinical Studies, Lond., 1911, 134.

7 Gates, Pub. Health, Lond., 1910-11, 226.

8 Rundle and Burton, Lancet, Lond., 1912, I, 720.

INFECTION BY AIR 281

Typhus Fever not Air-Borne. — As will be shown in
Chapter VIII, it is quite probable that typhus fever is
transmitted by insects exclusively. But whether this is so or
not, the successful management of this disease in the same
wards with other patients in Edinburgh, in Liverpool and in
Mexico City indicates that it is not an air-borne disease.

Pneumonic Plague. — The extensive outbreak of plague
in Manchuria in 1910 attracted much attention, since
rodents, while perhaps furnishing the original infection,
played little part in the extension, as it seemed to be spread
almost exclusively from person to person. Kitasato,1 if
reported correctly, concluded that the disease could not
be air-borne but is caused " by coming in close contact
with plague victims and by sputa." That droplet infection
is of great importance in this type of the disease cannot be
doubted and was amply demonstrated by Strong.2 In a
preliminary note he reports experiments in which 15 of 39
plates held in front of coughing patients were infected
with virulent bacilli. He says that the air throughout the
ward was infected, though the experiments, as reported, do
not seem to furnish proof of this. Respirators are said
to have been used successfully in avoiding infection by
physicians and others in close contact with the patients.
No real evidence, however, seems to have been advanced to
show that the disease was air-borne other than by droplets.

Infection by Air not Impossible. — Since the development
of bacteriology, evidence has been accumulating to show
that the air as a vehicle of infection is of less importance
than was formerly believed. Very little evidence has been
found in support of the theory and much against it. It has
been definitely proved that some diseases are not air-borne
and the evidence against certain others is very strong.
While the tendency is thus away from air infection we must

1 Kitasato, U. S. Pub. Health and Mar. Hosp. Serv., Pub. Health
Rep., 1911, XXVI, 567.

* Strong, J. Am. M. Ass., Chicago, 1911, LVII, 1270.

282 THE SOURCES AND MODES OF INFECTION

be on our guard lest our generalization carry us too far.
It may be a fact that most diseases are not air-borne,
and yet further investigation may show that certain other
diseases concerning which we are still in doubt may be
usually transmitted in this way.

Winslow has recently been making some interesting ex-
periments in New York concerning the presence of living
streptococci in dust. He finds that certain types of strep-
tococci characteristic of human secretions are found in a
viable condition in enormous numbers in both house and
street dust, notwithstanding the fact, which also has been
demonstrated, that these bacteria tend to die off with great
rapidity when thrown off from the body. Several observers
in different parts of the country have recently noted the
presence of streptococci in outbreaks of severe sore throat,
and Winslow asks whether it may not be possible that street
dust as well as indoor dust may be a factor in the spread of
this affection.

Infection by Air at Short Range. — Most of the discussion
thus far has related in large part to alleged extension of dis-
ease through the air from apartment to apartment, or from
house to house, or from hospital to adjoining districts.
Transmission over such considerable distances might be
rare or even impossible, and yet nevertheless the air might
be the chief vehicle in ordinary institution or family infec-
tion. That it is believed to be so is indicated by the com-
mon regulation that a wet sheet shall be hung before the
door of the sick-room to prevent the escape of germs, by the
wearing of caps by visitors and the disinfection of inacces-
sible portions of the room. If we ask the reason for this
universal belief in the importance of air-borne infection, we
shall find that it is based entirely on theory, and that there
is no clinical evidence at all to indicate that such a mode of
infection is of any great importance. What evidence we
have indicates rather that infection by means of the air is
of comparatively little importance.

INFECTION BY AIR 283

Reasons for Belief. — The real reasons why people gener-
ally attach so much importance to this mode of infection are,
first, the hearty belief in the general theory of aerial infection
which has prevailed from remote antiquity, and, secondly,
because infection so often takes place when there has not been
any known contact. Contact is the most certain and obvious
mode of infection, and other modes should not be assumed
without good reason. The burden of proof rests on those
who make the assumption.

Air and Aseptic Surgery. — The history of aseptic surgery
is very suggestive in connection with the theory of air-borne
infection. Lister at first unquestionably considered the air
to be the chief source of wound infection, and this view for
awhile dominated surgical practice. Then gradually more
and more attention was given to contact infection, contact
with instruments, contact with the patient's own skin, con-
tact with the operator's hands. As danger from these sources
was more and more perfectly guarded against, surgery became
more and more successful and aerial infection was less and
less dreaded. A successful surgeon of my acquaintance tells
me that he can operate with as little wound infection in a
tenement house as in the best operating room. Ochsner1
says, "Air infection is not impossible, but practically no wound
infection is to be considered except from contact." Most
surgeons at the present time consider aerial infection of very
little importance. Nevertheless it has been shown, as we
shall see, that the air, even of well-constructed operating
rooms, contains considerable numbers of living pus-forming
bacteria. Yet they are usually not numerous enough nor
virulent enough to infect, and unless there is some other
source of infection the wound heals aseptically. Pus-forming
bacteria are quite resistant, and are by far the most numerous
and ubiquitous of all disease germs and more likely to be
found in the air, and the fact that they generally fail to infect
should give pause to any claim that the much less numer-
1 Ochsner, Clinical Surgery, Chicago, 1902, 26.

284 THE SOURCES AND MODES OF INFECTION

ous air-borne germs of other diseases are the chief source
of infection.

Sewer Air and Bacteria. — In this connection the relation
of sewer air to the transport of bacteria should be referred to.
Winslow 1 in a very valuable paper gives a brief history of
the theories on this subject, and shows how the former ideas
as to the great danger to be apprehended from sewer air
gradually gave way before increasing knowledge of bacteria,
and particularly after it was determined' that these organisms
are not readily detached from moist surfaces. From that
time it was generally believed that sewer air had little or
nothing to do with the extension of the infectious diseases.
The subject, however, was reopened by Andrewes 2 and Hor-
rocks.3 The latter showed that under natural conditions
B. prodigiosus and also the bacillus of typhoid fever might
be carried by the sewer air long distances and escape at man-
hole and soil-pipe openings. Lewis 4 also showed that sewage
bacteria could be found in the air passing over a sewage farm,
and also in that blowing at low tide over a beach where
sewage was discharged. Winslow by a series of careful experi-
ments confirmed the work of Horrocks, but went further and
determined the number of bacteria that are transported in
this way. He found, as have others, that mechanical splash-
ing may produce a slight local infection of the air in immediate
contact with the spray, but such infection extends for only
a very short distance and persists for not more than a minute
or two. A careful quantitative study of the air of nineteen
different plumbing systems in various parts of Boston showed
that very few sewage bacteria are found in such air. These
bacteria were found only four times in 200 liters of air,

1 Winslow, Rep. to San. Com. Nat. Ass. Master Plumbers, 1907-09;
Abst. Am. J. Pub. Hyg., Bost., 1909, V, 640.

2 Andrewes, Rep. Med. Off. Local Gov. Bd., Lond., 1906-07, XXXVI,
183, and 1907-08, XXXVII, 266.

8 Horrocks, Pub. Health, Lond., 1907, XIX, 495.
4 Lewis, Scot. M. & S. J., Edin., 1907, XX, 487.

INFECTION BY AIR 285

and then in the presence of mechanical spraying of sewage
at the point of collection. To illustrate the paucity of dan-
gerous bacteria in sewer air he says:

" In a surface water of good quality, like that of New
York City, the colon bacillus can almost invariably be iso-
lated from ten cubic centimeters. This means a slight degree
of intestinal pollution, but experience has shown that the
chance of infection from such a water is but slight; and we
drink it without serious alarm. If one were to breathe for
24 hours the undiluted air of a house-drainage system, at
any point not immediately infected by mechanical splashing,
it appears that less than fifty intestinal bacteria would be
taken in; for the daily consumption of air is about 10,000
liters, and in 200 liters I obtained negative results from air
of this sort. In drinking New York water twice as many
colon bacilli are ingested every day, for 1000 cubic centi-
meters is a small amount for daily consumption. So there
would be less danger of contracting disease from continually
breathing the air of a vent pipe, or of a soil pipe, except
where liquid is actually splashing, than from drinking New
York water."

Anthrax. — Some time since, while considering this subject,
it occurred to me that anthrax ought to be air-borne more
often than any other disease. The spores are extremely resist-
ant, and are found in great numbers in hair, wool, etc., and
the manipulation of these materials is quite likely to raise a
considerable amount of dust. If this be so, and if floating
germs are carried to the alveoli of the lungs, as is alleged, the
pulmonary type of this disease ought to be very common.
Formerly this seems to have been the case. According to the
report of the Local Government Board,1 of thirty-two cases
occurring in the woolen industry of Bradford during nine
months, twenty-three were of the internal type. Since then
great effort has been made to eliminate dust as much as
possible from the woolen and other industries in which dry
1 Rep. Med. Off. Local Gov. Bd., Lond., 1882-83, XII, 98

286 THE SOURCES AND MODES OF INFECTION

infected material is handled. As a consequence, as stated in
the last report of the factory inspector which I have at hand,1
of four hundred forty-four cases of industrial anthrax in Eng-
land from 1899 to 1907 only twenty-one were of the pul-
monary type, and all of these twenty-one were in the dusty
woolen industry. The disease is not nearly so common in
the United States, owing to the fact that less infected material
is imported, but of fifteen cases in Philadelphia two only were
internal. Even now, under the best conditions, there must
be considerable infected dust caused by opening and separat-
ing the bales, and the comparative rarity of the pulmonary
type of the disease indicates that it is not very easily air-
borne, though it appears almost certain that some cases
develop in this manner. Furthermore, it appears that this
disease may be transmitted by the air even out of doors.
Legge states that he has seen two horses infected by feeding
where the dust from the blower of a wool-sorting room was
discharged, and Silberschmidt 2 reported a similar infection of
eight out of twenty-two horses near a hair factory at Zurich.
Slight Evidence that Disease is Air-borne. — It is thus
seen that clinical and epidemiological evidence of the spread
of contagious diseases through the medium of the air is
scanty. No proof of extension through the external air is
presented for any important disease except smallpox, and
this is far from conclusive. I have never seen any good clin-
ical evidence that diseases are air-borne, even indoors. On
the contrary, there is much evidence that this mode of infec-
tion is not a common one. The reasons for 'the widespread
belief in the transmission of disease through the air seem to
be entirely theoretical, and to have been developed simply
because no other satisfactory explanation was at hand. Let
us now consider laboratory and experimental evidence.

1 Report Chief Inspector of Factories and Workshops, 1904, 49;
1905, 49; 1906, 38; 1907, 56; also Legge, Lancet, Lond., 1905, 1, 841.

2 Silberschmidt, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1896,
XXI, 455.

INFECTION BY AIR , 287

Bacteria not given off from Moist Surfaces. — Among
the many new conceptions which resulted from the study of
bacteriology, one of the most novel was that, contrary to all
previous ideas, bacteria are not given off from and are not
readily detached from moist surfaces or liquids in a state of
rest. This was first shown by Tyndall and has been con-
firmed by Nageli,1 Buchner,2 Wernich,3 Huhs 4 and others.

When this became known, numerous experiments were
undertaken to determine whether the expired air was free
from germs, as in accordance with the newly discovered facts
it should be. Tyndall was the first to show that bacteria
are not found in expired air, and Gotschlich 5 cites a dozen
or so workers, all of whom obtained only negative results
from an examination of the expired air. According to Fliigge,6
Cadeac and Malet, Grancher and Gennes, and Miiller were
unable to find tubercle bacilli in the ordinary expiration of
phthisical patients, and these early observations have been
amply confirmed by others.

It has been shown not only that air currents are incapable
of removing bacteria from liquids but also that such currents
do not remove them from the surfaces of solids. Most mate-
rials which contain pathogenic bacteria, such as culture media,
saliva, mucus, pus, excreta, etc., present, when dry, a some-
what hard and often glazed surface, so that it is not sur-
prising that exceedingly strong air currents, even of sixty
meters per second, do not remove the contained germs. This,

1 Nageli, Die niederen Pilze, Miinchen, 1877, 107, Untersuchugen in
die niederen Pilze, 1882.

2 Nageli u. Buchner, Sitzungsber. d. Bay. Akad. d. Wiss., Miinchen,
7 June, 1879.

3 Wernich, Virchow's Arch, f . path. Anat. [etc.], Berl., 1880, LXXIX,
424.

4 Huhs, Ztschr. f. Tuberk. u. Heilstattenw., Leipz., 1906, IX, 396.

5 Gotschlich, Kolle u. Wassermann, Handbuch [etc.], Jena, 1902, I,
171.

6 Fliigge, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXX,
107.

288 THE SOURCES AND MODES OF INFECTION

according to Gotschlich,1 has been demonstrated by Nageli,
Buchner, Wernich, Hamburger and Stern, and Flugge. Got-
schlich says that Honssell found it impossible to detach
bacteria from infected clothing.

Infection in Dust. — So firmly was the theory of aerial
infection intrenched in the minds of men, that search was
made for some other means than the expired breath by
which bacteria might get into the air. It was very early
found that many bacteria could withstand drying for con-
siderable periods of time, and it was at once suggested that
material containing disease germs might become dry and pul-
verized, and that the resulting infective dust might readily
be transported by currents of air. Indeed Koch2 was one
of the earliest, as well as one of the strongest, supporters of
this view.

Resistance of Bacteria to Drying. — If dust is a vehicle
for the transport of the germs of disease, and bacteria are
air-borne on or in bits of dust, or float as separate particles,
they must withstand a considerable amount of drying. That
some species do retain their vitality and virulence after
becoming quite thoroughly dry, has been demonstrated. Sys-
tematic studies of the effects of drying and of light on dif-
ferent disease-producing bacteria have been made, and almost
every germ has been examined from this standpoint by men
particularly interested in working out its biological characters.
In some instances, as, for example, the tubercle bacillus, the
experiments and observations are very numerous. In the
chapter on fomites infection, the resistance of the different
pathogenic organisms to dry ness and to light was considered.
While some species were shown to have very little resisting
power, the germs surviving for a few hours or a few minutes
only, others, like the bacilli of typhoid fever, diphtheria and

1 Gotschlich, Kolle u. Wassermann, Handbuch [etc.], Jena, 1902, I,
170.

J Koch, Mit. a. d. k. Gesundheitsamte, 1884, II, Trans. Sydenham
Soc. Pub., CXX.

INFECTION BY AIR 289

tuberculosis, may, when dry, retain their vitality for months.
In addition to the general knowledge of the resisting power of
bacteria, which has been accumulating, special and systematic
attempts have been made to determine directly the chances
of infection by air-borne dust.

Studies on the Drying of Germs. — Among the earliest
work of this kind was that by Germane.1 He employed room
dust and different kinds of earth, which after sterilization
were inoculated with cultures of bacteria grown in various
media. He found that generally the bacteria perished sooner
in room dust than in other materials, and that bacteria
which, like the typhoid bacillus, might survive for months
even when dried on clothing or solid material, would speedily
die in a very short time in dust. After a large number of
experiments he concluded that cholera, plague, typhoid fever,
influenza and gonorrhea could not be dust-borne; that under
certain circumstances, with strong air currents, streptococcus
and the germs of pneumonia and of diphtheria might be air-
borne, and that, besides the spores of anthrax and tetanus,
many of the pus organisms, meningococcus and the tubercle
bacillus, might be transported in dust. It is suspected that
he did not employ the true coccus of cerebro-spinal meningi-
tis, for recent workers are agreed that this bacterium has very
weak powers of resistance. His results with the pneumo-
coccus also are surprising, as this too is rather feeble.

Neisser 2 in 1898 made a careful study of the strength of
air currents necessary to move dust infected with various
pathogenic bacteria. He showed that currents of from 1
to 4 mm. per second are sufficient, to transport room dust,
and it was chiefly with such currents that he worked. He
used from twenty to thirty drops of an agar culture mixed
with 30 c.c. of sterile dust. He drew dust through narrow
tubes in a rather complicated apparatus, and it appears that

1 Germane, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1897,
XXIV, 403; 1897, XXV, 439; 1897, XXVI, 66, 273.

1 Neisser, Ueber Lauftstaub-Infection, Inaug. Dis., Breslau, 1898.

290 THE SOURCES AND MODES OF INFECTION

the conditions were too far removed from the natural to
render the results of much value. Neisser's conclusions were
that diphtheria, typhoid fever, plague, cholera, pneumonia,
are not dust-borne diseases, but that spores, most of the pus
organisms, tubercle bacilli, and the germs of cerebro-spinal
meningitis may be so carried. It will be seen that to a large
extent he agrees with Germano, but it is to be noted, however,
that his tests, like those of Germano, were all culture tests,
except those for tuberculosis, in which alone animal inocula-
tions were made. Besides such systematic work as that of
Germano and Neisser, much has been done in the study of
special diseases by men particularly interested therein.

Drying of Typhoid Bacilli. — Firth and Horrocks 1 found
that the typhoid bacillus would live for 23 days in sand dry-
enough to be blown by the wind. Harrison and Harrison,*
working in India, recovered the organism after 118 hours
when kept in diffused light in very dry dust. Aldridge3
moistened sand with urine containing typhoid bacilli on
three successive days, and dried it on the fourth day, and
blew it with a bellows over sterile bouillon. He recovered
the bacillus on the 1st, 4th and 9th day thereafter.

Horrocks 4 showed that the micrococcus of Mediterranean
fever would survive in dry soil for about 3 weeks.

Drying of Diphtheria Bacilli. — Fliigge 5 says .that diph-
theria bacilli perish when dry enough to be blown about in
dust. This is confirmed by Pernice and Scagliosi and Reyes.8
Reyes found they would live for 14 days in dry sand.

Drying of Plague Bacilli. — Tidswell,7 experimenting with
dust of various kinds, could not recover the bacillus of bu-

1 Firth and Horrocks, Brit. M. J., Lond., 1902, II, 936, 1094.

2 Harrison and Harrison, J. Roy. Army Med. Corps, Lond., 1904,
II, 721.

Aldridge, Indian M. Gaz., Calcutta, 1903, XXXVIII, 249.
Horrocks, J. Roy. Army Med. Corps, Lond., 1905, V, 78.
Fliigge, Ztschr. f . Hyg. u. Infectionskrankh., Leipz., 1895, XVII, 401.
Cited by Germano.
Tidswell, Rep, on Plague in Queensland, 1902, 67,

INFECTION BY AIR 291

bonic plague after 11 days when the dust was dried under
natural conditions, and it usually died within 3 or 4 days.
When dried very slowly it lived about twice as long. Rose-
nau1 found that this bacillus did not live in dried bone dust
over 6 days, and he did not recover it from dry and sterile
garden soil after 1 day.

Drying of Tubercle Bacilli. — More attention has been
given to the tubercle bacillus than to any other pathogenic
organism. Besides other experiments referred to elsewhere
in these pages it may be mentioned that Kirstein2 experi-
mented with various kinds of dust, and could not find living
tubercle bacilli after 8 days. The dust was artificially
infected and exposed to diffused light. He thinks droplet in-
fection far more important than dust infection. Cadeac 3
was unable to reduce sputum to dust until it had been dried
10 to 12 days, while the tubercle bacilli had nearly died out
on the 6th day. Even when pulverized sputum is injected
into animals, tuberculosis rarely develops, and it must be
still rarer as the result of inhalation. Sticher 4 also and
Beninde 5 found it difficult to demonstrate living bacilli
in dried and pulverized sputum under natural conditions.
Nevertheless most observers do find living tubercle bacilli in
dust, though usually with weakened virulence.

Drying of Cholera Spirilla. — According to Germano, chol-
era spirilla may sometimes survive in dust for 3 days, but
oftentimes they die in 1 day. He says that Honssell was
never able to obtain living spirilla from infected dust,
though Uffelmann was able to do so for a short period.

1 Rosenau, U. S. Pub. Health and Mar. Hosp. Serv. Hyg. Lab. Bull.
No. 4, 1901.

2 Kirstein, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1905, L, 186.

3 Cadeac, Lyon M<§d., 1905, CV, 893; also Lyon Med:, 1908, CXI,
532.

4 Sticher, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXX,
163.

5 Beninde, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXX,
193.

292 THE SOURCES AND MODES OF INFECTION

William1 believes that cholera cannot be a dust-borne
disease.

Tubercle Bacilli in Dust. — Of more practical importance
than experiments with artificially infected dust is the deter-
mination of the presence or absence of disease-producing
bacteria in supposedly infected localities. The chief interest
in such investigations has centered in tuberculosis. Cornet 2
has made a more extensive study than any one else of the
natural distribution of tubercle bacilli in dust. He examined
147 specimens of dust collected from hospital wards, dis-
pensaries, private houses, streets, etc. No bacilli were found
in the street or in places not occupied by the tuberculous,
and even in the environment of the consumptive germs were
found only when the patient was careless in the disposal of
sputum. Cornet's observations have been confirmed by Rem-
bold, Kriiger, Kastner, Ballinger, Kusterman, Le Noir and
Camus and Enderlin, besides others elsewhere referred to.
Gotschlich 3 examined one hundred and nineteen specimens
of dust from streets and public places and was not able to
demonstrate the presence of tubercle bacilli. While most
observers have been unable to find the tubercle bacillus in
street dust, it is said to have been found by Manfredi * and
Schnirer.5 Prausnitz 6 and Petri 7 obtained tubercle bacilli
from the dust in railway carriages, and Bissell 8 found them
in tram cars in Buffalo.

1 William, Ztschr. f . Hyg. u. Infectionskrankh., Leipz., 1893, XV, 166.

2 Cornet, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1885, V, 98;
Nothnagel's Encyclopedia of Pract. Med., Phila. & Lond., 1907, Tuber-
culosis, 85.

8 Gotschlich, Die Verbreitung der Tuberkelbacillen in Staub von
Raumen mit starkem Menschenverkehr, Inaug. Dis., Breslau, 1903.

4 Manfredi, Jahresb. u. d. Fortschr. . . . d. path. Mik., Baumgarten,
1891, VII, 570.

6 Schnirer, Wien. med. Presse, 1891, XXXII, 3.

8 Prausnitz, Arch. f. Hyg., Miinchen u. Leipz., 1891, XII, 192.

7 Petri, Arb. a. d. k. Gesund.-Amt., Berl., 1894, IX, 76.

8 Bissell, N. York M. J., 1895, LXII, 783.

INFECTION BY AIR 293

Some observers have not found the bacilli so numerous as
did Cornet. Thus Hill,1 following Cornet's methods, obtained
virulent bacilli in but 5 of 496 swabbings from private
houses where there were cases of the disease, and in 3 of
180 swabbings from hospital wards.

Heymann 2 has criticised Cornet's findings on the ground
that the latter recovered the dust by means of moist swabs,
thereby perhaps taking up many bacilli which were attached
to the floor, table, etc., and which would not therefore have
any part in air-borne infection. He does not find living
bacilli so numerous in dust as did Cornet. Heymann in 120
tests found them only one-third as often, but Coats,3 follow-
ing Heymann's methods, was able by inoculation tests to
demonstrate the presence of tubercle bacilli in 66 per cent of
specimens of dust obtained from fourteen rooms occupied by
tuberculous patients. On the whole, it appears that virulent
tubercle bacilli are quite commonly found in the dust of
rooms occupied by careless tuberculous patients.

In this connection may be mentioned an experiment of
Heymann's in which he rubbed and shook a sputum-infected
handkerchief, after two days' drying, in a closed box, and
found floating bacilli after the lapse of an hour.

Meningococcus in Dust. — Jaeger4 claimed to have found
the meningococcus on the floor of barracks, and Netter5 the
pneumococcus in the dust of a sick-room four weeks after the
case was removed, but from what is now known of the resist-
ance of these bacteria this is highly improbable. Wash-
bourn and Eyre6 found the pneumococcus in dust from a
ward and laboratory at Guy's Hospital, but failed to find it in

1 Hill, Am. Pub. Health Ass. Rep., 1902, XXVIII, 209.

2 Heymann, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1901,
XXXVIII, 21.

3 Coats, Trans. Brit. Cong, on Tuberculosis, 1901, I, p. 88.
« Jaeger, Deutsche med. Wchnschr., 1899, XXV, 472.

5 Netter, Compt. rend. Soc. de biol., Par., 1897, IV, 538.

6 Washbourn and Eyre, Lancet, Lond., 1902, II, 1440.

294 THE SOURCES AND MODES OF INFECTION

\

street dust. Washbourn says that it has also been found in
dust by Emmerich, Maximowitch and Netter.

Diphtheria Bacilli in Dust. — These bacilli have been found
in the dust of a scarlet-fever ward,1 and are said to have been
found in a diphtheria ward by Richardiere and Tallemer,2
but Schlichter 3 could not find them in a hospital in Vienna.
Cobbett4 exposed plates in a diphtheria ward, but could
obtain no bacilli except when the plates had been touched
by the fingers of the patients.

Dust and Other Diseases. — There has been much dis-
cussion of late in England about the necessity for disinfecting
schoolrooms, and examination of school dust has shown the
presence of colon bacilli 5 and the pus organisms.6

General Conclusions. — While these facts relating to the
resistance of disease germs to drying, and their presence in
supposedly infected localities, afford no definite information
as to the danger to be apprehended from infection by air-
borne dust, they do lead to some tentative conclusions. It
is scarcely possible that gonorrhea, influenza, cerebro-spinal
meningitis and pneumonia can be dust-borne. It is possible,
perhaps, but highly improbable, that plague and cholera can
be so borne. The bacteriological evidence indicates that the
germ of typhoid fever, dysentery, Mediterranean fever, diph-
theria, tuberculosis and suppuration may be carried by float-
ing dust, and it is certainly possible for anthrax and other
spores.

Danger from Dust Slight. — Bacteriology also teaches what
is often forgotten, that pathogenic bacteria, with the ex-
ception of spores, die off quite rapidly when dried, and that
the survivors usually have a weakened virulence. If in a

Renney, Pub. Health, Lond., 1905, XVII, 706.

Richardiere et Tallemer, Gaz. d. mal. enfant [etc.], Par., 1899, X.

Schlichter, Arch. f. Kinderh., Stuttg., 1892, XIV, 129.

Cobbett, J. Royal San. Inst., Lond., 1904, XXV, 405.

Hewlett, Lancet, Lond., 1909, I, 741, 815, 889.

Kerr, Med. Off. Educa., Lond,, 1908, 31.

INFECTION BY AIR 295

sick-room or hospital ward the germs of disease are scattered
so freely on the floor or room contents that enough of them
can survive drying and pulverization to float in the air and
cause disease, the opportunities for contact infection with the
comparatively fresh infective material must be very great, so
great, it seems to me, that infection by air under such condi-
tions must be very insignificant or entirely negligible as com-
pared with infection by contact.

Droplet Infection. — Another way in which living bacteria
may be carried by the air is in tiny floating particles of liquid.
Flugge1 was the first to call attention to the fact that during
speaking, and especially during loud talking, coughing and
sneezing, tiny droplets of saliva are thrown off from the
mouth. Indeed such droplets may be readily seen in the
proper light, and it hardly needed special experiment to prove
their existence. Nevertheless, Fliigge2 and Laschtschenko,3
by infecting the mouth with B. prodigiosus, showed that
germ-carrying droplets are, during coughing, borne to a dis-
tance of nine meters in front of the mouth. These droplet
experiments have been repeated with confirmatory results by
Goldie, Esmarch, B. Frankel, Moller, Hiibner, Weismayr and
Koniger, and the last mentioned has shown that the droplets
may be found two meters behind the person coughing.4
Goldie showed that in fourteen per cent of the cases tubercle
bacilli could be caught on plates after a single act of coughing.
Every patient examined at one time or another gave positive
results. No bacilli were found, even as near as six inches,
during deep breathing, but after coughing they could be
recovered from all parts of the room.

1 Flugge, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1897, XXV,
179.

2 Flugge, Ztsch.f. Hyg. u. Infectionskrankh., Leipz., 1899, XXX, 107.

3 Laschtschenko, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899,
XXX, 125.

4 Koniger, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1900, XXXIV,
119.

296 THE SOURCES AND MODES OF INFECTION

Amount of Droplet Infection. — Since it has been shown by
Fliigge that droplets from speaking may float for from five
to six hours, and be transported by air currents of only one
mm. per second, it is not surprising that they should be car-
ried such distances. Nor is it surprising that Hutchinson1
was able to prove that a fine spray of a culture of B. pro-
digiosus was carried fifty-five meters along a corridor, and up
two flights of stairs, and also a considerable distance out of
doors. Others have shown that the bacteria of the mouth
may be carried by the air during speaking over a large room
or hall.2 Leon3 showed that in speaking three hundred
words 250,000 bacteria were thrown off from the mouth, and
Ziesche4 found over 20,000 tubercle bacilli on a plate 324
sq. cm. exposed for half an hour. But it has further been
shown by Kirstein5 and Koniger6 and Laschtschenko7 that
the size of the droplets and the distance they can be carried
depend to a large extent upon whether the liquid is thin and
watery or a thick mucus. Hence we should expect that
droplets of thick sputum would not be carried nearly so
far as droplets of more liquid saliva, and according to
Goldie8 droplets of the saliva rarely carry bacilli but only the
droplets of sputum. I

Quantitative Experiments. — Since the above was written
Winslow and Robinson9 have published a very interesting

1 Hutchinson, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1901,
XXXVI, 223.

2 Gordon, Rep. Med. Off. Local Gov. Bd., Lond., 1902-03, XXXII,
421.

3 Leon, Arch. f. klin. Chir., Berl., 1903-04, LXXII, 904.

4 Ziesche", Ztschr. f . Hyg. u. Infectionskrankh, Leipz., 1907, XLVII, 50.

5 Kirstein, Ztschr. f . Hyg. u. Infectionskrankh., Leipz., 1900, XXXV,
123.

6 Koniger, Ztschr. f . Hyg. u. Infectionskrankh., Leipz., 1900, XXXIV,
119.

7 Laschtschenko, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899,
XXX, 125.

8 Goldie, Canadian Pract. & Rev., Toronto, 1899, XXIV, 433.

8 Winslow and Robinson, Jour. Infect. Dis., Chicago, 1910, VII, 17.

INFECTION BY AIR 297

paper on this subject giving an excellent resume of previous
investigations. They repeat the experiments of some of the
European writers, and like them they find that if the mouth
of a speaker is infected with a specific germ, as B. prodigiosus,
agar plates exposed in different parts of the room show nu-
merous colonies of the bacillus. They also, by the exposure
of plates in the room with the speaker, recovered Gordon's
Streptococcus salivarius, which is a normal inhabitant of the
mouth. The authors emphasize the distinction noted by
others between the larger droplets of mouth spray which con-
tain the most bacteria and which settle out of the air in the
space of a few feet from the mouth and the smaller droplets
which float for a longer time and may pass to some distance
from the speaker, and which alone may be considered as prop-
erly constituting an infection of the air. The chief interest
in their studies attaches to their quantitative work carried
out on the lines devised by Winslow for his investigations of
sewer air. Out of 140 liters of air taken at various points in
the room immediately after 10 to 50 minutes' loud speaking
by a person whose mouth was infected with B. prodigiosus,
the bacillus was found seven times. Of 74 liters examined for
Streptococcus salivarius none were found to contain this nor-
mal inhabitant of the mouth. The authors consider that an
artificial infection of the mouth may give too high an index
of air contamination, while the normal germs of the mouth
may be thrown off in smaller numbers than are the disease
germs from sick persons. The authors conclude that these
experiments furnish "no basis for a belief that tuberculosis
or any other disease is contracted to an appreciable extent
through the inspired air " and are " in harmony with the
conviction now generally gaining ground that aerial infection
of any sort is a minor factor in the spread of zymotic disease."
Lepra Bacilli. — Schaffer l was able to recover bacilli from
a leprous patient by holding cover glasses a short distance in
front of the face while the patient was speaking and coughing.
1 Schaffer, Arch. f. Dermat. u. Syph., Wien, 1898, XLIV, 159.

298 THE SOURCES AND MODES OF INFECTION

Pneumococcus. — Wood ! found that pneumococci did not
retain their vitality in floating droplets over one hour, and
not half an hour in diffused light.

Influenza Bacilli. — According to Gotschlich,2 droplets con-
taining influenza bacilli will float for five hours. Very little*
has been done to demonstrate the existence of infected drop-
lets in any other diseases.

Bacteria found in Air. — Having shown that bacteria may
float in the air on particles of dust and in droplets of liquid,
we must next inquire whether pathogenic germs have actually
been found in the air. Graham-Smith3 examined the air of
the House of Commons for pathogenic bacteria with negative
results, as did Andrewes4 and Gordon5 the air in the streets
of London. Little light is thrown on our present problem
by these and similar negative tests of outdoor air or of air
away from the vicinity of the sick. Far more interest and
value attach to the examination of air in the vicinity of
cases of infectious sickness.

Tubercle Bacilli in Air. — Heymann6 was able to recover
virulent tubercle bacilli from the air of a small chamber in
which was placed a coughing tuberculous patient. Similar
results were obtained by Laschtschenko.7 Corbett8 recovered
acid-fast bacilli from the ventilating shaft of a hospital, but
made no inoculation tests. Klein9 infected guinea pigs by
exposure in the vent shaft of Brompton Hospital. According

1 Wood, J. Exper. M., N. Y., 1905, VII, 592.

2 Gotschlich, Kolle u. Wassermann, Handbuch [etc.], Jena, 1902, 1, 175.

3 Graham-Smith, J. Hyg., Cambridge, 1903, III, 498.

* Andrewes, Rep.Med. Off. Local Gov. Bd., Lond., 1906-07, XXX VI,
187.

5 Gordon, Rep. Med. Off. Local Gov. Bd., Lond., 1902-03, XXXII,
421.

6 Heymann, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1901,
XXXVIII, 21.

7 Laschtschenko, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899,
XXX.

8 Corbett, St. Paul M. J., St. Paul, Minn., 1904, VI, 735.

8 Stevenson and Murphy, Hygiene and Public Health, II, 212.

INFECTION BY AIR 299

to Cornet and Meyer,1 Williams, Celli and Guarnieri, Wehde,
and Baumgarten have examined air for tubercle bacilli with
negative results. Recently Le Noir and Camus2 demon-
strated by inoculation tubercle bacilli in the dust of a hospi-
tal ward, but they could not obtain them by the filtration
of even 53,000 liters of the air. Viewing the human nose as
a filter, they took swabbings from the nose of physicians and
attendants of phthisical patients, but could not demonstrate
tubercle bacilli by inoculation, though they found them in
the nose of the patients themselves. Cornet considers that
the germs are so sparsely distributed that one ought not to
expect to obtain them by the filtration of even 1000 liters
of air. He says that the finding of tubercle bacilli in settled
dust has as much bearing on air infection as finding it in
the air and is a much easier operation. It does not appear
that Cornet is correct in his contention that the examination
of dust is of more practical importance than the examination
of air. The examination of dust can throw no light on the
number of bacteria floating in the air at any one time, and
as Winslow in his work on sewer air previously referred
to has so clearly pointed out, a quantitative examination of
the floating bacteria is necessary if we wish to determine the
real danger from the inhalation of the air. No such enumer-
ation of tubercle bacilli seems to have been made, and the
difficulty of finding them suggests that they are not very
numerous, even in the vicinity of patients, and that perhaps
the air of a room is not always dangerous to breathe even if
tubercle bacilli can be found in the settled dust.

Pus-forming Bacteria in Air. — Numerous observers are
referred to by Gotschlich,3 Friedrich4 and Noeggerath5 as

1 Cornet and Meyer, Kolle u. Wassermann, Handbuch [etc.], Jena,
1903, II, 143.

2 Le Noir and Camus, Comp. rend. Soc. de biol., Par., 1908, LXV,
464, 622; Ann d'hyg. et de med. colon., Par., 1908, 4 s., IX, 74.

3 Gotschlich, Kolle u. Wassermann, Handbuch [etc.], Jena, 1902,1, 176.

4 Friedrich, Arch. f. klin. Chir., Berl., 1898, LVII, 288.

5 Noeggerath, Deutsche Ztschr. f . Chir., Leipz., 1900-01, LVIII, 277.

300 THE SOURCES AND MODES OF INFECTION

having found various pus-forming bacteria in the wards and
operating rooms of the hospitals. Among more recent Ameri-
can writers may be mentioned Robb,1 Rosenow,2 Monks 3 and
Harrington.4 The latter found that Petri dishes exposed in
an operating room in Boston always contained pus organisms.
The maximum was 131 per square inch per hour. Gordon 5
found staphylococci in an operating room in England, also in
a barber's shop, and they have been demonstrated in the air
by Hamilton.6 Gordon could not find them in the open air,
but they have been found in the air of streets by others.7

Other Bacteria in Air. — Concornotti 8 reports that he
found pneumonia germs in the air of the Hygienic Institute
of Cagliari.

Beck 9 found swine plague bacilli in the air of a laboratory
where experiments with that germ were being carried on.

Bruce 10 could not find the germ of Mediterranean fever in
air-borne dust in Malta.

Actual Danger of Infection by Air. — Pathogenic bacteria
may withstand drying and the pulverization of the dried
material, and they may be actually found floating in the air,
yet they may not after all be dangerous, either because they
have wholly or partially lost their virulence, or because they
are too few in number, or for some other unknown reason.

Little Infection of Wounds by Air. — It has been shown
that notwithstanding the presence of considerable numbers

1 Robb, Am. J. Obst., N. Y., 1909, LX, 451.
8 Rosenow, Am. J. Obst., N. Y., 1904, L, 762.

3 Monks, Ann. Surg., Phila., 1904, XL, 466.

4 Harrington, Ann. Surg., Phila., 1904, XL, 475.

6 Gordon, Rep. Med. Off. Local Gov. Bd., Lond., 1904-05, XXXIV,
387.

6 Hamilton, J. Am. M. Ass., Chicago, 1905, XLIV, 1108.

7 Newman, Bacteriology and the Pub. Health, Lond., 1904, 78.

8 Concornotti, Centrlbl. f. Bakteriol. [etc.], Jena, 1899, XXVI, 492.

9 Beck, Jahresb. u. d. Fortschr. . . . d. path. Mik., Baumgarten,
1891, VII, 567.

10 Nature, Lond,, 1908, LXXVIII, 40.

INFECTION BY AIR , 301

of bacteria in the air of operating rooms, the aerial infection
of wounds is of no practical importance. One reason for this
failure to infect has been shown by Friedrich1 and Noeg-
gerath.2 The conclusion of these authors is that drying and
exposure to light so weaken the bacteria that they are not
able to withstand the actively hostile influences of the tissues
of the human body, though they may be able slowly to vege-
tate on the more favorable culture media of the laboratory.
This lowering of virulence by drying and exposure to light
may be of great practical moment in preventing infection by
air. So also, though other pathogenic bacteria may be
demonstrated in the air, it may be that they are usually
too few in number to infect.

Experiments with Tuberculosis. — The experiment of
Bernheim,3 in which he was not able to infect animals with
mouth spray over 25 cm. from the mouth, but was able to
collect tubercle bacilli on agar plates at the distance of a
meter, is most suggestive of the importance of the number
Of bacteria as a factor in infection. The proper way to de-
termine the infectivity of the air is by animal experiment
or, better still, by carefully conducted observations on human
beings. Except in tuberculosis very few experiments of this
kind have been made. Much, however, has been done with
that disease.

Tappeiner4 had, even before the discovery of the tubercle
bacillus, shown that tuberculosis could be produced in dogs
by causing them to breathe dry and pulverized tuberculous
sputum. Bertheau, Veraguth, Weichselbaum and Frerich,
like Tappeiner, succeeded in infecting animals by causing
them to inhale pulverized sputum containing tubercle bacilli,
while Koch, Cornet, Gebhardt and Preyss accomplished the

1 Friedrich, Arch. f. klin. Chir., Berl., 1898, LVII, 288.

2 Noeggerath, Deutsche Ztschr. f. Chir., Leipz., 1900-01, LVIII, 277.

3 Bernheim, Clinique, Brux., 1905, XIX, 346.

4 Tappeiner, Virchow's Arch. f. path. Anat. [etc.], Berl., 1880,
LXXXII, 353.

302 THE SOURCES AND MODES OF INFECTION

same results by the use of dried bacilli obtained from cul-
tures. More recently Cornet l reports a still more striking
experiment. In a room of seventy-six cubic meters capacity,
48 guinea pigs were exposed in cages at various heights
above the floor. Sputum was placed on a carpet, and after
it was dry the carpet was shaken so that the dust rose up
in clouds. This was repeated on four days. The result was
that 47 of the 48 animals developed tuberculosis within two
months. Kuss 2 carried on experiments very similar to those
of Cornet and with similar results. Kohlisch,3 while admit-
ting that tuberculosis may be caused by the inhalation of dust,
claims that his experiment shows that enormously larger
quantities must be inhaled than are necessary when a spray
is employed. While B. prodigiosus is not pathogenic, it has
been used by various workers for studying the penetration
of bacteria into the respiratory tract. Nenninger,4 using
both infected dust and a sprayed culture, found that the
germs were quickly carried to the smallest bronchioles.

Dust Infection Questioned. — The contention that pul-
monary tuberculosis may be caused by the inhalation of dust
containing tubercle bacilli was not to go unchallenged. Sirena
and Pernice, de Toma, Celli and Guarnieri, and Cade*ac and
Malet were unsuccessful in their attempts to produce the
disease in this way. But perhaps Flugge5 more than any
other has cast discredit on this theory of the origin of pul-
monary tuberculosis. He was unable to induce infection by
causing animals to inhale tuberculous dust, and states that the
dust is not carried to the alveoli. He also developed the
theory of droplet infection, which has been received with much

1 Cornet, Verhandl. d. Berl. med. Gesellsch., 1899, XXX, 2 Th., 91.
a Kuss, Sixth Internat. Cong, on Tuberc., Wash., 1908, I, 101.

3 Kohlisch, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1908, LX,
508.

4 Nenninger, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1901,
XXXVIII, 94.

6 Fliigge, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXX,
107.

INFECTION BY AIR 303

favor, and did much to break down the almost universal
view that dust is the chief vehicle of infection in this disease.
Again when Koch in 1901 by his pronunciamento, that human
tuberculosis is never caused by the milk of tuberculous ani-
mals, stimulated a great number of workers to attempt to
prove the contrary, evidence began to accumulate that tuber-
culous infection of the lungs might be brought about in vari-
ous ways, and facts came to light which told against the view
that direct infection by means of dust is the only manner
in which pulmonary tuberculosis can be caused.

Dust versus Droplets. — According to the school of Fliigge,
infection by droplets is much more likely to take place than
infection by dust, and his pupils have demonstrated the
presence of the bacilli in the lungs immediately after inhala-
tion. Findel,1 working in his laboratory, has shown that the
inhalation of even so small a number as 62 germs is sufficient
to cause the disease, and he asserts that several million times
as many bacteria are necessary to infect when taken by the
stomach. Laschtschenko2 and Heymann,3 working under
the direction of Fliigge, were able to infect guinea pigs with
tuberculosis by causing them to breath directly in front of the
mouth of phthisical patients while the latter were coughing.
The pigs were not infected when distant over one meter.
Fliigge 4 himself infected 6 of 25 guinea pigs in this manner,
holding them distant from twenty to forty-five cm. from the
mouth of the patient. Pfeiffer and Friedberger 5 sprayed
guinea pigs with a culture containing 35,000 tubercle bacilli to

1 Findel, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1907, LVII,
104.

2 Laschtschenko, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899,
XXX, 125.

3 Heymann, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXX,
139.

4 Fliigge, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1899, XXX,
107.

6 Pfeiffer and Friedberger, Deutsche med. Wchnschr., 1907, XXXIII,
1577.

304 THE SOURCES AND MODES OF INFECTION

the c.c. The animals were held for ten minutes at a distance
of eight to twenty cm. in such a manner that their bodies
were perfectly protected. Those held at the greater distance
did not contract tuberculosis, but the others developed what
appeared to be primary tuberculosis of the lungs. Bartel
and Neumann,1 after spraying guinea pigs with tubercle
bacilli, found the germs immediately in mouth, throat and
lungs. Bernheim2 infected guinea pigs at the distance of
not over twenty-five cm., but he was able to collect tubercle
bacilli on agar plates at the distance of one meter from the
mouth. So also Kuss and Lobstein,3 carrying out very care-
ful inhalation experiments with a sprayed culture of tubercle
bacilli, very easily and constantly developed pulmonary
disease. Kovdcs,4 after inhalation experiments with tubercle
bacilli, could immediately recover them from the lungs. But
he thinks that they may also pass from the mouth to the
lungs via the cervical and bronchial glands. Cobbett,5 em-
ploying sprays both of B. prodigiosus and of the tubercle
bacillus, reports that the bacteria are carried to the periph-
ery of the lungs by the inspired air.

