SB lift 5fi5

ITALITY

licrobes in Oysters and
other Shellfish.

f ni

r

THE LIBRARY

OF

THE UNIVERSITY
OF CALIFORNIA

PRESENTED BY

PROF. CHARLES A. KOFOID AND
MRS. PRUDENCE W. KOFOID

EXPERIMENTS ON OYSTERS AND
OTHER SHELLFISH.

MY Committee has charged Professor Klein with
conducting experiments on oysters and other shell-
fish in order to ascertain the vitality of the typhoid
bacillus and other sewage microbes in them. The
results obtained with oysters, cockles, and mussels
are herewith published by the Fishmongers' Company.

J. WRENCH TOWSE,

Clerk of the Worshipful Company
May, 1905. of Fishmongers, London.

Report of Experiments and Observa-
tions on the Vitality of the Bacillus
of Typhoid Fever and of Sewage
Microbes in Oysters and other
Shellfish.

By E. KLEIN, M.D., F.R.S.,

Lecturer on Advanced Bacteriology in the Medical School of
St. Bartholomew's Hospital, London.

ALREADY in 1893 the then Medical Officer of the Local
Government Board, Sir Kichard Thorne, in his Summary
•(Keporfcs and Papers on Cholera in England in 1893, Local
Government Board) makes the following trenchant remarks
on page 29, in reference to a number of cholera attacks in
which the history pointed to infection by means of oysters
and shellfish which had been procured from, and specifically
fouled at Cleethorpes and Grimsby, viz. : — " But one thing
is certain, oysters and shellfish, both at the mouth of the
Humber and at other points along the English coastline, are
at times so grown and stored that they must of necessity be
periodically bathed in sewage more or less dilute ; oysters
Jiave more than once appeared to serve as the medium for
communicating disease, such as enteric fever, to man; and
so long as conditions exist such as those with which the
oyster trade of Cleethorpes and Grimsby is shown to be
associated, conditions which may at any time involve risk
of the fouling of such shellfish with the excreta of persons
suffering from diseases of the type of cholera and enteric
fever, so long will it be impossible to assert that their use
as an article of diet is not concerned in the production of
disease of the class in question."

B 2

M374361

In the subsequent two or three years several outbreaks
of typhoid or enteric fever having been demonstrated to be
caused by the consumption of oysters derived from sewage-
polluted layings in America, in France, arid in England (see
Eeport of the M.O., " Oyster Culture and Disease," 1894-
1895, Appendix 3 and 4), the Local Government Board had
instituted in 1894-1895 a careful survey of all oyster beds-
and oyster ponds along the whole coast of England and
Wales ; the results of those investigations were published
in 1896 by the Medical Department of the Board as a separate
volume (" Oyster Culture and Disease "), and it will be seen
therein to what a large extent oysters are laid down and
stored in several places in England in a manner which must
be considered not only objectionable qua cleanliness per se>
but also must be instrumental in conveying occasionally
dangerous infection to those consuming them.

Without intending to cite all those cases of typhoid fever
in which in single instances and in a small group of indi-
viduals who had partaken of oysters, mussels, or cockles,
derived from polluted localities (such as have been described
by various health officers — Dr. Newsholme, Dr. Thresh,
Dr. Nash, Dr. Allen, and others), typhoid fever has been
demonstrated to have been caused by such shellfish, I will
mention the two instances only in which in recent years, to wit,
November 9, 1902, infection with typhoid fever by polluted
oysters has manifested itself in a somewhat dramatic fashion
and on a considerable scale — I refer to the now historic
mayoral banquets at Winchester and Southampton. The
demonstration of this infection, of the derivation of the
typhoid oysters from sewage-polluted ponds at Emsworth,
are well known ; they have been well described by Dr.
Bulstrode in the annual report of the Medical Officer of the
Local Government Board, 1902-1903, pp. 129-189. Even
subsequent to these outbreaks, viz., during 1903 and
1904, cases of enteric fever have been traced in numerous
single instances to the consumption of polluted oysters or
polluted cockles (Dr. Collingridge, Dr. Allen, Dr. Buchanan,
and others), and it is common knowledge that, with the

exception of a few isolated instances in which since 1896
an improvement in oyster layings and oyster storage has
been effected, the general system obtaining in a good many
instances, viz., of exposing oysters to " be periodically bathed
in sewage more or less dilute," is still the same as it
was in 1895, that is to say, " conditions which may at any
time involve risk of the fouling of such shellfish with the
excreta of persons suffering from diseases of the type of
cholera and enteric fever." As a matter of fact, I have in
several instances discovered the B. typhosus in shellfish
coming from polluted sources. These are the instances : —

1. In a sample of oysters derived from Grimsby in 1895.

2. In a sample of oysters brought over direct from
America, 1903.

3. In a sample of mussels gathered from a polluted place
on the shore of Southend-on-Sea, 1904.

4. In a sample of oysters gathered from a place in Lang-
stone Harbour, about 600 yards distance from the Portsmouth
sewer outfall, 1904.

I shall have presently an opportunity of showing that the
identification of this microbe in such shellfish is a matter of
no small difficulty, owing to such shellfish always containing
a large amount, and in preponderance, of sewage microbes,
greatly impeding the identification. In order to detect
the B. typhosus in shellfish or other materials (water, milk)
exposed to sewage pollution, and therefore harbouring sewage
microbes, the former must be present in appreciable numbers,
and if found would a priori conclusively indicate that the
specific pollution (i.e., with typhoid excreta) must have been
considerable. Until the last few years such identification
was an almost hopeless task, but at present the task has been
facilitated to a considerable degree by the discovery by
Drigalski and Conradi of a culture medium, on which the
presence of the B. typhosus can be easier detected than by
the former methods. I say easier, although I should not omit
to add that also by this method its presence must be in fair
proportion. The first case in which the typhoid bacillus was
found in the* Grimsby oysters was described in the Eeport of

the Medical Officer of the Local Government Board (" Oyster
Culture in its Eelation to Disease," 1894-1895, pp. 114 and
115), and it was discovered by the older method : phenolated
broth and phenolated gelatine, so that in this instance it
must have been present in very considerable numbers. In
the other instances (2, 3, and 4) above mentioned the identi-
fication was somewhat easier, viz., by the Drigalski-Conradi
medium, of which presently more will be said. In the last
instance, viz., oysters from Langstone Harbour, the typhoid
bacilli were met with in one oyster to the amount of several
dozen per oyster. When I speak of the B. typhosus having
been identified, I mean the microbe had responded to all and
every test — of which there are many, as will presently be
described — which denote the known characters of the-
microbe of typhoid or enteric fever.

From the foregoing it will appear interesting to inquire
whether and to what extent under the circumstances of
actual specific fouling of shellfish, such as must occasionally
occur in estuaries^ and on the shores of the coast, where
oysterlayings and oysterponds, mussels and cockles, are
exposed to almost continuous sewage pollution, the obnoxious
and dangerous microbes which have found entrance into the
shellfish are readily, and by what methods, removable from
such shellfish; or whether having once gained entrance
remain in it and make as it were therein a home for them-
selves. It is clear that if the latter should be the case no
remedy would be available to render such shellfish —
principally oysters — fit for consumption ; whereas, if the
former should be the case a remedy would be available.
With reference to oysters this question is of greater import-
ance, inasmuch as the great majority of these shellfish are
consumed in raw state, whereas in the case of mussels and
cockles — leaving out the instances in which the enthusiastic
gatherers eat them raw — some kind of heating process,
though as we shall see this is not always effective and reliable,
is employed previous to their being eaten, and therefore the
majority of dangerous microbes presumably are in many
cases devitalised.

The dangerous microbe in shellfish with which we are
chiefly concerned is, of course, the Bacillus typhosus. It is
not, as we shall see later, the only dangerous microbe, but it
certainly is the chief one, because infection with it, as
mentioned above, has hitherto been of somewhat conspicuous-
frequency, and it is chiefly this microbe, i.e., the microbe of
typhoid or enteric fever, that need occupy us here.

In the experiments which I conducted for the Local
Government Board, " Oyster Culture in Eelation to Disease,"
1894-1895 (pp. 116-120), oysters were kept in sea water
infected with the B. typhosus, and it was found that this
microbe was recovered from the interior of some of the
oysters as late as 18 days after infection.

Professor Herdman in 1895 states (Eeport on the
Lancashire Sea Fisheries) that in the case of oysters grown
in water infected with B. typhosus it was found that there
was no apparent increase of the organisms, but that they
could still be identified in cultures taken from the water of
the pallial cavity and rectum 14 days after infection.

Dr. Chantemesse (Proceedings of the Academic de
Medicine of Paris, June (?) 1896) placed oysters for 24 hours
in sea water intentionally infected with B. typhosus, then
kept them for 24 hours out of this water ; examining them
after the lapse of this time, he found in the liquor and in
their bodies B. typhosus.

Herdman and Boyce in a series of experiments conducted
with oysters infected with B. typhosus (Oysters and
Disease, Thompson Yates Laboratories, Vol. II, 1898-1899,
Lancashire Sea Fisheries, Memo. No. 1) summarise, p. 54,
their results thus : " In our experimental oysters inoculated
with typhoid we were able to recover the organism from the
body of the oyster up to the tenth day. We show that the
typhoid bacillus does not increase in the body or in the
tissues of the oyster, and our figures indicate that the bacilli
perish in the intestine."

Most observers are agreed that the typhoid bacillus does
not multiply within the oyster, and is gradually destroyed or
eliminated when the oyster is placed in clean sea water,

8

although as to the time required for this process of cleansing
the various observers are not agreed ; thus, while my experi-
ments in 1894-1895 would indicate the duration of vitality of
the B. typhosus in the oysters to be three weeks, Chante-
messe at first recommended about the same period (quelques
semaines), later he reduced the time to 8 days (I.e. 9 June,
1896). Professors Herdman and Boyce, I.e. p. 54, say:
" In our experiments in washing infected oysters in a stream
of clean sea water .... there was a great diminution or total
disappearance of the typhoid bacilli in from one to seven days."

In order more accurately to determine the vitality of the
typhoid bacilli in oysters, and to ascertain how and in what
period a given number of B. typhosus introduced in or in-
gested by oysters, disappears from their (oysters) interior
under conditions resembling those obtaining in nature more
or less, the Worshipful Company of Fishmongers have
charged me with undertaking the required experiments, not
only with oysters, but also with mussels and cockles. Such
numerical determination is at present possible, and a matter
comparatively easy to achieve, as will presently appear when
describing the method used.

It must be obvious that as regards oysters the problem
resolves itself into the following questions, viz. : 1. Given
clean oysters, i.e., oysters which are laid down in clean water,
what power, if any, have such oysters to destroy or eliminate
a definite number of B. typhosus injected into or ingested by
them, and further in what time can they do this, supposing
that they are afterwards kept under conditions best for such
destruction or elimination, i.e., kept in clean water daily
renewed ? The answer to this question would demonstrate
in an absolute way whether or not such power is inherent
in the oyster. 2. Is there any difference in these respects
between oysters which are derived from clean layings and
oysters derived from initially sewage polluted beds ? 3. Is
there any difference, and what, between oysters previously
infected with the B. typhosus, which are then kept in clean
sea water continually changed, and such as having been
infected are not placed under these favourable conditions ?

The first part of this Eeport deals with experiments
capable of furnishing definite answers to these questions, and
these answers will enable us to draw conclusions with some-
thing like exactness as to the means required for dealing
with oysters presumably dangerously polluted. The second
part of this Report deals in similar fashion with mussels and
cockles. In a third part observations and experiments are
described which deal with the general question of identifi-
cation in oysters of microbes derived from sewage, a question
which at present is still imperfectly understood, and on
account of this not unfrequently misinterpreted.

SERIES A. — EXPERIMENTS WITH THE B. TYPHOSUS IN
OYSTERS.

Before we enter on a description of the details of these
experiments and the methods by which the experiments
were carried out, it may not be out of place to give in a
general way a summary of the present knowledge concerning
this microbe.

The typhoid bacillus — Bacillus typhosus — is the essential
cause of typhoid or enteric fever ; that is to say, when intro-
duced into a susceptible individual — generally by way of the
digestive organs — it is capable of setting up, after an incuba-
tion period of from 10 to 14 days, the clinically and patho-
logically well-defined acute febrile disease known as enteric
or typhoid fever.