The Inhalation of Bacteria. — On the other hand, Hewlett
and Thompson6 found that inhaled bacteria were not carried
even as far as the trachea, as Hildebrandt 7 had found before.
Hartl and Herrmann8 showed that inhaled germs decreased
very rapidly back from the nose, and they consider that the
upper passages are frequently the place of entrance for the
infection, and that bacteria are rarely carried by the air to

1 Bartel and Neumann, Wien. klin. Wchnschr., 1906, XIX, 167, 213.

2 Bernheim, Clinique, Brux., 1905, XIX, 346.

3 Kuss and Lobstein, Bull, med., Paris, 1907, XXI, 821.

4 Kovacs, Beitr. z. path. Anat. u. z. allg. Path., Jena, 1906, XL, 281.
8 Brit. M. J., Lond., 1909, II, 867.

6 Hewlett and Thompson, Lancet, Lond., 1896, I, 86.

7 Hildebrandt, Beitr. z. pathTAnat. u. z. allg. Path., Jena, 1887, II,
411.

8 Hartl and Herrmann, Wien. klin. Wchnschr., 1905, XVIII, 798.

INFECTION BY AIR 305

the lungs. Vallee l carried on spray experiments without
result, and Calmette and Guerin,2 in a few experiments made
with infected dust, did not find that the bacilli reached the
alveoli. Weleminsky3 could find no bacilli in the lungs of
guinea pigs immediately after they had been subjected to
inhalation experiments.

Lack of Agreement among Investigators. — Thus it is seen
that the school of Cornet claims that pulmonary tuberculosis
is almost always caused by the passage of infected dust to the
alveoli, and that infected droplets are of little moment; while
Fliigge and his pupils attempt to show that dust rarely reaches
the alveoli but that the inhalation of droplets is the easiest
method of causing the disease. The unprejudiced reader
must conclude that infection in either way is possible, but
the conditions of the experiments are so far removed from the
natural that there must be much hesitation before assuming
that this work indicates in any degree the common mode of
infection in human beings.

Bacilli may pass from Stomach to Lungs. — It is com-
monly assumed that because pulmonary tuberculosis begins
in the apices of the lungs the bacilli must be carried directly
to the alveoli by the inspired air. Moreover most patholo-
gists consider that evidence points to the alveolar surface as
the starting point of the disease, and that the relative age of
the lesions in the lungs and the glands indicates that it is
primarily a pulmonary disease. The writer is not competent
to discuss the findings of the pathologists, but there are a
large number of careful observers who believe on pathological
and experimental evidence that the bacilli which cause pul-
monary tuberculosis may find their way through the lymph
and the blood from any part of the alimentary tract to the
lungs. Some pathologists believe that the disease begins
in the capillaries rather than in the alveoli. Calmette and

1 Valtee, Ann. d. 1'Inst. Pasteur, 1905, XIX, 619.

2 Calmette and Gue"rin, Ann. de 1'Inst. Pasteur, 1905, XIX, 601.
8 Weleminsky, Berl. klin. Wchnschr., 1905, XLII, 743.

306 THE SOURCES AND MODES OF INFECTION

Guerin l say that primary tubercle of the lungs always be-
gins in the capillaries, and never in the alveoli, and Aufrecht 2
from his pathological studies comes to the same conclusion.

Even pulmonary anthracosis may be caused by the inges-
tion of particles of carbon. Vansteenburgh and Grysez3
caused animals to breath air loaded with soot from a smoky
lamp, and though the nose was filled with soot, none was found
in lungs or even trachea. If the experiment was prolonged,
carbon appeared in the lungs, but not if the esophagus was tied.
When one bronchus was occluded, the carbon appeared in the
parenchyma of the lung just the same, provided the esoph-
agus was open. Whitla4 fed animals with carbon and bacilli,
and both were found in the lungs in from four to twenty-four
hours. Feeding was done with a tube, with great care to
prevent inhalation. Hutchens 5 fed guinea pigs with coal
dust, and also injected it into the abdomen, and found it in
the lungs in three days. Grober,6 after injecting India ink
into the tonsils, was able to find the pigment in the lungs
in a short time. He thought that he could trace a continu-
ous lymphatic route from the tonsils to the costal pleura,
whence the particles passed directly to the parietal pleura
and the lungs. But such a connection has been denied by
Beitzke,7 Wood 8 and others.

Schultze 9 denies that pulmonary anthracosis can be pro-

Calmette and Guerin, Ann. de Tlnst. Pasteur, 1906, XX, 609.
Aufrecht, Deutsches Arch. f. klin. Med., Leipz., 1908, XCIV, 308.
Vansteenburgh and Grysez: Ann. de 1'Inst. Pasteur, 1905, XIX, 787.
Whitla, Brit. M. J., Lond., 1908, II, 61.
Cited by Oliver, Brit. M. J., Lond., 1908, II, 481.
Grober, Die Tonsillen als Eintrittspforten fiir Krankheitserreger,
Abdruck aus dem klin. Jahrb., Berl., 1905, XIV, 547.

7 Beitzke, Virchow'sArc^.f. path. Anat. [etc.], Berl., 1906, CLXXXIV,
1; Berl. klin. Wchnschr., 1908, XLV, 1235.

• 8 Wood, Rep. Henry Phipps Inst. Study . . . Tuberculosis, Phila.,
1906, IV, 163.

9 Schultze, Miinchen med, Wchnschr,, 1906, LIII, 1702.

INFECTION BY AIR 307

duced by feeding, as is claimed by the French experimenters,
and insists that during the feeding some inhalation takes place,
and that if the animals are fed through a gastric fistula with
adequate precautions, the fragments are not -carried to the
lungs.

A large number of workers have certainly demonstrated
that pulmonary tuberculosis may be caused by infection
through different portions of the alimentary canal.

The work of Calmette and Guerin l shows that tubercle
bacilli are speedily carried from the intestines to the lungs.

Schroeder and Cotton2 have shown that pulmonary disease
develops no matter in what part of the body tubercle bacilli
are inoculated, and in Bulletin 88 they have shown that
marked pulmonary invasion occurs constantly in hogs fed on
tuberculous material. So also Ravenel 3 has recovered tu-
bercle bacilli from the lungs within a few hours after placing
them in the stomach by celiotomy. Beitzke4 in a review of
the subject states that Schlossmann and St. Engel5 found the
bacilli in the lungs six hours after injection into the stomach
by laparotomy, as did also L. Rabinowitsch, and that the
latter and also Uffenheimer, Bacharach, and Stein and Orth
found them after injection into the intestine. Ficker, he says,
could not do this with adult dogs or cats, and Herrmann could
not with monkeys. Vallee,6 Calmette and Guerin,1 Bonome 7
and Arloing 8 also have demonstrated pulmonary infection

1 Calmette and Gue"rin, Ann. de Plnst. Pasteur, 1905, XIX, 601;
1906, XX, 353, 609.

8 Schroeder and Cotton, U. S. Dept. Agric. Bu. An. Ind. Bull. No. 86
and No. 93.

Ravenel, Cleveland M. J., 1909, VIII, 179.
Beitzke, Berl. klin. Wchnschr., 1908, XLV, 1235.
Schlossmann and St. Engel, Deutsche med. Wchnschr., 1906,
XXXII, 1070.

Vall4e, Ann. de 1'Inst. Pasteur, 1905, XIX, 619.
Bonome, Gazz. d. osp., Milano, 1907, XXVIII, Nos. 37-84, abstract
in J. Am. M. Ass., Chicago, 1907, XLIX, 888.

8 Arloing, Sixth Internat. Cong, on Tuberc., Wash., 1908, IV, 666.

308' THE SOURCES AND MODES OF INFECTION

via the alimentary tract, and very many of the experiments
have shown that the bacilli easily pass through the intestinal
wall without leaving a trace of their passage, so that the
absence of lesion of the alimentary tract is no proof that
infection has not taken place through it.

Unnatural Conditions of Experiment. — Although, there
has been, as is here shown, a vast amount of experimental
work on infection with tuberculosis, there has been very little
in which the conditions at all approached the natural. Usu-
ally there is an excessive amount of exposure, or an excessive
number of germs in spray or dust. Thus in Cornet's nota-
ble experiment, where 47 of 48 guinea pigs were infected by
breathing dust, the carpet had been smeared with large
quantities of sputum, and it was forcibly beaten so that
clouds of dust rose up directly in front of the animals. It is
surprising that so few have thought it worth while to see
how infection takes place in animals kept under conditions
as nearly as possible like those under which human beings
live.

Infection in Laboratory. — Animals kept in laboratories in
company with others that have been purposely infected rarely
contract the disease. Koch1 in his original paper says that
among the hundreds of animals so kept the disease did not
develop for three or four months, and then only exception-
ally. He reports autopsies on 17 guinea pigs and 8 rabbits
thus naturally infected, which sho.wed in all cases a pul-
monary tuberculosis resembling that which occurs in man.
But apparently no effort was made to determine whether
infection was by the air, by the food, or by contact through
the hands of attendants, though this laboratory infection has
been urged as evidence that the disease is air-borne. Of many
hundreds of guinea pigs confined in a room with various
tuberculous animals in the Department of Agriculture Ex-
periment Station at Washington, only one contracted tuber-

1 Koch, Mit. a. d. k. Gesund.-Amt., 1884, II, Trans. Sydenham Soc.
Pub., CXV, 129.

INFECTION BY AIR 309

culosis, and that was in a cage with several infected animals.1
In laboratories floating bacteria are not likely to be very nu-
merous, owing to the cleanliness practiced and the fact that
most of the laboratory animals do not excrete many bacilli.
Experiments under Natural Conditions. Tuberculosis. —
Schroeder and Cotton2 experimented under more natural
conditions. Seven cows were exposed in adjoining stalls to
3 tuberculous animals, and all but 1 of them contracted the
disease. As all but 2 were moved about from stall to stall,
exchanging with the infected animals, infection may readily
have been by contact. At the same time 100 guinea pigs
were exposed in the stalls, one half in cages below the man-
gers where food could sift through from the mangers, and
one half on the walls. The exposure lasted several months,
and only 1 of the pigs, in a cage under the manger, became
infected. In a subsequent experiment 35 guinea pigs were ex-
posed for one hundred and thirty-five days on the walls of
the stalls. Two developed generalized tuberculosis. Of 42
animals kept for fifty-one days under the manger of infected
cows, 6 developed a more or less generalized type of the
disease. There was no direct evidence that tubercle bacilli
were in the air of the stalls, but as the infecting animals
were excreting large numbers of bacilli, there is little doubt
of it. There is no record that contact infection from the
hands of attendants was strictly guarded against, and it may
be that the 2 out of 135 guinea pigs, and the 2 cows supposed
to have succumbed to air-borne infection, were really infected
by contact. Swenson, quoted by Aufrecht, exposed five
calves in a stable with tuberculous cows in such a manner as
to preclude contact infection, and they all developed the
disease, as Swenson thought, by dust infection. Klein3

1 U. S. Dept. Agric., Rep. Bu. An. Ind., Wash., 1906, XXIII, 31.

2 Schroeder and Cotton, U. S. Dept. Agric., Rep. Bu. An. Ind.,
Wash., 1903, XX, 61; 1904, XXI, 44, reprinted as Circ. No. 83.

1 Stevenson and Murphy, Treatise on Hygiene and Public Health,
Lond., 1893-96, II, 212,

310 THE SOURCES AND MODES OF INFECTION

exposed guinea pigs in the vent shaft of the Brompton Hos-
pital, and most of them contracted tuberculosis.

Experiments in Tuberculosis Houses. — Bartel and Spieler,1
realizing that most experiments are under unnatural con-
ditions, exposed 12 guinea pigs in cages in a house occupied
by tuberculous patients, and allowed 16 to run at large and
be handled by the children. The exposure was from two to
three weeks, and of the 12 cage pigs 3 developed tubercu-
losis, and of the 16 free pigs 10 developed tuberculosis of
various glands and other organs. No statement is made
that care was taken to prevent contact or mouth-spray in-
fection of the animals in the cages. These authors 2 exposed
8 guinea pigs, running free, in a house where the tuberculous
patient was taking fairly good care of the sputum. Only 1
of the guinea pigs developed tuberculosis.

At my suggestion Dr. M. S. Packard of Providence carried
on an experiment for the health department of that city on
the mode of infection of guinea pigs under natural conditions.
A fairly clean laborer's house was chosen, where there was a
consumptive whose sputum contained large numbers of ba-
cilli, and who was taking no care whatever in regard to its
disposal. Thirty-six small guinea pigs were exposed in cages
placed in a dark place in the room in which the patient usually
sat. The cages were much crowded. Of the animals, 16
were fed and cared for by the consumptive, and the others
by an employee of the department free from disease. These
latter animals were locked in a box covered with wire net-
ting, fourteen meshes to the inch. There could be no ques-
tion of contact infection for these pigs. The exposure was
from February 11, 1908, to May 14. All but 21 of the
animals died of non-tuberculous disease, or were starved or
killed by rats after removal from the house. Of the 11
surviving animals7 exposed to air infection alone, 8 were

1 Bartel and Spieler, Wien. klin. Wchnschr., 1905, XVIII, 218.

2 Festschr. enthalt. Arb. u. Tuberk. ... VI Internal. Tuberk. Konf .
[etc.], Wien u. Leipz., 1907, 71.

INFECTION BY AIR 311

shown to be tuberculous and 3 not tuberculous. Of the
other lot, 7 were shown to be tuberculous and 3 not. It
seems impossible that the test animals could have been in-
fected otherwise than through the air. That it was mouth
spray rather than dust infection seems likely, for the patient
took much interest in the animals, and was often seen with
his face close to the netting, talking to them, and coughing at
them only a few inches distant. The guinea pigs were pur-
chased, but I was unable to learn that they had ever been
exposed to the disease.

Since writing the above I note that Le Noir and Camus l
have undertaken a similar experiment. They exposed guinea
pigs in cages in a ward for phthisical patients. Four pigs
were placed in a cage on the floor and the patients fed these.
One of them developed tuberculosis. Five pigs were kept
for six weeks in a cage on the floor but protected so that the
patients could not reach them. One of these also developed
tuberculosis. Another pig with three little ones was placed
in a cage near the ceiling so that they had to be fed from a
ladder. One of the little ones died of an intercurrent affec-
tion, but the other two contracted tuberculosis.

Since it is claimed by the majority of those interested in
tuberculosis that the disease is spread chiefly by means of
dust, it is highly desirable that a sufficient number of well-
conducted experiments under truly natural conditions be
made to determine how important this mode of infection
really is.

Except for tuberculosis very little animal experimentation
has been done to determine whether diseases are air-borne.
One reason for this, of course, is that so many of the com-
moner diseases affecting human beings are not easily con-
tagious for the lower animals.

Experiments with Mediterranean Fever. — Horrocks,2
experimenting with monkeys, could cause Mediterranean

1 Le Noir and Camus, Presse meU, Par., 1909, XVII, 761.

2 Horrocks, Report of Royal Commission on Malta Fever.

312 THE SOURCES AND MODES OF INFECTION

fever by making the animals inhale dust artificially infected
with large numbers of M . melitensis, but he was not successful
with dust naturally infected with urine, though the organism
of this disease is very resistant to drying. Monkeys kept in
cages near infected animals did not contract the disease,
unless there was actual contact with infected material, i.e.,
there was no air-borne infection. Epidemiological evidence
is against this being a dust disease, for it prevails far less
during the dry than during the wet season. That it is in
reality exclusively milk-borne seems now to have been
demonstrated.

Experiments with Anthrax. — Buchner 1 was the first to
attempt to infect animals by making them breathe dried
anthrax bacilli or spores. He easily succeeded in thus caus-
ing primary pulmonary disease. Miiskatbluth 2 showed that
the spores would readily pass the lungs without causing local
disease. Morse, Hildebrandt, Tschistovitsch and Gramat-
schikoff 3 could not cause the disease by inhalation even
when the spray was sent directly into the trachea. Baum-
garten4 believes that these various experiments show that
infection takes place by the tonsils rather than directly by
the lungs.

Experiments with Plague. — Bubonic plague is often con-
sidered an air-borne disease, but the experiments of the present
Indian Plague Commission5 have shown that this is not the
case. Both monkeys and guinea pigs have in considerable
numbers been put in flea-proof cages, but exposed freely to
the air, and the cages placed in dwellings known to be in-
fected, and in no instance did they contract the disease. In
other instances the animals were exposed freely to the air, but

1 Biichner, Centrlbl. f. Bakteriol. [etc.], Jena, 1890, VII, 733;
VIII, 1. [

a Miiskatbluth, Centrlbl. f. Bakteriol. [etc.], Jena, 1887, I, 321.

3 Cited by Sobernheim in Kolle u. Wassermann Handbuch [etc.],
Jena, 1903, II, 49.

4 Lehrbuch der Pathologischen Mycologie, Braunschweig, 1890.

5 Journal Hygiene, 1905, 835; 1906, 445-471; 1907, 432, 835, 979.

INFECTION BY AIR 313

were protected from fleas by " tangle-foot/' with like results.
These experiments were made at various times and places,
but the animals never succumbed to air-borne infection. In
all instances control animals not .protected from fleas usually
contracted the disease.

Conclusions. — After the foregoing survey of the subject
we are, I think, justified in the following conclusions:

1. The theory of the aerial transmission of disease was
developed as the most reasonable way of explaining the phe-
nomena of infection.

2. Contact infection with carriers and missed cases affords
a better explanation of the phenomena.

3. The best medical thought has been steadily restricting
the supposed sphere of aerial transmission.

4. Only a few authorities now assert that disease is carried
by the atmosphere outside of dwellings, and this assertion is
made only in regard to smallpox.

5. Bacteriology teaches that former ideas in regard to the
manner in which diseases may be air-borne are entirely erro-
neous; that most diseases are not likely to be dust-borne,
and they are spray-borne only for two or three feet,*a phe-
nomenon which after all resembles contact infection more
than it does aerial infection as ordinarily understood.
Tuberculosis is more likely to be air-borne than is any other
common disease.

6. Surgeons at first developed aseptic surgery on the theory
that air infection was of the highest importance. They have
gradually learned to pay less attention to it, until at present
some of the best surgeons consider it a negligible factor.

7. Animal experimentation indicates that tuberculosis and
anthrax may be air-borne, and that plague and some other
diseases are not.

8. Pathology has not determined, as is sometimes alleged,
that even pulmonary consumption is an air-borne disease.

9. There is no good clinical evidence that the common
diseases are air-borne.

314 THE SOURCES AND MODES OF INFECTION

10. There is considerable clinical evidence that scarlet
fever, diphtheria, smallpox, measles, whooping cough, typhoid
fever and plague are not easily transmissible through the air.

11. Scarlet fever and diphtheria can be cared for in the
same ward with other diseases without extension, if clean-
liness be maintained and infection by contact avoided.

In reviewing the subject of air infection it becomes evident
that our knowledge is still far too scanty, and that the avail-
able evidence is far from conclusive. Yet it is of the greatest
practical importance that we should know definitely just what
danger there is of air-borne infection and in what diseases
it is to be feared. Infection by air, if it does take place, as is
commonly believed, is so difficult to avoid or guard against,
and so universal in its action, that it discourages effort to
avoid other sources of danger. If the sick-room is filled with
floating contagium, of what use is it to make much of an effort
to guard against contact infection ? If it should prove, as I
firmly believe, that contact infection is the chief way in which
the contagious diseases spread, an exaggerated idea of the
importance of air-borne infection is most mischievous. It is
impossible, as I know from experience, to teach people to
avoid contact infection while they are firmly convinced that
the air is the chief vehicle of infection.

While it is not possible at present to state with exactness
the part played by aerial infection in the transmission of the
different infectious diseases, we are by the evidence forced to
the conclusion that the current ideas in regard to the impor-
tance of infection by air are unwarranted. Without denying
the possibility of such infection, it may be fairly affirmed
that there is no evidence that it is an appreciable factor in
the maintenance of most of our common contagious diseases.
We are warranted, then, in discarding it as a working hy-
pothesis and devoting our chief attention to the prevention
of contact infection. It will be a great relief to most persons
to be freed from the specter of infected air, a specter which
has pursued the race from the time of Hippocrates, and we

INFECTION BY AIR 315

may rest assured that if people can as a consequence be
better taught to practice strict personal cleanliness, they will
be led to do that which will more than anything else prevent
aerial infection also, if that should in the end be proved to be
of more importance than now appears.

CHAPTER VII.

INFECTION BY FOOD AND DRINK.

Infection by Water.

Broad Street Well. — From time immemorial water has been
believed to be the bearer of disease, but it is only since the
middle of the nineteenth century that the subject has been
scientifically studied. One of the first instances of a clear-
cut demonstration of the causation of sickness by infected
water was that of the now famous Broad Street well, so ably
studied by Snow.1 During the outbreak of cholera in Lon-
don in 1854 there was an enormous concentration of cases
in a very limited area just east of Regent Street, there having
been reported, during a period of about six weeks, over 600
fatal cases. A careful study of the site, soil, subsoil, streets,
density and character of population, dwellings, yards, closets,
cesspools, vaults, drains, conditions of cleanliness and atmos-
pheric conditions, revealed nothing of interest. But a study
of the water supply discovered most interesting facts. Nearly
all of the cases were nearer a certain public pump in Broad
Street than any other well, and most of them gave a definite
history of getting water from this pump. Of the very few
cases (ten at the time of the investigation) outside of the
area supplied by this pump, half were known to drink water
from Broad Street. There were also several cases of cholera
in distant parts ipf London in persons who drank water from
this well. In the workhouse with 535 inmates, in the midst
of this district, but with its own well, there were only 5
deaths, less than one-tenth the rate in the neighborhood, and

1 Sedgwick, The Principles of Sanitary Science and the Public
Health, New York, 1902, 170.

316

INFECTION BY FOOD AND DRINK 317

in a brewery with 70 employees and using its own well there
was not a single case. It was also shown that a privy vault
and cesspool in the adjoining house discharged through a leaky
drain which ran within two feet of the well. There were 4
fatal cases of cholera in this house at the time of the outbreak
and obscure earlier cases which were not unlikely cholera
also.

North Boston Well. — In the United States the outbreak
of typhoid fever at North Boston, N. Y., in 1843, referred to
by Flint in his popular text-books, did much to call attention
to drinking water as a factor in the spread of disease. A
young man from Massachusetts went to the hamlet sick with
typhoid fever, and died there. He lodged at the tavern where
there was a well of water used by 6 of the neighboring fam-
ilies. One near-by family and 2 distant families did not
use the water and had no illness. In the other families which
did use the water there were 28 cases with 10 deaths.1

Recent Studies. — Such striking demonstration in a few
instances of the spread of disease through the medium of
drinking water led to unwarranted generalization, and during
the last half of the nineteenth century it was the common
belief of health officials and medical men that infected water
was the chief factor in the causation of typhoid fever and
cholera, and that it was of great importance in malaria, yellow
fever, dysentery, diarrhea, and was perhaps of moment in all
the " zymotic " diseases. The discovery of the specific organ-
isms of these diseases, and the application of more scientific
methods to their study, has of late shown that water, while
a factor of very great importance, is not so important as was
at one time supposed. Let us consider its relation to specific
diseases.

Typhoid Outbreaks. — For dwellers in temperate regions

typhoid fever is the most important water-borne disease.

The demands of modern civilization require for the constantly

increasing urban population a pipe-distributed municipal

1 Am. Pub. Health Ass. Rep., 1873, I, 167.

318 THE SOURCES AND MODES OF INFECTION

supply which in numerous instances must be taken from
streams or lakes more or less contaminated with sewage.
The plentiful use of water rendered possible by a municipal
supply produces a large amount of sewage, which in the past
has usually discharged into the nearest watercourse, thus
carrying danger to any other community which might become
a user of the water. From what is known of the life history
of the typhoid bacillus it is not surprising that in some cities
great explosive outbreaks of the disease have been caused by
a temporary pollution, and other cities, owing to continuous
pollution of their source of supply, have suffered from a con-
tinuous high death rate from this disease. The former class
of outbreaks, while in the aggregate less destructive, are most
impressive lessons for the public and have been exceedingly
instructive to the epidemiologist. The demonstration of a
water-borne outbreak depends primarily upon statistics. If
there is an excess of typhoid fever among the users of a certain
water over the neighboring population living under the same
conditions but using a different water, and if other sources,
as food and milk, are excluded, the outbreak is probably
water-borne. The more closely the users and non-users of
the water are commingled, and the greater the difference in
the incidence of the disease upon them, the more certain is the
demonstration. During the outbreak in Scranton, 1906-07,1
there were four separate sections of the city not supplied
from the infected reservoir in which there was very little
typhoid, and most of the cases which did occur in these dis-
tricts were shown to have used the infected water in other
places, or to have been subject to contact infection. In
Paris,2 owing to partial failure of the good supply, the im-
pure Seine water was turned on to one arrondissement after
another, with the result that a well-defined epidemic followed
in each instance. In Philadelphia the gradual introduc-
tion of filtered water has resulted in a decrease in typhoid

1 Wainwright, N. York M. J., 1907, LXXXV, 1027.
* Jordon, J. Am. M. Ass., Chicago, 1907, XLVIII, 1669.

INFECTION BY FOOD AND DRINK 319

fever, limited to those districts to which the water has been
supplied.

These water-borne outbreaks are usually explosive in char-
acter and are frequently of short duration. An inspection of
the supply not rarely shows that the feces of typhoid cases
have entered the water shortly before the outbreak. Such
outbreaks are apt to occur in the spring when the freshets
wash the surface of the ground into the streams. The bacilli
have rarely been found in the incriminated water.

While studying this subject in 1888,1 I was able to find
thirteen recorded instances of the recovery of typhoid bacilli
from water which had presumably been the cause of an out-
break of the disease. In the outbreak in Providence in that
year Prudden and Ernst reported finding the bacilli in house
filters. Methods of identifying the bacillus at that time were,
however, not entirely satisfactory, and it is not certain that
the bacteria isolated were really typhoid bacilli. By better
methods the germs have since been occasionally found in
infected waters,2 but as might be expected, it is more often
from wells than from rivers or lakes. According to Got-
schlich,3 typhoid-fever germs have been identified in drinking
water by the agglutination test only four times. Since then,
however, they have been isolated by Wesbrook4 from the
Mississippi water at Minneapolis, and by Fox from the reser-
voir which was the source of the Scranton outbreak.5 The
bacillus isolated by Wesbrook was used for years as a test
organism in his laboratory and also by Harris at Johns Hop-
kins. Anderson and Hutchings and Wheeler, as will be re-
ferred to, determined the presence of typhoid bacilli in ice
presumed to be the cause of the disease.

1 Chapin, Boston M. & S. J., 1889, CXX, 604.

2 J. Mass. Ass. Bds. Health, Bost., 1904, XIV, 66.

8 Gotschlich, Kolle u. Wassermann, Handbuch [etc.], Jena, 1903,
I, 191.

4 Wesbrook, Brit. M. J., 1897, II, 1774.
6 Rep. St. Bd. Health, Penn., 1907, 410.

320 THE SOURCES AND MODES OF INFECTION

Temporary Typhoid Infection. — These explosive and tem-
porary outbreaks of typhoid fever among the users of muni-
cipal supplies are sometimes due to an accidental pollution
with sewage, but are more often due to the overflow of privies,
or the placing of the discharges of typhoid patients on the
ground where they can be washed into the streams. In our
lake cities they have sometimes resulted from the wind tem-
porarily blowing the sewage of the city towards the intake of
the waterworks,1 or the dumping of dredged mud near the
intake.2 Outbreaks are sometimes due to the careless tem-
porary use of polluted water, owing to failure of the good
supply, as at Newburyport,3 or the accidental sucking in of
polluted water owing to some derangement of valves or other
mechanism, as in Lowell,4 Baraboo and Millinocket.5

Continuous Typhoid Infection. — When considerable
amounts of sewage are discharged into a stream or lake, the
water, owing to the prevalence of typhoid fever, must be
continuously charged with the bacilli. The users of the
water under such circumstances generally suffer from a con-
tinuously high death rate from this disease. Among cities
which have so suffered may be mentioned Chicago, Pitts-
burg, Cincinnati, St. Louis, Philadelphia, Newark and Al-
bany. The unenviable position of many Pennsylvania cities
in this respect is well set forth by Morris.6 Among Euro-
pean cities which have had a bad water supply are Berlin,
Hamburg, Paris, Frankfurt, Altona, Breslau and Zurich.
Under these circumstances the typhoid rate may, while re-
maining high, vary considerably. Sometimes the incidence
of the disease may reach enormous proportions, as in Chicago

1 Whipple, Typhoid Fever, New York, 1908, 167-168.

2 Whipple, Typhoid Fever, New York, 1908, 167; also U. S. Geol.
Survey, Wash., Water Supply and Irrigation Papers, No. 194, 138.

3 Rep. St. Bd. Health, Mass., 1892, 701.

4 Whipple, Typhoid Fever, New York, 1908, 174.

5 Whipple, Typhoid Fever, New York, 1908, 178-179.

6 Morris, Sanitation, Phila., August, 1904, 47.

INFECTION BY FOOD AND DRINK 321

in 1891, when it was 173.8 per 100,000, and .in Pittsburg in
1900, when it was 144.2. Sometimes epidemic waves of the
disease can be followed down a river from one municipality
to another, as in the Merrimac, Kennebec and Hudson.1

Cities with Infected Water. — The connection between an
excessive typhoid death rate and the sewage contamination
of municipal water supplies is thus shown by a large amount
of epidemiological evidence, but it is even more certainly dem-
onstrated by the improvement which almost always follows
when a pure water is substituted for the impure. Formerly
typhoid fever was very common in English cities, but now
the disease is far less prevalent, due in large measure to the
almost universal use of clean water. Many continental cities
present marked illustrations of the decrease in typhoid fever,
due to improvement in water supply. Among such may be
mentioned Paris, Berlin, Altona, Hamburg, Zurich, Breslau
and Frankfurt. The United States also presents numerous
examples, as Lawrence, Lowell, Albany, Buffalo, Newark,
Jersey City, Cleveland, Chicago and Philadelphia. (For more
detailed information see Whipple, " Typhoid Fever;" Fuertes,
"Water and Public Health;" Hazen, "The Filtration of Public
Water Supplies;" and Hazen, "Pure Water and How to Get
It.") It has in one or two instances happened that the purifi-
cation of a polluted water supply has not resulted in any very
marked diminution in the typhoid death rate. This was so
at Youngstown, Ohio, and notably at Washington.2 At
Youngstown the disease was shown to be spread chiefly by
contact, or by the use of infected wells. The conditions at
Washington are not so well understood, and a special com-
mission 3 has been able to find no very definite cause for
the disease. Levy and Freeman4 after a very careful study

1 Whipple, Typhoid Fever, New York, 1908, 149, 154, 236.

2 Whipple, Typhoid Fever, New York, 1908, 248.

3 U. S. Pub. Health & Mar. Hosp. Serv., Hyg. Lab., Bull. Nos. 35,
44 and 51.

* Levy and Freeman, Old Dominion J. M. & S., 1908, VII, 315.

322 THE SOURCES AND MODES OF INFECTION

of conditions in Washington and Richmond, conclude Aat
Washington must be put in the category of southern cities
in which an excessive typhoid-fever rate depends chiefly on
climatic conditions operating more or less directly, and that
perhaps to a considerable extent the disease depends on the
presence of flies. However that may be, Washington affords
a practically unique example among large cities of a mistake
in attributing the excess of typhoid fever to the pollution of
the water supply.

Amount of Sickness Due to Water. — To how great an
extent pollution of public water supplies is responsible for the
excessive typhoid mortality in the United States it is difficult
to determine. Whipple l says that the average typhoid death
rate in American cities is about 35 per 100,000, while cities
with a good water supply average not over 20. He would
thus attribute to the pollution of public water supplies about
40 per cent of the typhoid fever of the cities of the United
States. We must remember, however, that an excess of
typhoid fever in a city with a questionable water supply is
not always due to the water, as witness Washington and
Youngstown, and Dr. H. W. Hill writes me that Duluth is
another example. But a majority of the people of this coun-
try live beyond the reach of municipal water supplies, and
typhoid fever is even more prevalent among this rural popu-
lation. So that if the whole country is considered, it appears
that 10 or 15 per cent, rather than 40 per cent, of our typhoid
fever is the proper proportion to attribute to the infection of
municipal supplies.

Amount Due to Wells. — Doubtless the majority of the
inhabitants o^f the United States obtain their drinking water
from wells. Formerly polluted wells were believed to be the
chief source of typhoid fever. This was particularly so while
Murchison's "pythogenic" theory of the origin of the disease
prevailed. When wells were a principal source of supply in
cities where privy vaults and cesspools abounded, probably
1 Whipple, Typhoid Fever, New York, 1908, 132.

INFECTION BY FOOD AND DRINK 323

they were often the source of typhoid fever. With the sub-
stitution of municipal pipe-supplies, the well has become a
factor of comparatively little moment in city typhoid fever.
As it now appears that the bacillus of this disease does not
survive long in soil or water, and that each case must have
some connection with a previous case of human typhoid
infection, it seems improbable that country wells, largely iso-
lated as they must be from danger of indiscriminate human
contamination, play an important part in the propagation of
the disease. This is the view held by Hill, who has had a
large experience in the study of this disease in Minnesota.
He states, however, that sometimes in loose coarse gravel,
or in seamy shale and limestone formations, wells may become
infected from sources of pollution situated at long distances,
and have then been known to be the cause of much sickness.
Instances of Infected Wells. — That wells have actually
been the source of typhoid fever there is much evidence.
Instances of outbreaks due to infected wells at Basingstoke,
England, Newport, R. I., and Trenton are narrated by Whip-
pie,1 and a dozen or more instances in England are referred to
by Poore,2 and some of the 179 milk outbreaks of typhoid
fever tabulated in " Milk and Its Relation to the Public
Health"3 were due to infection of the milk by water from a
contaminated well. Outbreaks due to well water continue
to be reported from time to time. During twenty-five years
no instance of typhoid fever from infected well water has
come under my personal notice except at a summer hotel on
the shore of the bay, in which case the well was probably
infected by a broken drainpipe near by. An interesting case
was noted in Washington a few years since.4 Typhoid fever
occurred in four houses on a certain street, all supplied from

1 Whipple, Typhoid Fever, New York, 1908, 183-188.

2 Poore, The Earth in Relation to the Preservation and Destruc-
tion of Bacteria, Lond., 1902, 135.

3 U.S. Pub. Health & Mar. Hosp. Sen'., Hyg. Lab. Bull. No. 41, 50.

4 District of Columbia, Report of Health Officer, 1903, 29.

324 THE SOURCES AND MODES OF INFECTION

wells. A little further up the street there had previously
been some cases, the feces from which had been thrown into
a privy box after what was probably only partial disinfection
with carbolic acid. At any rate carbolic acid could be tasted
in the water in the wells below. While in this instance the
infecting material probably percolated through the soil, as
it did also in the outbreak at Trenton, Whipple states:1
" Sandy soil is a good filtering material, and when a well in
such soil stands at the center of a circle twenty-five or fifty
feet in radius in which there are no privies, cesspools, sink
wastes, or other sources of contamination, the water can
usually be depended upon as fit for domestic use, — provided,
of course, that the top of the well is properly guarded against
surface wash." Sometimes, as in the Trenton incident, the
infecting material passes a considerable distance apparently
through channels or crevices. The writer once knew of a well
which became polluted (though without causing sickness)
from a privy 300 feet distant, after there had been much
blasting near by, which presumably opened seams in the
somewhat slaty rock. As Whipple says, most wells which
have been the cause of typhoid outbreaks have been contam-
inated from the top.

Infected Springs. — Springs are essentially natural wells,
and there is no reason why they should not occasionally be
contaminated. A typhoid outbreak caused by spring water
occurred at Mount Savage, Md., in 1904,2 another at Spring-
field, Mass.,3 one at Northampton, Mass.,4 one at Ridgway,
Penn.,5 and two others in Maryland.6

Typhoid Fever from Bathing. — It has been suggested
from time to time that typhoid fever may result from bathing

Whipple, Typhoid Fever, New York, 1908, 85.

Whipple, Typhoid Fever, New York, 1908, 188.

Rep. Bd. Health, Springfield, 1903, 16.

Rep. St. Bd. Health, Mass., 1900, 844.

Rep. St. Bd. Health, Penn., 1907, 927.

J. Am. M. Ass., Chicago, 1905, XLIV, 595.

INFECTION BY FOOD AND DRINK 325

in water containing typhoid bacilli, the infection presumably
being due to the accidental swallowing of some of the water.
Reece 1 has recently reported an outbreak, presumably due
to this cause, at the naval recruiting station at Walmer. The
swimming tank was infected with sewage containing typhoid
excreta, and the earlier cases seemed to be closely connected
with the use of the tank. From the report, however, it does
not appear that contact infection outside the tank could be
rigidly excluded.

Cholera from Water. — Cholera was early shown to be
a water-borne disease. The longevity of the spirillum in
water is probably not so great as that of the typhoid bacillus,
but it is sufficient to permit of the disease being easily trans-
mitted in this manner. Whenever the excreta of human
beings infected with cholera gain access to cisterns, wells,
streams or other sources of drinking water, cholera is pretty sure
to follow. The cholera-infected Broad Street well in London
has been referred to, and although for many years the views
of health officials were somewhat colored by the miasm theory
of Pettenkofer, the transmission of the disease by water was
kept in mind, and many outbreaks due to this cause were
reported. Radcliffe, Simon and others 2 showed that infected
water played a large part in the outbreaks of cholera in Lon-
don in 1848-49, 1853-54 and 1866. Shakespeare in his very
valuable Report on Cholera in Europe and India3 gives nu-
merous instances of water-borne cholera infection, as from
wells at several places in France (pp. 76-81), and from muni-
cipal supplies in Genoa and Naples (p. 151). In Spain many
towns and villages suffered from water-borne outbreaks,
caused often, doubtless, by the universal custom of washing
soiled linen in running streams. According to Shakespeare,
water plays an important part in the spread of cholera in

* Rep. Med. Off. Loc. Gov. Bd., 1908-09, XXXVIII, 90.
2 Sedgwick, Principles of Sanitary Science and the Public Health,
New York, 1902, 182.

8 U. S. Government Printing Office, 1890,

326 THE SOURCES AND MODES OF INFECTION

India, and Koch1 reported finding the spirilla in a tank used
as a village water supply. A number of instances of the
finding of cholera spirilla in water that was presumably the
cause of the disease are noted by Kolle.2 Cases are recorded
of explosive outbreaks of cholera on shipboard, probably due
to infected water.3 The most striking instance of the trans-
mission of cholera by water is the celebrated outbreak in
Hamburg and Altona in 1892. These two municipalities
form practically one city, though each has its own water
supply. In the outbreak Hamburg, which used unfiltered
and polluted Elbe water, suffered severely, while Altona, the
water for which was filtered, was almost exempt. The line of
demarcation between the two water supplies could be recog-
nized by the incidence of the disease. Our officers in the
Philippines attribute great importance to water as a vehicle
for the diffusion of cholera, particularly in the villages and
smaller towns, where the streams are used for washing clothes
and for sewers and at the same time as sources of domestic
supply. The municipal supply of Manila has been kept free
from infection during the last outbreaks, though wells have
to some extent been the cause of sickness. . Woodruff4 states
that in one town a spring supplied part of the people, who
were quite free from the disease, which nearly decimated those
using the river, and in those towns where the people had well-
protected rain-water cisterns they were quite immune. While
we are not in a position to determine the relative importance
of water in the causation of cholera, it appears certain that it
is a factor of great consequence, and that people using a
sewage-polluted water are, on the advent of cholera, liable to
suffer from severe epidemics of the disease.

Dysentery from Water. — The bacillus of dysentery is found
in the discharges from the bowels, and has about the same

1 Koch, Brit. M. J., Lond., 1884, II, 403, 453.

» Kolle u. Wassermann, Handbuch [etc.], Jena, I, 191; III, 61.

» Wendt, Asiatic Cholera, New York, 1885, 113.

« Woodruff, J. Am. M. Ass., 1905, XLV, 1160.

INFECTION BY FOOD AND DRINK 327

resistance as the typhoid bacillus, so that we should expect
this form of dysentery to spread in much the same way as
typhoid fever does, and that infected water would prove a
factor of importance. Shiga1 reports outbreaks in Japan from
the use of well and river water. Eldridge2 states that dysen-
tery is a rural disease in Japan, and the use of human feces
as a fertilizer, and the frequency of the infection of the nu-
merous small streams and wells, renders it preeminently a
water-borne disease.

Little is heard about water-borne outbreaks of dysentery
in England or the eastern United States, and it is highly
probable that such are not very common. Dr. H. W. Hill
reports four such outbreaks in Minnesota, three of which
are briefly referred to in the report of the Minnesota state
board of health for 1911. At Hibbing in 1907 about 2000
out of 8000 persons were attacked, although only 5 deaths
occurred. The infection was probably caused by work-
men going down into the well to make repairs. After the
institution of the hypochlorite treatment of the water the
outbreak abruptly ceased. At St. Peter there were out-
breaks in 1908 and in 1909. Both were due to a .polluted
river water overflowing into a reservoir. At the Agri-
cultural College a deep well was infected by the backing up
of a sewer, and almost every person who drank the water
developed dysentery. At Mankato, Minn., there was an
outbreak of 405 cases of typhoid fever due to sewage leak-
ing into the well which supplied the town. Preceding the
typhoid cases there were 4000 to 6000 cases of diarrhea
and dysentery, nearly half the population, the first cases
appearing June 25, immediately after the infection of the
water and lasting until early in July.3

There is every reason to believe that dysentery amebse

* Shiga, Mod. Med. [Osier], Phila. & N. Y., 1907, II, 781.

* Eldridge, U. S. Pub. Health & Mar. Hosp. Serv., Pub. Health
Rep., 1900, 1.

3 J. Infect. Dis., Chicago, 1911, IX, 410.

328 THE SOURCES AND MODES OF INFECTION

are carried in water. Musgrave and Clegg1 claim to have
found the amebse in the water supply of Manila, but there
is some doubt as to the pathogenicity of the forms studied
by them. Allan2 reported a small outbreak of amebic
dysentery in North Carolina, due to an infected well.

Diarrhea from Water. — Reincke3 states that infantile
diarrhea was greatly lessened after the improvement in the
water supply of Hamburg. Sir Shirley F. Murphy, Seaton
and Newsholme in the discussion of the paper took issue
with its writer as to the part played by water in the causa-
tion of this disease. Sedgwick4 says that there is no doubt
that drinking water is the ready vehicle of dysentery and
diarrhea. He refers to the Ninth Report of the Medical
Officer of the Privy Council, London, 1867, p. 16, and to
his own investigations as to the excessive prevalence of
typhoid fever and diarrhea in Burlington owing to the
sewage contamination of the water supply.5 Sedgwick
also, in the paper mentioned on another page, quotes
freely from Reincke's various writings to show that well-
defined outbreaks of diarrheal disease have occurred in
Altona, Hamburg and Berlin in connection with defective
working of the filters or with unusual pollution of the water
supply due to other causes.