The microbe is found in large numbers — multiplying
readily — in the interior of the ileum and in its swollen mucous
membrane, Peyer's glands ; it is found in great abundance in
the swollen and inflamed mesenteric glands, and in the
swollen spleen. In the intestine, and also in the typhoid
stools, its demonstration is somewhat made difficult by the
simultaneous presence — generally in predominating numbers
— of other bacteria similar to it, but not the same, to wit
Bacillus coli (see below), but it has been shown, and lay some
of the most modern methods it has become more easily to do
so, particularly during the second and third week of the illness,

10

viz., that in the contents of the ileum and in the fluid and semi-
fluid (typical pea soup) typhoid stools the typhoid bacillus does
occur in great numbers. In several instances of cultivation
by the Drigalski-Conradi medium it was possible to show that
in the fluid typhoid stools the B. typhosus was present to
the number of one to three millions per one cubic centimetre,
that is to say, about one to two hundred thousand per drop
(minim) of the stool. As regards the mesenteric glands and
the spleen of a case of typhoid fever, the demonstration by the
microscope and particularly by culture of the B. typhosus can
be demonstrated readily, and in culture made with a trace of
a particle of these organs crowds of typhoid bacilli — generally
in pure culture — can be obtained. It has been further
shown that in some cases even several months after con-
valescence has set in, typhoid bacilli are still present in the
stools of the patient, although as a rule a few weeks after
convalescence has commenced they have practically dis-
appeared from the bowels. It has further been shown that
in localised inflammations following upon the acute stages
of the disease — particularly in the lung — the typhoid bacilli
may be, and sometimes are, present in large numbers. One
of the most important results of research that has been
brought to light is this, that typhoid bacilli pass out of the
body by way of the kidney and urine in enormous numbers
during convalescence, so much so that it has been calculated
that in at least 25 per cent, of cases (according to more
recent observations over 25 per cent.) the convalescent voids
typhoid bacilli by way of the urine (Bacilluria), this fluid being
in some marked instances turbid by the number of typhoid
bacilli — up to 6000 millions per one cubic centimetre (Dr.
Horton Smith, Gulstonian lectures before the Eoyal College
of Physicians, 1900). It must be obvious that these results
are of the utmost importance to public health, and I will for
the sake of illustrating this importance take the following
into consideration : Up to the time and before the above fact
concerning the copious presence of B. typhosus in the urine
of a large percentage of convalescents was elucidated, the
attention of sanitarians, physicians, and nurses was chiefly

11

directed to the stools of typhoid patients as being capable of
conveying the disease germs during the first four to six
weeks, that is during the active phases of the disease and
the early stages of convalescence, the urine not being specially
attended to. But since we now know that the patient,
weeks after convalescence has set in, voids typhoid bacilli
by the urine, the presence in any locality of a convalescent
from typhoid fever, in whom the stool has perhaps ceased
to be infective, remains, nevertheless, a fruitful source of
typhoid bacilli. We may have a seaside place in whose
population no typhoid fever cases have occurred, but to
which seaside place a person convalescent from typhoid
fever has been taken for recuperating ; the sewage of this
seaside place — ostensibly free from typhoid fever — would
nevertheless contain plenty of typhoid bacilli which might
find access to shellfish laid clown or kept on or near the
shore of such a place.

The typhoid bacillus belongs to a large group of microbes
— coli-typhoid group — which in morphological, cultural and
physiological respects possess certain characters in common,
but the individual species constituting the group differ,
nevertheless, from one another in definite manner. As to
the B. typhosus its essential differential character is that it is
found, as described above, in definite distribution in typhoid
fever and in this disease only, and that its introduction into
the alimentary canal under suitable conditions, and its multi-
plication within the infected person, sets up the specific
disease typhoid fever, as has now amply been demonstrated
by indirect epidemio-logical evidence, as also unfortunately
in several direct instances amongst those who have worked
in the laboratory with cultures of the typhoid bacillus.
None of the other species belonging to the coli-typhoid group
are connected, as cause, with typhoid fever, although some,
like the Bacillus Gaertner, some virulent coli-like organisms,
the Bacillus paratyphosus and Bacillus dysenteric, are con-
nected with other acute intestinal diseases, but not of the
nature of true enteric fever.

The morphological and cultural characters by which the

12

B. typhosus is distinguished from other coli-like microbes
of the coli-typhoid group (all of which fail to liquefy gelatine
and are gram negative in staining) are these: — (a) Morpho-
logical : motile, cylindrical bacilli, multiflagellated, the thin,
long, wavy flagella distributed over the whole body; the
bacilli in culture are capable of forming shorter or longer
threads ; like other coli-bacilli they grow well at all tempera-
tures up to 38° or 40° C. (&) Cultural : on gelatine colonies
angular discs, with thicker centre, filmy margin, finely granu-
lar ; on gelatine streak translucent, filmy, dry-looking band,
slightly thicker in centre, irregular margin ; in ordinary, as also
in sugar gelatine shake culture, no gas formation, colonies
uniformly distributed throughout the gelatine ; grows always
slower than B. coli ; on agar not characteristic, except that
the growth is slower and more translucent than that of most
B. coli; in litmus milk acid production, slower than that
of most B. coli, milk remains fluid ; in phenol broth good
growth, no gas formation ; in ordinary broth rapid growth
and uniform turbidity, no indol formation ; in neutral red
broth no change of colour ; on potato colourless, filrny growth ;
on potato gelatine colonies smaller round and more trans-
lucent than those of most B. coli ; on potato agar and in urine
gelatine grows more filamentous than B. coli ; in Proskauer
and Capaldi medium I negative, in Proskauer and Capaldi
medium II positive — B. coli gives the reverse test; on
Drigalski-Conradi medium the colonies are characteristically
bluish in laterally reflected light, violet-blue in directly
reflected light; translucent, filmy, violet margin, thicker
more or less acuminated centre, finely granular ; the bacilli
composing the colonies are oval to cylindrical, motile; in
MacConkey fluid (litmus glucose taurocholate of soda, pep-
tone) acid formation but no gas ; in litmus lactose peptone
growth, litmus becomes bleached, no acid or alkali, no gas ;
blood serum of typhoid patients (Widal's test), or blood
serum of an animal previously injected (prepared) with
typhoid culture, acting on broth culture, or emulsion of gela-
tine or agar culture of B. typhosus, the bacilli become arrested
in their motility and agglutinated into large more or less

13

dense clumps in marked and rapid manner (Bordet-Gruber
test in vitro), and in higher dilutions than other allied
microbes.

Small doses of a recent culture of B. typhosus intraperi-
toneally injected cause death of guinea-pigs from acute
peritonitis in a short time, 20 to 36 hours, according to
size of dose, though it has to be remembered that when
subcutaneously injected it acts locally only, except very
virulent strains in fair doses produce sometimes general infec-
tion and death. The virulence differs with different strains ;
an animal previously prepared with subfatal doses of culture
of B. typhosus (of either living or dead culture) is immunised
and protected against an otherwise fatal intraperitoneal dose
of virulent B. typhosus, the animal in proportion to its
previous preparation suffers no ill effects, the peritoneally
injected bacilli rapidly undergoing granular degeneration
and change into granules and globules of dead matter
(Pfeiffer's phenomenon or test in corpore). Assuming that in
hanging drop and in staining a microbe shows the characters
above mentioned, produces in litmus milk acid but no coagu-
lation, gelatine streak and shake positive, neutral red negative,
in phenol broth or MacConkey fluid good growth, no indol
formation, in litmus lactose peptone negative, litmus bleached,
and further in Drigalski and Conradi plates, in Proskauer and
Capaldi medium I and II, in flagella stained specimens, and
in agglutination test in high dilution with typhoid serum (of
man, better of typhoid immunised animals), it answers
in positive fashion, we would consider these sufficient to
establish the identity of the microbe in question with the
B. typhosus.

From the foregoing it will have been gathered that, in
order to definitely identify a particular microbe as the
B. typhosus, a number of tests, morphological, cultural and
experimental, have to be employed, and it will also be
readily understood that if in any material, subjected to
analysis, the typhoid bacillus should be associated with other
microbes belonging to the coli-typhoid group, the difficulty
of isolation of the B. typhosus out of the mixture must be

14

correspondingly greatly increased. And it is precisely
materials which contain such mixtures (water, milk, shellfish
fouled by filth and excremental matters) that we are often
called upon to analyse for the presence of the B. typhosus.
In ordinary domestic sewage the number of B. coli com-
munis alone amounts to between 100,000 and one million,
or even more, per 1 c.c. ; in ordinary normal fcecal matter
B. coli communis alone amounts to something between
40 or 50 millions and 400 to 1000 millions per one gram,
in the fluid typhoid stool (pea soup stool) the number of
B. coli amounts to something like 14 to 20 millions per
1 c.c. The B. coli communis, as also other coli-like microbes
belonging to the coli-typhoid group, grow in all the media
in which the B. typhosus is capable of growing, and unfor-
tunately with greater ease and rapidity; but there is no
medium known in which the reverse is the case, and it will
therefore be readily understood that by cultivation — the only
method which as the first step in the analysis can be resorted
to — the isolation of the B. typhosus from amongst a number
— generally an overwhelming number — of coli-bacteria and
other microbes in a given mixture must be, in the nature of
things, an extremely difficult matter, unless the B. typhosus
should happen to be present in very large numbers. Add to
this the well-recognised fact that, taking the above tests
for differentiation of the B. typhosus from the other species
of the coli-typhoid group, the differences are small, and some
of them more or less those of degree only, and a negative
result qua isolation of the B. typhosus from the polluted
materials can be easily understood, although the polluted
material (water, shellfish, milk) had been proved to be speci-
fically polluted with typhoid excreta, having produced typhoid
fever in the consumers.

Under these circumstances, any method by which
even the favouring growth and the rapid recognition by
culture of bacteria belonging to the coli-typhoid group
could be effected is of advantage, though it is only a
first small step; this is achieved by Parietti's method
(adding a certain amount, 0 • 05 per cent., of phenol to the

15

culture medium), by which other bacteria not belonging to
the coli-typhoid 'group are kept back, while the bacteria of
the latter group grow undisturbed. The same, to a large extent,
is the case with Eisner potato gelatine. But it must be obvious
from what has been said above that such inhibition (for a
time, at any rate) of other bacteria not belonging to the coli-
typhoid group does not carry us very much further, because
the difficulty about the coli-bacteria amongst themselves is
still present. MacConkey's fluid is a further step in advance,
because in this medium the rapid production of acid — red-
dening of the litmus fluid — denotes already in 24 hours or
so, the presence of acid-forming bacteria, most probably
belonging to the coli-typhoid group, and possibly including
the B. typhosus. This medium keeps back non-acid pro-
ducers to a large extent, and, owing to the easily perceived
change in colour of the litmus to red, the presence of an
acid-producing microbe is at once made out. Unfortunately,
all coli-bacteria forming acid grow well in this medium, and,
therefore, it does not carry us much further in the isolation
of the B. typhosus than did Parietti's method. The same
remarks apply to all other media that have been hitherto
described, with the notable exception of the Drigalski-
Conradi medium, for by this medium used for surface plates,
as described by Drigalski-Conradi (" Zeitschrft f. Hygiene,"
vol. 39, p. 283) (Nutrose, lactose, litmus, crystal violet,
agar), we are at once placed into this advantageous posi-
tion, that we can not only keep back or exclude (by the
use of the crystal violet) bacteria other than coli-typhoid,
but we can in positive fashion and in isolated aspect recog-
nise at once those colonies which are not B. typhosus.
Any colony which after 24-36 hours' incubation at 37° C.
appears on the medium red in colour surrounded by a red
halo (reddening of the litmus constituent of the medium due
to rapid acid production from, i.e., fermentation of the lactose
by the bacteria constituting the said colony), cannot be one
of B. typhosus, but must be one of B. coli, probably B. coli
communis ; if the medium did nothing else than this, it alone
would for obvious reasons be a great help, for we could from

16

our further study and search for the B. typhosus safely
exclude and disregard those red-haloed red colonies. But
the medium does more than this, for every red colony in
general, and every colony which is neutral in colour, i.e.,
neither red nor blue, can likewise be excluded as being
certainly not a colony of B. typhosus, because the colonies
of B. typhosus after one, two, or, better, three days at 37° C.
appear bluish or more or less violet-blue. As will presently
be further stated, not every " blue " colony is necessarily one
of B. typhosus, but if we find in our plate colonies of blue
or violet-blue colour, we have a guide to which colonies we
have to direct our attention in our further study and tests.
How important a step in advance of all others this method
of Drigalski-Conradi is, can best be estimated by the following
illustration : if, for instance, we are working with a given
material (water, milk, shellfish), in which by previous
analysis by means of Drigalski plates we have ascertained
the number of B. coli and the absence of blue colonies in a
definite amount of that material, we can, after adding to the
material a trace of typhoid culture, without any difficulty
ascertain by Drigalski plates the number of B. typhosus in
any given amount of the material by merely counting the
blue or violet-blue colonies which have made their appear-
ance in the plates. Of course we would also be able, if
necessary, to make from these blue colonies the further tests
for B. typhosus. In the experiments to be presently described,
the enumeration by Drigalski plates of the number of B.
typhosus introduced into oysters or taken up by them while
in sea water, to which a small amount of a pure culture of
B. typhosus had been added, was easily carried out ; indeed,
such exact determination, as will presently be shown, was
made possible because we had this method at our disposal.