Water in Relation to Seasonal Distribution of Diarrhea. —
In an article just published, McLaughlin6 attributes very
considerable importance to water supplies in the causation
of infantile diarrheas. He recognizes the fact that these
diarrheas may include at times much typhoid fever and
much dysentery, and perhaps other varieties of intestinal

1 Bull. 18, Bu. Gov. Lab., P. I., 93; Rep. Bd. Health, P. I., 1904-5, 10.

2 Man, J. Am. M. Ass., Chicago, 1909, LIII, 1561.

s Reincke, Trans. Epidemiol. Soc., Lond., 1904, n. s., XXIII, 135.

4 Sedgwick, Principles of Sanitary Science and the Public Health,
New York, 1902, 217.

6 J. N. Eng. Water Works Ass., X, 167.

6 Pub. Health Rep., U. S. Pub. Health and Mar. Hosp. Serv., Wash.,
1912, XXVII, 579.

INFECTION BY FOOD AND DRINK 329

infection. He also admits that many factors other than
the water supply are often of far more importance than is
the water. He bases his contention as to the danger from
water on an excessive prevalence of diarrheal diseases in the
winter months, alleging that while ignorance, poverty and
" unsanitary conditions " may cause an enormous excess
in summer, in towns with pure water, they do not cause such
an excess in the winter. He gives some striking illustra-
tions of communities with a grossly polluted water supply
which have an excess of winter diarrhea as well as of winter
typhoid fever. Among these may be mentioned Niagara
Falls and Cohoes, N. Y., and Escanaba and Marquette,
Mich. The evidence from the larger cities of Cleveland,
Chicago, Milwaukee and Pittsburg is not so convincing. It
is true that these do show a very considerable excess of win-
ter diarrhea and that the curves for the three lake cities are
strikingly parallel, but if this winter diarrhea is really the
result of and proof of dangerous pollution of the water we
should expect a correspondingly great excess of typhoid
fever in these cities; yet the typhoid death rate in all of
them is below the average. Indeed there is no very marked
parallelism between typhoid fever and diarrhea either in
different cities or in the same city at clifferent seasons. In
some of McLaughlin's tables the curves of winter diarrhea
and winter typhoid fever correspond exactly and in others
they are months apart. While on guard not to attribute
too much importance to water as a source of diarrhea we
must admit that McLaughlin has presented new evidence to
show that water is at times a factor to be reckoned with.
The Mills-Reincke Phenomenon. — In 1893 Mills of
Lawrence, Mass., and Reincke of Hamburg, Germany,
noted that the purification of the water supplies of their
respective towns was followed by a decrease in the general
death rate as well as in the death rate from typhoid fever.
Hazen later called attention to these facts, which he found to
hold true for other American cities, and he was in accord

330 THE SOURCES AND MODES OF INFECTION

with the first-named observers in attributing the decrease
to the improvement in the water and maintained that
where one death from typhoid could be prevented by im-
proving the water supply two or three deaths from other
causes would be prevented by the improvement. Sedg-
wick and McNutt,1 in an elaborate paper, have presented a
large amount of material which they believe establishes the
fact of the phenomenon and also indicates which diseases
are chiefly affected by the water supply. The authors
consider that their data show that an improvement in a
poor water supply may be expected not only to reduce the
death rate from typhoid fever but to decrease infant mor-
tality and the death rate from gastro-intestinal disturbance
and from tuberculosis and pneumonia. It has been sug-
gested that these changes may be due to the removal of
disease germs from the water, or to increased bodily re-
sistance due to better water, or to both. As to the explana-
tion, however, Sedgwick and McNutt offer no opinion.

Water Purification and the General Death Rate. —
According to the theory under discussion the purification
of a polluted water supply causes a fall in the general death
rate below that due to the diminution of typhoid fever.
This decrease certainly took place in Lawrence and Lowell,
but the deaths in a city during a year are due to a great
number of diseases, and the prevalence of each disease and
each death from that disease are due to the joint action of
very many causes. To attribute a change in the death
rate to any one cause is extremely dangerous and in the past
has led to most serious errors. It can be seen from the
diagram shown by Sedgwick and McNutt that in both the
cities under consideration there had been a considerable
rise in the death rate just previous to changes in the water
supply and that a considerable part of the increase was due
to pneumonia as was also a considerable part of the de-
crease. Changes equally as great in the mortality from

1 Sedgwick and McNutt, J. Infect. Dis., Chicago, 1910, VII, 489.

INFECTION BY FOOD AND DRINK 331

pneumonia and in the general death rate have occurred in
Albany, as shown in Sedgwick and McNutt's diagram, and
doubtless in other cities, entirely independent of changes in
the water supply, and such changes may as well be assumed
to be due to variation in the prevalence of influenza or to
some other cause as to changes in the water supply. That
two cities close together had a decrease in the death rate, in
large part due to decrease in respiratory diseases, at the
same time that the water supply was changed might well
be due to chance. It is suggestive that the change was
most marked in the two cities in the same year although
the water in Lowell was only partially improved in that
year. It may also be noted that the decline in the death
rate in Albany started from a maximum in 1892, the year
before the maximum in the Massachusetts cities, and that
this maximum coincided, as in Massachusetts, with a maxi-
mum mortality from respiratory diseases. The rate of
decrease in total mortality and in respiratory mortality was
not as great after filtration as before, and it was about as
great as the decline in Lawrence and Lowell which was
supposed to be due to water.

Water Purification and Diarrhea. — The infant death
rate, in which the diarrheal diseases play an important part,
has been shown to have the closest sort of relation to kinds
of food and methods of feeding and to vary enormously in
different sections of the same city and among different
classes residing in the same section. A body of statistics,
far more voluminous than are marshalled in support of the
Mills-Reincke theory, indicates that very great reduction
can readily be made in the infant death rate by better
feeding and other care. Yet we are asked to believe that
the diarrheal deaths in Hamburg were cut in half by the
change in the water. One cannot help thinking either
that the old water supply was specifically infected with the
germs of diarrhea, which is not the case in most English and
American cities, or else that the changes in population and

332 THE SOURCES AND MODES OF INFECTION

in sanitation, after the great cholera outbreak, produced in
themselves great changes in the diarrheal death rate.
Indeed, the writer has been informed by a well-known
scientific man, then residing in Hamburg, that the cholera
outbreak, by its destruction of life, and consequent changes
in the population, profoundly modified mortality rates. If
we turn to Sedgwick and McNutt's diagram for Lawrence
we see no change in the diarrheal death rate between the
7 years before and the 7 years after filtration. In Lowell
the diarrheal death rate did decline decidedly after the im-
provement in the water. In Albany the decrease began
before the improvement in the water and continued at
about the same rate afterwards. jj

Water Purification and Tuberculosis. — As shown by
Sedgwick and McNutt, pulmonary tuberculosis had been
declining for many years in Lawrence and Lowell, as it had
in most cities, and the diagrams indicate a retarded rather
than an accelerated decline after the improvement in the
water. The authors, in order to show the alleged effect
of the water improvement, are obliged to compare these
cities with Manchester, which they assume differs from
Lowell and Lawrence chiefly in having a good water supply,
and which they consider a " normal " city for their purpose.
Compared with Manchester, Lawrence shows a greater de-
crease by 30 deaths and Lowell by 68. While the authors
state that the age, sex, occupation and nationality of the
population of the three cities is substantially the same, it
would appear rather risky to assume that all the conditions
affecting health are the same. Cities show great variation
as regards their mortality, even when the figures are cor-
rected for age and sex, and would doubtless do so if cor-
rected for nationality and occupation. One of the most
difficult problems in vital statistics is to make correct
deductions from a comparison of the death rate of different
cities, particularly in such a disease as tuberculosis. A
theorem which depends upon a comparison made as above

INFECTION BY FOOD AND DRINK 333

seems to require further study. In Albany following fil-
tration there was no apparent change in the steady decrease
in tuberculosis which had been going on for several years.
Notwithstanding the efforts that have been made by some
to show by statistics that the decline in tuberculosis has
been due to sewerage, and by others that it is dependent on
the^segregation of advanced cases, and by still others that
it is the result of the recent popular propaganda, and now
by the supporters of the Mills-Reincke theory that it is
dependent in a very appreciable measure on improvement
in water supplies, I cannot see that the evidence presented
along any of these lines is sufficient to warrant official action
or definite predictions. We know less about the causation of
tuberculosis than about almost any other infectious disease.

Water Purification and Pneumonia. — While Sedgwick
and McNutt say with reference to diphtheria that " the
mixed character of the title and the highly epidemic char-
acter " of the disease render no conclusions possible, they
believe that valid deductions can be drawn from a study of
pneumonia. This disease group has, however, during the
last 25 years, shown marked epidemicity, due, I believe, in
large part, to the effect of influenza. There was in New
England a notable increase of pneumonia in the early
nineties followed by a decline and then another increase a
few years later, though the curves have varied somewhat in
different cities. There was a sudden drop in pneumonia
mortality in Lawrence after filtration and a slight increase
in Lowell, but for the reason given no deduction is permis-
sible as to the influence of the water.

Results of Purification in Providence. — An excellent
sand filter was installed in Providence in January, 1906.
There was an appreciable decrease in the amount of typhoid
fever but it is doubtful if it was dependent on filtration.
There was not very much change in the general death rate
in Providence following filtration, though it was slightly
higher in the 2 years after filtration was begun than it was

334 THE SOURCES AND MODES OF INFECTION

in the 2 years before filtration. Tuberculosis, which had been
declining for many years, exhibited a somewhat retarded
decline after 1906. Pneumonia after 1905 showed a decided
increase over the preceding 10 years, though there was
little change during the 3 years before filtration and the
2 years after. Diarrheal diseases had been declining and
reached a minimum in 1905, the year before nitration, fol-
lowed by a considerable increase during the next 2 years.
Certainly there is nothing in these figures to indicate that
an improved water supply has caused a decrease in the
diseases mentioned.

Mclaughlin's Evidence concerning Mills-Reincke Phe-
nomenon. — McLaughlin.1 has also given consideration to
the Mills-Reincke theorem, and, as does the present writer,
hesitates to make deductions from such a complicated
phenomenon as changes in the general death rate. He
prefers to study single diseases, and considers, in the paper
cited, infantile diarrhea. His argument as to the effect on
the diarrheal death rate by improving the water supply is
based on the causative connection between water and diar-
rhea which he seeks to establish as referred to above. He
also gives tables of deaths in four cities which have recently
wholly or to a large extent changed their water supplies,
namely Cincinnati, Columbus, Pittsburg and Philadelphia.
All have shown a very great decrease in the death rate from
typhoid fever. In Cincinnati and Columbus there was no
decrease in diarrhea; in Philadelphia it was very slight, but
in Pittsburg it was quite noticeable. This maintenance
of the diarrheal death rate is rather remarkable in view of
the direct efforts that have been made during the last few
years to decrease infant mortality. The figures for pneu-
monia are not given by McLaughlin. Tuberculosis de-
creased in the four cities, but the decrease had been going
on for some time as in most cities.

1 Pub. Health Rep., U. S. Pub. Health and Mar. Hosp. Serv., Wash.,
1912, XXVLI, 597.

INFECTION BY FOOD AND DRINK 335

Conclusion concerning the Theorem. — No attempt is
here made to disprove the reality of the Mills-Reincke
phenomenon but only to call attention to certain difficulties
in accepting it. It is a question of great importance which
ought to receive extended study. A number of important
cities have recently improved their water supplies with a
corresponding decrease in typhoid fever. The effect upon
other diseases should be carefully examined. At present it
seems unwise to consider the proposition demonstrated and
to promise a lowering of the general death rate as a result
of the purification of water supplies.

Malaria from Water. — Until recently malaria' has been
believed to be transmitted by means of drinking water.
Numerous instances were reported, such as that of the ship
Argo, in which it was supposed that the evidence pointed
conclusively to this mode of transmission. The discovery
of the part played by the mosquito in the causation of this
disease led to a more critical consideration of the alleged
evidence in support of its water-borne character, and most
of this evidence was found to be worthless, or in the few
instances in which water still seemed to be at fault it ap-
peared probable that the water served as a breeding place
for mosquitoes, and did not directly cause disease by its
ingestion.

There was nothing in the earlier demonstrated facts of the
transmission of the disease by insects to preclude the pos-
sibility of its transmission by water, and indeed Laveran and
Manson thought this not improbable. Celli 1 attempted to
demonstrate this in various ways by administering water
from the most malarious regions of Italy to human beings,
daily, up to a month. He failed completely, as, according to
Craig,2 have all other experimenters, except Ross in one in-
stance, which, however, from the conditions of the experi-
ment, was far from conclusive. Celli considers the fact that

1 Celli, Malaria, Lond., 1900, 94.

2 Craig, The Malarial Fevers, New York, 1909, 82.

336 THE SOURCES AND MODES OF INFECTION

large numbers of railway employees and others living in
intensely malarious regions around Rome use an unques-
tionably pure aqueduct water, but are yet frequently at-
tacked by malaria, is corroborative evidence that water
cannot be an important vehicle of the disease. But prob-
ably the best evidence we have that drinking water plays no
part in the causation of malaria is the fact that preventive
measures all over the world, directed solely against the
mosquito and with no reference to the water theory, have
resulted in an enormous decrease and in some places in the
eradication of the disease.

Yellow' Fever from Water. — Yellow fever also was
formerly believed to be at times water-borne, but, as in
malaria, the epidemiological evidence therefor will not stand
criticism. The success of the preventive measures directed
solely against the mosquito indicates that drinking water
has no part in the diffusion of this disease.

Worms in Water. — The eggs and young of some of the
parasitic worms are often swallowed in drinking water, and
the Ankylostoma and Bilharzia, and others probably, enter
the skin from water in which they are contained, but the
consideration of the diseases caused by them is outside the
scope of this essay.

Conclusions. — Among the diseases which may be trans-
mitted by water, typhoid fever is doubtless by far the most
important in Europe and North America at the present
time. When water-borne disease is mentioned, typhoid
fever comes first to mind. Yet there are those who think
that the role of water in this disease has been exaggerated,1
and it is doubtless true that in Europe, and even in this
country, recent improvements in water supplies have greatly
diminished this disease, but there are still a great many
communities drinking polluted water and having a con-
sequently high typhoid death rate. It is probably true
that other fecal-borne diseases such as the diarrheal group
1 Houston, J. State M. Lond., 1912, XX, 21, 92.

INFECTION BY FOOD AND DRINK 337

are transmitted through the medium of water. Asiatic
cholera too may cause great devastation through infection
of 'water supplies, but for a number of years the western
world has been remarkably free from this disease. Typhoid
fever is, however, such a common and serious disease, caus-
ing probably 25,000 deaths annually in the United States,
as well as an enormous amount of disability at the most use-
ful period of life, that it is worth while to make large expend-
itures for its prevention. Such expenditures are to be still
further encouraged, since it is certain that the means taken
to prevent the diffusion of typhoid fever by water will also
prevent the spread of cholera and bacillary dysentery.

Protection of Water Supplies. — In the three diseases just
mentioned th'e sole source of infection of water is the excreta
of persons infected with the specific germs. If these excreta
can be kept out of drinking water, or if the germs can be
removed from it after it is infected, the problem is solved, and
this can be done at a not prohibitive expense. For a full
and practical consideration of the means for obtaining a pure
water supply reference should be had to such works as Hazen's
" The Filtration of Public Water Supplies" and "Pure Water
and How to Get It." Suffice it to say here that pure water
may be obtained by securing a clean source or by adopting
some method of purification.

Domestic Wells. — The larger part of our population ob-
tains its drinking water from shallow wells, and it is usually
neither difficult nor expensive so to locate them that they
will not receive drainage from privies or sink drains, and so
to protect them that they will not receive surface washings.
The federal Department of Agriculture and the state and local
boards of health should do and are doing much to instruct
farmers and villagers concerning the location and protection
of wells.

Municipal Supplies. — As regards municipal supplies it is
desirable if possible that the water be uncontaminated. This
may be secured by drawing from deep wells, or seeking a

338 THE SOURCES AND MODES OF INFECTION

source in a sparsely inhabited region. Some cities, as Liver-
pool, Boston and New York, have spent large sums to bring
clean water from long distances. Other cities, notably Chi-
cago, have secured good water by the construction of expen-
sive works to remove sewage from the watershed. Many
cities, when there is only a limited danger of pollution, ac-
complish much by a continuous patrol of the watershed
and the removal of minor sources of pollution under general
or special nuisance laws. In many states this duty of pro-
tection has been laid upon the state boards of health, and
in some instances cooperation between states becomes
necessary.

Purification of Water. — Unfortunately it often happens
that it is physically impossible for a municipality to obtain
a sufficient supply of water which is not subject to dangerous
pollution. Purification then becomes a necessity, and owing
to the labors of a long line of chemists, engineers and bacteri-
ologists, several efficient and economical methods for accom-
plishing this have been developed. First among these is
filtration. There are various methods of filtration, applica-
ble to different waters and different localities, and it is the
business of specialists to devise the best system for each city
confronted by the problem. Efficient filtration will remove
over 99 per cent of the contained bacteria, but cannot be relied
upon to remove amebse. Hence filtration will not protect
against amebic dysentery, but it will protect against the
bacillary form and against typhoid fever and cholera. That
it is efficient against typhoid fever is shown by the experience
of London, Berlin, Zurich, Hamburg, Lawrence, Albany,
Paterson, Philadelphia and many other cities, and the re-
moval of cholera spirilla was beautifully shown by the ex-
perience of Altona in 1893. The storage of water rapidly
removes bacteria by sedimentation and through the death of
the organisms, and is often a useful adjunct to filtration,
notably so in the case of some of the London water companies.
The boiling of water is an effectual means of destroying dan-

INFECTION BY FOOD AND DRINK 339

gerous bacteria, and as a domestic measure it is sometimes
extremely useful.

The use of hypochlorite of lime (commercial bleaching
powder) has within a short time come into use for the pur-
pose of sterilizing municipal water supplies. While this
method of protection had been used in England as a tem-
porary expedient, its first use as a permanent method of
treatment seems to have been by Leal, at Boonton, N. J.,
in 1908.1 A high degree of safety seems to be secured at a
low cost. The subject has been investigated also by Clark
and Gage,2 Phelps, Park, Pratt and others. The process has
been used at East Providence, R. L, Poughkeepsie, Chicago
and other places.3

Infection by Ice.

Danger Exaggerated. — The use of ice from polluted
sources has always been viewed with some alarm, and when
it became known from the work of Prudden4 and others that
the bacilli of typhoid fever live for some months when frozen
into ice, the alarm was believed to be well founded. But
further researches (see page 10) have shown that in the
freezing of ice under normal conditions a large proportion
of the bacteria in the ice are frozen out, and those that remain
tend to die off quite rapidly, somewhat as they do in water.
Hence theoretically little danger is to be apprehended from
the use of ice cut from water considerably polluted, even
so polluted that if the water itself were used it would
probably cause disease. There is very little epidemiological
evidence that ice has been the cause of sickness. The litera-
ture relating to the subject was collected by Sedgwick and

1 Leal, Fuller, Johnson, Papers read at the 29th Annual Con-
vention of the American Water Works Association, 1909.

2 J. N. Eng. Water Works Ass., 1909, XXII, 302.

8 Discussion of a paper by Phelps at a recent meeting of the Boston
Society of Civil Engineers, Abst. Eng. Rec., 1910, XLI, 80.
« Prudden, Med, Rec,, N. Y., 1887, XXXI, 341.

340 THE SOURCES AND MODES OF INFECTION

Winslow 1 in 1902. An outbreak the exact nature of which
was not determined occurred at Rye Beach, N. H., in 1875,
due presumably to the use of ice from a pond grossly polluted
with decaying vegetable matter. A single case of typhoid fever
in Connecticut appeared to be due to the use of specifically
infected ice. Alleged outbreaks of typhoid fever and dysen-
tery in the United States, and of typhoid fever in Evesham,
England, and Rennes, France, were also noted by the above-
named authors, but the evidence was far from conclusive.
More recently Hutchings and Wheeler 2 report an outbreak
of typhoid fever in a hospital at Ogdensburg, N. Y. Many
were attacked, and the disease was suspected to be due to
the use of ice. An examination of the ice in the ice house
showed that some of it contained visible dirt from which
colon and typhoid bacilli were isolated. A certain lot of
ice harvested the winter before began to be used on Sep-
tember 26. Eight cases of typhoid fever developed from
October 2 to 5 and subsequent cases followed until some
time in November, involving in all 39 persons. The earliest
cases had a temperature of 104° and 105° on the first day.
An examination of the ice in the ice house showed that it
contained visible dirt from which colon and typhoid bacilli
were isolated, the latter determined by agglutination.
Dr. H. W. Hill 3 suggests that the evidence in this case is far
from conclusive, because 6 days is a very short period of
incubation, and for 8 cases to develop in 9 days would be
still more unusual, and to have the cases, after so short an
incubation, develop a temperature of 104° and 105° on the
first day is even more remarkable. Hill thinks that this
indicates an infection before the use of the ice, and he calls
attention to the fact that typhoid outbreaks in hospitals

1 Sedgwick and Winslow, Mem. Am. Acad. Arts & Sc., 1902, XII,
No. 5, 472.

* Hutchings and Wheeler, Am. J. M. Sc., Phila., 1903, n. s., CXXVI,
680.

• Hill, Address, National Ice Association of Amercia, N. York, 1910.

INFECTION BY FOOD AND DRINK 341

for the insane due to other causes are by no means rare.
If typhoid bacilli really survived in the ice he says that it
would indicate a persistence of a month over what has ever
been observed in experiments. At Batangas in the Philip-
pines l an outbreak of bacillary dysentery at the army post
was said to be due to contaminated artificial ice, but no
data were given. Park, in the paper cited below, refers to
a small outbreak of typhoid fever due to ice cut from a pond
where sewage flowed onto ice already formed and then
became frozen. Anderson 2 reports an outbreak at Lexing-
ton, Va., possibly due to artificial ice made from spring
water polluted with typhoid excreta, from which water he
isolated typhoid bacilli, proved to be such by cultural and
agglutination tests. Hamer3 refers to a possible instance
of ice-borne typhoid in London in 1898. Judging from the
small number and indefinite character of these reports, it
appears that there is little direct evidence of danger from ice.
In the outbreak reported by Anderson and in the one in the
Philippines the ice was artificial. With such artificial ice the
danger must be very much greater than in the case of nat-
ural ice, for in the former the bacilli are not excluded dur-
ing the process of freezing and little time is allowed for
their natural death in the ice. The one noted instance
of infection by natural ice which has seemed most con-
clusive has been shown by Hill to have no significance.
There is on the other hand strong epidemiological evidence
that ice is not a factor of much moment in the causation
of disease. Hundreds of thousands of tons of ice are cut
from sewage-polluted waters in the United States, but
sickness does not appear to follow the use of this ice in
drinking water. Park 4 has studied this point in New York
City, where four-fifths of the ice is cut from the markedly

1 Rep. Bu. Health, P. L, 1908-9, 54.

2 Anderson, Med. Rec., N. Y., 1908, LXXIV, 909.

8 Hamer, Rep. Med. Off. Health, Lond., 1904, Appendix I.
< Park, J. Am. M. Ass., Chicago, 1907, XLIX, 731.

342 THE SOURCES AND MODES OF INFECTION

contaminated Hudson River. Vast quantities of this ice are
used in drinking water, and if it is the bearer of typhoid
fever there should begin a noticeable increase of the disease
when the use of the new ice commences in March, and this
increase should continue during April and May. During the
ten years studied by Park no such increase was noted, and
no outbreak occurred which could be attributed to ice.

Infection by Milk.

Character of Outbreaks. — It has long been known that
some of the common contagious diseases of human beings
may be transmitted by means of milk. The only diseases
that are definitely known to be so carried are typhoid fever,
scarlet fever and diphtheria. There is no reason why dysen-
tery and cholera, and perhaps other diseases, should not be
transmitted in the same way, but no evidence of it has come
to my knowledge, except that referred to by Kober,1 who
states that McNamara in Calcutta, in 1872, traced an out-
break of cholera to an infected dairy. Outbreaks of the
above diseases have been reported from time to time, and
have been studied by Hart, Kober and others, but the latest
tabulation and the fullest consideration of the subject are
to be found in Bulletin 56, Hygienic Laboratory, United
States Public Health and Marine Hospital Service, 1909.
Milk outbreaks of these diseases have an explosive character,
such as is shown by water-borne outbreaks of typhoid fever.
In scarlet fever and diphtheria the explosive character is owing
to the shorter incubation of these diseases, more pronounced
than it is in typhoid fever. A milk outbreak is determined
to be such from the presence of an excessive number of cases
on a particular milk supply which cannot be accounted for
in any other way. From three to five cases within a few
days on a route covered by one wagon in a city with an
average typhoid death rate is highly suspicious, and in most

1 Kober, J. Am. M. Ass., Chicago, 1907, XLIX, 1091.

INFECTION BY FOOD AND DRINK 343

instances proves to be the beginning of an outbreak. The
more the customers of a dealer are scattered, the stronger
is the evidence. Care must be taken to exclude all other
sources, such as neighborhood or family contact, water, ice,
shellfish or other foods. Milk outbreaks often prevail more
extensively among the well-to-do, owing to their greater con-
sumption of milk, and a typhoid milk outbreak usually shows
a high percentage of children attacked. Often the source of
infection can be found. In only two instances reported by
Konradi 1 and Shoemaker 2 has the typhoid bacillus been
isolated from milk, and in only four instances has the
diphtheria bacillus been recovered.3

Number of Outbreaks. — The number of outbreaks pretty
definitely traced to milk is quite large. Trask, in the Hygi-
enic Laboratory Bulletin referred to, tabulated 179 outbreaks
of typhoid fever, 51 of scarlet fever, 23 of diphtheria and
7 of sore throat. Besides these Hart reported 51 of typhoid
fever, and Busey and Kober 86 of typhoid fever, 59 of scarlet
fever, 21 of diphtheria, making in all 316 outbreaks of typhoid
fever, 125 of scarlet fever, 51 of diphtheria and 7 of sore
throat. In these are not included the 90 outbreaks tabu-
lated by Caroe. This is certainly an impressive aggregate,
but it must be remembered that these records cover a period
of perhaps half a century. To determine within any degree
of accuracy how large a part milk plays in the spread of
these diseases is difficult, but it seems to me that the tend-
ency at the present time is to exaggerate its importance.
Schiider 4 found that of 640 outbreaks of typhoid fever 462
were caused by water and 110 by milk, but it is highly im-
probable that cases caused respectively by water and by milk

1 Konradi, Centralbl. f. Bakteriol. [etc.], Jena, I Abt. Orig., XL, 31.

2 Shoemaker, J. Am. M. Ass., Chicago, 1907, XLVIII, 1748.

» Nuttall and Graham-Smith, The Bacteriology of Diphtheria,
Cambridge, 1908, 326.

4 Schiider, Ztschr. f. Hyg. u. Infectionskrankh., Leipz./ 1901,
XXXVIII, 343.

344 THE SOURCES AND MODES OF INFECTION

occur in any such ratio in the United States. Most of the
water-borne typhoid fever does not occur in outbreaks, but
is due to the continuous pollution of municipal supplies.
Milk-borne typhoid fever, on the other hand, probably does
occur chiefly in the form of outbreaks. During the last four
years, in Providence, 155 of the 600 milk dealers have had
one or more cases of typhoid fever on their routes. In only
9 instances have there been more than 2 cases on a single
route within a week. Investigation promptly showed that
in 3 instances the groups of cases were due to personal con-
tact, and in the other 6 instances an unmistakable milk
outbreak developed. It is true that some epidemiologists
assume that a considerable number of sporadic cases are due
to milk, but the evidence appears to be the other way. I
think that a study of the distribution of typhoid cases on
milk routes in Washington, as shown in diagrams in bulletins
35, 44 and 52 of the Hygienic Laboratory, United States
Public Health and Marine Hospital Service, bears out my
contention that outbreaks of typhoid fever due to milk are
not likely to escape notice if the cases are daily tabulated
according to milk supply. Furthermore, these charts show
that there is little excess of typhoid fever on individual milk
supplies except in those instances where there is a noticeable
outbreak.

Fewer Outbreaks in Large Cities. — It is also to be noted
that very few outbreaks are reported from the large cities
like New York, Chicago and Philadelphia, due perhaps to
the fact that the chemical and bacteriological changes in the
milk resulting from the long haul necessary for such cities
are unfavorable for the typhoid bacillus. In the country,
milk outbreaks are not likely to be a factor of moment, as
large milk routes are not common. It is in the medium-sized
and smaller cities and villages that most of the milk out-
breaks are reported. It will be noted that though scores of
outbreaks of typhoid fever may occur each year, they do not
include more than a small fraction of the total typhoid fever.

INFECTION BY FOOD AND DRINK 345

It may be claimed that only a small proportion of the milk-
borne outbreaks are recognized as such, but I cannot think
that this is so in the English cities and the better class of
American cities. In Providence during the last twenty-five
years there have been nine outbreaks of typhoid fever, includ-
ing 363 cases, or about 8 per cent of the total occurring during
that time. But the typhoid death rate in Providence is
below the average, so that 8 per cent in Providence would
probably be equivalent to not over 4 per cent in the average
American city with its polluted water supply and numerous
privy vaults. In only a few cities have Trask, Hart, and Busey
and Kober been able to find records of more than two, or more
rarely three, outbreaks ; yet we can scarcely believe that milk
outbreaks could have, except rarely, escaped detection in the
English cities, or in such cities as Boston, Worcester, Spring-
field, Rochester, Baltimore, Philadelphia and New York.

Amount of Typhoid Fever Due to Milk. — The city of
Washington is reported as showing an excessive amount of
milk-borne typhoid fever.1 In 1906, 79, or 9 per cent, of the
866 cases investigated by the commission, were traced to milk;
in 1907, 31, or 5 per cent, of 635 cases investigated; and in
1908, 54, or 8 per cent, of 679 cases. According to figures
furnished by Dr. Woodward, Health Officer of Washington,
the number of typhoid cases due to milk during the three
years was 139, which gives a percentage of 4.6. The com-
mission considers that 10 per cent of the typhoid fever in
Washington is due to milk. These percentages, however, seem
to me unfair, as doubtless all the outbreaks were reported,
and the percentages should, from the standpoint of the present
consideration, be based on the total number of cases reported
in the city, and not on the cases studied and of known origin.
This gives a percentage, as above stated, of 4.6 of the typhoid
fever in Washington during the three years 1906-08 as due
to milk-borne typhoid. The importance and value of the

1 U. S. Pub. Health & Mar. Hosp. Serv., Hyg. Lab., Bull. No. 35,
59; No. 44, 46; No, 52, 100.

346 THE SOURCES AND MODES OF INFECTION

reports on typhoid fever in Washington have, as it appears,
given a rather exaggerated importance to the danger from
milk. That there was a certain amount of typhoid fever in
Washington due to milk, during the three years in question, is
no reason for assuming a similar percentage of milk-borne
typhoid fever for the other cities of the country. Trask was
able to find during the years 1903-07 about 1900 cases in
the United States traced to milk. During this period there
were in the registration area 57,023 deaths, and the total num-
ber in the whole country must have been at least double, or
114,000. The fatality is probably not over 10 per cent, so
that there must have been at least 1,000,000 cases. Nine-
teen hundred is 0.19 per cent, which is very different from the
10 per cent now commonly assumed as due to milk. I do not
of course believe that all the milk outbreaks were reported,
but the number would have to be increased fifty-fold to
equal 10 per cent, and it can scarcely be believed that 98 per
cent of all milk outbreaks fail of recognition.

Scarlet Fever and Diphtheria. — Scarlet fever and diph-
theria are certainly much more rarely transmitted by means
of milk than is typhoid fever. Although the percentage of
typhoid fever, scarlet fever and diphtheria due to milk is
small, the danger is a real one and the aggregate of cases not
inconsiderable, and their occurrence should, if possible, be
guarded against.

Animal Sources. — The infection of milk with the above
diseases is almost invariably from human sources. There is
no evidence to show that cows may be infected with the
bacillus of typhoid fever, and the contamination of the milk
by drinking sewage-polluted water does not occur. Scarlet
fever also does not appear to be an animal disease. Power
and Klein in England believed that they had found the cause
of milk-borne outbreaks of scarlet fever in the sickness of the
cows supplying the milk. These findings have not been sub-
stantiated, and the micrococcus which Klein believed was
the cause of the disease has not been accepted as such by

INFECTION BY FOOD AND DRINK 347

bacteriologists. While many of the lower animals may be
successfully inoculated with diphtheria, the disease does not
appear often to occur spontaneously among them. Never-
theless two milk-borne outbreaks have been traced to diph-
theria of the cows' udder. One of these was reported by Dean
and Todd1 and the other by Ashby.2

Human Sources. — Besides the bacteriological evidence
that milk rarely acquires its infection of these diseases from
animals, we have the direct evidence that in a considerable
number of outbreaks the milk is known to have been handled
by persons sick with the disease or carrying its organisms.
Trask 3 says that in 113 of the 179 outbreaks of typhoid fever
reported by him an infected person was found to have
handled the milk, and in only 4 instances was the infection
reported to have come from polluted water used in washing
utensils, and in 4 others to the use of milk bottles from infected
houses. In 35 of the 57 scarlet-fever outbreaks an infected
person was found to have handled the milk, and 3 others were
due to the use of bottles from infected houses. Of the 23
diphtheria outbreaks, 18 were traced to handlers of the
milk. A study of the details of milk outbreaks shows that in a
large proportion the infecting case was not recognized as such,-
that the symptoms were mild or atypical, that infection oc-
curred during the incipient stage, or that the person was
merely a carrier. Infection may occur at any time during the
handling of milk, from the beginning of milking until the milk
is delivered to the consumer, and it may also be caused by those
who take care of the pails, cans, bottles, strainers, drums, etc.
Outbreaks are believed to have been caused by the tasting of
milk by infected dealers, and in the instance reported by
Shoemaker 4 a convalescent patient was in the habit of start-'

1 Dean and Todd, J. Hyg., Cambridge, 1902, II, 194.

2 Ashby, Pub. Health, Lond., 1906-07, XIX, 145.

3 Trask, U. S. Pub. Health & Mar. Hosp. Serv., Hyg. Lab., Bull.
No. 41, 25 et seq.

* Shoemaker, J. Am. M. Ass., Chicago, 1907, LXVIII, 1748.

348 THE SOURCES AND MODES OF INFECTION

ing a siphon from the milk drum by sucking the tube. It
is also believed by many that flies are an important factor
in the contamination of milk.

Protection of Milk. — There are various ways in which the
danger from infected milk may be avoided.

First. If all recognized cases of the diseases just considered
are promptly reported, and proper measures of isolation are
insisted on, there will be a considerable decrease in the
amount of milk-borne infection. When such a disease occurs
in the family of a milk producer or dealer the patient should
usually be removed from the premises, and the vessels, etc.,
if possibly infected, should be. disinfected. The greatest diffi-
culty is encountered in the management of convalescents,
and particularly carriers, and it is the view of most health
officers that the latter should be permanently excluded from
the milk business. Unfortunately there are many mild cases
of contagious disease which are never recognized and are not
likely to be. Moreover, persons in the incipient stage are
quite likely to transmit the disease, and we cannot hope for
much earlier recognition than we now have. Hence for the
prevention of outbreaks from these unknown sources we must
rely on other methods.

Second. It is possible to reduce the chances of infection
by insisting on cleaner methods in handling milk and requir-
ing the sterilization of all vessels, particularly of bottles. If
all handlers would wash their hands, and keep their hands
out of the milk and out of the vessels, there would be no milk
outbreaks. But we can hope for no very radical improvement
as regards the cleanliness of milkers, farm hands and peddlers.

Third. The pasteurization of the milk by the dealer before
delivery would also decrease to a large extent the amount of
milk-borne disease, for a study of outbreaks shows that in a
large proportion the infection of the milk takes place on the
farm. But there is also considerable risk of contamination in
the hands of the dealer after pasteurization. For this
reason much attention has been devoted to the problem of

INFECTION BY FOOD AND DRINK 349

pasteurization in bottles so that the milk may with cer-
tainty be delivered to the consumer free from infection.
Pasteurization of beer in bottles has long been practiced by
brewers, and investigators1 and milk dealers have adapted
the methods of brewers to the treatment of milk and some
large milk dealers are now delivering milk prepared in this
way.

Fourth. In the absence of pasteurization by the dealer,
doubtless the most successful way of preventing milk-borne
disease at the present time is for the public to consume no
milk except that which has been pasteurized or scalded in
the house. Many object to the trouble and do not like the
taste of heated milk, so that this practice is not likely to
become universal. Moreover outbreaks have occurred, as
at Bristol,2 where milk pasteurized in an institution had
been afterwards infected by a maid.

It therefore seems that while it is possible materially to
lessen by the above methods the amount of milk-borne
disease, we cannot hope at present entirely to do away with
the danger of milk-borne outbreaks of the diseases we have
been considering.

Ice Cream. — As milk is so frequently the bearer of dis-
ease, it is not surprising that outbreaks have at times been
attributed to food products derived from milk. Thus
Soper3 reported an outbreak due to the use of cream on
breakfast food, and one of the outbreaks reported by Trask
was due to whipped cream. Sedgwick and Winslow4 col-
lected records of four outbreaks of typhoid fever due to the
use of ice cream, and another has been reported by Barras.5

1 North, Med. Rec., N. York, 1911, LXXX, 111.

2 Davies and Walker, Proc. Roy. Soc. Med., Lond., 1808, Epi-
demiol. Sec., 175.

3 Soper, J. Mass. Ass. Bds. Health, Bost., 1904, XIV, 68.

4 Sedgwick and Winslow, Mem. Am. Acad. Arts & Sc., 1902, XII,
No. 5, 475.

6 Barras, Lancet, Lond., 1904, II, 1281.

350 THE SOURCES AND MODES OF INFECTION

A quite remarkable ice-cream outbreak of typhoid fever
occurred in Eccles and Manchester in November, 1910,
involving 108 cases.1 The makers and sellers of the ice
cream were two Italians in a tenement house in Manchester.
Professor Delepine, who investigated the outbreak bacterio-
logically, was of the opinion that the ice cream must have
been so grossly infected as only to be explained by a multi-
plication of the bacilli in the materials used after they had
been heated and before freezing. Of the 13 persons living
in the house with the Italians there were 6 whose blood gave
a positive Widal reaction. No examination of their excreta
appears to have been made.

Buchan2 refers to an outbreak of scarlet fever due to ice
pudding and reported by Buchanan in 1875. He also notes
two outbreaks due to the infection of ice cream with
B. enteritidis (Gaertner). Buchan made many examina-
tions of market ice cream in Birmingham and suggested a
bacteriological standard for ice cream and a set of excellent
rules for its manufacture.

Another typhoid outbreak due to ice cream has recently
been reported from Fort Sill, Okla., where 20 cases were
due to eating ice cream from the neighboring town of
Lawton.3

Butter. — While it has been shown that pathogenic bacilli
may live for some time in butter, only one outbreak seems
to have been reported as arising from the use of this sub-
stance. This is probably due partly to the fact that the
germs must die off pretty rapidly, leaving perhaps only a
few survivors, and partly to the difficulty of tracing an
outbreak to such a source.

An outbreak of diphtheria at Lewiston, Minn., was be-
lieved to have been caused by eating infected butter.
There had been no diphtheria in the place until a boy re-

1 Rep. Med. Off. Local Gov. Bd., Lond., 1910-11, XL, 18.
* Buchan, J. Hyg., Cambridge, 1910, X, 93.
8 Rep. Surg. Gen. Army, Wash., 1909-10, 45.

INFECTION BY FOOD AND DRINK 351

turned from the " Twin Cities " after an attack of diph-
theria. The milk from the farm where he lived was sent to
a creamery and every family in the place, in which there
was diphtheria, was found to have used butter from this
creamery. Experiments showed that diphtheria bacilli can
live in butter for a month.1

Epidemic Sore Throat or Tonsillitis. — According to
Winslow2 outbreaks of what is often called septic sore throat
have many times been reported in England and have been
shown to be associated with particular milk supplies.
Some of these outbreaks more or less resemble scarlet fever
or include some cases of typical scarlet fever. In others
there is no reason to consider the infection of a scarlet-fever
nature. Winslow, without a systematic search of the
literature, was able to find 12 outbreaks of this character.
In 4 of these outbreaks there was no evidence whatever as
to the original source of the disease. In 4 others there was
a dubious connection with some disease in cows. At Glas-
gow and Colchester, Winslow thinks that there was fairly
strong evidence connecting the outbreaks with an in-
flammatory condition of the cow's udder. At Rothesay
probability pointed to human infection and at Working
there had been both human quinsy and bovine udder in-
flammation on the farm. Another outbreak3 occurred in
Christ iania in 1908 and it was not determined whether it
had a bovine or human origin.

Sore Throat in Boston. — The first recorded American
outbreak occurred in Boston in 1911 and was reported by
Winslow as above. In Boston and the adjoining towns of
Brookline and Cambridge there were, during a short period
in May, probably not far from 2000 cases, of which 48
proved fatal and many more were quite severe. Many of

1 Rep. State Bd. Health, Minn., 1911, 203.
* Winslow, J. Infect. Dis., Chicago, 1912, X, 73.
3 Norsk Mag. f. Lagevidenskahen, 1908, LXIX, 585, 811: Ref.
J. M. Ass., Chicago, 1912, LVIII.

352 THE SOURCES AND MODES OF INFECTION

the cases developed abscesses, and not a few general septic
infection, as has been the case in European outbreaks.
The outbreak was quite clearly traced to milk and cream
supplied by the Deerfoot Farm Dairy in Southboro. This
dairy and the farms supplying it were under the constant
supervision of a trained bacteriologist and the cows were
frequently inspected by a veterinary physician; the milk
received as little handling as possible and indeed everything
was done to deliver milk of a high bacterial standard and
free from infection. It seemed improbable that there waa
any udder disease among the cows which could have given
rise to the trouble. The farms supplying the milk are
largely located some distance from Boston in Southboro
and to some extent in the adjoining towns of Westboro,
Framingham and Marlboro. As shown by the investiga-
tions of Winslow there was a marked prevalence of the same
disease in the towns of Marlboro, Southboro and Hudson
all through April and May. In these towns the disease
appeared to pass by contact from person to person in prose-
demic fashion, just as diphtheria ordinarily does, except
that there was marked and sudden excess in Marlboro and
Southboro in May, coincident with the Boston outbreak.
In both these towns much Deerfoot milk is sold, but none
in Hudson, where there was no May exacerbation. Wins-
low is of the opinion that the milk in some way became
infected from cases or carriers, though no such infection
could be demonstrated. As streptococci were isolated from
the internal organs of fatal cases these were thought to
be the infecting organisms. Other outbreaks have been
reported from Baltimore l and Chicago.2

Views concerning Tuberculosis. — Of all the animal
diseases which are transmissible to man tuberculosis has
received the most attention. After many years of discussion

1 J. Am. M. Ass., Chicago, 1912, LVIII, 1109.

2 Miller and Capps, J. Am. M. Ass., Chicago, 1912, LVIII, 1111.

INFECTION BY FOOD AND DRINK 353

and investigation there is now general agreement that while
the two types of tubercle bacilli, human and bovine, are dis-
tinct and quite permanent, they are reciprocally infective,
though not to the same degree. That human beings may
become infected with tuberculosis derived from cattle is now
generally admitted, but there is as yet no general agreement
as to the amount of such infection, though present views are
not so divergent as those of a few years ago. The question
of the mode of infection in tuberculosis and the source of the
bacilli is so largely dependent on pathological evidence that
it is very difficult for one who is not a pathologist properly to
weigh the evidence. Yet in view of the lack of agreement
among pathologists it becomes necessary for the health officer,
who must take definite action to combat the disease, to
attempt to arrive at some sort of conclusion. It seems fair,
from the evidence thus far available, to draw the following
conclusions :

Bovine Bacillus may infect Man. — First. The bovine
type of bacillus is capable of causing infection in human
beings. This is shown by the successful inoculation of human
beings with bovine virus. A considerable number of cases
of the accidental inoculation of veterinarians and laboratory
workers while manipulating animals dead with tuberculosis
have been reported by Pfeiffer, Tscherning, Naughton, Cop-
pez, Priester, Hartzell, Grothan, Jadasohn, Ravenel, de
Jong,1 Salmon 2 and others. In most cases the lesion was
local, and some have argued that this is strong evidence of
weakened virulence for man. The finding of the bovine type
of the bacillus in human beings is generally considered as
furnishing a more certain demonstration of the latter's suscep-
tibility to the disease. As will be again referred to, the Ger-
man and British commissions, and various Americans, have

1 Moss, Johns Hopkins Hosp. Bull., 1909, XX, 39; Cornet, Nothnagel's
Encyclopedia of Practical Medicine, Phila. & Lond., 1907, Tubercu-
losis, 77.

» U. S. Dept. Agric., Bu. An. Ind., Bull. No. 33, 16.

354 THE SOURCES AND MODES OF INFECTION

found the bovine bacillus in from 16 to 25 per cent of the
cases studied. These facts, while demonstrating the suscep-
tibility of human beings to the bovine form of the disease,
do not indicate definitely to what extent the bovine type
prevails in man, as the cases examined were usually specially
selected.

Feeding Experiments. — Second. Tuberculosis may be
produced experimentally in animals by the ingestion of mate-
rial containing tubercle bacilli. Milk known to be infected
with tubercle bacilli has been in numerous instances fed to
guinea pigs, rabbits, dogs, calves, swine and monkeys, and
has caused fatal tuberculosis in them. This has been demon-
strated by a large number of workers in all parts of the world.
A good summary of previous experiments is given by Mohler,1
who himself produced tuberculosis by feeding guinea pigs
with milk from tuberculous cows. Subsequent experiments
by the Bureau of Animal Industry 2 show that while guinea
pigs are with difficulty infected by feeding, hogs are very
easily infected, the primary infection apparently occurring in
the submaxillary glands, the lungs usually becoming second-
arily infected.3 That the ingestion of tuberculous material,
especially milk from creameries, is the chief cause of tuber-
culosis in swine, is the opinion of the officers of this depart-
ment and of Ravenel and Russell.4 The more recent work
of Calmette and the French school, while not always including
direct experiments with milk, show that material containing
tubercle bacilli, introduced in any way into the stomach or
intestines, results in the production of tuberculosis, and even
the pupils of Fliigge admit that tne ingestion of a sufficient
number of bacilli will cause the disease, though it is produced,
they say, much more easily by inoculation or inhalation. It

1 Mohler, U. S. Dept. Agric., Bu. An. Ind., Bull. No. 44, 13.
a U. S. Dept. Agric., Bu. An. Ind., Circ. 83, Bull. Nos. 86, 88, 93.
8 Bull. No. 88, 46.

* Ravenel and Russell, Am. Pub. Health Ass. Rep., 1906, XXXII,
Pt. I, 139.

INFECTION BY FOOD AND DRINK 355

is also generally admitted that bacilli may be absorbed
through the intestines without leaving any lesion. While
some pathologists think that it is usually possible to deter-
mine the route of infection by the age and character of the
lesions, other equally good men do not consider that any very
reliable conclusions can be arrived at in this manner. Under
these circumstances it appears that we cannot as yet appeal
to the pathologists for a definite decision as to the mode of
infection in this disease, and we may be permitted to assume
that a large amount of tuberculosis may be due to infection
through one part or another of the alimentary canal.1

Tubercle Bacilli in Milk. — Third. Tubercle bacilli are
frequently found in cows' milk. It was formerly believed
that tubercle bacilli do not occur in cows' milk unless the
udder is diseased. Nevertheless various observers from time
to time have reported finding the bacilli although no udder
disease was evident. Finally Schroeder and Cotton2 showed
that frequently the feces of cattle contain large numbers of
tubercle bacilli, although the animals present no visible symp-
toms of disease. They furthermore showed that the pollu-
tion of milk with infected feces is the most common source
of tubercle bacilli found in milk. As from 3 to 50 per cent
of the cattle in different parts of the United States react to
tuberculin, and probably 25 per cent of the cattle in Great
Britain are infected, it is not surprising that market milk in
these countries is frequently found to contain tubercle bacilli
in sufficient numbers to cause the death of test animals.
Anderson3 gives a resume of the literature of the subject.
Among the percentages of infected milk referred to are, in
Copenhagen, 14.3 per cent ; Boston, 21 per cent and 40 per cent ;
Liverpool, 5.2 per cent from city dairies, 13.4 per cent from

1 For other references to the causation of tuberculosis through the
alimentary canal see p. 255.

3 Schroeder and Cotton, U. S. Dept. Agric.^Bu. An. Ind., Bull. No. 99.

1 Anderson, U. S. Pub. Health and Mar. Hosp. Serv., Hyg. Lab.,,
Bull. No. 41, 163.

356 THE SOURCES AND MODES OF INFECTION

country dairies, though later,1 owing to the efforts of the
authorities in eradicating the disease, the percentage in city
dairies was 1.4 per cent and in country dairies was 7 per cent;
Genoa, 9 per cent; London, 22 per cent; and Berlin, 28 per
cent. Recently of 620 samples of milk consigned to London,
61, or 11.6 per cent, were found to contain tubercle bacilli.2
Anderson found 6.72 per cent of 223 samples of market milk
in Washington to contain virulent tubercle bacilli. Hess 3
found virulent tubercle bacilli in 16 per cent of 107 samples
of New York market milk (not bottled). He thinks the
reason that his percentages are higher than those of Anderson
is the employment of more test animals and the injection of
the cream as well as the milk. He also found virulent
tubercle bacilli in one of eight samples of commercially " pas-
teurized" milk. In Manchester, England, 7.7 per cent of 542
samples of milk were found to be infected.4

Danger Less than Supposed. — There seems to be no
doubt that a large part of the milk consumed in Europe and
the United States contains tubercle bacilli in numbers suffi-
cient to cause the disease in test animals. It also seems to
be certain that in the lower animals at least, particularly in
young individuals, the ingestion of this tuberculous milk will
cause not only tuberculosis of the alimentary canal, but will
produce pulmonary disease and generalized tuberculosis also.
It is probable that the ingestion of such milk by human be-
ings will produce similar results. There is, however, consider-
able variation in different kinds of animals as regards their
susceptibility to this sort of infection. Even in animals as
susceptible to the disease as are guinea pigs, infection by the
ingestion of milk under normal conditions is not very easy.
Thus Schroeder and Cotton5 fed 132 guinea pigs with tuber-
Rep, on Health of Liverpool, 1906, 189.
Rep. Med. Off. Health, Lond., 1908, 60.

Hess, Sixth Internat. Cong, on Tuberc., Wash., 1908, IV, 523.
Rep. on Health of Manchester, 1906, 187.

Schroeder and Cotton, U. S. Dept. Agric., Bu. An. Ind., Circ.
No. 83.

INFECTION BY FOOD AND DRINK 357

culous milk on the average for forty-seven days, and only 1,
which was fed for 357 days, contracted the disease. This
was only 0.76 per cent. At the same time 14.28 per cent of
the guinea pigs exposed in the mangers of the cows from which
the milk was obtained developed tuberculosis. Dr. Ver-
ranus Moore writes me that he knows of a small village
where most of the people received the milk from a herd
of cattle that later was found to be extensively diseased;
that is, about 77 per cent reacted to tuberculin and some
2 or 3 per cent showed the disease on physical examination,
and a considerable percentage of the guinea pigs inoculated
with the mixed milk from the herd died of tuberculosis.
This community had used this milk for a number of years.
After the facts as stated above were ascertained the condi-
tion was changed and up to this time there has not occurred
a single case of recognized tuberculosis. Hess l examined 18
children who a year previous had been known to be consum-
ing tuberculous milk. None of them showed any visible signs
of tuberculosis, though 4 reacted to the tuberculin test. Of
100 children at Randalls Island, N. Y., fed on milk from
tuberculous cows, none developed the disease.2 A recent
note,3 which I have not had the opportunity to verify, gives
the results of some observations made by Weber of the Ger-
man Imperial Health Office on the use of tuberculous milk.
From January, 1905, to April, 1908, 69 cows with tuberculosis
of the udder were kept under observation, the milk from
which was consumed raw by 360 persons, of whom 159
were children. Of these, 5, of whom 2 were between 1
and 2 years of age, were " indubitably infected with tuber-
culosis through the use of the milk." The children had
taken the milk of these cows from one to one and a half
years. In both cases the milk was used raw by all the mem-
bers of the family, the parents and several children; all these

1 Hess, J. Am. M. Ass., Chicago, 1909, LII, 1014.

2 Park, Sixth Internat. Cong, on Tuberc., Wash., 1908, I, 156.
» J. Am. M. Ass., Chicago, 1910, LIX, 886.

358 THE SOURCES AND MODES OF INFECTION

people remained healthy. In the sick children there was
merely an affection of the cervical glands in which tubercle
bacilli of the typus bovinus were demonstrated by bacteri-
ological tests. There were no other symptoms. Whitla,1
however, reports an instance where of 150 children fed on
milk known to be tuberculous, 15 contracted the disease.

Percentage Due to Bovine Infection. — It has thus far
been shown merely that human beings may contract tuber-
culosis by drinking the milk of tuberculous animals. It re-
mains to determine, if possible, how great this danger really
is. Various attempts have been made to estimate this. It
has been assumed by some that intestinal tuberculosis, and
to some extent generalized tuberculosis without preponderant
pulmonary involvement, is indicative of infection by ingesta,
and presumably by milk. The fact that these types of the
disease are more common in children, of whose diet milk
forms a relatively large part, has been thought to lend color
to this view. It is riot improbable, however, that the dif-
ferent form which the disease presents in early life may be
due to the characteristics of the age rather than to the mode
of infection. Children certainly consume a relatively large
amount of milk, but tuberculosis is not so common among
them as among adults. In Providence only about 15 per cent
of the tuberculosis is in children under five years of age, and
only about 4 per cent of the tuberculosis death rate is attrib-
uted to abdominal tuberculosis. In the registration area of
the United States abdominal tuberculosis accounts for about
3.5 per cent of all deaths from this disease. This form of
tuberculosis appears to be much more common in England.2
But careful observers find that contact with other cases is
probably responsible for a large proportion of tuberculosis in
children. Park 3 states that of 100 cases of glandular and
bone tuberculosis in St. Mary's Hospital, New York, 44 per

1 Whitla, Lancet, Lond., 1903, II, 135.

2 Bovaird, Sixth Internal. Cong, on Tuberc., Wash., 1908, II, 446.
8 Park, Sixth Internal, Cong, on Tuberc., Wash., 1908, I, 157.

INFECTION BY FOOD AND DRINK 359

cent had been in close contact with tuberculous cases.
LaFetra1 found 40.4 per cent of 131 cases of tuberculosis in
infants, probably due to family infection. Comby 2 considers
family infection of prime importance in the tuberculosis of
children. Floyd and Bowditch 3 found that 679 of 1000
tuberculous children had been in contact with the disease in
their homes. Approaching the subject in another way, they
found signs of the disease in 66 per cent of 746 children living
in tuberculous families among the poor. Miller and Wood-
ruff4 found the same true in 51 per cent of 150 children, and
Sacks 5 in 53 per cent of 322 children.

While there seems to be no doubt that tubercle bacilli may
remain latent in the body for some time, there are very few
who accept von Behring's view that most human tubercu-
losis is acquired in childhood from drinking tuberculous cows'
milk. As greatly discrediting von Behring's theory may be
mentioned the investigations of Speck,6 who found that of
8010 cases of tuberculosis only 27 per cent had been fed on
cows' milk in infancy. Von Ruck7 found that certainly not
over 25 per cent, and possibly not over 10 per cent, had been
brought up on cows' milk. Flick8 obtained similar evidence
at the Phipps Institute in Philadelphia. Heymann9 says
that in Christiania, where nearly all the infants are nursed,
tuberculosis is more common than in Bavaria, where artificial
feeding is very common, and that in Prague, where nearly all

1 LaFetra, Sixth Internat. Cong, on Tuberc., Wash., 1908, II, 361.

2 Comby, Sixth Internat. Cong, on Tuberc., Wash., 1908, II, 503.
» Floyd and Bowditch, Boston M. & S. J., 1908, CLIX, 783.

4 Miller and Woodruff, Sixth Internat. Cong, on Tuberc., Wash.,
1908, II, 487.

5 Sacks, Sixth Internat. Cong, on Tuberc., Wash., 1908, II, 479.

• Speck, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1904, XLVIII.27.

7 Von Ruck, J. Am. M. Ass., Chicago, 1905, XLIV, 1350.

8 Flick, Report of Henry Phipps Inst. Study . . . Tuberculosis,
Phila., 1906, IV, 49.

9 Heymann, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1904,
XLVIII, 45,

360 THE SOURCES AND MODES OF INFECTION

babies are breast fed, the tuberculosis death rate is as high
as in any city in Europe. I am inclined to the opinion that
such epidemiological facts as we have indicate that children
are more susceptible to the human than to the bovine type
of the disease, and there seems to be as much clinical evidence
that the tuberculosis of childhood is due to family infection
as is the tuberculosis of adult life.

Instances of Milk Infection. — Instances are not rarely
reported in which there is considerable evidence of the pro-
duction of tuberculosis in children by the use of milk from
tuberculous animals. One of the most convincing is that
narrated by Adams of Aberdeen and summarized by Hess.1
Two children of a farm laborer died of tuberculous meningitis
in January and March of 1907. There was no known expo-
sure to the disease except the use of milk from a cow with
tuberculosis of the udder. Bacilli identical in type with those
found in the cow were recovered from the cerebro-spinal fluid
of one of the children. Most of the cases reported are not
so convincing as this, and in many the evidence is decidedly
weak and would not stand critical examination. The evidence
afforded by these cases is not of much value, for though some
of the cases are doubtless caused in the manner alleged, they
merely indicate what is also suggested by animal experiments,
— the possibility of human infection from milk. An estima-
tion of the extent of this danger must be made in other ways.

Proportion of Human and Bovine Types. — Since the sig-
nificance of Theobald Smith's discovery of the difference
between the bovine and human types of the tubercle bacillus
has been recognized, the relative frequency of the two types
in human beings has been used as a measure of the impor-
tance of milk infection, for it is conceded that it is through
milk almost exclusively that bovine tuberculosis is trans-
mitted to human beings. Much care and labor are involved
in the differentiation of the two types, but the number of
observations made is considerable, among which those of the
1 Hess, J. Am. M, Ass,, Chicago, 1909, LII, 1015,

INFECTION BY FOOD AND DRINK 361

German and British commissions are notable. Moss1 sum-
marizes the cases previously reported. In all there had been
306 cases investigated, in 63, or about 20 per cent, of which
bovine bacilli were found. But it must not be inferred from
this that 20 per cent of all cases of human tuberculosis would
present this type. With the exception of 54 cases in adults
studied by the German commission, most of the cases have
been children with the intestinal or glandular forms of the
disease. In the 54 adults no bovine bacilli were found. The
bovine type of bacillus has rarely been found in pulmonary
tuberculosis, even among children. It may perhaps be in-
ferred that 25 per cent of bone and glandular tuberculosis in
children is due to the bovine bacillus. But Tendeloo2 calls
attention to the fact that if the bovine type of bacillus is a
permanent one, — and unless it is these observations are of
little value, — the above figures are probably misleading, for
the bovine bacillus if permanent will be transmitted from
man to man and does not always indicate a bovine origin.
As only about 15 per cent of the tuberculosis in the United
States is of the form in which the bovine bacillus is found,
and as only about 25 per cent of the cases show this type of
bacillus, we are justified, using the type as a test, in assuming
that only about 5 per cent of our tuberculosis is derived from
bovine sources. But this, while a small percentage, would
amount to almost 2500 deaths annually in the registration
area of the United States, — certainly a number which ought
to be considered.

Epidemiological Evidence. — It seems to be a fact that
tuberculosis, even those forms of the disease which are gen-
erally supposed to be caused by milk, is not appreciably less
in those parts of the world where cows' milk is little used, or
where it is usually sterilized, or where there is little tubercu-
losis among cattle. Kitasato3 says that there is very little

1 Moss, Johns Hopkins Hosp. Bull., 1909, XX, 39.
* Tendeloo, Sixth Internal,. Cong, on Ttiberc., Wash., 1908, I, 87.
3 Kitasato, Ztschr. f. Hyg. .u. Infectionskrankh., Leipz., 1904,
XLVIII, 471.

362 THE SOURCES AND MODES OF INFECTION

v*

tuberculosis among the cattle in Japan, and so few cows that
the daily milk supply does not average 3 c.c. per capita, yet
nearly 8 per cent of all deaths are due to tuberculosis, and in
a series of autopsies 17 per cent of the tuberculosis cases were
under eighteen years of age, and 10 per cent of the cases
showed primary intestinal infection. Cobb says that substan-
tially the same conditions prevail in China. According to
Heymann1 tuberculosis is very common in Greenland, where
no cows' milk is used. Fisch2 states that on the Gold Coast
no milk is used, yet 12 per cent of the sick have tuberculosis.
Tuberculosis prevails as extensively in Cuba as in the United
States, but Dr. Guiteras tells me that milk is practically
never used without scalding, and tuberculosis is comparatively
rare among cattle. If, as appears from a study of the relative
prevalence of the two types of bacilli, less than 5 per cent of
all tuberculosis is due to cows' milk, the entire elimination of
this factor would not, as, from the data furnished by Japan,
Cuba, and other places, it appears that it does not, have any
appreciable effect on the death rate from this disease. The
total elimination of bovine tuberculosis, so far from " stamp-
ing out the great white plague," as some assert, would prob-
ably not make any noticeable difference in the tuberculosis
death rate. Nevertheless it appears from present evidence
that in the aggregate a large number of deaths are due to
this cause, and if there is any practical method of reducing
this cause of death, effort should be made to apply it.

Protection against Tuberculous Milk. — Various ways are
suggested for eliminating this danger from milk. Among
these is the heating of milk to a sufficient temperature to kill
tubercle bacilli as well as typhoid bacilli and other pathogenic
organisms. Some have urged compulsory pasteurization by
dealers, and this practice is, without any legislation, making
rapid progress in cities. But, as is shown by Hess in New

1 Heymann, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1904,
XLVIII, 45.

2 Fisch, Cor.-Bl. f. schweiz. Aerzte, 1904, XXXIV, 761.

INFECTION BY FOOD AND DRINK 363

York, commercial pasteurization does not always destroy
tubercle bacilli, and the proposition that commercial pas-
teurization be under the control of health officials is a timely
one. As has already been stated, there are objections to the
compulsory pasteurization of all milk sold, hence the federal
Department of Agriculture1 has suggested a classification of
market milk into —

First. Certified milk, which presumably would be free
from danger and of high quality, and would be sold at a
higher price to those who desired it.

Second. Inspected milk, which also should be from cows
free from tuberculosis, but which would not show so low a
bacterial count as the certified milk and which could be sold
for not much if any more than the next class.

Third. All milk not produced under such rigid inspection
should be pasteurized under municipal supervision.

There certainly is a tendency, without any compulsion, for
municipal milk supplies to become classified in the way above
indicated. Another way of removing the danger is for the
consumer to heat the milk after it is received from the dealer.
In this way each consumer can protect himself from the dan-
ger of contracting tuberculosis, typhoid fever and similar
diseases. But most persons will not do this unless they are
induced to do so by a slow process of education, and this
education can scarcely be hastened without causing an
unreasonable fear of milk, which will result in lessening the
consumption of a cheap, digestible and nutritious article of
food. It seems to me that what is needed is a better and
more exact knowledge of the relations of milk to disease on
the part of health officers and physicians, from which ought
to follow a gradual education of the public, but without an
alarmist propaganda, and then there will probably gradually
develop a specialization in the milk business to meet the
necessarily different needs of different people.

1 U. S. Pub. Health & Mar. Hosp. Sen'., Hyg. Lab., Bull. No. 41,
559.

364 THE SOURCES AND MODES OF INFECTION

<%

Eradication of Bovine Tuberculosis. — Another way of
dealing with the tuberculous milk problem is to strive to eradi-
cate bovine tuberculosis, or at least to eliminate from herds
of milch cows all animals in the active stage of the disease.
This is urged by many, and has been and is being attempted.
It is claimed by the federal and state departments of agricul-
ture that the existence of bovine tuberculosis entails great loss
upon the farmers, and if this is true, these departments should
devise and urge means for eliminating the disease, but the
expense should be borne by the farmers and not by the public.
If, however, the restriction or elimination of bovine tubercu-
losis is urged as a public health measure, as it so often is, we
should count well the cost before placing it upon the general
taxpayer. We ought to have a finer sense of financial per-
spective in sanitary matters. The cost of efficient measures,
if efficient measures have yet been devised, for eliminating
bovine tuberculosis will be enormous, and we may well con-
sider whether as good results from a public-health standpoint
may not be secured, say by pasteurizing milk, and whether
the money might not be expended in other ways with far
greater results. It is very doubtful whether the expenditure
of $100,000 a year for several years in a state like Massachu-
setts would really result in stamping out the disease in cattle ;
but consider how much such a sum would accomplish if spent
in the establishment of dispensaries and the employment of
district nurses, with perhaps $5000 to $10,000 used for fur-
ther scientific study of the many unsolved problems connected
with the causation, prevention and cure of the disease.

Tubercle Bacilli in Butter. — From what is known of the
viability of the tubercle bacillus, there can be no doubt that
it will live for some time in butter. Schroeder and Cotton1
carefully investigated the subject and found virulent tubercle
bacilli in butter after three months, and Trask2 refers to a large

1 Schroeder and Cotton, U. S. Dept. Agric., Bu. An. Ind., Circ.
No. 127, 1898.

8 Trask, J. Am. M. Ass., Chicago, 1908, LI, 1491.

INFECTION BY FOOD AND DRINK 365

number of similar observations. Cases of tuberculosis do not
seem to have been traced to the consumption of such butter,
and it is very unlikely that they would be traced even if they
occurred. Owing to the small amounts of butter consumed,
it is probable that very few cases of tuberculosis are caused
in this way. Butterine also has been shown to be infected
from the milk, fat and butter of which it is made.1 Owing
to the mode of making and the time occupied in curing
cheese, virulent tubercle bacilli are probably rarely if ever
found in it.2

Mediterranean Fever from Milk. — Mediterranean fever,
which is an important disease around the shores of the
Mediterranean, has been considered on a previous page, and
reference made to the brilliant researches of Bruce, Horrocks
and others, who determined that it is primarily a disease of
goats, and only incidentally transmitted to man through the
medium of milk. The obvious remedy is to avoid the use of
goats' milk whenever these animals are known to be infected,
and in the Malta garrison the disease has by this means been
eliminated. It is rather surprising that in view of the known
facts the United States Department of Agriculture should in
1905 have imported a large number of goats from Malta,
but the importation resulted in a complete demonstration of
the inf ectiousness of the milk, for quite a number of persons on
the ship, and one in America, developed the fever as a result
of drinking the milk. The goats have now all been killed.

Anthrax from Milk. — Experiments have shown that
anthrax bacilli may be absorbed through the intestines,
and they are found in the milk of diseased animals.3 In-
stances of this mode of infection in man have been reported,
but they are certainly very rare.4

1 U. S. Dept. Agric., Rep. Bu. An. Ind., 1907, XXIV, 152.

2 Harrison, U. S. Dept. Agric., Rep. Bu. An. Ind., 1902, XIX, 217.
8 Horrocks, J. Roy. Army Med. Corps, Lond., 1908, XI, 46.

4 Teacher, J. Comp. Path. & Therap., Edinb. & Lond., 1906, XIX,
225; Meyer, Deutsche med. Wchnschr., 1908, XXXIV, 108; J. Hyg.,
Cambridge, 1909, IX, 279, 315.

366 THE SOURCES AND MODES OF INFECTION

Foot-and-Mouth Disease. — Foot-and-mouth disease is
said to be transmitted to human beings by means of milk,
and according to Salmon such cases are frequently reported
in European outbreaks,1 but none occurred during the last
outbreak in Massachusetts, perhaps owing to the fact that
suitable precautions were taken. An instance of this sort of
infection occurred near Boston in 1871. 2 Kober 3 refers to
other instances of the transmission of foot-and-mouth disease
by milk.

Rabies from Milk. — According to Kober,4 Nocard proved
experimentally that rabies could be conveyed in milk, and
Burdach infected animals with the milk of a woman sick with
the disease at the Pasteur Institute. Repetto5 was able to
kill rats by feeding them with rabies virus, and Remlinger*
did the same. Dr. Austin Peters of Boston writes me that
" Several times where cows have had rabies I have had milk
taken from them and brought to the Harvard Medical School.
Rabbits and guinea pigs inoculated with it have never devel-
oped rabies. From this I should say that there is very little
danger of rabies being carried through the milk."

Other Animal Disease and Milk. — As regards actinomycosis,
botryomycosis, tetanus, cowpox and many other animal
diseases, little or nothing is known about their transmission
by milk.

Diarrhea from Milk. — The diarrhea of infants is prob-
ably not a specific disease, but the group of disturbances
which pass under this name are likely to be due to infection by
a number of organisms, and perhaps to some extent to intoxi-
cation with the products of bacterial growth outside of the

1 U. S. Dept. Agric., Rep. Bu. An. Ind., 1902, XIX, 405.

2 Marion, J. Mass. Ass. Bds. Health, Bost, 1903, XIII, 11.

3 Kober, Milk in Relation to Public Health, Senate Doc. No. 441,
57th Congress, First Session, 1902, 154.

* Kober, Idem, 157.

6 Repetto, Compt. rend. Soc. biol., Paris, 1908, LXIV, 716; Abst.
Am. J. Pub. Hyg., Bost., 1909, XIX, 426.

• Remlinger, Compt. rend. Soc. de biol. Par., 1908, LXV, 385.

INFECTION BY FOOD AND DRINK 367

body. Among the bacteria suspected of causing diarrhea
are B. coli, B. enteritidis sporogenes, B. enteritidis (Gartner),
B. dysenteries, B. paratyphosus and Streptococcus enteritidis.

The summer diarrheas are confined principally to infantile
life, the chief incidence falling on the last half of the first
year. According to Newsholme1 they are associated with a
deficient rainfall, and their dependence upon a high tempera-
ture is particularly marked. A very large proportion of the
deaths occur during the hot summer months, and the hotter
the season, as a general thing, the higher the mortality. The
relation between temperature and this disease is probably
more or less indirect, and at present is not well understood.
Another marked characteristic of the summer diarrheas is
that they are intimately associated with the mode of feeding.
Breast-fed infants are very little affected, while artificially
fed infants suffer severely. As the artificial food of infants
is chiefly cows' milk, and as these diseases appear to be of
bacterial origin, and as the growth of bacteria in milk is
enormously facilitated by hot summer weather, it is generally
assumed that diarrhea is due to the growth of infective organ-
isms in the milk. Some, however, doubt the validity of
this assumption. Liefmann2 considers that artificial feeding
causes disturbances of digestion and nutrition, and that these
lay the foundation and open the way for new and injuri-
ous factors. The latter are mostly bacterial, and the bacteria
effect an entrance into the child's body in various ways,
usually by contact infection, and not so often by milk as has
been commonly supposed. He says that the use of sterilized
milk is disappointing, which indicates that milk is not the
chief vehicle of the infection.

Epidemiological Studies. — In 1901-02 a very valuable
series of observations was undertaken in New York by
Park and Holt, under the auspices of the Rockefeller Insti-

1 Newsholme, Pub. Health, Lond., 1899-1900, XII, 139.

2 Liefmann, Ztsehr. f. Hyg. u. Infectionskrankh., Leipz., 1908-09,
LXII, 199; Abst. Pub. Health, 1909, XXII, 430.

368 THE SOURCES AND MODES OF INFECTION

tute.1 It appeared to these observers that lack of care in the
feeding and general management of the babies had more to do
with the development of diarrhea than had the character of
the milk. It is true that exceptionally dirty store milk gave
worse results than any other, but the users of this milk were
poorer and dirtier and gave less intelligent care to their
children. A rather limited number of children were fed on
fairly good milk raw, and another group on the same milk
pasteurized, other conditions being nearly the same. Those
on pasteurized milk had much less diarrhea, but the authors
state that a considerable percentage do quite as well on raw
milk. Condensed milk was found to be associated with a
large percentage of diarrheal cases. Intelligent care and feed-
ing seem to the authors more important than the character
of the milk. They also consider the infection of the milk in
the home or store as a very important factor. Breast-fed
babies are not immune to diarrhea, and Newsholme,2 in a
very careful study of the subject at Brighton, found that 6.6
per cent of the deaths from diarrhea were in breast-fed infants,
who almost certainly could not have been infected by the
milk. Newsholme pertinently asks why, if these cases were
not due to milk, should it be assumed that all the artificially
fed infants who succumbed to diarrhea met their death because
of milk contamination. From a study of the kinds of milk used
in Brighton it was found that more diarrhea was associated
with the use of condensed milk than with that of any
other. Similar results from the use of condensed milk were
found by Richards in the neighboring city of Croydon,3 and
by Sandilands in Finsbury, London,4 and their findings do
not differ materially from those of Park and Holt. In the
English cities a very good brand of condensed milk, viz.
Nestles', was the one chiefly employed, and Sandilands found

1 Med. News, N. Y., 1903, LXXXIII, 1066.

2 Newsholme, J. Hyg., Cambridge, 1906, VI, 139.
8 Rep. Med. Off. Health, Croydon, 1904, 1908.

* Sandilands, J. Hyg., Cambridge, 1906, VI, 77.

INFECTION BY FOOD AND DRINK 369

it very free from bacteria. He also states that this condensed
milk is collected and put up in Switzerland in such a manner
as to make infection during the process much less likely than
in ordinary market milk. All the writers above referred to
consider that these facts indicate that the infective material
of infantile diarrhea gets into the milk and other food to a
large extent in the home. Newsholme especially insists on
this, and he attributes not a little to direct contact. Tomp-
kins,1 from a study of the very local distribution of the disease
in Leicester, concluded that it must have its source in local
conditions. Robertson and Niven, from a study of cases in
their respective cities of Birmingham and Manchester, believe
that the infection takes place in the city chiefly, either by
direct infection or through the milk.

Explosive Outbreaks of Diarrhea. — Sometimes what
appear to be explosive milk-borne outbreaks of diarrhea
occur. Several such are referred to by Newman.2 Recently
a very interesting outbreak of this kind has been reported
by Hay.3 Various bacteria, such as B. coli, B. enteritidis
sporogenes and B. enteritidis of Gartner, have been isolated
from the milk in such cases and also from the evacuations
of the patients. It is uncertain whether such outbreaks have
any connection with ordinary summer diarrhea, though that
they have is the view of Delepine,4 and he believes that the
contamination of milk usually takes place at the farm.

Prevention of Diarrhea. — It appears, then, that we have
very little accurate knowledge as to the causation of infantile
diarrhea. It certainly has some intimate relation to the diet,
but exactly what is not known. We do know, however, that
the most efficient means we have of combating it is correct
feeding. Breast feeding is far superior to anything else, but
good success can be obtained with the careful and scientific

1 Tompkins, Brit. M. J., Lond., 1889, II, 180.
1 Newman, Infant Mortality, Lond., 1906, 170.
» Hay, Pub. Health, Lond., 1910, XXIII, 180.
4 Delepine, J. Hyg., Cambridge, 1903, III, 89.

370 THE SOURCES AND MODES OF INFECTION

use of good cows' milk. Pasteurized milk appears to be better
than dirty raw milk of high bacterial count, but there does
not seem to be any definite connection between the disease
and an increasing bacterial content. The longer milk re-
mains exposed in shops and houses, the more dangerous it
seems to be. Almost all who have had to do with modern
milk stations, either those using pasteurized milk or those
dispensing clean raw milk, agree that the larger part of the
resulting good is due to the education of the mother in
the care of her child. The chief thing, then, is to teach the
mothers how to modify, keep and feed milk. It is of impor-
tance, too, that the mothers should be enabled to get fairly
clean and fresh milk. In some cities it is doubtless necessary
to establish milk stations for this purpose, which dispense
either pasteurized milk, or preferably clean raw milk. In
Providence stations were found to be unnecessary, as there is
no part of the city in which it is not possible to get 18 to 24
hours' old milk with less than 50,000 bacteria per c.c., and a
number of dealers furnish milk below 10,000. This milk is
sold by the producer in glass bottles and at the ordinary
market price. The milk is by no means equal to certified
milk, but the observations of Park and Holt, and our expe-
rience in Providence, seem to show that milk of the grade
above indicated is ordinarily as little likely to cause infantile
diarrhea as is certified milk. Only a small amount of the milk
supply of a city is used for feeding infants, and it seems un-
necessary to insist on a high grade of expensive milk for all
consumers. At present it is more economical to bring the
existing good supplies to the knowledge of the mothers
through physicians, the health department, district nurses
and milk stations. In cities up to 200,000 or 300,000 it is
probable that enough dealers can be found to supply the
necessary demand for a good though not certified milk.

INFECTION BY FOOD AND DRINK 371

Infection by Meat.

Inspection of meat is considered by the public a matter of
the greatest importance, but concerning many forms of meat
infection we have little definite knowledge, and the danger of
others has been greatly exaggerated, not only by the public,
but also by physicians and health officials. Various animal
parasites, such as tapeworms and trichinae, are derived from
the lower animals through the use of their flesh as food, but
a consideration of these is beyond our present purpose. One
of these animal parasites, the trichina, is quite common in
pork, and a considerable number of deaths are caused by it
each year.

Food Poisonings. — What are commonly called food poi-
sonings, when resulting from eating meat, are due to two
general causes. One class is the result of the action of various
kinds of putrefactive organisms which infect the food after
slaughter. These cases we need not consider. Another class
results from infection of the flesh during the life of the animal,
and is due, so far as known, to various members of the colon
group, such as B. paratyphosus, B. enteritidis (Gartner) and
B. morbificans. In many instances this unwholesome meat is
derived from diseased animals, usually showing some sort of
enteritis or septic infection. That the infection is always
derived from diseased animals seems unlikely, for several
observers have found in healthy animals the bacteria which
are believed to be the cause.1 Savage,2 however, was un-
able to isolate B. enteritidis (Gartner) from the intestines of
23 healthy animals which he carefully studied. But from
what is known of the relations of this group of bacteria to
human beings we should expect to find them occasionally in
healthy animals, convalescents and carriers. Unfortunately
the toxins produced by these bacteria are not always de-

1 Bolduan, Food Poisoning, New York, 1909, 33.
* Savage, Rep. Med. OS. of Local Gov. Bd., 1906-07, XXXVI, 253;
1907-08, XXXVII, 425.

372 THE SOURCES AND MODES OF INFECTION
^

stroyed by heat, so that while cooking may kill the bacteria,
cooked meat, which>*ias been the seat of bacterial growth,
has been known tci cause sharp outbreaks characterized by
acute gastrointestirjal symptoms. If living bacteria of this
group are present, infection with them may result, causing a
slow after development of symptoms somewhat akin to those
of typhoid fever. The cooking of meat cannot be relied upon
wholly to prevent sickness arising from this sort of food
infection. Government inspection is suggested as the only
method by which these diseases can be prevented. If, how-
ever, bacilli are found in healthy animals, it is questionable
whether any amount of inspection would entirely eliminate
the danger. How great the danger is it is difficult to deter-
mine. A good many outbreaks have been reported in Ger-
many, aggregating thousands of cases, and reports come not
rarely from England. No data are available for the United
States. For several years I have been on the lookout for
the reports of such cases in the medical press and in the
" Index Medicus," and one year I employed a press-clipping
bureau to secure cuttings from the lay press, but I have
notes of scarcely more than a dozen outbreaks. Doubtless
others occurred, but they cannot have been very numerous.
A good many reports of instances of " ptomaine poisoning "
find their way into the newspapers, which prove on investi-
gation to have no basis in fact.

Meat and Tuberculosis. — In the public mind the fear of
contracting tuberculosis by eating meat is very considerable,
and public sentiment is sufficient to support very stringent
regulation of the sale of meat from diseased animals. Yet
it does not appear that there is a single recorded instance of
the transmission of disease in this way. And we should
expect that if such were possible it would be exceedingly rare.
Tubercle bacilli have their habitat in lungs, liver, intestines,
glands and other viscera, and not usually in the muscle or
fat. The tubercle bacillus is easily killed by heat, and very
little of this class of food is eaten without cooking. Smoked

INFECTION BY FOOD AND DRINK 373

beef and ham are occasionally eaten without cooking, but even
then considerable time is consumed in the process of corning
and smoking, and in the rare cases in which a few bacilli
are contained in the meat they are likely to have lost most
of their virulence. As for the viscera, if they are used at all
for food they are generally pretty well cooked. Cornet states
that Schottelius studied the use of meat from tuberculous
animals in Wiirzburg, and could not find a case of the disease,
although the meat was eaten in every form. It has been
said that while tuberculosis has been decreasing the consump-
tion of meat has been increasing; and this is certainly an
indication that the use of meat cannot be a factor of any
great moment in the causation of the disease. At the most
it is scarcely possible for the disease to be derived from this
source except in rare instances.

Conclusions. — The diseases which it is alleged may be
transmitted by flesh foods are those caused by animal para-
sites, of which trichinosis is the most important, diseased
conditions produced by the colon group of bacilli, and tuber-
culosis. The latter is a negligible quantity, the second group
probably causes very few deaths in this country, while trichi-
nosis is doubtless the most important disease transmitted in
this manner.

Federal Control. — The federal government has instituted
an expensive meat-inspection service, ostensibly to guard the
health of the public, but as trichinosis, the most common and
serious of the animal diseases, though it is very rare in
man, is ignored, it is suspected that the system was insti-
tuted by Congress as the result of an ill-informed though
popular demand. Filthy conditions in the slaughterhouses,
and the killing of diseased animals, though their flesh may
not be injurious to health, are shocking to the esthetic sense
and the public demands reform. Including the expense of
inspection and the value of the meats condemned, the cost
to the country is $5,000,000 or $6,000,000 per annum. It is
true that the conditions of labor in the great packinghouses

374 THE ^SOURCES AND MODES OF INFECTION

have been improved, and that cleanlier methods of handling
meat have been enforced, but it is doubtful whether any sick-
ness among consumers has been prevented. I cannot help
thinking how much ultimate good might have accrued if a
tenth of the sum spent in meat inspection had been devoted
to the systematic study of the many unsolved problems of
sanitation, such as, for instance, the danger from fomites, the
part played by air in the spread of disease, the causes of
the decline of tuberculosis and the mode of extension of the
disease, the relation of food to health, the causes of infantile
diarrhea, the relation of water supplies to the general health,
or the meaning of bad air and its effect on health.

Infection by Shellfish.

Oysters. — Since oysters and other shellfish are often
eaten raw, and often live in sewage-polluted waters, they
might be suspected, and indeed were suspected by Sir Charles
Cameron as long ago as 1880, of being the cause of typhoid
fever, and in 1893 Thorne suggested that the sporadic cases
of cholera which appeared here and there in England were
due to the eating of raw shellfish infected at the mouth of
the Humber. The first demonstration of the relation of shell-
fish to disease was by Conn.1 He showed that 23 of 100
students who went to a certain banquet developed typhoid
fever, probably as a result of eating oysters. Of those who did
not eat raw oysters none were sick, and one man who did not
go to the banquet ate oysters at the dealers' and also was sick.
The oysters had been kept about 300 feet from a drain leading
from a house where there was typhoid fever. Similar out-
breaks have been reported by Chantemesse,2 Mosny,3 Chatin,4

1 Conn, Rep. St. Bd. Health, Connect., 1895, 253; Med. Rec., N. Y.,
1894, XLVI, 743.

2 Chantemesse, Bull. Acad. de med., Par., 1896, 3 a. XXXV, 588,
724.

3 Mosny, Rev. d'hyg., 1900.

4 Chatin, Semaine med, 1897, XVII, 91.

INFECTION BY FOOD AND DRINK 375

Thresh and Wood,1 a committee which studied the subject
at Atlantic City,2 Fraser,3 Soper 4 and Morse.5

Clams, Mussels. — Clams as well as oysters have been
believed to be the cause of outbreaks,8 and where mussels
and cockles are consumed raw in considerable quanti-
ties they are equally liable to carry the infecting organ-
isms. Three outbreaks in Norwich, England, in 1908 were
traced to mussels.7 Mussels also were believed to be the
cause of an outbreak of typhoid fever in North Ormsby,8
and have been an important factor in the causation of
typhoid fever in Belfast.

Shellfish in English Cities. — Not only have shellfish been
shown to be the cause of marked outbreaks of illness, as
just shown, but they are strongly believed by many, espe-
cially by English health officials, to be an important source of
the ordinary " residual " typhoid fever occurring in cities.
Newsholme9 is especially insistent on this. A careful study
of the matter in Brighton, where he was then health officer,
showed that in 1894-96, of 189 reported cases of typhoid fever,
41 were imported, and of the 148 remaining, 51, or nearly 30
per cent, had eaten raw oysters or mussels within the incuba-
tion period of the disease. These shellfish all came from
grounds which were contaminated with sewage. From 1894
to 1902, of 643 reported cases of typhoid fever, 158 were due
to eating oysters and 80 to other shellfish, making in all about
37 per cent due to this cause. Similar observations and con-
clusions may be found in the reports of the health officers
of Birmingham, Leicester, Southend, Manchester, London,

Thresh and Wood, Lancet, Lond., 1902, II, 1567.

Phila. M. J., 1902, X, 634.

Fraser, Lancet, Lond., 1903, I, 183.

Soper, Med. News, N. Y., 1905, LXXXVI, 241.

Morse, Rep. St. Bd. Health, Mass., 1900, 836.

Plowright, Brit. M. J., Lond., 1900, II, 681.

Rep. on Sanitary Condition of Norwich, 1908, 14.

The Medical Officer, 1909, II, 431.

Newsholme, J, San, Inst,, Lond., XVII.

376 THE SOURCES AND MODES OF INFECTION

Portsmouth and other places. The typhoid death rate in
English cities is low and health officials find it difficult to
account for the origin of the cases. The pollution of shellfish
with sewage is not uncommon around the English and Irish
coasts.1 A considerable percentage of the cases of typhoid
fever are known to have eaten shellfish, often from polluted
sources, within two to four weeks of the date of attack.
Perhaps the assumption is justifiable that some of these cases
at least are due to the shellfish. Typhoid fever has for some
time been rather prevalent in Belfast, and a special commis-
sion was appointed for its investigation. This commission
believes that the chief source of the disease is mussels and
cockles, picked up by the poorer people along the sewage-
polluted flats.2 Mair3 states that in Belfast it was impossible
to make a satisfactory canvass of the number of mussel users,
either among the general population or among the typhoid
patients. He bases his conclusions as to the part played by
mussels on a careful statistical study, and shows that the
disease in Belfast has varied according to changes in the
amount of mussels consumed. He also shows that Jews and
the wealthier classes, who use no mussels, had little typhoid
fever. Nash4 states that at Southend 54 per cent of the
typhoid-fever cases confessed to the eating of shellfish, while
only 0.4 per cent of 501 sick with other diseases confessed to
such eating. Since the consumption of raw shellfish has
decreased, typhoid fever has decreased also. In Leicester,5
50 per cent of the typhoid cases ate mussels, but they were
used in only 15 per cent of a small number of non infected
houses. Johnston6 found that 25.8 per cent of 62 persons

1 Rep. Med. Off. Loc. Gov. Bd., 1894-95, XXIV; Loc. Gov. Bd. for
Ireland, Rep. on Shellfish Layings, 1904.

2 U. S. Pub. Health & Mar. Hosp. Serv., Pub. Health Rep., Wash.,
1908, XXIII, 995.

3 Proc. Roy. Soc. Med., Lond., 1909, II, Epidem. Sect., 187.

4 Nash, Pub. Health, Lond., 1903-05, XVI, 80.

5 Rep. Med. Off. Health, Leicester, 1908, 31.

6 Johnston, The Medical Officer, II, 1909, 431,

INFECTION BY FOOD AND DRINK 377

with typhoid fever had eaten shellfish, mussels and peri-
winkles within a short time of their illness, while of 827
other persons only 7.3 per cent had eaten them during the
whole summer. In the United States, also, non-epidemic or
" residual " typhoid has been attributed to the use of raw
oysters, as in New York.1 The chief reason for this seems
to be that there is known to be a considerable consumption
of sewage-infected oysters.2 Most of the beds on which
oysters are grown are free from dangerous pollution, but it
is quite common to " fatten," i.e., freshen and swell them, in
estuaries near sewer openings.