The oysters, with the exception of those in Experiment IV,
selected for our experiments, as also the mussels and cockles,
were all clean, containing no microbes of the coli-typhoid
tyPe- The sea water used for the experiments was clean, and
free of any microbes of the coli-typhoid type — in fact, the sea
water had been previously sterilised. The infecting material

17

was a pure cul ture of our laboratory B. typhosus. Under these
conditions, therefore, the presence of any B. typhosus in the
oysters or in the sea water could be determined readily and
at once numerically by means of Drigalski plates, since
all colonies of the colour and appearance resembling those
of B. typhosus could, without hesitation, be declared as
those of B. typhosus. The other tests : microscopic exami-
nation in the hanging drop, agglutination with typhoid
serum, subcultures in MacConkey fluid, neutral red broth,
litmus milk, streak and shake gelatine culture, would fully
confirm the diagnosis. We shall have later an opportunity to
describe experiments with oysters of a certain locality —
experiments made not with B. typhosus but with B. coli — in
which the Drigalski plates revealed the presence of microbes
whose colonies in their blue colour and general appearance
bore a great resemblance to those of B. typhosus, but which
by microscopic tests and by subculture in the various
media could be recognised as different; but in our experi-
ments (except Experiment IV) of testing the vitality of the
B. typhosus in oysters, cockles, and mussels, no such dis-
turbing microbes were present at starting, and under the
conditions above stated none could have been afterwards
present to disturb the simplicity of the procedure. This can
in no way interfere with the general results obtained, since
our object was to determine how far and to what degree and
in what manner living oysters, cockles, and mussels as such
have the power to deal with the B. typhosus that have
had access to them. Whether other microbes are present
in the oysters or whether other additional microbes are
introduced with the B. typhosus are questions which do not
materially alter the simple and fundamental problem, viz. —
can, and to what degree do shellfish deal with the B.
typhosus? — and, therefore, the simpler the conditions for
elucidating it, the more accurate, it may be expected, will be
the result. There is one further point which, at the outset,
has to be stated here — this is the character of the B. typhosus
on Drigalski medium in surface plates. We mentioned
above that it was by this method that we analysed the shell-

C

18

fish and the water, and that by this method we were enabled
to determine the number of B. typhostis introduced into
the shellfish or into the surrounding water.

Now, what are the characters by which the B. typhosus
can be readily recognised by the Drigalski-Conradi plate
method ?

A given small amount of water or of substance of shell-
fish— up to O'l c.c.* — containing a limited number of B.
typhosus is uniformly rubbed, after the Drigalski method,
over the surface of the medium (Nutrose, litmus, lactose,
crystal violet, agar), previously set — about quarter-inch
depth — in a flat plate dish — the plate dish which I use is
four-and-a-half inches in internal diameter and seven- eighths
of an inch deep — and the plate is then placed in the incu-
bator at 37° C.

Inspecting the plate after 24 hours, the typhoid colonies
are at once recognised as isolated round translucent blue dots ;
when inspected with a glass in semi-reflected light, they are
violet-blue, and well differentiated from the purple medium ;
the colonies are moist looking, thin at their margin, a little
thicker in the centre ; after 48 hours, and better still after
72 hours, the colonies are several millimetres in breadth, bluish
in the middle, violet at the thin margin, which latter at the
same time has lost its regularly circular outline, being slightly
irregular ; in transmitted light the substance is distinctly
but finely granular, and, owing to the prominent thicker
centre, the colonies look more or less Hke limpets, being low
conical. When a trace of the colony is distributed in a little
sterile bouillon, the bacilli which constitute the substance of
the colony are seen to be shorter or longer cylindrical rods,
many of them actively motile. When tested according to
Koch-Drigalski's method, by mixing a few drops of the bouillon
•emulsion with a trace — a small platinum loop — of blood
serum of a typhoid-prepared animal, it will be seen that
arrest of motility and distinct agglutination into large

* This amount can be easily spread out and rubbed over the plate
surface, without leaving any excess fluid — even 0-15 c.c. can be so dealt
with ; more than that cannot be satisfactorily managed.

19

compact clumps occur within a minute or two. As wo
shall point out later in detail, colonies may be blue or bluish
or blue-violet, without being those of B. typhosus, but the
above differential characters, viz., conical in shape, with
prominent centre, flat thin margin, violet-blue in the middle
part, violet in the margin when viewed in reflected light on
black ground, finely granular, moist or glistening in aspect ;
the individual bacilli short cylindrical in shape (not fila-
mentous and not in chains), motile and quickly clumping
and in marked manner with typhoid serum, are sufficient
presumptive indications * of the colonies beiog those of B.
typhosus. Subcultures in the different media are made as a
matter of routine, so as to confirm the diagnosis.

We proceed now to describe in detail the experiments
which we made with oysters, cockles, and mussels.

In all our experiments with oysters, the method used
was this : the oyster, after the outside of the shell had been
thoroughly washed and brushed under the tap, was opened
with a sterile knife, the liquor was drained off as completely
as possible, the body of the fish with its mantle and branchiae
was then transferred to a sterile glass dish and herein cut up
(minced) with sterile scissors as finely as possible; after
thoroughly mixing the minced material, the fluid (thick
turbid) is removed with sterile glass pipette and measured.
From this fluid a definite amount, in no case more than
O'l c.c. or 0*15 c.c. (generally the former quantity), was
either directly transferred to a Drigalski plate, or, as in those
cases in which the presence of a large number of B. typhosus
in the oysters could be supposed, -fa c.c. of the oyster-mince
was first diluted by a measured amount of sterile sea water,
and of this dilution -fa c.c. was used and dealt with on
the Drigalski medium in the plate. If the number of

* Flat colonies, deep blue in reflected light, fringed at margin, dry
looking, are not B. typhosus; colonies bluish or pale blue in reflected
light, with greenish margin, uniformly raised, moist looking, are not B.
typhosus ; colonies bluish violet, strongly granular, with thin margin, but
composed of filamentous bacilli, are not B. typhosus, they do not grow on
Drigalski plate at 37° C. Very small, blue, uniformly-raised, colonies may
be those of streptococci or vibrios.

c 2

20

B. typhosus in the oyster could be expected to be small, as
for instance in the later oysters of a series, more than one
Drigalski plate was made directly with the fluid of the
minced oyster, each plate receiving 0 * 1 c.c. After having, by
means of the sterile bent glass rod, carefully and thoroughly
and uniformly rubbed the material over the surface of the
dry medium (all previous moisture having been previously
removed by allowing the plates to evaporate it spontaneously
for 2-3 hours in the incubator), the plates are transferred to-
the incubator at 37° C.