Danger Variable. — It seems reasonable to conclude that
the danger from eating sewage-infected shellfish is a real one.
Exactly how great it is, is difficult to determine. In England
anywhere from 15 to 50 per cent of the disease in cities is
attributed to eating raw mussels or oysters, but this is on the
supposition that every typhoid patient who has recently eaten
raw shellfish derived the disease from that source. In Provi-
dence raw oysters are very popular; they are consumed in
restaurants in large numbers, and form a course in a large
proportion of banquets and dinners. Many oysters are
grown in the upper part of the bay, in water grossly contami-
nated with sewage, and in the water and in the oysters
colon bacilli are found. Until within two or three years,
numbers of oysters from clean water have been "fattened"
near sewer openings, yet Providence has a typhoid death
rate less than half that of the average American city. Oys-
ters are not eaten to any extent in August, when typhoid
fever begins to increase, and they are largely consumed in
the winter and spring, when there is little of the disease.
During the years 1902-05, of 263 typhoid-fever patients who

1 Med. News, N. Y., 1904, LXXXIV, 325; 1905, LXXXV, 571.

2 Report of U. S. Commissioner of Fisheries for year ending Jan. 30,
19C4, Appendix, 189; Rep. Dept. Health, City of N. Y., 1904, I, 313;
Rep. St. Bd. Health, New Jersey, 1904, 226; Rep. Dept. Health, Balti-
more, 1907, 124.

378 THE SOURCES AND MODES OF INFECTION

replied definitely as to whether they had eaten oysters, only
26, or about 10 per cent, said that they had. Very few raw
oysters are eaten by laboring people, but at present laboring
people furnish fully their share of typhoid fever.

While the amount of typhoid fever due to the use of raw
shellfish is not very great, this danger ought to be eliminated
entirely, and state boards of health should have the authority
to forbid the sale of shellfish from polluted waters.

Crawfish and Typhoid Fever. — Dr. Bissell of Buffalo
wrote to me about an interesting local outbreak of typhoid
fever which was at first suspected to be due to milk. But
further investigation showed that it was confined chiefly to
boys, and that these boys were in the habit of catching craw-
fish from a lake grossly polluted with sewage. After par-
tially cooking the crawfish before an open fire in the field,
the boys would eat them.

Infection by Fried Fish.

In 1900, Earner 1 of London reported outbreaks of typhoid
fever in Southwark, Lambert and Kensal-town which seemed
to be confined in each case to the customers of certain fried-
fish shops. While the sickness was believed to be due to
the eating of fish, no conclusion was reached as to how the
fish became infected. It is scarcely possible that infection of
the fish before cooking should not be destroyed by the pro-
cess, and yet it seems unlikely ihat handling by carriers after
infection could cause such an outbreak.

Infection by Watercress.

In the summer of 1903 there was a very considerable out-
break of typhoid fever in Hackney, London.2 This was very
carefully investigated by Warry, and he eliminated all articles
of food and drink as sources of the infection, except water-
cress. Of the 110 cases 55.3 per cent ate watercress which

1 Earner, Special Rep. to Med. Off. Health, Lond., 1900.
8 Warry, Rep. Med. Off. Health, Lond., 1903, 35.

INFECTION BY FOOD AND DRINK 379

grew in sewage-polluted water. An inquiry showed that the
incidence of the disease on watercress eaters was three times
as great as upon those who did not eat it.

Infection by Celery.

Morse1 reports an outbreak of typhoid fever due to the
use of celery. There were 49 cases in an insane asylum.
Nearly all of these belonged to the class of pay patients, to
whom alone celery was served. Several other persons, how-
ever, had access to the celery and contracted the disease.
There were no cases except four contact cases among non-
users of celery. There had been typhoid fever in the insti-
tution some months before and the celery bed had received
the hospital sewage. The disease developed soon after the
celery came into use.

1 Morse, Rep. St. Bd, Health, Mass., 1899, 761,

CHAPTER VIII.

INFECTION BY INSECTS.

Importance of Subject. — The subject-matter of this chap-
ter is of the utmost importance in the practical work of pre-
ventive medicine, and it is of equal interest to the student of
scientific epidemiology. Our actual knowledge of the insect
carriers of disease has all been acquired during the last fif-
teen or twenty years, and marks as brilliant and successful
an epoch in the history of medicine as did the phenome-
nal development of bacteriology in the years immediately
preceding.

I hesitate very much to discuss the subject at all, as most
of the diseases considered are essentially tropical, and of
tropical diseases I have had no personal knowledge. But
even we who dwell in temperate regions are likely to meet
with isolated cases of tropical diseases, or to suffer from
occasional invasions of yellow fever, bubonic plague and
relapsing fever. In any event, it is important that both the
student of preventive medicine and the health officer keep
informed as to current progress in this line of research, and
ever bear in mind the possibility that insects may play a
part, at least, in the spread of those diseases with which he
is more familiar.

Modes of Transmission. — There are various ways in
which disease may be transmitted by insects. The most
interesting and to us novel manner in which this happens is
that the discovery of which we owe to Theobold Smith, in
which the insect, as well as the higher animal, serves as the
true host of the pathogenic organism which causes the<iisease.
Most of the diseases transmitted in this way are probably
caused by animal parasites, usually protozoa. As a rule,

380

INFECTION BY INSECTS 381

they propagate asexually only in man or the higher animals,
but develop sexual forms in insects, which thus become, zoo-
logically considered, their true hosts, while man, if it be a
human disease, is the intermediate host. The insects in such
instances are sometimes called biological carriers. In other
cases the parasites do not develop in the insect, which is then
merely a mechanical carrier, as would be a lancet or a hypo-
dermic syringe. The classification of insect-borne diseases is
far from definite, owing to the present fragmentary condition
of our knowledge. We will first consider those diseases which
appear to be biologically carried by insects.

First Proof of Transmission by Insects. — Although there
had previously been suggestions that disease might be trans-
mitted by insects, the first definite proof was the demonstra-
tion in 1893 by Smith and Kilborne l of the development of
Piroplasma bigeminum and the part played by ticks in the
transmission of Texas cattle fever. This discovery did much
to point out the lines on which experimental work should
proceed. These authors demonstrated the presence of the
Piroplasma in the blood of infected cattle and in the ticks
which fed upon them. They also showed that the ticks trans-
mit the germ through their eggs to their progeny. The
new generation of infected ticks then become attached to
cattle, and by their bites inoculate them and cause the disease.
It was demonstrated that there could be no infection of the
fields or fodder, either by the excreta of cattle or by dead
ticks or ova. The living tick is necessary for the transmis-
sion of the disease, and doubtless the Piroplasma passes
through definite phases of its development in the body of the
tick. Smith and Kilborne showed not only that ticks trans-
mit this disease, but also that it is transmitted only in this
way. They furthermore showed that apparently healthy cat-
tle might be carriers of the Piroplasma, and therefore in the
presence of ticks could cause the spread of the disease. Texas

1 Smith and Kilborne, U. S. Dept. Agric., Bu. An. Ind., Bull. 1,
1893,

382 THE SOURCES AND MODES OF INFECTION

cattle fever, though of immense economic importance in the
cattle-raising industry, is not transmissible to human beings.
The great importance of Smith and Kilborne's discovery, from
our present point of view, is the encouragement it gave to the
study of the transmission of human diseases by insect agen-
cies. Koch has demonstrated a similar connection between
ticks and a disease of cattle in German East Africa caused by
another species of Piroplasma, — P. parvum. Nuttall and
Graham-Smith l have investigated a similar disease in dogs,
also transmitted by ticks, and have described the devel-
opment of the parasite P. canis. Christophers also2 has
studied the development of this parasite. There appears
to be no doubt that piroplasmosis of horses and sheep is
transmitted by ticks.

Malaria.

Discovery of Insect Transmission. — While Nott,3 King,4
Laveran,5 Koch and others had suggested, on epidemiological
grounds, that this disease might be insect borne, it was not
until 1895 that Ross,6 stimulated by the work of Manson
on filariasis, watched the development of the malarial parasite
in mosquitoes which had been allowed to bite persons sick
with the disease. As Manson had already shown that in
filariasis it is only a particular species of mosquito which
can serve as host for the filaria, Ross suspected that the
same might be true of human malaria, and he finally deter-
mined that it was only in individuals of the genus Anopheles
that the malarial parasite can develop. Ross 7 now turned

1 Nuttall and Graham-Smith, J. Hyg., Cambridge, 1904, 1905, 1906,
1907.

2 Christophers, Brit. M. J., Lond., 1907, II, 1333.
Nott, N. Orl. M. & S. J., 1847-48, IV, 563.
King, Tr. Philos. Soc., Wash., 1883.
Laveran, Le Paludisme, Paris, 1891, 147.

Ross, Abst. by Manson, Lancet, Lond., 1896, I, 831.
Ross, Rep. on the Cultivation of Proteosoma Labbe in Grey Mos-
quitoes, Indian M, Gaz., Calcutta, 1898, XXXIII, 133, 401, 448.

INFECTION BY INSECTS 383

to the study of the malaria of birds due to a Proteosoma, and
he demonstrated that the disease could be transmitted from
bird to bird by the bites of mosquitoes. The parasites of the
disease are taken up with the blood by the insect in the act
of biting, and after undergoing sexual multiplication, spread
through the insect, and are found in the salivary gland,
whence they are injected into the next bird bitten. Mean-
while Grassi in Italy had come to the conclusion, from a study
of the distribution of different species of mosquitoes, that
Anopheles was one of the forms likely to transmit the disease.
Bignami had previously been unsuccessful in transmitting
malaria by the bites of mosquitoes, chiefly because he experi-
mented largely with Culex instead of Anopheles. Finally
Grassi, Bignami and Bastianelli in 1899 caused malaria in
human beings by allowing anopheles mosquitoes to bite them
some days after they had bitten other individuals sick with
malaria. To avoid the criticism that these experiments were
carried on in Italy, an intensely malarial country, where
natural infection »could not be absolutely excluded, Patrick
Manson 1 had a number of infected mosquitoes sent from
Italy to England in 1900, and caused two men, Dr. Thurburn
Manson and Mr. Warren, who had never been in a malarial
region, to be bitten. Both men developed malaria in exactly
eighteen days. In the same year Sambon and Low lived for
three months in the most malarial section of the Roman
Campagna, protecting themselves absolutely against the bites
of mosquitoes, but taking no other precautions. They did
not contract malaria, but of fifteen or sixteen police sent from
Rome, who spent one night unprotected in the same place,
every one developed the disease.2

Mosquito Sole Carrier of Malaria. — The observations
and experiments above noted prove conclusively that malaria
is transmitted by the bites of infected anopheles mosquitoes,
but do not indicate whether or not it may be spread in other

1 Patrick Manson, Brit. M. J., Lond., 1900, II, 949.
* Sambon and Low, Brit, M. J., Lond., 1900, II, 167&

384 THE SOURCES AND MODES OF INFECTION

ways. It was Manon's theory at first that the malarial para-
site is transmitted from adult to larvae by means of the water
in which the eggs are laid, and that this water is also the
means of infecting human beings. He suggested that germs
might be inhaled in dust from dried-up and infected pools.
As regards transmission by water the experiments of Celli l
and others have already been referred to, which indicate that
such a mode of transference is highly improbable, and there
is certainly no epidemiological evidence in its favor. That
the plasmodium of malaria grows outside of the animal body
is also improbable, for it has never been possible as yet to
cultivate it in any way.2 It has never been found in the water
of malarial regions.3 Even if there are sporelike forms resist-
ant to drying, there is strong evidence that such are not air-
borne, for mosquito netting would in no way interfere with the
free movement of particles small enough to be wafted by the
air; but abundant evidence has shown that efficient screening
against mosquitoes affords complete protection against this
disease, even in the most intensely malarial regions. More-
over the success attending methods of prevention based on
the mosquito theory point indubitably to the truth of the
theory. Experimental, clinical and epidemiological evidence
combine to demonstrate that malaria is a disease transmitted
solely by the bites of infected mosquitoes. There are several
forms of malarial fever, such as tertian, quartan, aestivo-au-
tumnal, apparently caused by different species of Plasmo-
dium, but there is no need here of considering these different
forms and their relation one to another. Suffice it to say that
the mosquito appears to be the true host of this parasite, and
in its body it undergoes sexual reproduction. From eight to
ten days after the mosquito is infected, its entire body, in-
cluding its salivary gland, becomes infested with the so-called
sporozoite forms, and when these are injected into human be-

1 See chapter on Infection by Water.

2 Craig, The Malarial Fevers, New York, 1909, 93.

3 Craig, idem, 83.

INFECTION BY INSECTS 385

ings they continue to undergo a vegetative or asexual prolifera-
tion, the successive phases of which give rise to the symptoms
of the disease. It is believed that no vertebrate other than
man serves as the host of the Plasmodium, and until recently
it was thought that no mosquitoes except those belonging to
the genus Anopheles could transmit the disease; but recent
observations in the Philippines l indicate that a mosquito
breeding in salt marshes, Myzomyia ludlowii, may become the
host of the Plasmodium. But it is certain that almost always
it is some species of Anopheles which is the carrier. Not all
Anopheles, however, can serve as the host, for one of our
common species in the northern United States, A. puncti-
pennis, does not.2

Habits of Mosquitoes. — Mosquitoes develop only in water.
The larval and pupal stages may occupy from eight days to a
number of weeks, or the larvae may even live all winter
under the ice in northern regions, or retain their vitality for
some time in the dried mud of the tropics. The mature
insects usually live only a few weeks, but in northern regions
hibernate during cold weather. The habits of life of the dif-
ferent species of Anopheles vary considerably, and should be
carefully studied when measures for extermination are under-
taken. Only the adult female bites, so she alone is the carrier
of the disease. Anopheles usually bite at night. Gorgas3 states
that they do not usually fly above 100 yards, and that clearing
and draining the area within this distance of houses affords
great protection. Nevertheless it sometimes happens that
under favorable conditions the insects may be carried a con-
siderable distance by the wind. Craig4 states that they may
be borne even two and one-half miles.

Must be Many Mosquitoes. — Mosquitoes are not infected
with the malarial parasite unless they bite infected human

1 Banks, Philippine J. Sc. [B. Med.J, 1908, III, 335.

1 Hirschberg, Johns Hopkins Hosp. Bull., Bait., 1904, XV, 53.

3 Gorgas, J. Am. M. Ass., Chicago, 1909, LII, 1967.

* Craig, The Malarial Fevers, New York, 1909, 69.

386 THE SOURCES AND MODES OF INFECTION

beings. The number of infected mosquitoes, then, varies with
the number of infected persons to whom they have access.
Craig l refers to the findings of himself and others as varying
from 35 per cent to 0.6 per cent. For the spread of malaria it
is necessary that there should be Anopheles and human beings
infected with the parasites. But if the mosquitoes are not
numerous and the cases of infection are not numerous, the
chances of a mosquito becoming infected and then biting
another victim after a proper interval are not great, so that it
may happen that there may be mosquitoes and human infec-
tion without extension of the disease. Thus it is not known
that cases of malaria have ever developed on the island of
Rhode Island, but I have found a few A nopheles breeding there,
and with the great numbers of summer visitors from all over
the world, and many soldiers, and sailors of the navy, it is
probable that the malarial parasite is frequently present on
the island. These epidemiological facts led me for some time
to suspect that the supposed mode of transmission of this
disease above outlined did not contain the whole truth; but
the directness and strength of the experimental proof, and,
above all, the success of mosquito control in checking the
disease render it certain that practically the only way in
which malarial disease is caused is by the bites of mosquitoes.
There are numberless interesting details connected with the
life history and habits of the mosquito which are of great
importance but which cannot be considered here. The way
in which these habits explain the peculiarities in the develop-
ment of malaria is well considered by Nuttall.2

Ways of Prevention. — There are several ways in which
the spread of malarial disease may be checked.

First. Quinia. — The administration of the various salts of
quinia prevents the development of the malarial parasite in
the blood, and this drug is an extremely valuable remedy in
the prevention and cure of malaria. Antitoxin and vaccines

1 Craig, loc. cit., 74.

• Nuttall, Johns Hopkins Hosp, Rep,, Bait,, 1899, VIII, 78,

INFECTION BY INSECTS 387

are of great value in the prevention and cure of certain of the
infectious diseases, as diphtheria, typhoid fever and cholera,
but they are of value chiefly to the individual. The difficulties
attendant upon their administration are so great that they
cannot be used except in rare instances to protect large bodies
of people. The antitoxins, too, probably do not prevent the
growth of the germs which are the cause of the disease, so
that in a way their use, as for instance that of diphtheria
antitoxin, may actually in some instances favor the spread of
the disease, by preventing the development of symptoms
when the bacilli are present, thus making a "carrier" instead
of a "case," and carriers are always difficult to control. It is
otherwise with quinia. This drug actually kills the Plas-
modium or drives it out of the circulation and out of the
way of the mosquito, and so not only protects the individual
but prevents him from becoming a focus of infection. Quinia
is not only universally used as a specific curative agent, but it
is also employed in large quantities as a prophylactic, and its
employment in this manner is considered a valuable means of
fighting the disease, reducing as it does the number of foci of
infection. Indeed Koch and some others claim that this is
the only way in which malaria can be eradicated, as mosquito
destruction is in many places impossible. All workers in this
field employ quinia for this purpose, though there is con-
siderable discussion as to its mode of administration, which
varies from 15 grains every eight days, as used by Koch, to
3 grains daily, as advised by Gorgas. This subject is fully
discussed by Craig,1 who states that in the vast majority of
instances 2J grains (0.15 grm.) daily is sufficient to prevent
malarial infection. It is difficult to estimate the value of a
particular prophylactic measure, for one measure is rarely
used alone, but there is little doubt of the great efficiency of
quinia in the fight against malaria. Craig states that at
Camp Stotsenburg, in the Philippines, measures directed
against mosquitoes had little result until supplemented by
1 Craig, The Malarial Fevers, New York, 1909, 343.

388 THE SOURCES AND MODES OF INFECTION

the daily administration of quinia. Koch and his followers
also report excellent results from the use of quinia alone.
Celli,1 while urging every method directed against the mos-
quito, considers that in Italy experience has shown that the
daily administration of quinia has been the most effective
measure in reducing the number of deaths from malaria,
which have fallen steadily from 13,861 in 1901 to 4871 in 1906,
during which year 20,723 kilos of quinia were sold.

Second. Isolation of Persons. — The attempt may be
made to isolate the infected persons so that the mosquitoes
may not pass from them to the healthy. This may be accom-
plished in various ways. In many tropical regions where
the native population, or at least the native children, are
always infected, the dwellings of the natives and the sus-
ceptible Europeans are separated a sufficient distance to
escape the ordinary excursions of the mosquitoes.

Sometimes the patients, if few in number, are screened from
mosquitoes in their own homes or are removed to screened
hospitals. This is very commonly done with the imported
cases which come to Havana,2 and in Panama many malarial
patients are treated in the screened hospitals. Others are
cared for in their own well-screened houses.

To assist in this measure it is desirable to kill all mosquitoes
in the house which may have bitten the patient before the
disease was recognized or reported to the authorities. Gor-
gas considers sulphur dioxid the best for this, but as this
cannot be used in an occupied room, pyrethrum may be
burned, which so stupefies the insects that they may be easily
swept up from the floor. Celli has given much study to this
class of culicides and recommends one or two tablespoonfuls
of a mixture of the unexpanded flowers of chrysanthemum
and valerian root.

Third. Screening. — Effort should be made in all malarious
regions 'to protect susceptible persons from the bites of mos-

1 Quoted by Craig, The Malarial Fevers, New York, 1909, 345.

2 Guiteras, J, Am. M, Ass,, Chicago, 1909, LIII, 1165.

INFECTION BY INSECTS 389

quitoes. The screening of the whole house is the most impor-
tant step. Copper netting is the cheapest in the end, and it
should, according to Craig, have 11 meshes to 2 cm. Gorgas
says that it is necessary that houses should be screened by
persons experienced in mosquito work, or openings are sure
to be left. The experiments of Sambon and Low proved the
efficacy of this method of prophylaxis, and Craig states that
the screening of the houses of railway employees on certain
Italian lines has been as effectual a method as the administra-
tion of quinia. Screening is carried on most extensively in
the Canal Zone. Craig says that where screening is impos-
sible the use of mosquito canopies over the beds at night is
a very useful measure, and head nets and gloves for men
obliged to be out at night are of much value. By the use of
these methods the percentage of malaria among troops in
Sardinia was reduced from 70 to 20, and in Formosa no mala-
ria developed among 115 soldiers protected by netting, while
of 717 not so protected 251 were attacked.

The removal of trees, vines and shrubs from the vicinity
of dwellings is considered by Gorgas an important measure,
as these serve as hiding places for the insects by day.1

A certain amount of protection may be secured by the
application to the skin of substances annoying to the mos-
quito, or of ointment which mechanically interferes with
biting. Oil of citronella or eucalyptus, and petroleum or
cocoanut oil or vaseline, are used, but such applications
are not by most writers considered of much value.

Fourth. Mosquito Prevention. — Theoretically the most
effectual way to eradicate malaria is to prevent the growth of
mosquitoes. While their complete destruction is not usually
possible, its attempt must always take first rank in preventive
measures. The fact that mosquitoes are a great nuisance,
as well as carriers of disease, makes still more desirable their
extermination wherever possible. Large sums, in certain
parts of the United States, are being expended in fighting the
\Gorgas, J. Am. M. Ass., Chicago, 1909, LII, 1967.

390 THE SOURCES AND MODES OF INFECTION

salt-marsh mosquito, which is harmless so far as carrying
disease is concerned.

Fifth. Draining. — When mosquitoes breed in pools and
swamps, as they so often do, draining and filling must be
done. If water cannot be entirely gotten rid of, the banks of
streams, ponds and ditches must be straightened and deepened,
and freed from vegetation, to remove hiding places for the
larvae and permit their destruction. Gorgas insists upon the
importance of having this work done by men especially trained
for it, and supervised by experts. Ordinary contractors and
laborers are sure to overlook breeding places and to neglect im-
portant details of the work. Many species of Anopheles breed
in cisterns or accidental receptacles of water near houses.
Hence all such should either be removed or carefully screened.

Sixth. Oiling. — The growth of larvae in water can be pre-
vented by the application of crude petroleum or kerosene to
the surface at the rate of an ounce to each 15 square feet of sur-
face. Various poisonous substances may be added to the water,
as hydrochloric acid, corrosive sublimate, formalin, chloride
of lime and various aniline dyes. Some of the last named
are strongly recommended by Celli,as they are harmless to the
higher animals. The use of quicklime causes a glaze on the
surface which is a protection, and sulphate of copper kills the
algae on which the mosquitoes live. Most fish are voracious
feeders on the larvae, and if pools and streams are supplied
with fish, and kept free from weeds, mosquitoes cannot thrive.
It is said that the introduction of a certain small fish into
the streams of Barbados caused the extermination of mos-
quitoes.1 According to Craig, the application of cultures of
Aspergillus niger and A. glaucus destroys the larvae.

Success of Measures. — There is probably scarcely a
locality in which the application of a single prophylactic
measure will be found effectual in " stamping out " malaria,
and sometimes all methods combined fail to bring success.
Thus at Mian Mir in India, according to Craig, the most
» Ballou, Nature, Lond., 1909, LXXX, 16.

INFECTION BY INSECTS 391

careful draining and oiling were followed by very discourag-
ing results. On the other hand, preventive measures have in
many localities given wonderfully good results. The success
of the work in Italy, as reported by Celli, has already been
referred to. In Havana the deaths from malaria previous
to the occupation by the United States averaged 350 per
annum, rising to 1907 in 1898. The measures carried out by
Gorgas to destroy the yellow-fever mosquitoes also practi-
cally eliminated the Anopheles, and the deaths now average
only about 40, and three-fourths of these are, according to
Kean,1 due to infection contracted elsewhere. At Ismailia 2
the cases were reduced from 2284 in 1900 to 37 in 1905,
chiefly by destroying the breeding places of mosquitoes. At
Klang and Port Swettenham in the Federated Malay States
99 per cent of the mosquitoes were eliminated, with a corre-
sponding decrease in malaria. Ross also refers to the success-
ful reduction of malaria at Port Said, Durban, Hong Kong,
Khartoum, and in Candia, St. Lucia, Greece and Algeria.3

Success in Panama. — To my mind the most wonderful
success of all has been attained by Gorgas in the Panama
Canal Zone. Here was a stretch of territory 45 miles long,
intensely malarious, nowoccupied by a force of 44,000 persons,
mostly foreigners. The rainfall is heavy, and the work of
canal building alters in every direction the configuration of
the land and the natural drainage. Malaria caused enor-
mous losses during the French occupation and was second
only to yellow fever in interrupting the work. Mosquitoes
have been entirely eliminated from Colon, a town of 15,000
inhabitants, and the death rate from malaria among employ-
ees was only 1.34 per thousand in 1908, having been reduced
from 5.57 in 1905. Considering the tremendous obstacles to
be overcome, this success is certainly astonishing.4

1 Kean, J. Am. M. Ass., 1909, LIII, 1166. *

2 Ross, Lancet, Lond., 1907, II, 879.

* Ross, Nature, Lond., 1909, LXXX, 415.

« Gorgas, J. Am. M. Ass., Chicago 1909, LII, 1967.

392 THE SOURCES AND MODES OF INFECTION

Yellow Fever.

History of Discovery. — Nott appears to have been the
first (1848) to attribute to mosquitoes a r61e in the causation
of yellow fever, but this was merely a suggestion which re-
ceived little attention. According to Cruz,1 Beauperthuy
published on the 23d of May, 1854, in the Official Gazette of
Cumana, Venezuela, an article which expounded with great
clearness a mosquito theory of this disease. But to Finlay
of Havana is due the credit for the first experimental work, a
report of which was presented to the Royal Academy of Medi-
cal, Physical and Natural Sciences in Havana in 1882. Finlay
continued to experiment and write until 1900, when the
American Commission, consisting of Reed, Carroll, Agra-
monte and Lazear, undertook their investigations at Havana.2
Finlay had come to the conclusion that Culex fasciatus, now
Stegomyia calopus, was the species of mosquito which was
likely to be involved in the transmission of the disease. After
the occupation of Cuba by the United States in 1898, the
war department undertook with great energy the extirpation
of yellow fever from Havana. At that time the disease was
believed to be essentially a filth disease, and the energy of
the government was directed towards making Havana clean,
and soon its " sanitary condition" vied with that of the very
best cities in the United States. At the same time the most
stringent measures of isolation, as isolation was at that time
understood, were applied. But yellow fever refused to be
stamped out, and in 1900 caused 1244 cases and 310 deaths,
many of them among the " best people " in the cleanest parts
of the city.3 It was becoming evident that the old theories
were not satisfactory, and the war department appointed the

1 Cruz, U. S. Pub. Health & Mar. Hosp. Serv., Pub. Health Rep.,
Wash., 1909, XXIV, 1741.

8 Lee, Am. Pub. Health Ass. Rep., 1905, XXX, 8.

* Series 4, Yellow Fever Pub., San. Dept., Havana, 1902, 10.

INFECTION BY INSECTS 393

commission above referred to, which proceeded to Havana
and in its earliest work put Finlay's mosquito theory to the
test. They were not a little influenced to this step by the
observations of Carter.1 He had determined, from a study
;of the disease in isolated farmhouses, that from nine to six-
teen days usually elapse, after the introduction of the disease,
before the house can infect second cases. It was thought
that this interval might depend upon a period of incubation
in the mosquito. Owing to the hearty cooperation of General
Wood, the governor, himself a medical man, every facility
was accorded the commission, and inoculation experiments
were made on volunteer human subjects, among whom was
Lazear, a member of the commission. Of 11 persons bitten
by mosquitoes which had some days before been allowed to
feed on a yellow-fever patient, 2 developed the disease, one
of whom was Lazear, who died as the result of the inoculation.2
Carroll too was bitten and developed the disease. As critics
suggested that natural infection could not be excluded, the
experiments were repeated in November and December,
1900, under more convincing conditions, and 6 of 7 persons
bitten developed typical yellow fever, and the transmission
of the disease in this manner was demonstrated beyond ques-
tion.3 Later Guiteras 4 succeeded in inoculating 8 more per-
sons, which made a total of 24 persons infected by the bites
of mosquitoes. Of these 3 died, and the post-mortem ex-
amination demonstrated the lesions of yellow fever. The
commission had by the direct transfer of blood shown that
the infective agent exists in that fluid, even when filtered.5
All the experimenters demonstrated that the mosquito could
not transmit the disease until six to eighteen days after

VI

New Orleans M. & S. J., 1900, LII, 617.

Am. Pub. Health Ass. Rep., 1900, XXVI, 37; Phila. M. J., 1900,
790.

Series 3, Yellow Fever Pub., San. Dept., Havana, 1902.
Series 6, Yellow Fever Pub., San. Dept., Havana, 1902, 26.
Am. Med., Phila., 1902, III, 301.

394 THE SOURCES AND MODES OF INFECTION

biting a yellow-fever patient, thus explaining the wonder-
fully accurate clinical observations of Carter. Although it
has been shown that filtered blood contains the infective
principle, thus far it has not been demonstrated by micro-
scopical or cultural methods. It seems in the highest degree
probable, from this very remarkable work, that the exciting
agent of yellow fever closely resembles that of malaria, in
that it is found in the blood, is taken up by a particular
species of mosquito, develops in the mosquito during a period
of some days, and is then transmitted to new subjects only
by the subsequent bites of the insect.

Transmission by Fomites Disproved. — Belief in the infec-
tiousness of the vomitus and excreta and in the important
part played by fomites in the extension of the disease was
so firmly fixed that it seemed necessary to test this theory
by actual experiment. The commission for this purpose
exposed a number of non-immune persons to the closest
possible contact, during a period of two weeks, to bedding,
clothing and other articles all grossly contaminated with
supposedly infective material, but none of them contracted
the disease. Later Guiteras, while carrying on his mosquito in-
oculations, incidentally exposed many non-immunes to fabrics
that had been in close touch with yellow-fever cases, but with
like negative results; and again Gorgas, at Las Animas Hos-
pital in Havana, put the fomites theory to test in the most
thorough manner by the exposure of non-immunes to close
and continued contact.1

Even before these demonstrations, some acute observers,
especially Carter,2 had concluded from epidemiological studies
that fomites played little part in the dissemination of this
disease. Carter showed among other things that countless
pieces of baggage, many of them certainly from infected
houses in Cuba and Vera Cruz, had passed to northern ports
without causing yellow fever in a single instance.

1 Series 3, Yellow Fever Pub., San. Dept., Havana, 1902, 22.
1 Carter, Med. News, N, Y., 1904, LXXXV, 878.

INFECTION BY INSECTS 395

Transportation of Mosquitoes. — Theoretically there is no
reason why infected mosquitoes might not occasionally be
carried in baggage, but admission of baggage without dis-
infection, without any known instance of the development of
yellow fever, shows that such a mode of transmission must
be extremely rare. Even the carriage of infected mosquitoes
in vessels must be unusual, though instances are recorded by
Carter l and others. Grubbs 2 found Stegomyia calopus on
three of sixty-five vessels entering the Gulf Quarantine Sta-
tion. Souchon,3 at New Orleans, found that 2.5 per cent of
the mosquitoes caught on incoming steamers from Havana
or the West Indies were Stegomyia calopus. Probably none
of these were infected.

The rapid and brilliant demonstration of the true mode of
extension of yellow fever was immediately followed by an
equally rapid and brilliant application of the new knowledge.
Two months after the termination of the commission's ex-
periments, Gorgas was as energetically applying the new
methods of yellow-fever control in Havana as he had pre-
viously devoted himself to cleansing the city, and within
eight months the Stegomyia calopus had been nearly exter-
minated, and Havana was free from the disease for the first
time in 150 years.

Habits of Yellow-fever Mosquito. — Different species of
mosquitoes have different habits of life, and Stegomyia calopus
is much more of a domestic mosquito than are the Anopheles.
It prefers to breed in clean rain water, but will grow in any
water that is not too muddy. It may even be found in cess-
pools and in gutters, but its favorite home is the rain-water
cistern, barrel, jar or other container which is usually pro-
vided for every house. It seems to prefer dark, covered recep-
tacles. The adult mosquito is rather feeble and rarely flies far.

1 Carter, Med. Rec., N. Y., 1902, LXI, 441.

2 Grubbs, Yellow Fever Institute, U. S. Pub. Health & Mar. Hosp.
Serv., Bull. No. 11.

1 Souchon, J, Am, M. Ass,, Chicago, 1903, XL, 1647.

396 THE SOURCES AND MODES OF INFECTION

Preventive Measures. — The defensive measures against
yellow fever must of course be somewhat like those directed
against malaria, but differ in some important particulars.
The most important of these is that there is no drug which
will prevent and cure this disease as quinia does malaria.
Another point of difference is that yellow fever is not nearly
so widespread a disease as is malaria, and that it is an acute
disease in which the patient remains infectious for only a
short time. Among the measures directed against yellow
fever are:

First. Quarantine. — While inland quarantine has never
been successfully administered, maritime quarantine for
yellow fever is certainly of value. The short period of
incubation of the disease, the absence of chronic cases
and the fact that it prevails in endemic form in only a
few places which are in communication with the rest of
the world chiefly by water, render it possible to intercept
at the port of debarkation a goodly number of incom-
ing cases. Quarantine, while often failing, has many times
proved its value in the United States; and in Havana, after
the disease was exterminated, it would certainly have many
times been introduced from Vera Cruz and Colon, and indeed
probably from New Orleans, if the careful inspection of
incoming passengers had not been maintained by the depart-
ment of health. The efficiency of quarantine is greatly
increased if an agent is stationed at the port of embarkation,
as is done at many ports by the United States Public Health
Service.

Second. Screening of Houses and Cases. — Houses and
their healthy inmates must be screened to protect them from
infected mosquitoes, and cases must be screened to prevent
mosquitoes from becoming infected. Extraordinary effort
must be made to get control of every case. The harsh
treatment formerly accorded these cases, the injury to
property caused by disinfection and the loss to business
by the old methods of quarantine put a premium on con-

INFECTION BY INSECTS 397

cealment; but there is much less tendency to hide cases now
than formerly. When reported, the case is protected from
mosquitoes either by efficient screening or, as is now usually
done in Havana and the Canal Zone, by removing the patient
to a well-screened hospital.

Third. Destruction of Mosquitoes. — Infected mosquitoes
in the house occupied by the patient and in neighboring
houses must be destroyed. Sulphur dioxid is the best culi-
cide for this purpose, but pyrethrum powder may be burned
while the room is occupied and the stupefied mosquitoes
swept up, as in malaria.

Fourth. Prevention of their Growth. — The most impor-
tant means of preventing yellow fever is to reduce the number
of stegomyia mosquitoes to a minimum by destroying their
breeding places in the same manner in which the breeding
places of malarial mosquitoes are destroyed, by draining, fill-
ing, ditching and the use of petroleum and culicides in the
water. As the stegomyia so often breeds in domestic recep-
tacles which in the absence of a municipal water supply and
sewers cannot be dispensed with, the efficient screening of
these receptacles is an important part of sanitary work in
the American tropics, and constant inspection is needed to
see that the screening is maintained.

Success of Measures. — The phenomenal success of these
measures in Havana has already been referred to, but the
eradication of the disease in the Canal Zone is a still more
remarkable achievement. Havana is a compactly built, well-
drained city, and was under military rule. The Canal Zone
is about forty-five miles long, with swamps and streams, many
straggling villages and camps, a heavy rainfall, continuous high
temperature, a constantly changing non-immune population,
stupendous engineering works altering the configuration of
the land, and a persistent yellow-fever infection for centuries.
This disease had been the chief obstacle in the construction
of the Panama railroad, and it was the chief obstacle to the
French in their canal work. Their loss by death from this

398 THE SOURCES AND MODES OF INFECTION

disease amounted probably to from 12 to 15 per cent per
annum.1 Sanitary work began on the isthmus in February,
1905, at which time yellow-fever cases were being constantly
reported. The cases soon began to decrease, and within a
year the disease was practically exterminated.2 In Rio Ja-
neiro the campaign against yellow fever began in April, 1903.
The undertaking was expensive, as it always must be if it is
to be successful. It required the expenditure of $1,650,000 a
year and the employment of thirteen hundred men. Deaths
from yellow fever, which had averaged twelve hundred a
year for nearly forty years, decreased rapidly, as is shown
by the following:

Tear. Deaths.

1903 584

1904 48

1905 289

1906 42

1907 39

1908 4

1909 0

In Vera Cruz, which was another stronghold of yellow
fever, the disease has been almost eradicated by the employ-
ment of anti-mosquito measures.3

Filariasis.

First Parasite Studied in Insects. — While the infection of
the blood of man with Filaria bancrofti is in most cases not
accompanied by noticeable symptoms, it is of interest in this
connection as being the first infection in which the parasite
was proved to be taken up into and undergo metamorphosis
in the body of an insect. This tiny worm is found swimming

1 Gorgas, J. Am. M. Ass., Chicago, 1909, LIII, 597.

2 Cruz, Pub. Health Rep., U. S. Pub. Health & Mar. Hosp. Serv.,
Wash., 1909, 1742.

3 Liceaga, Am, Pub, Health Ass. Rep., 1905, XXXI, 284.

INFECTION BY INSECTS 399

in vast numbers in the blood of those who are infected, but
is found only at night. During the day it disappears from
the peripheral circulation and is found only in the heart and
lungs. When these larval forms are sucked up by the mos-
quito they increase rapidly in size and become more highly
organized, and working their way to the proboscis of the
insect, infect the next person bitten. The worms, now pre-
senting sexual forms, reach the lymph ducts, where the eggs
are laid and whence the young larvae reach the blood
current. It is now generally believed that injury to the
worms in the lymph vessels causes obstruction of the latter
and the development of tropical elephantiasis. Culex fatigans
is probably the insect chiefly concerned in the spread of the
infection. The disease has not been actually transmitted to
man by the bites of infected mosquitoes, but the develop-
ment of the worm in the mosquito was observed by Manson
in 1878 l and by Lewis in 1879.2 A similar infection in dogs
has been actually transmitted by means of mosquitoes.3

Sleeping Sickness.

Discovery of Trypanosome. — The trypanosome (T. gam-
biensis), which is the cause of sleeping sickness, was found
in the blood of patients by Button and Todd in 1901. This
disease is very widely distributed in Africa, and it is esti-
mated to have caused over 100,000 deaths in Uganda during
the six years from 1901-1907. This trypanosome is patho-
genic for many of the lower animals, and in monkeys causes
symptoms resembling those produced in man. Bruce and
Nabarro4 showed that the disease may be transmitted by
insects, for in 1903 they succeeded in inoculating a monkey
by allowing it to be bitten by tsetse flies (Glossina palpalis)

1 Manson, Med. Times & Gaz., Lond., 1878, II, 731.

2 Lewis, Quart. J. Micr. Sc., Lond., 1879, XIX, 245.

* Braun, The Animal Parasites of Man, 3d ed., New York, 1908, 285.
4 Bruce and Nabarro, Rep. of Sleeping Sickness Commission, Roy.
Soc., 1903, No. 1.

400 THE SOURCES AND MODES OF INFECTION

which had twenty-four to forty-eight hours previously bitten
negroes suffering with the disease. While this much has
been established with certainty, there are important points in
the causation of the disease which are still under discussion.

While Glossina palpalis is known to transmit sleeping sick-
ness, and the distribution of this fly in a large degree corre-
sponds with the distribution of the disease, it is not certain
whether it may not be carried by other species of tsetse flies,
and perhaps by domestic insects and by mosquitoes,1 and
Koch 2 thinks it possible that it may be transmitted by the
sexual act. Comparatively little is known about the breeding
habits of the fly.

Does it infect Lower Animals? — There has also been much
discussion as to whether the specific trypanosome of this
disease naturally infects the lower animals. If they are gener-
ally affected, they become a vast reservoir from which the
trypanosomes may be transferred to human beings by the
fly, and the difficulties in the way of controlling the disease
become enormous. Even the crocodile has been suggested
by Koch as an important " reservoir." But Hodges 3 says
that actual observation shows that no animal except the
native dog, and that in only a few instances, has been found
naturally infected. A cause of the uncertainty is a wide-
spread distribution of several kinds of trypanosomes not easily
distinguished from T. gambiensis.

Is it carried Mechanically? — Another question of impor-
tance, especially from a scientific standpoint, is whether the
trypanosome passes through a cycle of development in the
fly. It was at first supposed that this was the case, but later
observations, especially those of Minchin,4 seemed to indicate
that this was not so. According to his observations flies are

1 Bull. Soc. path, exot., Par., 1908, I.

« Koch, Deutsche med. Wchnschr., 1907, XXXIII, 1889.

3 Cited in Supl. to Third Rep., Wellcome Research Lab., Khartoum,
1908, III, 172.

4 Minchin, Quart. J. Micr. Sc., Lond., 1908, n. s., LII, 159.

INFECTION BY INSECTS 401

infective at once after biting a subject, and the power to
infect does not last over forty-eight hours. More recently,
however, Bruce l has shown that the parasites do undergo
developmental changes in the fly, and that flies do not infect
until from 14 to 20 days have elapsed after feeding on a case.
Kleine 2 also has demonstrated the same for other forms of
trypanosomes. It is possible that the trypanosomes may be
occasionally carried mechanically on the proboscis of the
tsetse fly, in which case infection will take place at once,
and also that the parasites pass through certain metamor-
phoses in the fly, thus permitting the flies to remain infective
for some time.

Prevention of Sleeping Sickness. — The measures sug-
gested for limiting sleeping sickness are:

First. Quarantine, or the prevention of the entrance of
infected human beings into regions occupied by the fly.
While this may be of value at times, it scarcely seems possible
to an observer at this distance that it can be often employed
effectually in Africa.

Second. The sick may be segregated and kept out of
reach of the flies until the trypanosomes have disappeared
from the blood.

Third. The removal of the entire population from the
area occupied by the fly. The tsetse flies are found only from
thirty to one hundred yards from water, so that if all settle-
ments are removed this distance from rivers and lakes much
will be done to prevent the spread of the disease.

Fourth. As trees and shrubs afford a hiding place for the
flies, the thorough clearing of the land around villages and
near landing places and shore market places has been found
useful, just as similar clearing has been found to be an impor-
tant means of controlling malaria in the Canal Zone at
Panama.

1 Bruce, Bull, of Sleeping Sickness Bu., 1909, Nos. 6 & 7.
* Kleine, Deutsche med. Wchnschr., 1909, XXXV, 924.

402 THE SOURCES AND MODES OF INFECTION

Fifth. The administration of atoxyl (anilin meta-arse-
nate) is believed by many to shorten the life of the parasite
in the body, and if so, it becomes a valuable prophylactic
measure, similar to the use of quinia in malarial affections.

Kala-azar.

Protozoan Discovered by Leishman. — Kala-azar, or dum-
dum fever, occurs in various tropical countries, particularly
in India. It runs a chronic course and is characterized by
marked enlargement of the spleen. In 1900 Leishman dis-
covered in splenic pulp, from a case of the disease, certain
bodies the importance of which he did not appreciate, but
later, in 1903, he concluded that they might be protozoa.
The discovery was confirmed the same year by Donovan.
These bodies are now recognized as protozoa, and are known
as Leishmania donovani, and are probably the cause of the
disease. Similar bodies were found by Wright of Boston in
oriental sore, and by Nicolle and Cassuto in an infantile
splenic disease in Tunis. Rogers, Patton and others have
succeeded in growing the parasite in culture medium outside
of the body, and both have also obtained evidence to show
that the disease may be transmitted by the bedbug. Both
Patton and Rogers have shown that a well-defined cycle of
development takes place in the body of that insect, which
renders it very probable that it is the intermediate host.
Patton says, "There is no shadow of doubt that the bed-
bug transmits the disease," but this has as yet not been
definitely proved.1 Further studies by Nicolle and others
have shown that the dog can be inoculated with the type of
the disease occurring in Tunis, and furthermore that the
disease occurs spontaneously in dogs which are possibly the
source of the human disease. In Tunis only children have
so far been observed to be infected. The parasite at first
was thought to be somewhat different from L. infantum

1 Supplements to Third and Fourth Reps., Wellcome Research Lab.,
Khartoum, 1908, 95; 1911, 150.

INFECTION BY INSECTS 403

but the two forms are now generally believed to be identical.
Basile thinks the flea is the carrier, while Donovan suspects
some species of Connorrhinus. The disease has recently
been observed in Italy, Sicily, Malta and Portugal.1

Pellagra. — Although until recently the majority of
writers have been inclined to attribute this disease to the
consumption of spoiled maize, Sambon as early as 1905
seems to have suspected that it might be a true contagious
disease transmitted by insects. Sambon considers that the
topographical distribution of the disease in Italy corresponds
closely with the distribution of a small biting fly, Simulium
reptans. Other observers, as Alessandrini and Lavinder,1
agree that it is a rural disease and even in the country has
a " patchy " distribution as has malaria. Sambon and
others have noted its prevalence along watercourses and its
occurrence is confined to spring and autumn. Sambon
thinks that Simulium is the carrier because it has the same
geographical and seasonal distribution as pellagra and be-
cause it breeds in watercourses along which pellagra is
found. The insect does not enter houses, which, according
to Sambon, explains the limitation of the disease to out-of-
door workers. It is the only insect which has the same
distribution as pellagra.

Recently in South Carolina2 the relation of pellagra to
the Simulium vittatum has been noted. A similar relation
has been reported in Kansas by Hunter.3

Opilacao or Canguary. — These names are in certain
parts of Brazil applied to hook-worm disease, but in the
region where it was studied by Chagas they are names of a
quite fatal disease of children characterized by dropsy,
glandular swellings and various nervous disturbances,

1 Pub. Health Rep., U. S. Pub. Health and Mar. Hosp. Serv., 1911,

XXVI, 1459.

2 Pub. Health Rep., U. S. Pub. Health and Mar. Hosp. Serv., 1912,

XXVII, 255.

» Hunter, J. Am. M. Ass., Chicago, 1912, LVIII, 547.

404 THE SOURCES AND MODES OF INFECTION

especially convulsions. Chagas found the disease to be
characterized by the presence of a species of trypanosome
named, by him, Schizotrypanum cruzi. This trypanosome
is also found in various animals from which it is probably
transmitted to man by a large biting bug, Connorrhinus
megistus. The bugs only become infective on the eighth to
tenth day after feeding.1

Pappataci Fever. — This fever, which much resembles
dengue, occurs along the eastern coast of the Adriatic, in
Malta and Crete and probably in India. Doerr in 1908
definitely proved that it is transmitted by a species of
" sand fly," Phlebotamus papatassi. The cause of the dis-
ease is probably ultra-microscopic, as it is contained in the
filtered blood. The virus is transmitted by the female fly
to its offspring which become capable of causing infection
after reaching maturity. The habits of the fly have been
studied and plans devised for its control.2

Oriental Sore. — This is also variously called " Bagdad
boil," " Biskra boil," "Aleppo button" and " Bouton
d'Orient." It is also found in Panama and in Brazil, as well
as in Asia and Africa. The parasite, Leishmania tropica,
has been successfully cultivated. Inoculated into monkeys
local lesion may result, but these animals seem not to be
very susceptible. Although the parasites are not found in
the blood it has been thought that certain biting insects
may transmit the disease after becoming infected by feed-
ing on the sores. Anopheles mosquitoes and the itch
insect have been suggested. Wenyon3 has made careful
studies of this disease in Bagdad and has succeeded in
getting a growth of the parasites in bedbugs and mosquitoes
but not in ticks. As only the exposed parts of the body are

1 Second Review of Recent Progress in Tropical Medicine, Khar-
toum, 1911, 371.

* Second Review of Recent Advances in Tropical Medicine, Khar-
toum, 1911, 248.

« Wenyon, Parasitology, 1911, IV, 273.

INFECTION BY INSECTS 405

infected he thinks it improbable that bedbugs transmit the
disease but suggests a sand fly, a species of Phlebotamus, or
Stegomyia fasciatus. Wenyon also thinks that the virus
may sometimes be carried mechanically by flies.

Typhus Fever. — According to Wilder,1 Mexican physi-
cians have long looked upon insects as carriers of " tarba-
dillo" or Mexican typhus. Recently Sambon suggested the
louse as perhaps the transmitter of this disease, and hi 1906
Hay2 noted many epidemiological facts tending to impli-
cate the flea. The first experimental work was done by
Nicolle and his associates in Tunis, who succeeded in trans-
mitting the typhus fever of that region from man to an ape.
Ricketts and Wilder3 and Anderson and Goldberger,4 work-
ing with the Mexican type of the disease independently of
Nicolle, demonstrated conclusively that the body louse,
Pediculus vestimenti, can, under experimental conditions,
easily transmit the disease from man to monkey and from
monkey to monkey. They also showed that the virus is
contained in the blood and is filterable and that one attack
protects against another. Three of the investigators con-
tracted typhus fever while engaged in this work and Ricketts
died. One of the others, Dr. Goldberger, writes me that
while experimenting on typhus fever he was bitten by a
presumably infected louse. Five days later the symptoms
of typhus fever developed.

Some experimental evidence was secured showing that in
lice the infection is transmitted to offspring. Some work
was done with bedbugs and fleas but no evidence was ob-
tained except of a negative character. Wilder presents
considerable epidemiological evidence to show that bed-
bugs and fleas are not likely to be the carriers of this disease.

1 Wilder, J. Infect. Dis., Chicago, 1911, IX, 9.

2 Pub. Health, Lond., 1906-7, XIX, 772.

3 Ricketts and Wilder, J. Am. M. Ass., Chicago, 1911, LIV, 1304.

4 Anderson and Goldberger, U. S. Pub. Health and Mar. Hosp.
Serv., Pub. Health Rep., 1910, XXV, 177.

406 THE SOURCES AND MODES OF INFECTION

Epidemiological Evidence. — Typhus fever was formerly
considered one of the most contagious diseases, and at the
same time a typical filth disease. It was especially rife in
camps and prisons and on shipboard. Physicians and
nurses were frequently attacked. An impressive record of
its contagiousness is a tablet in Bellevue Hospital, where I
served as an interne, on which are inscribed the names of
the medical men in the hospital who had died from it. In
1864 of 21 members of the staff 14 contracted the disease.
Of late the contagiousness has apparently diminished. It
is now a rare disease in western Europe and North America
(except in certain parts of Mexico). Outbreaks are usually
easily controlled and hospital infection is comparatively
rare. Thus 600 cases were treated in the City Hospital,
Liverpool, without a single case of hospital infection.1
Wilder says that in the American Hospital in Mexico 144
cases were treated without transfer of the disease. Hay in
Aberdeen, while noting instances of infection in the hospital
during the outbreak of typhus in 1906, says that they were
very few, as compared with former times, and confined to
those who were brought in contact with the patients before
or at entrance into the hospital. Those who handled the
patients after they were cleansed escaped. In the typhus
outbreak in New York in 1864, as stated above, 14 of 21
members of the hospital staff contracted the disease, while
from 1881 to 1893, during which time there were 1897 cases,
most of which went to the hospital, only one hospital
physician was attacked, namely in 1884.

New Theory Explains the Facts. — The writer has always
considered the decrease in typhus fever extremely puz-
zling. The old-time hygienists were in the habit of attrib-
uting it to improvements in " sanitation/' meaning thereby
better water supply and drainage, improved scavenging and
better housing, but no one explained why all this had such
a marked effect upon typhus and so little upon smallpox,
1 Robinson and Potts, Brit. M. J., 1905, I, 1137.

INFECTION BY INSECTS' 407

scarlet fever and diphtheria. Again, if typhus is contagious
by the same means as are the diseases just named, why
should isolation " stamp out " typhus and do nothing of
the kind for scarlet fever and diphtheria? Much light is
thrown on the problem if, as now seems probable, the louse,
or some similar insect, is the chief factor in the transmission
of the disease. There is little doubt that there has been a
vast improvement in the attitude of the public as regards
these parasites, and the poverty and squalor which favor
their presence has vastly decreased in the most progressive
nations where the disease no longer prevails. In prisons,
hospitals and similar institutions lice and bedbugs formerly
abounded, but in all the better class of institutions a con-
stant warfare is now waged against them. Thus a descrip-
tion of Bellevue Hospital in 1837, where typhus fever was
then prevailing, reads " The patients were lying in their
filthy blankets, destitute of sheets and pillow cases and in
some chronic cases they had not had a change in three
months." Even when I was interne in 1879 the rooms of
the staff were infested with bedbugs, pediculosis was quite
common and the lodging house in the basement was out-
rageously filthy and full of vermin of all kinds. The con-
trast furnished by most modern hospitals is marvelous.
If typhus fever is spread by lice no wonder that it has
almost vanished from our institutions.

Seasonal Distribution. — Wilder calls attention to a
number of points in which this mode of extension fits in
with the epidemiology of the disease. Thus typhus is a
disease of temperate climates, and lice are not as numerous
in the tropics as in cooler regions, and on account of light
and loose clothing their parasitism is not as close. In addi-
tion he found some evidence that warm weather shortened
the persistence of the virus in the louse, also that the
seasonal distribution of typhus in Mexico corresponds to
the seasonal distribution of lice. In this connection it may
be noted that Hamer found in London that the maximum

408 THE SOURCES AND MODES OF INFECTION

seasonal prevalence of lice is in January and the minimum
in April, May and June.1 The seasonal distribution of
typhus in London is not so marked as for most infectious
diseases, but of 18,286 cases admitted to the London Fever
Hospital the maximum number was in January and 30 per
cent were admitted during the first three months of the
year.2

The marked predilection of typhus for poverty and filth
is well explained on this hypothesis. The only place in
America where typhus prevails is on the Mexican plateau
where Wilder says the masses are grossly lice infested.
Even in the better quarters of the City of Mexico, the lousy
lower classes, as servants and otherwise, are brought in
such frequent contact with more cleanly people that pedi-
culi and typhus are not rarely found among the latter.
Hay called attention to a number of facts which he thought
tended to implicate the flea. He found that in clean houses
and among clean people the disease did not spread. The
only hospital employees who contracted the disease were
in contact with patients before the latter had been cleaned.
Several were then attacked. No cases developed in the
wards after the patients had been cleansed. He was per-
haps led to suspect fleas because of the flea bites noted on
the patients, but fleas and lice are often together, and
Wilder and Anderson and Goldberger, judging from epi-
demiological as well as from experimental data, believe that
the louse is the chief factor in the spread of the disease.

Such recent writers as McVail3 and Ker4 consider typhus
above all things an air-borne disease, though they also state,
as do others, that its " striking distance " is remarkably
short. The short striking distance is easily explained if

1 Hamer, Rep. Med. Off. Health to London Co. Council, 1909,
Append. IV, 8.

2 Blyth, A Manual of Public Health, Lond., 1890, 401.

3 McVail, The Prevention of Infectious Diseases, Lond., 1907, 46.
* Ker, Infectious Diseases, Lond., 1909, 202.

INFECTION BY INSECTS 409

lice are the carriers. So also it is easy to understand how
clothing may at times transmit infection, as is said to be
the fact. Thus Wilson1 reports an instance in Belfast
where the disease was probably carried in this manner in
a shawl, yet according to the observations of Doty (page
216) such carriage is rare.

Identity of Typhus Strains. — There has always been
some doubt as to the identity of Mexican typhus and the
typhus of Europe, though very many think that they are
one and the same disease. In his early experiments Nicolle,
in Tunis, was not able to infect rhesus monkeys, but only the
higher apes, while the experimenters in Mexico found the
rhesus susceptible ; but later Nicolle, by using larger doses,
was able to cause infection just as could his American co-
workers. No experimental difference was finally noted by
the two groups of observers. Recently Brill2 has reported
cases of fever occurring in New York for many years which
in many respects resembled typhus fever but which ap-
peared to be much milder and less contagious. In New
York it has been commonly called Brill's disease. Ander-
son and Goldberger,3 by inoculation and immunizing tests
with monkeys, now seem to have shown that Brill's disease,
of New York, and Mexican tarbadillo are identical. If
this is so the New York type has been greatly modified or
else insect carriers are rare in New York, for the disease
shows no tendency to spread in either the family or the
hospital.

Dengue.

Transmitted by Mosquitoes. — The causation and mode
of transmission of dengue are somewhat uncertain. Graham
believes that he has demonstrated in the blood a protozoan

1 Wilson, J. Hyg., Cambridge, 1910, X, 155.

2 Brill, Am. J. M. Sc., Phil., 1911, CXLII, 196.

3 Anderson and Goldberger, U. S. Pub. Health and Mar. Hosp. Serv.,
Pub. Health Rep., 1912, XXVII, 149.

410 THE SOURCES AND MODES OF INFECTION

which he considers the cause of the disease, but his findings
have not been substantiated. Ashburn and Craig l demon-
strated that it is possible to transfer the disease to healthy
men by the inoculation of blood from the sick, and that the
virus is contained in the filtered blood. They also proved
that the disease can be transmitted by a mosquito, Culex
fatigans, and they consider that this is probably the most
common mode of transmission. Stitt also believes that the
mosquito is the bearer of the disease. Carpenter and Sut-
ton could not transfer the disease by mosquitoes, but they
did not experiment with C. fatigans. Ross has shown that
while dengue prevailed elsewhere in Egypt, there was none
in Port Said and Ismail ia, where the mosquitoes had been
exterminated. Balfour,2 from whose review this information
is chiefly taken, states that in Khartoum, which was com-
paratively free from mosquitoes, especially C. fatigans, there
was no dengue, though persons with the disease in all proba-
bility came there from Port Sudan and Haifa. It appears
probable, therefore, that the contentions of Ashburn and
Craig are correct.

Relapsing Fever.

Varieties. — According to Craig,3 there are at least four
forms of relapsing fever, each due to a particular kind of
spirocheta, and each apparently limited in geographical dis-
tribution. It is still uncertain whether these spirochetse are
bacteria, as was formerly believed and as is strongly urged
by Novy and Knapp, or protozoa, similar to the trypano-
somes, as is maintained by Schaudinn, Prowazek, Leishman
and others. Whatever they may be, there is no doubt that
at least one form of the disease is carried from person to
person by means of ticks.

1 Ashburn and Craig, J. Infect. Dis., Chicago, 1907, IV, 440; also
Philippine J. Sc. [B. Med.], II, 93.

2 Supplement to the Third Rep., Wellcome Research Lab., Khar-
toum, 1908, 37.

' Craig, The Malarial Fevers, New York, 1909, 445.

INFECTION BY INSECTS 411

Ticks. — The African form of relapsing or tick fever is
caused by S. duttoni, which was discovered by Button and
Todd, and also by Ross and Milne in 1904, the first named
of whom, Button, lost his life while studying the disease.
Button and Todd demonstrated beyond question that the
disease may be carried from man to man by a tick, Ornitho-
dorus moubata, and their observations were confirmed by Ross
and Milne and by Breinl and Kinghorn. It was also shown
that the disease is hereditary in ticks, according to Mollers,1
even to the second generation. Unlike the malarial parasites
and the trypanosomes, these spirochetae soon disappear from
the blood, and chronic latent infections do not seem to exist.
Koch, however, suggests that the spirochetae may ultimately
be found in rats, and that these rodents may be the real
source of this disease, as they are of plague. The tick, how-
ever, is said to be exclusively a human parasite. It is found
in the floors and crevices of houses and native huts, where it
hides during the day and feeds at night. The spirochetae
undergo a certain development and multiplication in the tick,
but whether they pass through a definite cycle of develop-
ment, like the protozoa, is not known. Ticks have been
known to be infective for a year and a half.

Bedbugs. — Even before the discovery of the spirocheta it
was suggested that the bedbug might be the carrier of the dis-
ease, but, according to Balfour,2 Breinl and Kinghorn, and
Todd have by experiment shown that this is probably not so.

Less is known about the other forms of relapsing fever.
Tictin in 1897 believed that the recurrent fever of Europe is
transmitted by the bedbug, and he claimed to have demon-
strated its possible transmission in this way, but the experi-
ments were not entirely satisfactory and have not been
substantiated.

1 Mollers, Ztschr. f. Hyg. u. Infectionskrankh., Leipz., 1908, LvTII,
277.

2 Supplement to the Third Report, Wellcome Research Lab.,
Khartoum, 1908, 190.

412 THE SOURCES AND MODES OF INFECTION

The Indian form of the disease is caused by S. carteri, and
has been much studied. Mackie could demonstrate the spi-
rochetse in bedbugs fed on relapsing-fever patients, up to
the sixth day, but he found no evidence of their increase.
Of fifty-three bugs found in a fever ward, only one contained
spirochetae. He also placed infected bedbugs in cages with
six monkeys, and only one of the monkeys contracted the
disease. Other sources could not be excluded.

Lice. — Recently the same author has found the spiro-
chetae in from two to fourteen per cent of body lice found in
fever wards, and he showed that the parasites increased
in number in the lice. Sergent and Foley, in Algeria, have
made observations and experiments tending to implicate the
clothes louse, Pediculus vestimenti.

Rocky Mountain Fever.

Transmitted by Ticks. — Rocky Mountain tick fever prob-
ably appeared in Idaho about 1885, and is now found in
Idaho, Montana, Washington, Oregon, Wyoming and Utah.
It is an acute fibrile disease with a high mortality, usually
exhibits an eruption on the skin, and is followed by immunity.
According to the investigations of Ricketts,1 the disease is
transmitted by means of ticks, probably from some of the
small wild animals of the country. He showed that it is
possible to transmit the disease to monkeys, guinea pigs,
gophers, chipmunks, rock squirrels and other animals by
allowing ticks (Dermacenter andersoni) to bite first an infected
animal, and later a healthy animal, which soon becomes
infected. There is evidence that the virus of the disease
develops in the body of the tick and is transmitted to its
offspring. Ricketts has shown by inoculation experiments
that the virus is contained in the blood serum of infected
animals, and he recently2 reports the discovery of a bacillus
which he believes to be the cause of the disease.

1 Ricketts, Rep. St. Bd. Health, Montana, 1907-08, 88.
* Ricketts, J. Am. M. Ass., Chicago, 1909, LII, 379.

INFECTION BY INSECTS 413

Biological Carriers Distinguished from Mechanical Carriers.

— In most of the diseases thus far considered in this chapter
the parasite which causes the disease passes through a defi-
nite cycle of its existence in the insect as well as in man.
The insect is in most cases the true host of the parasite and
has been called a " biological carrier " of disease. Most of
the parasites transmitted in this way belong to the protozoa.
On the other hand, insects may be merely " mechanical
carriers " of disease. The blood which contains the germs of
the disease may be carried on the mouth parts of the insect
and inoculated into the next person bitten, just as it might
be carried on the point of a lancet.

Plague.

Discovery of Bacilli in Fleas. — According to Nuttall,1
Ogata in 1897 found plague bacilli in fleas taken from the
body of a rat dead of bubonic plague, and this observation
was confirmed by the German Plague Commission the same
year. In 1898 Simond noted that persons who handle dead
rats not rarely die of the plague, but he found that infection
never occurred if the rats had been dead over twenty-four
hours. He also found that persons with plague occasionally
presented on their skin what appeared to be flea bites, and
he was able to demonstrate plague bacilli in these supposed
bites. He was able also to demonstrate plague bacilli in
fleas taken from rats dead of the plague.

Plague derived from Rodents. — Simond's observations
led to much study and experiment and to still more contro-
versy. It has been shown by large numbers of observations
that plague is primarily a disease of the lower animals, par-
ticularly of rodents, many species of which easily become
infected. It has been shown also that plague in man is gen-
erally derived from the same disease in rats, or more rarely
mice, and recently in California from ground squirrels (Citel-
lus beecheyi).

1 Nuttall, Johns Hopkins Hosp. Rep., Bait., 1899, VIII, 1.

414 THE SOURCES AND MODES OF INFECTION

Epidemiological Evidence. — The very careful observa-
tions of Thompson1 in Australia showed that outbreaks of
human plague were preceded by outbreaks of rat plague,
that human cases were connected in one way or another with
foci of rat infection, and that the dispersal of human cases
did not result in the dispersion of plague. He showed that
while human plague had a local association with rat plague,
the association was not a direct one, but was such as to render
necessary the assumption of some such intermediary as the
flea. It also seemed to him that the location of the buboes
was in accord with the theory of flea infection.

Experimental Work. — Objection had been made by Nut-
tall and others to the flea theory on the ground that rat and
human fleas are of different species, but Tidswell in Sydney2
showed that rat fleas when hungry will bite man. This has
been substantiated by many other observers, some of whom
have also shown that Pulex irritans, the human parasite, will
occasionally bite rats. It has also been shown that fleas
speedily leave a dead animal, and then, impelled by hunger,
they attack the nearest victim. As was shown by Nuttall,
plague bacilli may live in fleas for several days. The English
Plague Commission found as many as 5000 bacilli in the
stomach of a single flea, and they believe that they increase
in number for a while and sometimes live for fifteen days,
but in starved fleas do not live over a week.3 Verjbitski 4 has
also confirmed these observations. He has shown, too, as did
Nuttall, that bedbugs may become infected and harbor the
germs for some days. According to Tidswell,5 rats are fre-
quently infested with bedbugs (Cimex lectularius) , but there
is no evidence as yet that these play any important part in

1 Thompson, Reports of Plague in Sydney, 1902 to 1907.
8 Tidswell, Rep. on Second Outbreak of Plague in Sydney, 1902,
73-75.

» J. Hyg., Cambridge, 1907, VII, 419.

4 Verjbitski, J. Hyg., Cambridge, 1908, VIII, 162.

• Loc. cit.

INFECTION BY INSECTS 415

the causation of the disease. Verjbitski showed that if the
biting insect is crushed on the skin, either the bite or neigh-
boring scratches are easily infected.

Demonstrations in India. — Finally, the very beautiful
experiments of the English Plague Commission1 showed con-
clusively that infection by fleas was practically the only way
in which experimental animals could be infected under natural
conditions. These experiments have been referred to on
pages 246 and 312. In one experiment, a plague-infected vil-
lage was cleared of its human inhabitants and guinea pigs
placed in the houses, and it was demonstrated that infection
took place only through the medium of fleas. The experi-
mental and epidemiological evidence is that the bubonic form
of human plague, which is by far the most common form, is
caused by the transference of the bacilli from rodents to man
by means of fleas.

Plague rarely spread except by Fleas. — While it is true
that the flea is the chief agent in the transmission of plague,
the disease may possibly be, and doubtless occasionally is,
spread in other ways. In experimental animals it is possible
to induce it by the ingestion of infected food, and the contact
of the abraded or injured skin with any substance containing
plague bacilli may result in infection. The pneumonic form
of plague, when no precautions are taken, is highly contagious,
for the sputum contains the bacilli. In some outbreaks of
plague the pneumonic form prevails. In some localities the
overcrowding, filth and habits of the people may cause con-
tact infection or infection by food to assume some impor-
tance, but in Western Europe, Australia and America there
can be little doubt that plague is chiefly a flea-borne disease.

Measures against Plague. — Measures against the plague
ought, then, to involve the isolation of the sick, particularly
those sick with the pneumonic form. In India evacuation

'< * J. Hyg., Cambridge, 1906, VI.

See also Summary of Work of Commission, 1908, Calcutta, Supt. of
Gov. Printing.

416 THE SOURCES AND MODES OF INFECTION

of infected villages, that is, the removal of the population
from the vicinage of the plague rat, is a measure of value.
Disinfection, if it includes the killing of fleas and rats, is an
important measure. Haffkine's serum has been found to pre-
vent attacks for several months, and its general use is often
of value in preventing the development of individual cases.
But the chief anti-plague measures must be directed, not
against the flea, but against rats and other rodents. The most
obvious procedure is to attempt the destruction of the rats,
but thus far this has proved to be a rather hopeless under-
taking. Poisoning, trapping and destruction by infectious
disease have proved only palliative or adjuvant measures.
In Japan, in Australia and in California the most energetic
rat destruction has not materially reduced their numbers.
Thompson1 says: "Prevention of epidemic plague conse-
quently lies in maintaining a distance between the rat and
man. The requisite separation between rat and man will be
better secured by improving the construction of buildings
than by attempts to exterminate the rat." In Sydney large
sums are being spent in rebuilding wharves and warehouses
so as to render them rat-proof. In San Franscisco 2 every
effort has been made to prevent the growth of rats by cutting
off their food supply, by improved scavenging, and by pro-
viding metal garbage pails and rat-proof receptacles for all
kinds of food and grain likely to attract rats. But the chief
effort was directed to rat-proofing, by the use of concrete
construction, of stables, markets, provision and produce ware-
houses and wharves.

Anthrax.

Nuttall, in his valuable discussion and bibliography of the
subject of insect carriers of disease,3 cites a large number of

1 Thompson, Internat. Cong. f. Hyg. u. Demog., Berl., 1907, III, 672.

* Eradicating Plague from San Francisco, Rep. of Citizens' Health
Com., 1909.

« Nuttall, Johns Hopkins Hosp. Rep., Bait., 1899, VIII, 1; see also
Rep. to Local Gov. Bd., Lond., on Pub. Health, 1909, n. s., 16.

INFECTION BY INSECTS 417

authors who suggest that anthrax may at times be trans-
mitted by any biting insects which pass from cattle to man.
Numerous instances are given of persons who date the begin-
ning of their pustule from the sharp sting of an insect, but,
as was well said, it is probable that often the first noticeable
sensations in the development of the pustule are wrongly
considered as due to the bite of an insect. In a few cases a
fly was felt to bite and was killed, and on the spot a ma-
lignant pustule developed. Most of the score or more of
writers quoted urge this mode of infection on purely a priori
grounds, and it is certainly highly probable that where
anthrax is at all prevalent among animals, and biting insects
pass from the lower animals to human beings, the disease is
sometimes transmitted in this way; but, as has been sug-
gested, if this happened often, the disease would be more
common than it is. Certainly in England and the United
States practically all human anthrax results from handling
infected material.

Nuttall 1 allowed bedbugs to suck blood from mice infected
with anthrax, and then caused them immediately to bite
healthy mice. In all, 136 infected bugs were used, but in no
instance did they transmit the disease. In six instances the
bugs were crushed while biting, but these experiments were,
like the others, without result. Nuttall also made a few
similar experiments with fleas, with like negative results. He
found that anthrax bacilli remained alive in bedbugs only
for from two to four days, and died off even more quickly
in fleas.

Nuttall carried on similar experiments with bedbugs and
chicken cholera, but these were equally unsuccessful.

Septic Infection.

Nuttall cites Faure, Paltauf, Chrzaszczewski, Joseph and
Berry as having noted instances of septic infection following
the bites of insects, and others are occasionally noted in medi-
1 Nuttall, loc, tit,, 13.

418 THE SOURCES AND MODES OF INFECTION

cal literature. That this accident should sometimes happen
seems very probable.

Typhoid Fever.

Button1 caused bedbugs to bite a typhoid-fever patient
and after twenty-four hours' fasting to bite two other
healthy persons, both of whom developed typhoid fever in
twelve and fourteen days respectively.

Insects carry Germs on Bodies. — The insects thus far
considered inoculate disease by means of their bites, but it is
evident that some diseases may be carried by them in other
ways. If infective material exists in considerable quantity,
and is accessible to insects, they are almost certain to get it
on their legs and bodies, and may then carry the germs to
other human beings. The chance of their doing this varies
greatly with varying conditions. If the insects fly, the
danger would seem to be much greater. If the infective
material is large in amount and freely exposed, as typhoid
excreta in privy vaults, the danger of fly infection is in-
creased, and it is also greater if there is a possibility of
infecting food which insects love to visit. Flies, particularly
the common house fly, Musca domestica, have been most
often accused of thus carrying infection, but other insects
have not escaped suspicion.

Cockroaches and Other Insects. — Engelmann2 reports
cockroaches extremely numerous in certain houses in
Chicago where typhoid fever prevailed, and she attributed
the spread of the disease to these insects. Weber3 accuses
various species of Psocidce of carrying tubercle bacilli from
cow to cow, and states that he demonstrated the germs in one
of these insects found in a manger. In fact any "domestic"
insect with power of active locomotion, and of filthy habits,
may be considered as a possible carrier of disease germs.

1 Button, J. Am. M. Ass., Chicago, 1909, LIII, 1248.

2 Engelmann, Med. News, N. Y., 1903, LXXXII, 225.

3 Weber, N. York M. J. [etc.], 1906, LXXXIV, 884.

INFECTION BY INSECTS 419

But the house fly is undoubtedly the most numerous and
ubiquitous and the most " domestic " in its habits, and it is
rightly suspected of playing a larger part as the mere passive
carrier of disease germs than is any other insect.

Mayer l states that some ants were seen to pass between
cages containing mice, some of which were sick with mouse
typhoid and some well. The well mice soon developed the
disease, and agar plates, so placed that the ants ran over
them, showed numerous colonies of the B. murium typhi.

Flies seen to carry Infective Material. — A large number
of observers, noting the passage of flies from infected matter
to human beings or to food, have become convinced that
these insects are important carriers of disease. According
to Nuttall, Budd as long ago as 1862 considered it proved
that Egyptian ophthalmia is carried from child to child by
the flies that can be observed in great numbers crawling
over the face and eyes. Laveran noted the same conditions
in Biskra, and he and many others believe that oriental sore
is spread in the same way. Nuttall also cites several of the
earlier writers as believing that anthrax, cholera and plague
are transmitted by flies.

Experimental Work. — With the development of bac-
teriology, experiments began to supplement conjecture, and
it was definitely determined that bacteria might be carried
on the bodies of insects, which indeed is self-evident, and
also that they might in a living state pass through the
stomach and be voided in the feces. According to Nuttall,
Raimbert, Davaine and Ballinger demonstrated living
anthrax bacilli on flies of various kinds that had been feed-
ing on infected material. Nuttall himself showed that
house flies could carry for 48 hours living plague bacilli
which they had derived from material they had fed on, and
Hankin showed the same for ants. Nuttall, ir. the mono-
graph previously cited, reviews the early literature, and
many useful references as well as accounts of original work
1 Mayer, Miinchen med. Wchnschr., 1905, LII, 226.

420 THE SOURCES AND MODES OF INFECTION

are to be found in Howard's1 recent work and in the valuable
reports to the Local Government Board.2

Cholera Spirilla on Flies. — In 1886 Tizzoni and Cattani
caught flies in a cholera hospital in Bologna and found that
cultures made from them showed cholera spirilla. Sim-
monds made similar observations in Hamburg in 1892.
Macrae in India, in 1894, exposed boiled milk in different
parts of a jail where cholera prevailed, and found that it
became infected with cholera germs. Tsuzuki, in 1904,
captured cholera-infected flies in a house in Tientsin where
there were cases of the disease. Chantemesse and Gagnon
have shown that flies may carry virulent germs for 17 and
24 hours respectively. Maus3 says that cholera spirilla
were found on several bluebottle flies caught in infected
houses in the Philippines. Graham-Smith4 in eight experi-
ments could recover cholera spirilla from the legs and wings
of flies only once later than 5 days after infection. Even
in the crop and intestines the numbers rapidly diminished,
all cultures after 48 hours yielding negative results.

Typhoid Bacilli on Flies. — The earlier observers were
merely content to show that flies could transfer pathogenic
germs, and made little attempt to show in what way and
for how long they could be carried. Thus Manning5 and
Buchanan6 caused flies to walk over typhoid feces and then
over Petri plates with the result that some of the plates
became infected. Firth and Horrocks7 showed that typhoid
bacilli could be carried on the legs, wings, head and bodies
of flies and that they could be found in their excreta after
the insects had been fed on infected material, but they did

1 The House Fly, Disease Carrier, N. York, 1911.

2 Reports on Pub. Health and Med. Subjects, Nos. 5, 16, 40, 53.
8 Maus, Abst. Med. News, N. Y., 1902, LXXI, 318.

4 Rep. to Local Gov. Bd., Lond., 1910, No. 40.
6 J. Am. M. Ass., Chicago, 1902, XXXVII, 1291.

6 Lancet, Lond., 1907, II, 216.

7 Brit. M. J. Lond., 1902, II, 936.

INFECTION BY INSECTS 421

not consider that they had definitely proved that the germs
passed through the intestinal tract. Faichnie l believes that
typhoid bacilli are more often carried in the intestines of
flies than on their legs. He bred flies in earth infected with
typhoid feces. From the intestines and from the excre-
ment of flies so bred he recovered typhoid bacilli up to the
sixteenth day. Similar breeding experiments were carried
on in earth impregnated with feces containing B. paraty-
phosus A. and bacilli were obtained from the flies up to the
tenth day. Graham-Smith2 fed flies on syrup containing
typhoid bacilli and endeavored to determine the persistence
of the germs on their feet, in their intestines and in the
" specks " or excreta. Bacilli were not found on the feet or
in the specks longer than 48 hours but were recovered from
the intestines for at least 6 days. Similar experiments
carried on with B. enteritidis (Gartner) showed that while
the bacilli could rarely be found on the legs they could be
recovered from the " crop " and intestines up to 7 days and
probably longer. Hewitt3 allowed flies to walk over thin
smears of typhoid feces and then kept them in cages for
24 hours and after that allowed them to walk over culture
media. No typhoid colonies were found and only a few of
B. coli. Similar experiments were made with B. prodigiosus
with negative results, though the germs were recovered
from the feet of the flies after 12 hours.4

Typhoid Bacilli on Flies from Infected Places. — Hamil-
ton6 in Chicago, and Fricker in Hamburg, in the same year,
recovered typhoid bacilli from flies in houses where there
were cases of the disease. Klein6 also recovered typhoid
bacilli from flies from houses where there was typhoid fever.

1 J. Roy. Army M. Corps, Lond., 1909, XIII, 672.

2 Rep. to Local Gov. Bd., Lond., 1910, No. 40.

3 Hewitt, Quart. J. Micr. Sc., 1909, n. s., LIV, 394.

* Ibid., 403.

6 Hamilton, J. Am. M. Ass., Chicago, 1903, XL, 576.

• Klein, Brit. M. J., Lond., 1908, II, 1150.

422 THE SOURCES AND MODES OF INFECTION

Faichnie, above cited, proved the presence of typhoid
bacilli on flies caught in barracks at Ramptee, India, where
there was an outbreak of the disease, and later he reported
finding typhoid-infected flies on nine other occasions in
various localities. Bartarelli,1 in studying a house outbreak
near Turin, found typhoid bacilli on 8 of 120 flies caught in
the house and on 2 of 35 caught in the yard. On the other
hand, as might be expected, some have failed to find the
bacilli under similar circumstances, as Buchanan at Glas-
gow, Gorham at Providence and Gunn at Orange, N. J.

Tubercle Bacilli carried by Flies. — Spillmann and
Haushalter found tubercle bacilli in the intestines and in
the dejecta of flies that had fed on tuberculous sputum.
Hoffmann found tubercle bacilli in 2 of 4 flies captured
in a room occupied by a tuberculous patient, and also in
the excreta of flies scraped from the walls and furniture of
the room. Lord2 found virulent bacilli in flyspecks, but
could not induce the disease by causing guinea pigs to
breathe air drawn over infected specks. Hay ward3 demon-
strated living bacilli in flyspecks after the flies had been
fed on sputum in such a way as to preclude the possibility
of the infection of their feet or bodies. Andre4 demon-
strated tubercle bacilli in the excreta of flies by inoculation
tests, and he found that they appeared in the feces about
six hours after feeding and continued for five days. Flies
may also become infected by eating tuberculous dust.

Graham-Smith5 after feeding flies for several days on
syrup infected with a culture of human tubercle bacilli de-
termined that the bacilli may " be present in the crop for
3 days. In the intestines, however, they may be found for
much longer periods, being present in considerable numbers

Centralbl. f. Bakteriol. (etc.), I, Abt., Jena, 1910, LIII, 486.
Lord, Boston M. & S. J., 1904, CLI, 651.
Hayward, N. York M. J. [etc.], 1904, LXXX, 643.
Andr6, 6th Internat. Cong, on Tuberc., Wash., 1908, I, 162.
Graham-Smith, Rep. to Local Gov. Bd., 1910, No. 40.

INFECTION BY INSECTS 423

for at least 6 days. Subsequently their numbers diminish,
but they may be discovered by careful search for 12 days
or even longer. In the faeces they are numerous up to the
fifth day, and occasional specimens may be found in faecal
material deposited between the sixth and fourteenth days
after infection." Flies fed on sputum, only, contained the
bacilli for 4 days, but flies cannot stand forced feeding with
sputum as they do with infected syrup.

Anthrax Bacilli carried on Flies. — As has been stated,
several early observers noted the ability of flies to transfer
anthrax germs. Graham-Smith, however, has shown the
period during which the flies may remain infective. He
found that after flies have fed on the blood of an animal
dead with anthrax the spore-free bacilli do not remain alive
on the external parts of the fly for more than 24 hours.
They may, however, persist in the intestine for 3 days and
in the crop for 5 days. When flies are fed on anthrax spores
the latter do not develop in the fly. They tend to disappear
gradually from both the exterior of the body and from the
intestine, and in one experiment none were obtained from
legs or wings after the tenth day and from the crop only
once after that, namely on the sixteenth day. In another
experiment the spores persisted both outside and inside
the fly up to the twentieth day, when the experiment
terminated.

Other Disease Germs on Flies. Dysentery. — Auche*1
allowed flies to become contaminated with cultures of the
dysentery bacillus, and found that they could carry the
germs for hours, and that the flies would take up the bacilli
from feces as well as from cultures.

Graham-Smith experimented by feeding flies on an emul-
sion of a diphtheria culture and showed that the bacilli
seldom remain alive on the legs and wings for more than a
few hours, but may live in the alimentary canal for 24
hours or even longer.

* Auch6, Compt. rend., Soc. de biol., Par., 1906, LXI, 450.

424 THE SOURCES AND MODES OF INFECTION

He also made use of the Danysz rat virus: " A number of
flies were allowed to feed for 1 hour on a broth culture of
virus, recently recovered from the body of a rat, and were
then transferred to a fresh cage. A piece of bread, soaked
in milk, was put into the cage daily and the flies allowed to
settle and feed on it. After 1 hour the bread was removed
and given to a mouse.

" A mouse was also fed on bread soaked in milk contain-
ing an emulsion of faeces, passed about 48 hours after in-
fection, scraped from the walls of the cage. The animal
died in 2 days, and the organism was isolated from it.

" These experiments show that flies which have fed on
virus are capable of infecting food on which they settle and
feed to such an extent that mice fed on it become infected."

According to Nuttall and Jepson, Welander found that
flies could carry living gonococci on their feet for 3 hours.

Nuttall showed that plague bacilli could be carried by
flies, and Graham-Smith states that Yersin in Hong-Kong
observed dead flies lying about where he made his autopsies
on animals which had died of the plague and demonstrated
that the flies contained virulent plague bacilli.

Although no specific germ of poliomyelitis has been dis-
covered the infectivity of the fluids and other tissues can
be tested on monkeys, and modes of infection can in this
way be studied. Flexner and Clark1 allowed flies to feed
on the fresh spinal cord of a monkey which had the disease,
and determined by inoculation that the virus would persist
on or in the flies for at least 48 hours. Experiments after a
longer interval were not tried.

Mechanism of Infection by Flies. — Graham-Smith has
shown that flies often discharge the contents of their crops,
especially when feeding on dry substances. The regurgi-
tated material frequently contains pathogenic bacteria for
some days after these have been taken up by the flies. Such
bacteria are also discharged in the excrement. The germs
1 J. Am. M. Ass., Chicago, 1911, LVI, 1717.

INFECTION BY INSECTS 425

usually do not live more than a day or two, perhaps less,
on the legs and wings but persist much longer in the ali-
mentary tract. Flies allowed to walk over food material
may infect it not only by means of the germs on their feet
but also by the matter which they disgorge when they stop
to feed and also by their excrement. There is less chance
of infecting liquids, as then the flies do not need to disgorge
to soften their food. Many of Graham-Smith's experi-
ments were made with B. prodigiosus because of the ease
with which it may be recognized. Yet in his experiments
he could not once infect milk with this bacillus either by
allowing infected flies to feed on it or to drown in it.

These experiments seem to show that while infection by
flies is possible it is not as easy as might at first sight appear.
Nevertheless, given much infected material and large num-
bers of flies, it is entirely conceivable that a large amount
of disease may be caused by fly-borne infection. That food
may, under natural conditions, become infected is shown
by the observations of Macrae in regard to cholera. In
1899 the writer, with the assistance of Prof. F. P. Gorham,
exposed agar plates and dishes of milk in the pantry of a
house, and poured large amounts of a culture of B. pro-
digiosus into the near-by privy vault. Of nine tests, four
showed infection with prodigiosus, while all of five plates
exposed near the vault were so infected. Plates covered
with wire netting were not infected. Flies were very
numerous and were constantly walking over the plates.

Such a mode of infection is also indicated by the observa-
tions of Vaughan,1 who reports that during the Spanish War
flies covered with a white coating of lime were often seen
crawling over the food, the lime showing that they had just
come from the latrines where lime had been thrown over
the fecal matter.

Wild Flies carry Fecal Bacteria. — Flies caught in the
open have been shown to be at times loaded with fecal bac-
1 Vaughan, J. Am. M. Ass., 1900, XXXIV, 1456.

426 THE SOURCES AND MODES OF INFECTION

teria, thus indicating at once their habits and the possibility
of their carrying certain diseases. Jackson1 found as many
as 100,000 fecal bacteria on a single fly, and as a general
thing the nearer the flies were to the sewer outlets the more
numerous were this class of bacteria. Graham-Smith2 found
fecal bacteria on 23.6 per cent of 148 flies examined in Cam-
bridge, England. For a part of them the surface of the body
only was examined, and for a part the intestine only. Sur-
face infection was about two and a half times as frequent as
intestinal infection. The highest degree of infection was
found among flies caught near decaying animal matter, and
the next highest among those caught near manure. Nash3
recovered B. coli from a fly caught in a hospital ward.
Delepine 4 found bacilli of the colon type in four of thirty-
five collections of flies caught in Manchester.