As mentioned above, the typhoid colonies are noticeable
already after 24 hours, and a preliminary counting can now
be made, but it has to be controlled after the plate has been
placed back in the incubator for at least another day, gene-
rally two more days, because the character of the colonies-
can by this time be considered fully established. By this
time, colonies, which after the first 24 hours' incubation
might be doubtful typhoid colonies, can with certainty be
declared to be or not to be typhoid, and all those which
show the above differential characters in the same manner
may be taken to be typhoid colonies ; agglutination test and
subcultures on gelatine, and if necessary in other media, are
made, selecting at random from different quarters of the
plate one colony for the purpose. Since after two, or better,
three, days' incubation at 37° C. the character of all the
colonies in the plate — typhoid and not typhoid — is fully
established, and since the typhoid colonies are distinct and
different from all others by colour, size, general aspect, and
shape, there is no difficulty in at once diagnosing them and
to recognise their identity,* and it is not therefore necessary
to test more than a few of them for agglutination and sub-
culture on gelatine. The growth on this gelatine subculture
(after 24 hours' incubation at 20° C.) is inspected, examined
in the hanging drop and tested for agglutination, and must,
if B. typhosus, comply with the required tests : on gelatine
translucent filmy growth, composed of cylindrical motile
bacilli, agglutinating markedly and instantaneously with a

* See photograms accompanying this Report.

21

trace of typhoid serum of a prepared rabbit, just like a
similar gelatine culture of the laboratory B. typhosus kept
and tested for control.

Subcultures from the Drigalski plate in the different media
(neutral red broth, litmus milk, MacConkey fluid, lactose
litmus peptone, shake gelatine, phenol-broth, Proskauer and
Capaldi I and II) are made ; if there were any doubt about
a colony, further similar subcultures are made in each
series of oysters from the Drigalski plates of the first and
last oyster, and occasionally in addition from one or the other
in the middle of the series. Intraperitoneal injection of
guinea-pigs with a given dose (J-^o or ^ess) °^ a ^4 hours'
old agar culture was practised for testing the virulence of
the microbe, generally only of the first and last oyster of the
series from which the microbe was recovered, and in all cases
it was found to have retained its full virulence.

EXPEEIMENT I.

Clean Burnham Oysters. — These were obtained from one
of the foremost oyster shops in the City. The oysters were
thoroughly cleaned on the outside of the shell, and one was
used for a preliminary test by means of a Drigalski plate for
the presence of B. coli and for that of blue colonies, the rest
were placed in a clean wooden tub * in 2000 c.c. of clean sea
water. The tubs used were oval in shape and could hold
easily 12-16 natives in one layer ; the sea water was always
sterilised by heat (90°-100° C.), and after cooling well and
repeatedly shaken up with air. As a matter of routine we
always prepared the 2000 c.c. of sterile sea water the day
previous to using it. Having ascertained after 24 hours'
incubation at 37° C. by the Drigalski plate that the test
oyster was clean, we now proceeded to infect twelve of the

* Each tub after having been used was well brushed under the hot
water tap, and was then kept filled with the hot water 80° C. for some
time (hours). This process was, as a rule, repeated on two or even three
subsequent days. In the later experiments the tubs were even " steamed "
before being used again.

22

oysters which, had ,been in the sterile sea water for the 24
hours, in the following manner : a thick zinc wire, bent at
right angle at the last inch, is carefully inserted between the
two halves of the shell, the oysters in the sea water having
their shells spontaneously opened ; immediately as the wire
end is inserted the shell closes so tightly on it that the oyster
can now be lifted out of the water on to a glass plate ; the
end of the cannula of a hypodermic syringe filled with turbid
emulsion of B. typhosus is easily introduced close to the
wedged-in wire, and the desired amount — (in our case 1 c.c.)
— of culture slowly injected. When finished the cannula of
the syringe is withdrawn — the cannula being thinner than
the wedge — and by gentle action the wire is removed ; the
shell immediately closes again tightly. The whole pro-
ceeding need not and does not occupy more than a few
minutes, and, as subsequent observation showed, no harm had
thereby been done to any part of the oysters. In the above
manner 12 oysters were injected with the B. typhosus, each
with 1 c.c. of the emulsion.

The emulsion injected was obtained by distributing in
sterile sea water the growth from the surface of a pure agar
culture 24-48 hours old of our laboratory B. typhosus.
Before using the emulsion for the injection, a Drigalski plate
was made with a definite amount of a definite dilution of it,
and after 48 hours' incubation at 37° C. the number of
colonies of B. typhosus was counted, and thereby their
number per 1 c.c. of emulsion ascertained. It was thus
found that the number of B. typhosus injected into each
oyster was 162,500,000.

The determination was made thus : 0 • 1 c.c. of the typhoid
emulsion was added to 100 c.c. sterile sea water, well shaken ;
of this dilution 0*01 c.c. was carefully and well rubbed over
the surface of Drigalski plate. After incubation for 48 hours
the Drigalski plate showed numerous colonies of B. typhosus
and no others, all being of the same colour and aspect ; a
careful count showed 1625 colonies in the plate, that is to
say, 1625 x 100 x 1000 = 162,500,000 per 1 c.c.

Although the amount actually injected within the cavity

23

of the shell was 1 c.c., it has to be mentioned that while
the injection was proceeding a little fluid, about the same
quantity that was being injected, was escaping from the oyster
near the lock, so that although this escaped fluid appeared
to be the clean water from within the shell, and although
the injection was made fairly gradually, we cannot suppose
that none of the injected bacilli escaped with the water. At
any rate, the above amount of B. typhosus was injected into
the shell of each of 12 oysters. Of these six were put back
into the sterile sea water, the other six were transferred to a
plate and placed in the cold chest. The first six will be
designated as " wet oysters," the other six as " dry oysters."
The analyses of the latter would show whether any, and what,
changes took place in the number of B. typhosus as com-
pared with the wet oysters. It is well known that oysters
after they are removed from their ground are, in many
instances, not consumed at once, but are occasionally kept
for days at the wholesale dealer's or the retailer's in a " dry "
state, in barrels, bags and the like ; in fact, oysters imported
from a distance must of necessity be so kept. As regards
the "wet" oysters the sea water (2000 c.c.) was changed
after 24, 48, 72, 96 and 144 hours, that is to say, after one,
two, three, four and six days.

The oysters analysed were taken in this order : Oyster 1,
wet, after having been one day in clean water ; oyster 2, dry,
having been kept dry one day ; oyster 3, wet, after two days
in clean sea water ; oyster 4 having been kept dry for two
days, and so on.

Oyster 1 was taken out of the sea water after one day, its
outer surface well brushed under the tap, then dried with a
clean cloth, opened with sterile knife, the liquor drained off
as well as possible, then minced with sterile scissors in a
sterile glass dish, well mixed and the turbid fluid measured.
It amounted to just one cubic centimetre. From this fluid
made two Drigalski plates, each with yj^ c.c., i.e., 10 cubic
millimetres.