More Fecal Germs in July and August. — Recent obser-
vations by Torrey5 indicate that flies may be more dangerous
late in the summer. He found that fecal bacteria of the
colon type were first encountered in abundance in the early
part of July. The bacteria in the intestines of the flies
were 8.6 times as numerous as on the surface of the insects.
On the surface of the flies the colon-group of bacteria consti-
tuted 13.1 per cent of the total; and within the intestine
37.5 per cent of the total. Of the lactose fermenters, iso-
lated and identified, 79.5 per cent belonged in the colon-
aerogens group and 20.5 per cent in the acidi lactici group.
Fifteen cultures of streptococci, isolated and identified,
were distributed among the equinus, fecalis and salivarius
groups. There were none of the pyogenes type. Most

1 Jackson, Rep. to Com. on Pollution [etc.], of the Merchants' Ass.
of N. York, 1907, 16.

* Graham-Smith, Further Preliminary Rep. on Flies [etc.], Rep. to
Loc. Gov. Bd. on Health, n. s., 1909, 16.

3 Nash, J. Hyg., Cambridge, 1909, IX, 141.

4 DelSpine, Rep. on Health of Manchester, 1906, 85.

5 Torrey, J., Infect. Dis., Chicago, 1912, X, 166.

INFECTION BY INSECTS 427

important was the isolation of three colonies of B. paraty-
phosus, A. Bacteria of the paracolon type causing a final
intense alkaline reaction in litmus milk and fermenting only
certain monosaccharids were frequently encountered during
August.

Range of Flight. — As stated by Howard, probably the
distance traversed by the house fly varies greatly under
different circumstances, being greater if with the wind and
if suitable food or breeding places are not at hand. Howard
cites Forbes as observing a flight of half a mile from the
tuberculosis hospital to the general hospital of Cook Co.,
111. He also cites Hine as finding marked flies at a distance
.of 40 rods on the third day but never as far as half a mile.
Austen 1 after liberating marked flies found them at dis-
tances varying from 300 to 1700 yards. Under the direc-
tion of Delepine 300 flies were liberated in the administra-
tion building of a hospital, and within 5 days 4 were caught
in wards distant from 30 to 190 yards. At the Worcester
State Hospital Orton and Dodd2 observed flies carry B.
prodigiosus from the laundry to the kitchen and to all of
the wards although the hospital was well screened.

Seasonal and Local Distribution of Flies. — There is a
disposition in many quarters to rely upon a study of the
relation of fly distribution to disease to throw light upon
the causative relation of the insects to the disease under
consideration. Such studies are not likely to prove very
convincing and they are of little value unless extensive.
Proper studies of the seasonal distribution of flies have been
undertaken in only a few localities and for a few seasons.
There have been still fewer observations on the local dis-
tribution of the insects in relation to the local distribution
of disease. Even if a quite constant correlation of seasonal
fly distribution and disease distribution is determined, the
causative relation between the insects and the disease is

1 Rep. to Local Gov. Bd., Lond., 1911, No. 53.

2 Boston M. & S. J., 1910, CLX1II, 863.

428 THE SOURCES AND MODES OF INFECTION

far from established, as both phenomena might be de-
pendent on some other cause.

Species of Flies. — Various species of flies are found in
and about houses which may possibly serve as the mechani-
cal carriers of disease. Of these by far the most common in
dwellings is the house fly, Musca domestica, though other
flies are not infrequently found. In New York Jackson
found that 98 per cent of all flies captured were M. domestica.
In London in 1908 Austen l found that next to the house fly
the most common species were Homalomyia canicularis,
Calliphora erythrocephala and Muscina stabulans. In Provi-
dence in 1909 Sykes2 found all the above, and also consider-
able numbers of Lucttia ccesar, Sarcophaga sarracenice and
Stomoxys calcitrans, and a few Scenopinus fenestralis, but
99 per cent of the flies caught indoors were M. domestica.
The bluebottle fly, Lucilia ccesar, has been accused by
Button3 of transmitting typhoid fever. Maus in the Philip-
pines blames bluebottles for the spread of cholera.

Transmission of Disease by Flies. — Of this we have
very little actual knowledge. Because flies are seen to pass
from sputum and feces to human lips and to food, and
because it is shown that flies may carry disease germs on
their bodies, it is assumed that these insects are the frequent
carriers of disease. Again, because of a correspondence
between 4;he curves of fly distribution and disease preva-
lence, though this correspondence in some instances is
found not to exist, it is assumed that flies are a prolific
cause of sickness. There has been almost no experimental
work under natural conditions. All that I have been able
to find are the two negative experiments of Peters, to be
referred to, in regard to summer diarrhea. To determine
with any degree of certainty the part really played by flies
in the transmission of disease, we need a large amount of

1 Austen, Rep. to Local Gov. Bd. on Pub. Health, n. s., 1909, No. 5.

2 Sykes, Rep. Supt. of Health, Providence, 1909, 13.

3 Button, J. Am. M. Ass., Chicago, 1909, LIII, 1561.

INFECTION BY INSECTS 429

epidemiological evidence such as would be afforded by
changes in disease incidence following control. of the fly
nuisance. At present there is very little of such evidence
and none at all until very recently. We may now consider
what is said and known about the fly carriage of certain
diseases.

Murrina.

Due to Flies. — Darling l has reported a disease occurring
among horses and mules on the Isthmus of Panama. It is
apparently caused by a trypanosome, T. hippicum. Darling
has given careful attention to the cause of the disease and
is satisfied that it is transmitted by a biting fly, Stomoxys
calcitrans. He has never been able to demonstrate the
growth of the trypanosome in the body of the fly, and he
believes that the disease is carried mechanically. He does
not think that it is introduced by the bite of the insect but
only through abraded surfaces on the bodies of the animals.
Only animals having superficial sores become infected.

Tuberculosis.

Due to Flies. — Cobb 2 watched flies passing from tuber-
culous sputum to food, and he attributes to these insects
a large proportion of tuberculosis in man. Many health
officials and anti-tuberculosis leagues have issued lurid
bulletins illustrating the agency of flies in transmitting the
germs of " the great white plague." While it is certainly
possible for the disease to be transmitted in this way, there
is absolutely no evidence that it is a factor of appreciable
importance. The opportunities for infection with tubercle
bacilli in other ways are so much greater and more numerous
that we are not justified in considering infection by flies of
much importance.

1 Darling, J. Infect. Dis., Chicago, 1911, VIII, 467.
« Cobb, Am. Med., Phila., 1905, IX, 475.

430 THE SOURCES AND MODES OF INFECTION

Cholera.

Due to Flies. — Cholera has of late been considered to be
to a considerable extent a fly-borne disease. This has been
the view of Chantemesse and Borel,1 as well as of many of
our officials in the Philippines,2 as Heiser,3 Woodruff 4 and
McLaughlin,5 though all of the latter seem to consider con-
tact infection a good deal more important than infection by
means of flies. Nuttall cites Flugge, Macrae, Buchanan
and Tsuzuki as considering flies of great importance in this
disease. Heiser 6 says that cholera was spread in Bilibid
prison in Manila by flies passing from the imperfectly dis-
infected latrines to the food. At mealtime they were seen
to pass in swarms to where the food was served. After
screening cholera ceased.

Dysentery.

Due to Flies. — In a recent outbreak of dysentery in an
asylum at Danvers, Mass., involving 156 cases, and carefully
studied by Ryder,7 flies were believed to be the chief factor
in the extension of the disease. In 1910 Orton and Dodd 8
reported an outbreak of 136 cases and 22 deaths, occurring
in July and August, which they believed must be due to flies,
as all frank cases were well isolated, screened and disin-
fected. They believed that contact infection from these
cases was out of the question, but the possibility of infec-
tion from carriers does not seem to have been eliminated.
The buildings were well screened, but many flies found their

1 Chantemesse and Borel, Bull. Acad. de me"d., Par., 1905, 3 s., LIV,
252.

2 Maus, Abst., Med. News, N. Y., 1902, LXXXI, 318.

3 Heiser, J. Am. M. Ass., Chicago, 1907, XLVIII, 856.

« Woodruff, J. Am. M. Ass., Chicago, 1905, XLV, 1160.

6 McLaughlin, J. Am. M. Ass., Chicago, 1909, LII, 1153.
• Heiser, BuU. State Bd. Health, N. Y., Sept., 1911.

7 Ryder, Boston M. & S. J., 1909, CLXI, 681.

s Orton and Dodd, Boston, M. & S. J., 1910, CLXIII, 863.

INFECTION BY INSECTS 431

way in nevertheless. Dodd 1 reports much the same con-
ditions, though with only 99 cases and 14 deaths, in 1911.
Dick2 states that dysentery, in which both Flexner and
Shiga types of bacilli were present, was endemic for several
years in an asylum at Dunning, 111. Flies were thought
to play an important part, and the disease disappeared
from one of the wards after screening and the practice
of strict medical asepsis. Hamer 3 refers to an outbreak of
dysentery in an asylum near London in 1906 alleged to
have been caused by flies.

Diarrhea.

Due to Flies. — Many of the English have been inclined
to attribute considerable influence to the fly as a factor in
the causation of the summer diarrheas of infants. News-
holme, who believes that diarrhea is due chiefly to infection
of the milk in the house, says that opened cans of condensed
milk are often seen to be black with flies, attracted by the
sugar in the milk, and attributes to them no inconsiderable
share in the causation of the disease. Sandilands,4 like
Newsholme, finds diarrhea especially prevalent among the
users of condensed milk, and he thinks that the milk is
probably infected by flies after the can is opened. Nuttall
cites Copeman as seeing in flies the possible cause of an
outbreak of diarrhea investigated by him. Nash5 in 1903
stated his belief that the house fly played the chief part
in the epidemiology of summer diarrhea. This also is the
view of Sandwith.6

Seasonal Distribution of Diarrhea and Flies. — Obser-
vations made in widely different localities have shown a

1 Dodd, Boston M. & S. J., 1912, CLXVI, 211.

2 Dick, J. Infect. Dis., Chicago, 1911, VIII, 386.

8 Rep. Med. Off. Health, Co. of Lond., 1907, Append. II.
4 Sandilands, J. Hyg., Cambridge, 1906, VI, 77.
6 Nash, J. San. Inst., Lond., 1903, XXVI, 495.
8 Sandwith, Clin. J. Lond., 1911, XXXIX, 19.

432 THE SOURCES AND MODES OF INFECTION

certain relation between the seasonal distribution of flies
and summer diarrhea. Stations are established at various
places in a town and flies are caught in traps or by means
of fly paper, and the number of flies is compared with the
number of deaths from diarrhea. Niven1 in Manchester,
in 1903, found that the " fly curve " and the curve for
diarrheal deaths corresponded very closely, and the same
agreement was noted in 1905 and 1906.2 The table given
by him shows both the date of inception and the date of
death of the fatal cases of diarrhea. In both years the max-
imum number of cases occurred at almost exactly the same
time as the maximum number of flies, and the maximum
number of deaths about a week later. In 1905 the maxi-
mum was about August 1, and in 1906 about September 5.
Robertson in Birmingham3 found a similar correspondence
in that city. Jackson in New York4 showed a similar close
relation between flies and diarrhea, as has Ainsworth 5 for
Poona in India. Nash,6 showing the seasonal distribution
of flies, states that in 1902, at Southend, there were few
flies and little diarrhea in August, and that diarrhea in-
creased in September as the flies increased. Both 1902 and
1903 had cool summers with few flies and little diarrhea,
while 1904 and 1906 had plenty of flies and plenty of
diarrhea. In 1904 there was a heavy local incidence near a
large dump where flies abounded. Hamer7 made a careful
study of the seasonal distribution of flies in London, and he
shows that while there is an apparent agreement between
the fly curve and that of diarrheal deaths, the latter begins
to fall before the former, which he thinks that it would not

1 Niven, Rep. on Health of Manchester, 1903, 123.

2 Rep. on Health of Manchester, 1906, 82.

3 Robertson, Rep. Med. Off. Health, Birmingham, 1910, 111.

4 Jackson, Rep. to Com. on Pollution [etc]., of the Merchants' Ass.
•of N. Y., 1907, 17.

6 Ainsworth, J. Roy. Army Med. Corps, Lond., 1909, XII, 485.

6 Nash, J. Hyg., Cambridge, 1909, IX, 141.

7 Hamer, Rep. Med, Off. Health, Co. of Lond., 1907, Append. II.

INFECTION BY INSECTS 433

be likely to do If flies were the chief cause of the disease.
Hamer1 in a later report shows that the relation between
the fly curve and the diarrhea curve varies considerably in
different years. He thinks that if flies have any part in
the causation of diarrhea it is much obscured by other
influences and that too great importance should not be
attached to correspondence in any one year. He also says
that the period of maximum infections (not dates of attacks)
in typhoid fever in both England and America is some three
weeks later than in diarrhea and that this is not reconcilable
with the theory that the seasonal distribution of both dis-
eases is determined solely by the distribution of the fly.

Maximum of Diarrhea before Maximum of Flies. — As
regards this phenomenon Peters2 has suggested that it is
exactly what might be expected. If flies are the chief
factor in the spread of diarrheal disease the diarrheal curve
ought to lag a little behind the fly curve in the early part of
the summer outbreak, for the increase in flies approaches
an ordinary arithmetical progression while the diarrheal
increase, he says, starting from an amount of infection
relatively small, but proceeding more by geometrical pro-
gression, later on rises faster than the fly curve and reaches
its maximum at about the same time. The fall of the
diarrhea curve before the fly curve is due, he believes, to the
exhaustion of infection, or of material to infect, and would
doubtless occur even if the fly prevalence continued with-
out a decline. A general agreement between the fly curve
and the diarrhea curve, but with a certain definite lack of
correlation, is, on theoretical grounds, to be expected, and
Peters states that just this correlation exists in each of the
eight seasonal charts of fly and diarrhea distribution pre-
pared by Niven and Hamer before referred to.

Problem still Unsolved. — Peters in this paper gives the
details of his exceptionally careful and valuable epidemio-

1 Hamer, Rep. Med. Of. Health, Co. of Lond., 1909, Append. IV.

2 Peters, J. Hyg., Cambridge, 1910, X, 602.

434 THE SOURCES AND MODES OF INFECTION

logical study of diarrheal disease in Mansfield. He found
that many of the local conditions were well explained by the
fly theory of transmission and that no facts were met with
which were directly contradictory, yet he arrives at no more
positive conclusion than " that the whole question merits
the most thorough and laborious investigation." He sug-
gests that a " positive experiment " in which " households
duly protected against all other chance of infection, and in
which flies from infected houses have been liberated, de-
velop abundant diarrhea " would afford convincing evi-
dence. Such an experiment he tried on two occasions
without positive results. He also suggests " negative
experiments " to determine whether screened houses in
the midst of infection remain free.

Parallel Seasonal Distribution not Conclusive. — Al-
though there is considerable evidence of a close parallelism
between the seasonal distribution of flies and of diarrhea,
this parallelism is no proof that the latter is dependent
upon the former. Even as corroborative evidence it has
little value. The return of the sun from the equator results
in a great variety of phenomena, many of which, though
closely related in time, have no causative relation one with
another. The presence of flies seems to be closely corre-
lated with the temperature, though it may be that the con-
nection is far from direct. So also are a vast number of
other events correlated with the temperature, though they
have no causative connection with flies. A few years ago it
was discovered that summer diarrhea developed only when
the temperature of the soil had reached a certain point, and
this was by many considered a demonstration that the
germs of the disease developed in the soil. Now it is the
fashion to say that diarrhea is due to flies because at times
the increase in flies precedes by a little the increase in
diarrhea. While it may be that flies have much to do with
the causation of infantile summer diarrhea, as yet we have
no proof of it.

INFECTION BY INSECTS 435

Local Distribution of Flies and Diarrhea. — Jackson in
his report has a map on which are located all the deaths
from intestinal disease. This map shows that the great
bulk of these are at no great distance from the water front,
where the flies swarm about the sewer outlets. It is just
there that the tenements crowded with poor children are
situated, and to maintain his argument it would be neces-
sary to show an excess over and above the excess of diarrhea
always found in tenements, and to eliminate the other
unfavorable factors associated with poverty.

A close relation between an excess of flies in a neighbor-
hood and an excess of diarrhea has not been found by
Davies in Woolwich, Dudfield in Kensington, Harris in
Islington and Porter in Finsbury.1 Hamer also calls atten-
tion to the fact that among English towns many which have
a very high typhoid death rate have a low diarrheal death
rate and vice versa. If both diseases are largely dependent
on flies this needs explanation. In our own country Fall
River, which, for the five years 1904-1908, had a death rate
from typhoid fever of 13.2 (less than half the average of
the registration cities), had a diarrheal death rate of 380
(over three times the average of the registration cities and
the highest of all of them).

Typhoid Fever.

Flies in Spanish War. — More attention has, however,
been bestowed on the relation of the fly to typhoid fever
than to any other disease. Sedgwick 2 was the first that I
know of to call attention to the importance of flies as a
means of spreading typhoid fever, but the excessive typhoid
death rate in the home camps of our soldiers during the
Spanish-American War did more than anything else to call
attention to the possible importance of the fly as a distrib-
uter of the germs of this disease. Surgeon General Stern-

i Rep. Med. Of., Co. of Lond., 1909, Append. IV, 7.
> Sedgwick, Rep. Bd. Health [etc.], Mass., 1892, 736.

436 THE SOURCES AND MODES OF INFECTION

berg had issued orders to guard against flies, but they had
little effect. Veeder * clearly set forth the possibility of the
transfer of fecal matter to food by means of flies, and he
declared that flies were the principal cause of the prevalence
of the disease in the camps. The report of the commission
to investigate the cause of the fever, although laying chief
stress on contact infection, forcibly emphasized the part
played by flies in the spread of the disease.2 Vaughan, a
member of the commission, personally urged the importance
of flies in the spread of this disease.3 Ever since, great
popular and scientific attention has been bestowed upon the
fly in the United States and also in other countries.

Other Reports of Typhoid Fever due to Flies. — Nuttall
cites Quill, Tooth and Calverly, Smith, Austen, Straton and
Jones as attributing much of the army typhoid in South
Africa and India to the presence of flies. Numerous writers
on typhoid fever in civil life have referred to the fly as a
most active agent in the transmission of this disease. In
" The House Fly at the Bar," published by the Merchants'
Association of New York in 1909, are printed opinions of
seventy or more health officials and others, all but nine or
ten of whom are emphatic in their statements that flies are
very important carriers of disease germs. Some few, how-
ever, consider that the case against the fly has not been
proved, though some circumstances are suspicious. Several
give instances of fly infection that have come to their notice.
Hurty writes of an outbreak in an asylum following the brief
use of the privy by a walking case of typhoid fever. Flies
abounded and there was no chance for contact. Taylor, of
Denver, reported typhoid fever at a dairy. The milk be-
came infected, he thinks by flies, for the privy was near the
milk house, and gelatine cultures exposed near the privy

1 Veeder, Med. Rec., N. Y., 1898, LIV, 429.

2 Abst. of Rep. on the Origin and Spread of Typhoid Fever in U. S.
Military Camps during the Spanish War, 1898, 183.

3 Vaughan, J. Am. M. Ass., Chicago, 1900, XXXIV, 1451, 1496.

INFECTION BY INSECTS 437

and in the milk house showed colonies of typhoid bacilli,
presumably deposited by flies! Dr. H. W. Hill, the exceed-
ingly cautious epidemiologist of the Minnesota State Board
of Health, writes me that from his observations in that state
he "firmly believes that flies are an important factor in
the spread of typhoid in such places as mining and lumber
camps, and that the small country village with its exposed
outdoor closets parallels the camp conditions exactly."
He further states that Wald noted that in a certain camp
the Italians did not contract the disease because, as he
thought, they did not eat between meals, while the Finns
suffered severely because they kept their food laid out on
the table all the time, where it was exposed to innumerable
flies, and helped themselves during the day at random.

Pease 1 refers to an outbreak of 50 cases at Castleton from
June to September all in the vicinity of a yard where typhoid
excreta were thrown and where flies abounded. Button2
accused bluebottle flies of carrying infection from the feces
of a carrier to 4 other members of the family. In 1906 an
outbreak of typhoid fever near London was thought by
Ransome and Young to be due to swarms of flies which
bred in a manure dump near by.3 In none of these alleged
instances of fly-borne infection are other modes excluded,
and often it is quite as probable that the cases were due to
contact, and in scarcely any of them is there more than a
possibility that the disease was caused in the manner
alleged.

Numerous bulletins and circulars have been issued by
state and municipal health officials, some of which depict,
in exaggerated language and with extravagant illustration,
the danger to be feared from this household pest, and a dis-
tinguished entomologist has urged that the familiar name
of M . domestica be changed from house fly to typhoid fly.

1 Long Island M. J., Brooklyn, 1910, IV.

2 Button, J. Am. M. Ass., Chicago, 1909, LIII, 1561.

3 Rep. Med. OS. to Lond. Co., Council, 1907, Append. III.

438 THE SOURCES AND MODES OF INFECTION

Flies suspected from their Habits. — The chief reason
why the fly is believed to be the carrier of diarrhea, cholera
and typhoid fever are that flies are seen to pass from feces
to food; that they have been shown to carry fecal bacteria,
and in several instances the specific bacteria of cholera and
typhoid fever; and that they often swarm in unusual num-
bers at times and in places where there is an exceptional
prevalence of disease. The very definite observations of
Hamilton, Fricker, Klein, Simmonds, Bartarelli Faichnie,
Mann and Tsuzuki as to the finding of infected flies in in-
fected houses, and the accounts given by Veeder, Vaughan,
Tooth, Smith, Hill and others, of the contamination of food
by flies in camps, military and civil, certainly render it
probable that these insects do at times cause sickness, and
may under certain conditions, such as prevail in camps, be
an important factor in outbreaks. But these observations
are far from a demonstration of what is now generally
claimed, that flies are the chief factor in the spread of
typhoid fever, and perhaps of summer diarrhea, in well-
ordered civil communities.

Dr. Howard's Views. — Howard in his admirable work
on the house fly, before cited, devotes 174 pages to the
" Carriage of Disease," a large part of which is taken up by
a consideration of typhoid fever; he quotes many of the
writers referred to in the last few pages of this book and also
some others as Wanhill in Bermuda, and Washburn in
Minnesota, who, noting the filthy habits of flies and their
seasonal distribution in relation to typhoid fever, are con-
vinced that flies are an important factor in the causation
of this disease. Very little convincing epidemiological evi-
dence is offered. Howard apparently bases his own view
on this matter upon the habits of the fly, for he says that
" the correlation or non-correlation of the curve of house
fly abundance and the abundance of typhoid has prac-
tically no effect upon our conclusions as regards the pos-
sible transfer of the disease by flies." Howard cites as

INFECTION BY INSECTS 439

significant the experience of Palmer in Georgia who offered
to care for, without charge, any typhoid patient living in
a fly-proof house, but none applied. He also found no
secondary cases in families which controlled flies as directed.
This does not appear to throw much light on the subject,
for, if there were no secondary cases, his families must have
been wonderfully well trained in avoiding contact infection;
and this is doubtless the complete explanation.

Seasonal Distribution of Flies and Typhoid Fever. —
As the seasonal distribution of flies has been studied with
reference to diarrhea, so has it been studied with reference
to typhoid fever. Jackson, in his report to the Merchants'
Association in New York before referred to, gives a chart
showing the seasonal distribution of flies in New York and
of deaths from typhoid fever, the latter being set back two
months to allow for the assumed time between the date of
infection and the date of death. Hamer thinks two months
is too long a time. He thinks that it is nearer five weeks.
The apex of the typhoid curve corresponds fairly well with
the apex of the fly curve, but there is a rise in the typhoid
curve in February, two-thirds as high as that of August,
which certainly cannot be attributed to flies. Ainsworth
in India 1 states that at Poona the maximum admissions of
typhoid, fever to the hospital occur about one month after
the maximum fly catch. Judging from Ainsworth 's paper,
there appears to be great seasonal variation in the number
of flies in Poona, correlated closely with the temperature,
as with us.

In Jacksonville2 it is said that the fly curve and the
typhoid-fever curve correspond closely, though Dr. Terry
gives no figures or diagrams. According to Howard a
similar correlation has been worked out by Purdy in New
Zealand (1910) and Osmond in Cincinnati (1909).

1 Ainsworth, J. Roy. Army Med. Corps, Lond., 1909, XII, 485.
* Am. J. Pub. Health, 1912, II, 14.

440 THE SOURCES AND MODES OF INFECTION

Evidence against Theory. — Niven 1 found that, while in
Manchester in 1903 the maximum number of flies was
caught about August 29, the maximum number of cases of
typhoid was not reached until November, during the whole
of which month the morbidity remained high. In 1906 2
there was, as usual, a well-marked maximum of flies about
September 1, and while the typhoid morbidity was high in
September, being 65, it was somewhat higher in October,
and there were 46 cases in November and 50 in January
and only 21 in July and 23 in August. In Washington,3
in 1908, the maximum fly catch was for the week ending
June 24, after which it steadily and rapidly decreased.
The maximum of typhoid cases, according to date of at-
tack, was in the week ending July 22, but instead of falling
rapidly, as did the fly catch, it continued high until the
middle of September. There certainly is little parallelism
between the two curves. The commissioners appointed to
study the disease, and who make the report, say that the
evidence is quite strong that flies play a relatively small part
in the spread of typhoid fever in Washington.

In Providence the seasonal distribution of typhoid fever
cases, according to date of first symptoms, for the six years
1904-1909, was as follows. All cases due to milk and all
certainly contracted out of the city are excluded.

Jan. Feb. Mar. Apr. May June
45 36 26 48 49 56

July Aug. Sept. Oct. Nov. Dec.
80 161 133 117 96 59

According to observations of Sykes, the maximum fly catch
out of doors was about the last of July, when it began to fall

1 Niven, Rep. on Health of Manchester, 1903, 123.

s Rep. on Health of Manchester, 1906, 63, 84.

8 U. S. Pub. Health & Mar. Hos. Serv., Hyg. Lab. Bull., No. 52, 30.

INFECTION BY INSECTS 441

off quite rapidly. The observations as to the indoor catch
were not very numerous, but it seems probable that the flies
go indoors in greater numbers as the weather grows cooler in
September. There is certainly not a very close agreement
between the number of flies and the prevalence of typhoid
fever. It is scarcely possible that the typhoid fever develop-
ing in November is due to flies, and yet there were 96 cases
in November, and the November rate is higher than that of
any other month except August, September and October.
If we cannot attribute the excess in November to flies, why
should it be thought necessary to attribute the excess of
the preceding three months to flies? I can see nothing in
the seasonal distribution of typhoid fever in Providence to
warrant the assumption that flies are an important factor
in its causation.

In London,1 while the typhoid curve and the fly curve
corresponded fairly well in 1907, in 1908 the greatest preva-
lence of flies was in the second week in August, while the
maximum of typhoid fever notifications was in the second
week in November. In 1909 there was only a slight autum-
nal rise of typhoid fever, the maximum being in January,
while the maximum fly prevalence was in August.

Correlation of Typhoid Fever and High Temperature. —
I have been able to find only the very few observations noted
above as to the correlation of fly distribution and typhoid
fever, and these few do not indicate that there is any reason
to assume that the flies stand in causative relation to the
fever. The very careful work of Sedgwick and Winslow2
shows that throughout the world, and in both hemispheres,
there is a distinct relation between this disease and the sea-
sonal rise in temperature. These authors, though they
recognized a certain amount of infection by flies, explained
the autumnal increase in typhoid fever as due to the direct

1 Rep. Med. Off. Health, Co. of Lond., 1907, 1908, 1909.

2 Sedgwick and Winslow, Mem. Am. Acad. Arts & So., 1902, XII,
No. 5.

442 THE SOURCES AND MODES OF INFECTION

and favorable action of high temperature on the life of the
bacillus outside the body. This view hardly seems tenable
from what has since been learned of the life habits of the
bacillus, and I think is now no longer held by the authors.
It is the custom now for most writers to attribute to flies
the chief role in the autumnal excess of typhoid fever, but
from the evidence at hand it seems wiser, while admitting
the fact of seasonal distribution, as determined by Sedg-
wick and Winslow, to await further study before attributing
this definitely to any one cause or group of causes.

Epidemiological Evidence, Minnesota. — More direct
evidence of fly carriage is offered by Hill.1 He reports three
large insane hospitals, with 1400 to 1800 patients each,
having outbreaks of typhoid fever at about the same time
in the summer. In one, the wards of the institution, though
screened, were full of flies, which followed the food from the
kitchen, which was unscreened and swarming with the in-
sects. In a second hospital the disease was confined almost
exclusively to the men, whose dining rooms, and wards too,
were swarming with flies, while the women's side was com-
paratively free. In both these institutions the outbreak
continued until cold weather. In another hospital the flies
were so numerous that when they settled on the slender
wire supporting the electric lamps they gave it the appear-
ance of an inch rod! The management determined to
exterminate the flies, which they did, and though it was
two months before cold weather, no more cases developed
after the period of incubation had passed. Such a bit of
evidence is not conclusive, but, if observed often enough,
would be cumulative and would indicate a real causative
relation between typhoid fever and flies as can no amount
of watching the habits of flies or comparing fly curves and
typhoid curves.

Evidence from Jacksonville, Richmond and Asheville. —
The health officers of the two cities first named have very
1 Hill, Rep. State Bd. Health, Minn., 1911, 206.

INFECTION BY INSECTS 443

forcibly called attention to the much greater liability of
infection by flies in the comparatively more " unsanitary "
cities of the south and have criticized the present writer for
an apparent disregard of this, and it is true it must be ad-
mitted that a large part of what is here written about the
comparative unimportance of nuisances in general and the
fly nuisance in particular is based upon conditions studied
in the cities in the northern states and in England. Ac-
cording to Terry,1 Jacksonville, with its large numbers of
poorly made privies and enormous swarms of flies, has in
the past been no better than an old-time army camp as
regards conditions favoring the spread of typhoid fever.
Late in 1910 a most energetic campaign was begun against
the fly and for the construction of fly-proof privies. In 1911
there were reported only 142 cases of typhoid fever as
against 321 for the same period of 1910. Formerly 60 per
cent of the cases had been in the privy section and 40 per
cent in the sewered section, though the distribution of the
population was the reverse of these figures. After the fly
campaign there was little difference in the incidence of the
disease in the sewered and unsewered portions. The high
typhoid rate of several Georgia cities showed no falling off
in 1911. It would have been well if Terry had given the
typhoid fever figures for a number of years and had also
shown the seasonal distribution so that it might be seen
whether the decrease was confined chiefly to the fly season.
Still the figures as they stand are striking and call for con-
tinued effort along the same lines. Just as this is going to
press the Report of the Board of Health of Jacksonville for
1911 has come to hand and on page 16 is a diagram which
shows that reduction in typhoid fever incidence in 1911 was
confined to the May- August period, the time of fly prevalence.
In Richmond, Levy2 has been carrying on an energetic
and well-directed campaign against typhoid fever since

* Terry, Am. J. Pub. Health, 1912, II, 14.
1 Levy, Rep. Health Dept., Richmond, 1910.

444 THE SOURCES AND MODES OF INFECTION

1907, with the result that the death rate per 100,000 fell
from nearly 50 in the preceding five years to 24.1 in 1909,
21.9 in 1910 and 17.8 in 1911. In 1910 strong effort was
made to prevent fly infection by screening privies and
patients, and by reducing the number of flies. In a letter
Dr. Levy states that the number of flies is still large but
that the screening has been very effective. I judge that he
attributes a good deal of the typhoid reduction to this, but
it is difficult to disentangle this factor from supervision of
cases, instruction of attendants, disinfection of excreta and
better notification. It is worth noting that the decrease
does not seem to be greater for the summer months than for
the winter when flies can scarcely be a factor.

In Asheville, N. C., so Dr. McBrayer writes me, there
was a reduction in typhoid fever cases from 119 in 1910
to 60 in 1911, due chiefly, he believes, to strict control of
privies which are now required to be made fly-proof.

Privies, Flies and Typhoid Fever. — That an excess of
typhoid fever is found in cities or sections of cities where
there are many open privies has often been noted. Of late
some have attributed this to the transference of infected
matter by flies. There is little warrant for this assumption,
though Terry says that in Jacksonville after the screening
of the privies there was no excess of typhoid fever in the
unsewered parts of the city. But meanwhile an active
campaign against typhoid fever was being waged, and more
cleanly privies and better care of cases would do much to
prevent contact infection. That the excess of typhoid
fever in privy districts is doubtless largely due to contact
is shown by the fact that, as has recently been noted in
Providence, localized outbreaks among careless people using
privies have occurred out of fly time. In Washington 1
investigation has shown that in the fly season the users of
privies furnish a higher percentage of typhoid fever as com-

1 U. S. Pub. Health and Mar. Hosp. SerV. Lab., Bull. 78, 1911, 66,
160.

INFECTION BY INSECTS 445

pared with the users of water-closets than they do in the
winter time. The percentage of cases furnished by the
users of privies in fly time was 13.3, in the winter 7.7, but
the figures were not large. It was also found in Washington
that during 1908 and 1909 the 32 per cent of the population
living in unscreened houses furnished about 59 per cent of
the cases occurring in the summer and fall. But, as the
report suggests, people living in unscreened houses are more
likely to have other unsanitary surroundings and to be of
careless habits.

Sykes' investigations in Providence showed that flies are
25 to 30 times more numerous in the uncleanly parts of the
city than they are in the clean parts, and the differences in
the indoor catch is even greater. Though, as shown above,
there appears to be some excess of typhoid fever in the fly-
infected and privy-using parts of cities, the difference bears
no comparison to the difference in fly distribution.

Unwise to Claim Too Much. — While the fly is a nuisance
and it is highly desirable to get rid of it, and while it seems
likely that it is to some extent a means of spreading disease,
it is extremely unwise to make definite statements that it is
the chief source of diarrhea, of cholera or of typhoid fever,
unless we have very exact proof, and it is unwise, unless
such proof is at hand, to urge large expenditures to get rid
of flies, promising thereby the eradication of typhoid fever
and other diseases. If it should chance that a mistake has
been made, and that the fly is not the chief disseminator of
typhoid fever, and the disease still persists after the fly
has disappeared, we need not be surprised if the public fail
to take us seriously when we advise, on knowledge that is
well established, that mosquitoes are the sole carriers of
malaria and yellow fever; that escape from the tsetse fly
means escape from sleeping sickness; and that a rat-proof
city will be a plague-proof city.

Control of Flies. — As flies are a great nuisance and a
possible source of danger it is very desirable that they

446 THE SOURCES AND MODES OF INFECTION

should be eliminated as far as possible, but too much money
should not be expended in the experiment and too large
results should not be promised. Each individual family
can very effectually control these pests by good screening
and the use of fly paper. It is desirable that people should
be educated to dislike flies. Chantemesse says that the
housewife should think it as much of a discredit to have
flies in her house as bugs in her bed. Circulars of informa-
tion may be distributed, but care should be taken to avoid
exaggeration and not to promise too much either as to the
effectiveness of remedies or the resulting decrease in disease.
We know even now far too little about the habits of flies
and the best ways of attacking them. Packard and How-
ard gave us our first definite knowledge, but this has been
much extended by Newstead,1 who showed that the fly is
far less exclusive in its choice of breeding places than was
supposed.

Howard in his " The House Fly — Disease Carrier " gives
a summary of our present knowledge which contains much
information. The recent papers by Terry and by Dodd
previously cited recite interesting and suggestive experi-
ences. It appears that effective scavenging is the most
important means of getting rid of flies. If yards, streets
and vacant lots were kept clean, market refuse removed
promptly, and all garbage kept covered, there would be an
enormous reduction in the number of flies about dwellings.
The most practicable way to prevent the breeding of flies in
stable manure is to compel its removal once a week. Wher-
ever it is possible the privy vault should be abolished.
When this is impossible, the fly-proof privy has been urged
and the Federal Government has issued a detailed account
of how one may be economically constructed.2 This

1 Newstead, Rep. on the Habits, etc., of the House Fly, to the Health
Committee, Liverpool, 1907.

1 Stiles and Lumsden, U. S. Dept. Agriculture, Farmers Bull. 463,
1911.

INFECTION BY INSECTS 447

has seemed somewhat visionary, but the energetic work
done by the health officers of Richmond, Jacksonville
and Asheville show that by means of constant supervision
it is possible to make such privies effective. The question
of covering food in markets and shops to protect it from
flies as well as from dust has been much discussed. Slack l
in an excellent paper concludes that the danger from this
source in a well-ordered city is not very great, but that for
aesthetic reasons the public might well demand, through
ordinances or otherwise, that food be so protected. In this
opinion the writer heartily concurs.

Summary. — After this brief examination of the evidence
in regard to the role of insects in the transmission of disease
we are justified in the following conclusions:

First. It is certain that yellow fever and malaria are
transmitted solely by certain mosquitoes, and that by con-
trolling the mosquitoes it is possible, even under very un-
favorable conditions, to eradicate or reduce to a minimum
these two diseases.

Second. It is highly probable that the relapsing fevers
are transmitted solely by certain ticks, sleeping sickness by
the tsetse fly, filariasis by the mosquito, pappataci fever by
a fly and typhus fever by lice.

Third. The bubonic type of plague in human beings
is usually transmitted from rat to man by the flea.

Fourth. It is probable that under certain conditions, as
in military and civil camps and in filthy communities with-
out sewerage, insects, especially flies, may be an important
factor in the spread of the fecal-borne diseases, but there is
no evidence that in a well-sewered city with few privies the
house fly is a factor of great moment in the dissemination
of disease.

1 Slack, Am. J. Pub. Hyg., 1909, V, 159.

INDEX

Actinomycosis 366

African cattle fever 122

Air, see Sewer air.

bacillus influenzae not carried by 99-100

tuberculosis in 298-299

bacteria found in 282, 298-301

danger of infection by 300

expired, free from germs 287

infection by, thought important 137

conclusions 313-315

experiments with anthrax 312

Mediterranean fever 311-312

plague 281, 312-313

not impossible 281

of wounds by 300-301

pneumococcus in 300

pus-forming bacteria in 299-300

swine plague bacilli in 300

Air-borne anthrax 285-286

chicken pox 270, 280

disease in hospitals 270-271, 279-280

slight evidence of . . '. 286

infection indoors 282

in surgery 283-284

reasons for belief in 259-260, 283

influenza is not 277

malaria is not 384

measles 272-280

rubella 280

scarlet fever 270-273, 280

out of doors 272-273

small-pox, cases traced to other sources 268-269

conclusions 269-270

evidence against 266-269

of, in United States 264-265

from hospitals 260-270

infection of, denied 263-270

objections to theory 265-266

tuberculosis 194, 202-203

typhus fever 281

whooping cough 280

Amebae dysenteriae, carriers of 118

grown on culture media 24

in soil 24

in well persons 118

on vegetables 24

449

450 INDEX

Amebae, intestinal parasites 112

species of 24

Amebic dysentery 24, 118

American hospitals 142-145, 279

Anchylostoma, see Uncinaria, Hookworms.

duodenalis 186

Anchylostomiasis, due chiefly to contact 186-187

Animals, diseases of, affecting man 371-373

spread tuberculosis by licking one another 204

Anopheles mosquitoes carry malaria 382-383

species and habits 385-386

Anthracosis 306-307

Anthrax, a dust-borne disease 285-286

an air-borne disease 285-286

and fomites 226-227

insects 416-417, 423

bacillus, see Bacillus anthracis.

due to food infection 2, 365

experiments in infection by air 312

from hair, hides 227

morocco factories 3

wool 227

wool refuse

perpetuated by direct contact

soil in relation to 2

spores in dust, hair, hides, wool 226-227

spread by unrecognized or concealed cases

Antisepsie medicale in French hospitals 198, 278-279

Antitoxin may favor spread of disease 386-387

Ants carry bacillus murium typhi 419

plague bacilli 419, 424

Aspergillus destroys mosquitoes 390

Asylums, bacillary dysentery in 183-184, 430-431

percentage of bacillus diphtherias carriers in 90

typhoid fever in 51-52, 442

Atoxyl to prevent sleeping sickness 402

Atypical cases, laboratory evidence of 123

not easily distinguished from carriers 130-131

Babies' Hospital, gonorrhea in 165-166

Bacillary dysentery, carriers of 67-69

caused by contact 183-184

culture of bacillus in eye 183-184

in camps 183

in institutions 183-184

mild cases of 68-69

Bacillus anthracis, see Anthrax spores.

carried by flies 419, 423

growth in ponds

relation to the soil 2, 4

saprophytic existence of

Bacillus coli communis cause of diarrhea 367

in dust of schoolrooms 294

on the hands 179, 192

on roller towel 192

INDEX 451

PAGE

Bacillus diphtherias carried by flies . 423

carriers in family 89-90

insane asylum 90

schools 90-94

investigation of, in Massachusetts. . . 83

of 82-99

found intermittently 98

in bread 26

carriers, varieties of 83

virulence of .82, 84-86, 91-92

clay 233

convalescents 82, 84-85, 98-99

dust 290,294

holy water, on cups, pencils, drinking glass . 192

membrane 240

milk 25,26

otitis 94-95

pupils cause disease in teacher 98

rhinitis 94-95

rooms 233, 241

scarlet fever 86-88, 91, 96

school children 82-86, 88, 90-94

soil 26

sore throat 93-^94, 98-99

well persons 82-83

life of, in butter 251

may be air-borne 289

not in normal throats 84-85

on coat 227

drinking glass 192

fomites 192-193, 227, 240-241

virulence of 240

pencils 193, 233

resistance to drying 233, 290

saprophytic existence of 25

Bacillus dysenteriae, carried by flies 423

cause of diarrhea 367

in carriers 67-69

f eces 24

milk 15

soil 23

resistance to drying 234

saprophytic existence of 23

Bacillus enteritidis, cause of diarrhea 367

(Gaertner) in ice cream 250

in food poisoning 58, 371

sporogenes, cause of diarrhea 367

Bacillus influenzae in droplets 298

not carried by air 99-100

saprophytic 26

persistence of 100

resistance to drying 238

widely distributed 100-101

Bacillus leprae in droplets 297

452 INDEX

PAGE

Bacillus morbificans in food poisoning 371

Bacillus paratyphi, cause of diarrhea 367

in feces 16

food poisoning 371

not found in healthy persons 59-61

Bacillus paratyphi A in carriers 58

Bacillus paratyphi B in carriers 59

Bacillus pestis carried by ants 419-424

in bodies of fleas. 236

convalescent animals 69-70

convalescents 70

cow dung 23

dust 290-291

sputum 71

infection of soil by. 22

not found on floors 235

on cotton goods 228

resistance to drying 235-236, 290-291

saprophytic existence of 22

Bacillus prodigiosus on flies 425

in droplets 295-297

sewer air 284

Bacillus pullorum, carriers of 71

Bacillus tetani in blank cartridges 6, 226

dirt of floors 239

gelatine 6, 284

intestines of animals 5

soil ' 6,239

latency of 106

on lamp wick 6, 226

resistance of 6, 225-226

saprophytic existence of 6

Bacillus tuberculosis, carriers of 103-105

in air 298-299

butter. 364-365

communion cups 195

droplets 295-296

dust 292-293

flies, and fly specks 422^23

lymphatic glands 203

milk 355-356

nose and mouth 103-104, 193, 203

railway carriages 292

room 235, 292

tonsils 203-204

tramcars 292

latency of 104-105

may be dust-borne 289-290

pass from stomach to lungs 305-308

not found in street dust 292

in expired air 287

on spirometer 195

saprophytic 26

so resistant as believed 234-235

INDEX 453

Bacillus tuberculosis, on a glass 195

dishes 195

fomites 194-195, 241-242, 292, 293

hands 193-194

napkin rings 195

sidewalks 194

telephones 195

tongue 203

perishes in light 234-235

proportion of human and bovine types. . . 360-361

resistance to drying 234-235, 241-242, 291

swept from sidewalk by dresses 194

Bacillus typhi carried on fingers 169-170

flies and 420-422

in blood 34-35

bones 35

butter 16

cerebro-spinal fluid 35

dead animals 8

dust 290

feces 9, 34-36

gall bladder 35, 42

ice 13, 319, 340-341

kidneys 35

milk 15

ovaries 35

oysters '. 13

privy vaults 9-10

- prostatic fluid 37

seltzer water 13

seminal vesicles 37

septic tank effluent 10

sewer air 284

soft drinks 13

soil 7, 10

epidemiological evidence of growth 16

spleen 35

sputum 37-38

tidal mud 8

tonsils 34

urine 37

persistence of 10

water 10-13

epidemiological evidence of growth 17

modes of contamination 320

recovered from 319

increase in milk 17

intermittent excretion of 46-47

moisture necessary to growth of 16

on blankets 227

vegetables 8

path of entrance 34, 178

resistance to drying 232-233, 290

saprophytic existence of 7, 16

454 INDEX

PAGE

Bacteria and sewer air 284-285, 288-289

cultured less resistant than uncultured 11

effect of drying upon 231-232

found in air 298-301

inhalation of 304r-305

not found in expired air 287

given off from moist surfaces 287, 288

of suppuration, see Pus bacteria.

on cups 192-193

f omites 239-242

hands 192-194

money 223-224

pencils 192-193

Bagging, yellow fever from 213

Balantidium 112

Ballast, yellow fever from 213

Bathing, typhoid fever from 324-325

Baths, public, encourage cleanliness 211

Bed clothing, bacillus diphtherias on 240

Bed, smallpox from 214

Bedbugs and anthrax 417

kala-azar 402

plague 414

relapsing fever 411-412

typhoid fever 418

Bedpan carries dysentery amebae •. . . . 183

Betel nut, cholera spread by 185

Biological carriers distinguished from mechanical carriers 413

Black-leg 4

Blank cartridges, bacillus tetani in 6

Blankets, bacillus typhi on 227

Blood, bacillus typhi in 34r-35

infection of, cause of typhoid fever 34-35

meningococcus in 73

yellow fever virus in 393-394

Boer War, typhoid fever from dust in 275-276

flies in 436

Books, bacillus tuberculosis on 241

scarlet fever from 215

Boophilus annulatus and cattle fever 112

Botryomycosis 366

Bovine tuberculosis bacillus may infect man 353-354

Bread, bacillus diphtherias in 26

Brill's disease 409

Broad Street well 316-317, 325

Brush, bacillus diphtherias on 240

Bubonic plague, see Plague.