After incubation for 48 hours the colonies, all of the colour,
aspect, and nature of B. typhosus — there were no others —

24

were carefully counted. They were recounted after three
days' incubation, and were found to amount to the average
of 700 per TJo c.c., that is to say, the whole oyster con-
tained 70,000 B. typhosus. Exactly the same procedure was
followed with oyster 2 dry. The amount of fluid was also
just 1 c.c. The average number of typhoid colonies present
in the two Drigalski plates (each inoculated with TJ ^ c.c.)
amounted to 1,200,000 B. typhosus for the whole oyster. Of
the subsequent wet oysters the amount of fluid, after mincing,
was practically the same, viz., just 1 c.c. Of the dry oysters
the amount was less, but it was always brought up to just
1 c.c. by adding sterile water.

Oyster 3, wet — 2 days in clean sea water showed 9100 B. typhosus

per oyster.

,, 4, dry — 2 days dry showed 175,000 B. typhosus per oyster.
„ 5, wet — 3 days in clean sea water showed 1100 B. typhosus
per oyster.

Seeing from the result in oyster 5 that the number of
B. t. is rapidly diminishing, I used for the Drigalski plate of
oyster 7 and oyster 9 TTQ c.c. of the oyster, and in the case
of oyster 11 I made three plates, each with Y1^ c.c. of the
oyster.

Oyster 6, dry — 3 days dry showed 42,000 B. typhosus per oyster.
,, 7, wet — 4 days in clean sea water showed 320 B. typhosus

per oyster.

,, 8, dry — 4 days dry showed 3700 B. typhosus per oyster.
„ 9, wet — 6 days in clean sea water showed 0 B. typhosus

in T^ part of oyster.

„ 10, dry — 6 days dry showed 40,000 B. typhosus per oyster.
„ 11, wet — 7 days in clean sea water showed 0 B. typhosus

per T3o of oyster.
„ 12, dry — 7 days dry showed 1220 B. typhosus per oyster.

All the oysters had their shell well and tightly closed,
and on opening were found to be quite normal in appearance,
plump and juicy.

In all the preceding experiments the Drigalski plates
contained practically no other colonies except those of

25

B. typhosus, and these, after 48 hours and 72 hours, could
readily be identified as such. It should also be added that
the counting of the colonies, except in the case of dry
oyster 2, presented no difficulties, and was always repeated
to control the first counting. As was mentioned already, in
all instances an accurately measured quantity of the turbid
fluid part of the minced oyster, TJo or -^ c.c. as the case
required, was used for Drigalski plates, and the total quantity
of the minced oyster was kept at just 1 c.c. Where originally
deficient, it was brought up to 1 c.c. by the addition of
sterile sea water.

In all plates the colonies were found isolated, not in fused
groups, thus proving that the bacilli were fairly uniformly dis-
tributed in the fluid of the minced oyster, and had not formed
nests, as it were, in the oyster tissues ; that is to say, had not
multiplied and made aggregations within the tissues of the oyster.
As was mentioned on a former page, colonies were taken
at random, and the required tests — examination in the
hanging drop, agglutination test, subculture on gelatine,
agglutination of this, and ultimately, if required, in other
media — were carried out. After a little practice the recog-
nition of the typhoid colonies on Drigalski plates in all
these and the subsequent experiments was merely a matter
of patient examination under a magnifying glass.

Tabulating the results of the preceding Experiment I. we
obtain this : —

TABLE I.

Oysters injected with 160 millions B. typhosus each :
WET — i.e., kept in clean sea water frequently changed.

Oyster 1 — after 1 day in water, 70,000 B. typhosus per oyster.
„ 3— „ 2 days „ 9100

» 5— » 3 „ „ 1100

» 7— „ 4 „ „ 320

„ 9 — „ 6 „ „ 0 „ per TXF part

of oyster.
» 11 — » 7 „ „ 0 „ per -f$ part

of oyster.

26

DRY — i.e., kept out of sea water.
Oyster 2— after 1 day . . 1,200,000 B. typhosus per oyster.

. 175,000

„ 6— „ 3 „ . . 42,000

„ 8— „ 4 „ . . 3700

„ 10— „ 6 „ . . 40,000

„ 12— „ 7 „ . , .;-.:, 1220

We learn from this experiment that oysters infected
with huge numbers of B. typhosus, then kept in clean sea
water changed frequently — practically every day — were able
to clean themselves and to get rid of them in a compara-
tively short space of time; in four days the number of
B. typhosus decreased to an enormous extent (320), and
after six days none could be found in ^ part of the oyster,
that is to say, less, than 10, if any, in the whole body of the
fish. At the same time we learn the important fact that oysters
of the same kind kept out of the water retained the injected
B. typhosus to a markedly greater extent (40,000 after six
days), although also under these conditions their number
considerably decreased. This part of the experiment, while
pointing out the danger attached to specifically infected
oysters being kept out of the water, shows at the same time
that the body of the oyster per se is not a soil in which the
typhoid bacillus is capable of multiplying ; on the contrary,
the tissues of the oyster distinctly acting inimically on the
microbe. Those oysters which were kept after infection in
fresh sea water might, one would perhaps be inclined to
conclude, have cleaned themselves of the extraneous B.
typhosus — extraneous to the oyster — on account of being
kept in changing water, but this evidently does not apply to
those oysters that were kept out of the water ; consequently
we are justified in, in fact are driven to, concluding that the
tissues of the oysters per se are endowed with the faculty of
devitalising this microbe. Considering that we started with
160 millions of B. typhosus per oyster we come down to
1220 in the course of seven days, during which time the
oysters were left entirely to themselves and without any
influence the surroundings could have exerted on them.

27

EXPEEIMENT II.

Clean Colchester Oysters. — One oyster well brushed on out-
side and prepared in the manner already described, -^ part of
the oyster contained no microbes capable of growing at 37° C.
on Drigalski medium, that is, it contained no B. coli or any
microbes forming blue colonies on that medium. The rest
(25) had been also thoroughly brushed on the outside under
the tap and then placed in clean tub in sea water (4000 c.c.)
to which previously, while sterile, emulsion of a pure culture
of B. typhosus had been added to the extent that each cubic
centimetre contained 744,000 B. typhosus. This determina-
tion was made in the following manner : to the 4000 c.c. of
sterile sea water were added 4 c.c. of a turbid emulsion of
B. typhosus, prepared by well shaking up the growth covering
the surface of 48 hours' old agar culture of B. typhosus with
10 c.c. of sterile sea water. Immediately after the addition
of the 4 c.c. of the typhoid emulsion to the 4000 c.c. of
sterile sea water and well shaking it up, 1 c.c of the infected
water was added to 99 c.c. of sterile distilled water, and of
this dilution -j^ c.c. was used for one Drigalski plate. After
48 hours' incubation the colonies, all of B. typhosus, were
carefully counted and found to amount to 744, so that
1 c.c. of the infected sea water contained 744 x 10 X 100 =
744,000 B. typhosus.