Butter, a source of diphtheria 350

bacillus tuberculosis in 364-365

typhi in 16

life of bacillus diphtherias in 251

Calliphora erythrocephala 428

Camps, bacillary dysentery in . . . . „ 183

INDEX 455

Camps, filthy condition of 180

typhoid fever in 170-171, 435-437

Caps and gowns to prevent infection 218

Carbolic acid in well water 324

Carriers a recent discovery 33, 38

source of protozoan disease 126

and mild cases, number of 124-125, 133-135

cause cattle fever 112

cerebro-spinal meningitis 74-81

cholera .65-67, 185

diphtheria 94-99

glanders 99

malaria 114-115

nagana 113

typhoid fever 47-54

definition of and classification of 38-39

during incubation 39

evidence against 55

of infectivity 127-131

explain spread of cerebro-spinal meningitis 78-81

in bacterial diseases 127

in institutions 50-52, 65, 67

infect food 47-51, 60, 65

milk 52-54

intermittent excretion in 46, 62

laboratory evidence of 123

less infective than the sick 155

more dangerous than things 225

not always dangerous 129

of amebic dysentery 118, 183

bacillary dysentery 67-69

bacillus pullorum 71

cattle fever 112

cerebro-spinal meningitis 72-81

cholera 61-67

diphtheria, importance of 149-152

dourine 113

glanders 99

gonorrhea 102

influenza 99

lepra bacilli 105

malaria 114-115

measles Ill

Mediterranean fever 72

nagana 113

paratyphoid fever 58-61

plague 69-71

pneumococcus 101-102

poliomyelitis ', 118-121

pus organisms 105-106

scarlet fever 109

sleeping sickness 115-116

smallpox 109-110

tetanus bacilli 106

456 INDEX

PAGE

Carriers of tubercle bacilli 103-105

typhoid fever : 38-55

among different classes 39-44

cannot be isolated 152

duration of infection 45

Vincent's angina 117

relation to inactivity 125

shade into missed cases and atypical cases 130-131

virulence of germs in 126

Cartridges, bacillus tetani in 226

Cattle fever 112, 122

Celery, typhoid fever caused by 379

Cerebro-spinal meningitis an accident of infection 81

carriers of 73-81

explain spread of 78-81

contagiousness of 79-80

due to pneumococcus 81

infection by carriers 74-81

isolation a failure in 81, 146

path of infection 73

prompt isolation in, a success 146

secondary cases 79-80

Charbon symptomatique 4

Charts, bacillus tuberculosis on 241-242

Chicken pox air-borne 270, 280

Chicks, white diarrhea of 71

Children not cleanly 190

Cholera and contact 184-185

flies 420, 430

fomites 215

atypical cases 65

caused by washing soiled linen in running streams 325

from betel nut 185

carriers 65-67, 185

clothing 215

handling food 185

rags 222

shellfish 374

soil 2&

unrecognized cases 20, 65

water 325-326

wells 325-326

infection of nurses 185

not dust-borne 289-290

on shipboard due to infected water 326

outbreaks in various places 20

spirillum, see Spirillum cholerae.

Christmas presents, diphtheria from 215

Cimex lectularius, see Bedbug.

Citellus beecheyi 413

Clams, typhoid fever from 375

Clay, bacillus diphtherias in 233

Cleanliness, education in, needed 208-209

municipality should encourage 210-211

INDEX 457

PAGE

Cleanliness, neglected 206-208

rare 178-180

in children 190

teaching of, in school 209-210

versus disinfection 257

Cloth, yellow fever from 213

Clothing as fomites 216

bacillus tuberculosis on 242

cholera from 215

infection by, rare 217-219

leprosy in 216

scarlet fever not from 217

typhoid fever from 219

typhus fever not from 216-217

Coat, bacillus diphtheriae on 227

scarlet fever from 215

Cockles, typhoid fever from 375-376

Cockroaches and disease 418

Colon bacillus, see Bacillus coli communis.

Communion cups, bacillus tuberculosis in 195

individual, adopted 210

Conductors not infected by money 224

Connorrhinus 403, 404

Contact by drinking cups 189-190, 192-193

chief mode of infection 195-206

indirect, vehicles for 188-190

infection and fingers 188-189

disregard of 207-208

importance of 206

in amebic dysentery 183

bacillary dysentery 183-184

cerebro-spinal meningitis 73-74, 76-77, 80

cholera 184-185

diarrhea 185

diphtheria 196-202

gonorrhea 165-167

influenza 277

measles 199

Mediterranean fever 312

scarlet fever 196-202

syphilis 164, 167-169

tuberculosis 204-205

typhoid fever 17, 47-50, 169-183

amount of 177-178

evidences of •. 177

in civil life . 171-173

in hospitals 198-202

opportunities for 179-180

role of privies 180-182

Spanish war 170

why disregarded 179-181

uncinariasis 186

less easy in some diseases than in others .... 187-188
most obvious 164

458 INDEX

PAGE

Contact infection, opportunities for 190-192

mode of infection between families 196-198

with missed cases, importance of 190-192

well carriers, importance of 191

Contacts, diphtheria carriers among 89

typhoid carriers among 43

Contagiousness, factors involved 153-155

less outside family 154-155

than believed 153-155

of carriers less than that of sick 155

Convalescents, cholera spirilla in 61-65

diphtheria bacilli in 84, 85

dysentery amebse in 118

bacilli in 68-69

malarial parasites in 114-115

meningococci in 74-77

plague bacilli in 70-71

pneumococci in 101

typhoid bacilli in 40-41

Cornet, diphtheria from 215

Cotton goods, bacillus pestis on 228

smallpox from 214

Cow dung, bacillus pestis in 23

Cows, diphtheria in 347

scarlet fever said to occur in 346-347

Crawfish and typhoid fever ' 378

Cream, typhoid fever from 349

Cubicles for isolation in French hospitals 198-200

Culex, experiments with 383

f asciatus and yellow fever 392

fatigans and dengue 410

filariasis 399

Culicides 388-389

mosquito destruction by 390, 397

Culture method of diagnosis, introduction of 137-138

Cups, bacteria on 192-193

Danysz rat virus carried by flies 424

Dead animals, bacillus typhi in 8

bodies, spirillum cholerae in 19

Dengue transmitted by mosquitoes 409

Dermacenter andersoni 412

Diarrhea, bacteria which cause 367

due to condensed milk 368-369

contact 185, 367

dust 276-277

flies 431-435

milk 366-369

not always 185

water 327-329, 331, 334

explosive outbreaks 369

nature of 366-367

prevention of 369-370

INDEX 459

PAGE

Diarrhea, relation of feeding to 367-369

white, of chicks 71

Diphtheria, atypical 92-94, 133

bacillus, see Bacillus diphtherise.

carried to homes by discharged scarlet fever cases ... 98

carriers, importance of 149-152

isolation of, often impossible 150-152

chronic , 99

cultures introduced 137-138

value of findings 84r-85

disinfection unnecessary after 247-250

extension in dwellings 197-198

family infection 89-90

from butter 350

carriers 95-99

Christmas presents 215

comet 215

drinking glass 192-193

fomites 215-227

milk 96-97, 346-347

pitcher. 97

sewer air 273

teacher 97

tools 215

in cows 347

Owatonna, Minn 144

Providence 136, 138

Willard State Hospital 144

isolation, duration in Providence 143

in institutions, failure of 143-145

principles of 149-152

mild cases found by cultures 138

not air-borne in hospitals 273-274

dust-borne 290

from soil 28

of extra-corporal origin 29

similarity to scarlet fever. 106-107

warning sign in 151

with recurrent ear discharge causes infection 98

Disease rarely carried by physicians 217

Dishes, bacillus tuberculosis from 195

Disinfection, American views on 256

desirable at times 256-257

English views on 255-256

French views on 255

in Providence 137

abandoned 247-252

schools 254

objections to 257

of little value 247-256

unnecessary after diphtheria 247-250

in other diseases 252-254

scarlet fever 250-252

versus cleanliness 257

460 INDEX

PAQB

Dog drinks from the drinking glass on train 190

Dourine, carriers of , 113

Drainage, mosquito destruction by 390

Dresses, bacillus tuberculosis on 194

Drinking cups, contact infection by 189-190

glass, bacillus diphtheriae on 192

common, abolished 210

Droplet infection 295-298

Droplets, bacillus influenzas in 298

leprse in 297

prodigiosus in 295-297

tuberculosis in 295-296

pneumococcus in 298

streptococcus salivarius in 297

versus dust 303-305

Drying, effect upon bacteria. 231-232, 288-289

resistance of bacillus diphtherias to 233, 290

dysenteriae to 234

influenzae to 238

pestis to 235-236, 290-291

tuberculosis to .... 234-235, 241-242, 291

typhi to .- 232-233, 290

gonqcoccus to 239

meningococcus to 238-239

micrococcus of Mediterranean fever to 233

pneumococcus to 238

protozoa to 242

pus-forming bacteria to 236-237

smallpox virus to 244

spirillum choleras to 237-238, 242, 291-292

spirochete of syphilis to 239

spores to 232

vaccine virus to 242-244

Dust, anthrax spores in 227

bacillus coli communis in 294

diphtheriae in 290, 294

pestis in 290-293

tuberculosis in 292-293

typhi in 290

danger from slight 294-295

infection 288

meningococcus in 289-290, 293

micrococcus melitensis in 21, 300

pneumococcus in 293-294

pus organisms in 294

streptococcus in 282

versus droplets 303-305

Dust-borne disease 288-290

anthrax 285-286

diarrhea 276-277

influenza -. 277

Mediterranean fever 312

poliomyelitis 278

tuberculosis 288, 291-293, 301-311

INDEX 461

PAGE

Dust-borne disease, typhoid fever 275-276

Dwellings, extension of diphtheria in 197-198

scarlet fever extension in 196-197

Dysentery, see Bacillary dysentery.

amebic 118, 183

due to chronic carriers 183

infection from bedpan 183

produced in men by amebae grown in culture . 24

monkeys by amebse 24

transmitted by contact 183

bacillus, see Bacillus dysenteriae.

due to flies 430-431

infection by water 326-329, 331

Ear discharge causes diphtheria 98

Elephantiasis 399

Endemic diseases 29-30

English hospitals 139-142

isolation in 200-202

Entameba coli and entameba histolytica, distinction between. . .24, 118
Erysipelas from rags 222

Family, carriers in, cerebro-spinal meningitis 75

diphtheria 89

infection of, in house 28, 196-198

isolation in 160-161

typhoid fever by contact in 176

Feces, bacillus dysenteriae in 24

paratyphi in 16

typhi in. 9, 35-36

persistence of 10

intermittency of 46-47

hookworms in 187

spirillum choleras in 19

Filaria bancrofti 398

Filariasis and mosquitoes 398-399

Filth theory of disease 27

Filtration of water 338

reduces typhoid fever 321

Fingers and contact infection 188-189

Fish, infection by fried 278

Fishermen, influenza among 277

Fleas and anthrax 417

plague 23, 312-313, 413^16

bacillus pestis in 236, 414-415

Flies a nuisance 445-446

and anthrax 419

cholera 430

diarrhea 431-435

dysentery 430

Egyptian ophthalmia 419

murrina 429

plague 419, 424

tuberculosis . . 429

462 INDEX

Flies and typhoid fever 169-171, 182-183, 435-445

epidemiological evidence 442-445

evidence against theory • . . . . 440-441

reports of outbreaks 435-439

statistical evidence 439-445

bacillus tuberculosis in 422-423

carry bacillus anthracis 419, 423

diphtherise 423

dysenterise 423

prodigiosus 425

typhi 420-422

Danysz rat virus 424

fecal bacteria 425-427

germs 419-437

gonococci 424

lime from privies 425

poliomyelitis 427

spirillum cholerse 420

control of 445-447

difficult to infect by 425

habits bring them in contact with excrement 438

infected, in typhoid houses 421-422

local variation in distribution 435, 445

mechanism of infection 424

range of flight 427

reduction of and decrease of typhoid fever 442-444

seasonal and local distribution 427

species in houses 428

tuberculosis due to 429

Flock cough from rags 222

Floors, bacillus pestis not on 235

tetani in dirt of 239

Fly specks, bacillus tuberculosis in 422-423

Fomites: Bagging 213

Ballast 213

Barbers' utensils 168

Barracks 76, 216

Bedding 168, 214

Blankets 227

Books 215, 241

Brush 240

Cartridges 6, 226

Charts 241, 242

Christmas presents 215

Clay 233

Cloth 213

Clothing 169, 215-216, 219, 242

Coat 215, 227

Communion cups 195

Cornet 215

Cotton 214

Cotton goods 228

Cups 192-193

Dentists' tools 168

INDEX 463

Fomites: Drinking glass 168, 189-190, 192-193, 195, 240

Floors 235, 239

Furniture 241

Gelatine 6, 226

Glass blowers' tubes 168

Grain 213

Hair 215, 226, 227

Handkerchief 168, 234, 240, 293

Hides 226,227

House, see Room.

Lamp wick 6, 226

Letters 215

Linen, soiled 175, 214, 240

Lumber 214

Mattress 213

Merchandise . 213

Mirror 240

Money 168, 223, 224

Napkin rings 195

Nursing bottles 168

Oyster buckets 213

Pencils 168, 193, 233

Pins 168

Pitcher 97

Rags 169, 222

Railway carriages 292

Roller towel 210

Room 214, 216, 219-222, 233, 235, 241, 292, 294

Rugs 223

Shawl 409

Shoes 218, 240

Spirometer 241

Spoons> etc 168

String 168

Surgeons' instruments 168

Telephones 195, 241

Thermometer 174

Tools 215

Toys 240

Tramcars 292

Wind instruments 168

Wool 226, 227

Fomites and anthrax 226-227

cholera 215

diphtheria • 215, 227

plague, experiments in India 246-247

observations in Sydney 245-246

scarlet fever 214-215

smallpox 214

tetanus. 225-226

tuberculosis, lack of experiments 247

typhoid fever 227

typhus fever 216-217, 409

yellow fever 213-214, 394

464 INDEX

PAGE

Fomites and yellow fever experiments in Havana 245

bacillus diphtheria? on 240-241

tuberculosis on 241-242

bacteria of suppuration on 25

on 239-240

clothing as 216

conclusions concerning 258

definition of term 212-213

evidence of infection by, unsatisfactory 224-225

infection, bacteriological evidence 230-231

evidence against 229-230

few instances of 228

reasons for belief in 228-229

thought important 137

Food, carriers infect 47-51, 60, 65

handling by contacts 162-163

infection and cholera 185

tuberculosis 364, 372-373

typhoid fever 169-170, 364, 374-379

poisonings 38, 60, 371-372

spirillum cholerae in 20

Foot-and-mouth disease and milk 366

French hospitals 198-200

"Antisepsie medicale" in 198

cubicles for isolation in 198-200

screens for isolation in 198-200

Furniture, bacillus diphtherias on 241

Gall stones and typhoid fever 36, 42

Gelatine, bacillus tetani in 6, 226

Glanders, carriers of 99

Glands, bacteria of suppuration in . . 25

cervical and mesenteric, bacillus tuberculosis in 203

Glass, bacillus diphtherias on 192, 240

tuberculosis on 195

contact infection by drinking from 189-190

Glossina morsitans 113

palpalis 115, 399-401

Goats and Mediterranean fever 21, 72, 365

Gonococci carried by flies 424

Gonococcus, not saprophytic 26

resistance to drying 239

Gonorrhea, air-borne infection not possible 166-167

believed to be spread by contact 164-167

carried by nurses 166-167

in Babies' Hospital. 165-166

infection by f omites impossible 166-167

latency of 102-103

not dust-borne 289

persistence of 102-103

Gown and cap for physicians and nurses in contagious cases. ... 218

Grain, yellow fever from 213

Ground squirrels and plague 413

Guinea pigs, bacillus pestis in 69-70

INDEX 465

PAGE

Hair, anthrax spores in 226-227

scarlet fever from 215

Handkerchief, bacillus diphtherias on 240

tuberculosis on 234, 293

Hands, bacillus coli communis on 179, 192

tuberculosis on 193-194

pus bacteria on 192-194

should be washed 179-180

Hazens theorem 329

Hides, anthrax spores in 226-227

Holy water, bacillus diphtherias in 192

Homalomyia canicularis 428

Home isolation effective 160

Hookworms, see Uncinaria, Anchylostoma 186-187

enter through skin 187

grow in soil 186

in feces. 187

Hospital isolation, failure of 139-142

in 161

Hospitals, air-borne disease in 270-271, 279-280

bacillary dysentery in 183-184

contact infection in ' 198-202

diphtheria not air-borne in 273-274

in America 142-145

England 139-142

France 198-200

scarlet fever, not air-borne from 272-273

smallpox, air-borne infection from 260-270

typhoid fever by contact in 173-174

value of 160

House, infection from family to family 28

of, cause of scarlet fever 214-215

tuberculosis infection in 205, 220-222

Houses, species of flies in 428

Ice, bacillus typhi in 13, 319, 341-342

infection by 339-342

Ice cream, a source of disease 349-350

control of 350

Incubation stage infective 39-40

Infection, former theories of 1

not so easy as believed 84, 153, 191-192

Influenza among fishermen 277

lighthouse keepers 277

atypical cases 99

bacillus, see Bacillus influenzae.

from rags 222

not air-borne .••••; 277

carried across Atlantic 277

dust-borne 277, 289

spread only by contact 277

Inhalation of bacteria 304-305

Insect-borne disease usually due to protozoa 380-381

Insects as biological and mechanical carriers 380-382

466 INDEX

PAGE

Insects carry germs on bodies 418

first proof of transmission of disease by 381-382

infection by, importance of subject 380

Texas cattle fever first disease proved to be carried by. . 381-382

Institutions, carriers in 50-52, 65, 67

Intermittent excretion in carriers 46, 62, 68, 76, 98

Isolation a failure in cerebro-spinal meningitis 146

measles 145-146

smallpox 147

causes of failure 147-149

duration of 160-162

effective if prompt 146, 157-158

in Michigan 156-158

rare diseases 158-159

hospitals, failure of 139-142

in American cities 142-145

English hospitals 200-202

family 160-161

effective 160

hospital 161

Monsall Hospital 200-202

Providence 136, 160

villages. . .155-158

not effective in extensive outbreaks 158

to be too strict 151-152

of cerebro-spinal meningitis 146

diphtheria 136, 137-139, 143, 149, 156

duration in Providence 143

malaria 388

measles 145, 156

plague 415

rare diseases 158

scarlet fever 136, 139, 156

school children 162

sleeping sickness 401

smallpox 139, 147, 156

typhoid fever 152, 156

wage earners 152, 162-163

yellow fever 393

should vary 155

too rigorous 152-153

useless if many carriers 159

value of 151-152

Jews do not eat shellfish 376

Kala-azar and insects 402

Kissing means of spread of tuberculosis 202-203

spreading syphilis 168

Laboratory infection, bacillary dysentery 183-184

typhoid fever 175

tuberculosis infection in 308-309

Laniblia.. 112

INDEX 467

Lampwick, bacillus tetani on 6, 226

Latency of infection common 106, 122

Laundress contracts smallpox 214

typhoid fever 175

Laundries and disease 218-219

tuberculosis 219

Leishmania donovani 402

tropica. . . 404

Leprosy, bacilli remain latent 105

in clothing 216

isolation of 156

Letters, scarlet fever from 215

Lettuce, bacillus typhi on 8

Lice and relapsing fever 412

typhus fever and 405-409

Light, bacillus tuberculosis perishes in 234-235

Lighthouse keepers, influenza among 277

Linen, soiled, smallpox from 214

typhoid fever from 175, 219

Lockjaw, see Tetanus.

Lucilia csesar or "blue-bottles" 428, 437

Lumber, smallpox from 214

"Lung Blocks" in New York 220-222

Lupus due to inoculation with saliva 203

Malaria, an endemic disease 29-30

and insects 382-392

water 384

anointing the skin in 389

clearing of space about dwellings 389

discovery of insect transmission 382-383

experiments on infection of human beings 383

isolation sometimes necessary 388

killing of mosquitoes in houses 388

latency of 1L3-115

explanation 114-115

in children 114-115

percentage of 114-115

mosquito sole carrier of 383-385

not air-borne 384

due to soil infection 30

from water 335-336

on ship Argo 335

screening of patients 388

success of mosquito destruction 390-391

ways of prevention 386-390

Malta, Mediterranean fever exterminated at 72

Malta fever, see Mediterranean fever.

Mattress, yellow fever from 213

Measles an air-borne disease 278-280

carriers of Ill

disinfection after 253

isolation in, a failure 145-146

not from soil. . 28

468 INDEX

PAGH

Measles, number attacked Ill

per cent of children attacked 145

Meat and tuberculosis- 372-373

infection by 371-374

inspection, federal control of 373-374

Mechanical carriers distinguished from biological carriers. ...... 413

Mediterranean fever an endemic disease 21

and contact infection 312

dust 312

goats 21,72,365

caused by milk from infected goats 21, 365

experiments in infection by air 311-312

exterminated at Malta 72

goats carriers of 72

human carriers of 72

in United States 365

micrococcus of, resistance to drying 233

on a steamship 72

spread by urine 22

Membrane, bacillus diphtherias in 240

Meningitis, cerebro-spmal, see Cerebro-spinal meningitis.

Meningococcus causes rhinitis 73

found intermittently 76

only near sick 77

in ah* 77

contacts 73-74, 76-77, 80

dust 289-290,293

families 75-80

normal nose 73

nose in sickness 72

infection in barracks . . . 76

not saprophytic 26

persistence of infection 74—75

resistance to drying 238-239

Merchandise, yellow fever from 213

Michigan, isolation effective in -. 156-158

Micrococcus albus 24

aureus 24

citreus 24

melitensis, discovery and study of 21

in dust 21, 300

resistance to drying 233

saprophytic existence of 21

meningitidis, see Memingococcus.
pneumoniae, see Pneumococcus.

Mild cases of infectious disease not recognized 134—135

Milk and disease 342-370

bacillus diphtherias in 25-26

dysenteries in

tuberculosis in 355^356

typhi in 15-17

carriers infect 52-54

classification of 363

condensed and diarrhea 368-369

INDEX 469

PAGE

Milk, diarrhea from 366-369

diphtheria from 96-97, 345-346

foot-and-mouth disease from 366

handling by contacts 162-163

increase of bacillus typhi in 17

infection by water 323

from human sources 347-348

Mediterranean fever from 365

mode of infection of 347-348

outbreaks, characteristics 342-343

few in large cities 344-345

frequency of 343-344

source of infection 346-347

pasteurization of 348-349

protection against tuberculosis from 362-363

of 348-349

rabies from 366

scarlet fever from. ; 346-349

spirillum cholerae in 19

sterilization of vessels 348

streptococcus in 352

tonsillitis from 351-352

tuberculosis from 354-362

amount of 356-362

epidemiological evidence 361-362

instances of 360

tuberculous, consumed by children 357-360

typhoid fever from 51-54, 343-345

Mills-Reincke phenomenon 329-335

Miners, typhoid fever among 182

Minnesota, smallpox isolation abandoned in 147

Mirror, bacillus otiphtheriae on 240

Missed cases a recent discovery 33

Moist surfaces, bacteria not given off from 287-288

Money and disease 223-224

i bacteria on 223-224

Monkeys, dysentery in 24

Monsall Hospital, isolation in 200-202

Montana, smallpox isolation abandoned in 147

Mosquito destruction by culicides 390, 397

drainage 390

oiling 390

prevention 389-391

Mosquitoes and dengue 409-410

filariasis 398-399

malaria * 382-383

conditions for carrying 385-386

yellow fever 392-398

destruction of 397

habits 395

destruction of, in malarious houses 388

habits of 385

species which carry malaria 382-383, 385

success of measures for extermination 390-391

470 INDEX

Mosquitoes, transportation of 395

Mouth, bacillus tuberculosis in 103-104, 193

pneumococcus in 101-102

Mucous surfaces, bacteria of suppuration on 24

Mud dumped near intake cause of typhoid fever 320

Municipal versus personal prevention 208

Murium typhi, bacilli of, carried by ants. 419

Murrina, flies and 429

Musca domestica (see Flies) 418, 428, 437

Muscina stabulans 428

Mussels, infection by 375-377

Myzomyia ludlowii 385

Nagana 113

Night soil as a fertilizer 9

North Boston well 317

Nose, bacillus tuberculosis in 103-104, 193, 299

meningococcus in, in sickness 72-73

in normal 73

tuberculosis of 204

Nurses, gown and cap for, in contagious cases 218

infect patients with typhoid fever 174

typhoid fever among 173-174

Oiling, mosquito destruction by 390

Ophthalmia, Egyptian, carried by flies 419

Opilacao due to connorrhinus 403

Opsonic index in typhoid carriers 42

Oriental sore and insects 404

Ornithodorus moubata 411

Otitis, bacillus diphtherias in 94-95

Owatonna, Minn., diphtheria in 144

Oyster buckets, yellow fever from 213

Oysters, bacillus typhi in 14

"fattening" of 377

infection by 374-378

Pail-closets cause typhoid fever 181-182

Pappataci fever and insects 404

Paratyphoid bacilli, see Bacillus paratyphi.

fever 58-61

Pediculis vestamenti and typhus fever 405-409

Pellagra and insects 403

Pencils, bacillus diphtherias on 193, 233

bacteria on 192-193

Personal versus municipal prevention 208

Persons, not things, are dangerous 225

Phlebotamus. 404, 405

Physicians, gown and cap for, in contagious cases 218

rarely carry disease 217

Pigment to lungs from stomach 306

Piroplasma bigeminum 112, 381-382

Placards for contagious diseases 151

INDEX 471

PAGE

Plague, air-borne 281

and bedbugs 414

fleas 23, 312-313, 413-416

epidemiological evidence 414-415

experimental work 414-415

flies 419,425

fomites, experiments in India 246-247

observations in Sydney 245-246

ground squirrels 413

other rodents 413-416

rats 23, 6^-70, 413-416

rugs 223

atypical and chronic in animals 69-70

human 70

bacillus, see Bacillus pestis.

experiments in infection by air 312-313

isolation of 415-416

not dust-borne 289-290

Plasmodium 113, 384-385

Pneumococcus cause of meningitis 81

in air 300

convalescents 101

droplets 298

dust.. 293-294

normal mouths 101-102

may be air-borne 289

not dust-borne 290

saprophytic 26

persistence of 101

resistance to drying 238

virulence of, in carriers 101-102

Pneumonia and water 333

Poisoning by food 371-372

Poliomyelitis a dust-borne disease 278

atypical cases 121-122

carried by flies 427

carriers of . . : 118-121

Privies and water-closets usually filthy 179-180

encourage uncleanliness 182, 211

infect water supplies 320

removal of, causes decrease of typhoid fever 181-183

typhoid fever from . 169-170, 179-183

Privy vaults, bacillus typhi in 9-10

Prosodemic infection, typhoid fever by 172

Prostatic fluid, bacillus typhi in 37

Proteosoma in birds 383

Protozoa, difficult to cultivate 27

not saprophytic 26

resistance to drying 242

Protozoan disease, carriers a source of 126

latency in 111-112

Psocidse and disease 418

Pulex irritans 414

Purification of water 338-339

472 INDEX

PAGE

Pus bacteria 24

distribution of 240

in air 299-300

dust 294

glands 25

schoolrooms 294

skin and mucous surfaces 24

tonsils 24

water 25

may be dust-borne 289-290

on fomites 25

resistance to drying 236-237

saprophytic existence of 24-25

Quarantine and sleeping sickness 401

yellow fever 396

Quinia prevents malaria , 386-388

Rabies from milk 366

Radishes, bacillus typhi on 8

Rags and disease 222-223

cholera from 222

erysipelas from 222

flock cough from 222

influenza from 222

scarlet fever from 222

septicemia from 222

smallpox from 222-223

typhoid fever from 222

Railway carriages, bacillus tuberculosis in 292

Rat-proofing of buildings 416

Rats and plague 23, 69-70, 413-416

carriers of plague 70

chronic plague in 69

destruction of 416

Relapsing fever and bedbugs 411-412

lice 412

ticks 410-411

infectious during afebrile period 116

types and parasites 116

Resistance of bacteria less in cultures 11

Rhinitis, bacillus diphtheriae in 94-95

caused by meningococcus 73

Rhode Island, death rate from typhoid fever 18

Rice, spirillum choleras in 20

Rocky Mountain fever and ticks 412

Rodents, plague derived from 413-416

Roller towel, bacillus coli communis on 192

should be abolished 210

use of, forbidden 211

Room, bacillus diphtherias in 233, 241

tuberculosis in 235

infection cause of typhoid fever in barracks 216

Rooms, infection of 205, 214-215, 219-222

INDEX 473

Rubella, air-borne 280

Rugs and plague 223

Saliva, inoculation with, cause of lupus 203

transfer of 189

vehicle of infection 188-190

Sarcophaga 428

Scarlet fever, air-borne 270-273

out of doors 272-273

and books 215

clothing 217

coat 215

fomites 214-215

house infection 214-215

letters 215

milk 346-347

rags 222

atypical. 107-109

disinfection unnecessary after 250-252

due to ice cream 350

extension in dwellings 196-197

hospital temporarily closed in Leicester 141

hospitals in England 139-142

in cows. 346-347

in Providence 136

like diphtheria 106-107

not air-borne from hospitals 272-273

from soil 28

virulence varies hi 140

Scenopinus fenestralis 428

Schizotrypanum cruzi 404

School children, isolation of 162

disinfection 254

teaching cleanliness in 209-210

Schoolrooms, bacillus coli communis hi 294

pus organisms in 294

Screening for protection against malaria 388

of cases of malaria 388

yellow fever 396-397

Screens for isolation in French hospitals 198-200

Sedgwick, typhoid fever outbreak in Bondville, Mass 169

Seltzer water, bacillus typhi in 13

Seminal vesicles, bacillus typhi in 37

Septic infection and insects 417—418

tank effluent, bacillus typhi in 10

Septicemia from rags 222

Sewer air and bacteria 284-285

diphtheria 273

• typhoid fever. 274-275

bacillus prodigiosus in 284

typhi in 284

infection by 297, 299

Sex of smallpox cases 267

Sheet hung before door to prevent infection 282

474 INDEX

Shellfish, infection by 374-378

Shipboard, influenza on 277

malaria on 335

typhoid fever by contact on 176

Ships for smallpox hospitals 262-263

Shoes as carriers of infection 218

bacillus diphtherise on 240

Shuttle, infection from 192

Sidewalks, virulent bacilli tuberculosis on 194

Simulium • 403

Skin, bacteria of suppuration in 24

hookworms enter through 187

Sleeping sickness, an endemic disease 30

and tsetse flies , 399-^01

atoxyl to prevent 402

discovery of trypanosome 399—400

in lower animals 400

is it carried mechanically? 400-401

modes of transmission 399-401

not due to soil infection 30

per cent of carriers 115-116

persistence of infection 115-116

prevention of 401-402

quarantine '. 401

Smallpox, air-borne, conclusions . 269-270

evidence against 266-269

infection of, denied 263-270

in United States 264-265

objections to theory 265-266

sex of cases 267

and fomites 214

rags 222-223

atypical cases •. 109-110, 133

carriers 109-110

hospital ships 262-263

isolation in, often a failure 147

mild cases of 109-110

not from soil 28

virus, resistance to drying 244

Soft drinks, bacillus typhi in 13

Soil, amebse dysenteriae in 24

bacillus cliphtherise in 26, 28

dysenteriae in 23

pestis in 22

tetani in 239

typhi in 7, 10

hookworms grow in 186

infection, cholera from 29

in common diseases 28

typhoid fever from 29

melitensis micrococcus in 21

not infected with cattle fever 30

malaria 30

measles 28

INDEX 475

Soil, not infected with scarlet fever 28

sleeping sickness 30

smallpox 28

Texas cattle fever 30

yellow fever 30

spirillum cholerae in 19-20

Sore throat and milk 351-352

Spanish- American War, typhoid fever in 17, 56, 435

from dust in 275-276

flies 435-436

Spirillum cholerae carried by flies 420

found in water 326

grow only in human body 20

in convalescents 61-62

cooked food 20

dead bodies 19

feces 19

gall bladder 62

healthy persons 62-65

milk 19

soil 19,20

water 19

intermittent excretion 62

resistance to drying 237-238, 242, 291-292

saprophytic existence of 19

Spirocheta carteri 116, 412

duttoni 116, 411

Spirochete of syphilis, latency of 116

not saprophytic 26

resistance to drying 239

Spirometer, bacillus diphtheriae not on 241

tuberculosis not on 195

Spitting should be forbidden 211

Spores may be dust-borne 290

resistance to drying 232

Springs as sources of infection 324

Sputum, bacillus pestis in 71

typhi in. . 37-38

Squirrels, see Ground squirrels (Citellus beecheyi).

Stamp out disease, failure to 135-136

Stegomyia calopus .392, 395, 405

Sterilization of municipal water supplies . 339

Stomach, bacillus tuberculosis may pass to lungs from 305-308

Stomoxys calcitrans 428

Storage of water 338

Street dust, bacilli tuberculosis not found in 292

Streptococcus enteritidis, cause of diarrhea 367

in dust 282

milk 352

may be air-borne 289

py9genes 24

salivarius in droplets 297

Suppuration, bacteria of, in healthy organs '. 105-106

latency of bacteria of 105-106

476 INDEX

Suppuration, result of infection by bacteria 24

Surgery, air-borne infection in 283-284

Swine infected by bacillus tuberculosis through the tonsils .... 203-204

plague bacilli in air 300

Syphilis, air-borne infection not possible 167

articles infected with 168-169

believed to be spread by contact 164-169

droplet infection in 167

fomites infection in, not important 169

infection by kissing 168

non-sexual contact 168-169

latency of 116

not persistent 168-169

sexual act chief mode of infection 168

spirochete of, see Spirochete of syphilis.

spread solely -by contact 167-168

Tabanus lineola 3

Teacher caused diphtheria 97

contracts diphtheria from pupil 98

Telephones, bacillus diphtherias not on 241

tuberculosis on 195

Tetanus and fomites 225-226

bacillus, see Bacillus tetani.

decrease of, in Havana 226

from lamp wick used for tying umbilical cord 5, 226

in Red Bank, N. J 5

on Long Island 5

Texas cattle fever, an endemic disease 30

carried by tick 112

first disease proved to be carried by insects . 381-382

riot due to soil infection 30

persistence in blood 112

Thermometer transfers bacillus typhi 174

Things not so dangerous as persons 225

Ticks and relapsing fever 410-411

Rocky Mountain fever 412

Texas cattle fever 112, 122, 381-382

Tidal mud, bacillus typhi in 8

Tongue, bacillus tuberculosis on 203

Tonsillitis and milk 351-352

Tonsils, bacillus tuberculosis in 203-204

typhi in. 34

bacteria of suppuration in 24

tuberculosis infection through 306

of 204

Tools, diphtheria from 215

Towel, see Roller towel.

Toys, bacillus diphtheria on 240

Tramcars, bacillus tuberculosis in 292

Treasurer of United States, investigation of money as carrier of

disease by 224

Treponema pallida in gummata 116

life of, short 167, 169

INDEX 477

PAGE

Trier, typhoid fever in 17, 57

Trypanosoma bruccei 113

equiperdum 113

gambiense 115, 399-401

Tsetse flies and sleeping sickness 115, 399-401

nagana 113

Tuberculosis and flies 429

fomites, lack of experiments 247

laundries 219

meat .272-273

bacillus, see Bacillus tuberculosis.

bovine, eradication of 364

contact infection in 204-205

experiments concerning air-borne infection 301-311

feeding experiments 354-355

from flies 429

house infection 220-222

milk 354-368

amount of 356-362

epidemiological evidence 361-362

instances of 360

protection against 362-363

shuttle 192

water. ;••••: 332-333

human and bovine reciprocally infective 353

infection by air 194, 202-203

dust questioned 302-303

in the home 205

stomach 305-308

experiments under natural conditions. . .309-311

in alimentary tract 203-204

animals through mouth and pharynx 204, 306

laboratory 308-309

unnatural conditions of experiment 308

mode of infection, evidence from pathology 205

of the nose 204

t9nsils 204

proportion of human and bovine types of bacillus 360-361

spread by animals licking one another 204

kissing 202-203

through tonsils 204, 306

Tuberculous milk consumed by children 357-360

Typhoid bacillus, see Bacillus typhi.

carriers 18, 38-55, 180

cannot be isolated 152

cause disease. 47-54

inf ective during incubation stage 39-40

number of 180

opsonic index in 42

percentage of 40-44

feces in yard 170, 173, 182

fever among nurses 173-174

an infection of the blood 34-35

and gall stones 36, 42

478 INDEX

Typhoid fever and high temperature 374-375

tonsillitis 34

atypical 56-59

believed to be an intestinal disease 178

by bathing. 324-325

contact 17, 47-51, 169-183

in an almshouse 175

civil life 171-173

hospitals .' 173-174, 436

South Africa 171, 436

Spanish War 170-171, 435-436

the family 176

on shipboard 176

food infection 169-170

prosedemic infection 172

carriers of, among different classes 39-44

duration of infection 45

causes of excess in cities 321

contact outbreaks in cities 172-173

contracted by laundress from soiled linen 175

death rate in Rhode Island 18

decreases with removal of privies 181-183

filthy habits cause of 170, 173, 182

from bedbugs 418

celery 379

clothing 219

crawfish 378

cream 349

dust 275-276

flies 169-171, 182-183, 435-447

correlated decrease 442-444

seasonal distribution 439-442

statistical evidence 439-445

fomites 227

handling typhoid cultures 175

ice cream 349-350

infected bedding 175

milk : 51-54, 345-346

mud dumped near intake 320

pail closets 181-182

privies 169-170, 179-183

rags 222

sewer ah- 274-275

shellfish 374-378

soil infection 29

soiled linen 219

springs 324

thermometer 1?4

unrecognized cases 133

water 317-325

watercress 378-379

wind blowing sewage to water works 320

houses, infected flies in 421-422

in barracks due to room infection 216

INDEX 479

Typhoid fever in mines 182

soldiers' home due to contact infection from

physician 175

Spanish War 17, 56, 435

Trier 17

waitress due to contact infection 175

infection stopped by strict cleanliness 174

less with water closets 181-182, 144

mild cases in Panama 57

Spanish War 56

Trier 57

not an intestinal disease 34

dust-borne 289-290

outbreaks continued by contact infection 172-173

in Providence, R. 1 18

prevalent in country 173

spread among patients by nurses 174

by kitchen help 175

waitress 175

"Typhoid Mary" 47

Typhus fever and flies 408

lice 405^09

carried in shawl 409

considered air-borne 408

from clothing 216-217

identity of strains 409

not air-borne 281

of extra-corporal origin 29

Uncinaria, see Hookworms, Anchylostoma.

americana 186

due chiefly to contact 186-187

Urine, bacillus typhi in 37

persistence of 10

cause of Mediterranean fever 22

Vaccine, resistance to drying 243-244

transport across ocean 242-244

Vaginitis in Babies' Hospital 165-166

Vegetables, amebse dysenteriae on 24

bacillus typhi on 8

Venereal diseases, see Gonorrhea; Syphilis.

Vincent's angina 117

Wage earners, isolation of 152, 162-163

Waitress, typhoid fever spread by 175

Warning sign in diphtheria 151

Washing soiled linen in running streams cause of cholera ^ 325

Water and cholera V325-326

diarrhea 327-329, 331

dysentery 326-329, 331

malaria. 335, 384

pneumonia , 333

480 INDEX

PAGE

Water and the general death rate 331

tuberculosis 332-333

typhoid fever 317-325

worms 336

yellow fever 336

bacillus typhi in 10-13, 319

recovered from 319

bacteria of suppuration in 25

closets, typhoid fever less with 181-182

filtration of 338

infection by 316-339

of 18

milk by 323

micrococcus, melitensis in 21

mode of infection with typhoid bacilli 320

municipal supplies infected 320, 326

outbreaks, characteristics 317-319

protection of 337-338

purification of 338-339

spirillum cholerae in 19

sterilization of municipal 339

storage of 338

Watercress, typhoid fever caused by 378-379

Wells, amount of typhoid fever due to 322-324

cholera infection from 325-326

conditions of safety 324

infection from 316-317, 321-325

protection of 337

White diarrhea of chicks 71

Whooping cough, air-borne 280

not of extra-corporal origin 29

Willard State Hospital, diphtheria in 144

Wind blowing sewage to water works cause of typhoid fever 320

Wool, anthrax from • 227

spores in 226-227

Worms derived from water 336

Wounds, infection of , by air 300-301

Yellow fever an endemic disease 30

and bagging 213

ballast 213

cloth 213

fomites 213-214, 394

experiments in Havana 245

grain 213

mattress 213

merchandise 213

mosquitoes 293-298

oyster buckets 213

quarantine 396

discovery of causation 392-394

latent period in outbreaks 393-394

mild and atypical cases 117

mosquito destruction 397

INDEX 481

PAGE

Yellow fever mosquito, habits of ^ 395

not due to soil infection 30

from water 336

preventive measures 396-397

screening of cases and houses 396-397

success of mosquito destruction 391, 395, 397-398

virus filterable . . . , 393-394

in blood.. ..393-394

7^>

570076

Engineering
Library

TY OF CALIFORNIA

QEFARTJv.ENT OF CIVIL. ENGINEERING
LEY, CALIFORNIA1

UNIVERSITY OF CALIFORNIA LIBRARY