The oysters having been kept in the infected sea water
for 24 hours were taken out, well rinsed on the outside and
drained, were divided in two lots, one lot (12) were placed in
the cool chest dry, the other lot (12) were transferred to a
clean tub sterilised by steam, supplied with 2000 c.c. of
sterile sea water, and the remaining oyster 1 was, after well
brushing it under the tap, used for analysis. The oyster
having been opened with a sterile knife, and the fluid drained
off as carefully and as well as possible, the whole fish was
finely minced with sterile scissors. Total amount of turbid
fluid drained off of the minced material was 1 • 75 c.c. ; from
this made two Drigalski plates each with 150 cubic milli-
metres. After incubation the enumeration of the two plates,

28

that is, 0*3 c.c. of the oyster, showed, as near as could be
counted, 6400 colonies, that is, about 40,000 B. typhosus for
the whole fish. As regards the oysters kept in sea water in
the tub, the tub and the sterile sea water were changed after
one, two, three, five, six, and seven days ; oysters of the wet
lot were analysed one day after change, two, five, six, seven,
and nine days ; of the dry lot we analysed after one day dry,
two, three, five, six, seven, and nine days dry.

One day after change, oyster 3 (wet) — total amount of
fluid 3 c.c., ^Q c.c. for one Drigalski plate, contains 46 colonies
of B. typhosus— this would amount to 46 x 30 = 1380 B.
typhosus per whole oyster.

One day dry, oyster 2 — total amount of fluid 4 c.c.,
^0- c.c. for one Drigalski plate, contains 1000 colonies of
B. typhosus— this would amount to 1000 x 40 = 40,000 B.
typhosus per whole oyster.

Two days after change in sterile sea water, oyster 5 (wet)
— total fluid 2 c.c. ; two days, oyster 4 (dry) — total fluid
1 c.c. ; of each oyster -fa c.c. for one Drigalski plate.

Total number of colonies of B. typhosus in Drigalski
plate of oyster 5 were 22, that is, 22 x 20 = 440 B. typhosus
per whole oyster ; Drigalski plate of oyster 4 contained 377
colonies of B. typhosus — this amounts to 3770 B. typhosus
for the whole oyster.

Three days dry, oyster 6 — total fluid 1 c.c. ; -fa c.c. was
used for one Drigalski plate, which developed 70 colonies of
B. typhosus — this amounts to 700 B. typhosus for the whole
oyster.

Five days after change in sterile sea water, oyster 7 —
total fluid a little under 2 • 8 c.c., of this made two Drigalski
plates each with -fa c.c.

Total number of colonies of B. typhosus in the two plates
was six (two in one, four in the second plate) — this would
amount to about 82 B. typhosus for the whole oyster.

Five days dry, oyster 8 — total amount 1 • 5 c.c. ; -fa c.c.
produced in a Drigalski plate 10 colonies of B. typhosus —
this would amount to 150 B. typhosus per whole oyster.

Six days after change in sterile sea water, oyster 9 — total

29

amount 3*3 c.c.; made two Drigalski plates, each with 150
cubic millimetres (0'15'c.c.). Both plates together had 4
colonies (three in one, one in the other) of B. typhosus — this
would be 44 B. typhosus for the whole oyster.

Six days dry, oyster 10 — total amount 0*4 c.c.; -fa c.c.
produced in a Drigalski plate 70 colonies of B. typhosus —
amounts to 280 B. typhosus per whole oyster.

Seven days after change in sterile sea water, oyster 11 —
total amount 3 c.c. ; T^ c.c. produced in a Drigalski plate
0 colonies of B. typhosus.

Seven days dry, oyster 12 — total amount of fluid 1*5 c.c. ;
T^ c.c. produced in a Drigalski plate 34 colonies of B. typhosus

— this is equal to 510 B. typhosus for the whole oyster.

Nine days after change in sterile sea water, oyster 13 —
total amount 2 c.c. ; -fa c.c. produced in a Drigalski plate
0 colonies of B. typhosus.

Nine days dry, oyster 14 — total amount 1-5 c.c. ; fa c.c.
produced in a Drigalski plate six colonies of B. typhosus

— this amounts to 90 B. typhosus per oyster.

Tabulating the results of Experiment II we find thus : —

TABLE II.

Oyster 1 — after 24 hours in typhoid-infected sea water contained

40,000 B. typhosus.

„ 3 — after 1 day in clean sea water 1380 B. typhosus.

„ 5— „ 2 days „ „ 440

55 ' 55 5 55 55 55 82 ,,

J) " 55 " ,, 5, ?> 44: „

55 11 55 « 55 55 55 0 „

55 13 - ,5 9 „ „ „ 0 „

Oyster 2— after 1 day dry ..... 40,000 B. typhosus.

„ 4— „ 2 days „ ..... 3700

„ 6— „ 3 „ „ ..... 700

55 8 — „ 5 „ „ ..... 150 „

„ 10— „ 6 „ „ . . . . . 280

„ 12— „ 7 „ „ ..... 510

55 14— „ 9 „ „ ..... 90

30

Allowing for the much greater number of B. typhosus
introduced into the oysters of the first experiment, the
results of Experiment II harmonise well with those obtained
in Experiment I, namely: the rapid decrease and equally
rapid total disappearance of B. typhosus from the oysters
which, after infection, were kept in clean sea water repeatedly
changed; while in oysters at the same time and manner
infected, but afterwards kept out of the water (dry), the
decrease, though taking place, is much slower: after the first
day dry (oyster 2) no decrease being noticeable, the oyster
containing the same number of B. typhosus as the oyster 1
immediately after infection, viz., 40,000 ; whereas in oyster 3
that had been kept 24 hours in clean sea water the number
had decreased considerably — to 1380.

' Also from this experiment we are justified in concluding
that the decrease of the B. typhosus in the oysters could not
have been due to a simple " washing out " process, but must
be due to the capability of the oyster to directly devitalise
the B. typhosus, being something alien to the tissues of the
oysters and not capable of maintaining its existence therein ;
the dry oysters are clear proof for this conclusion.

That this function of the destruction of the B. typhosus
by the tissues of the oysters per se would be more marked
and extensive in those that were kept after infection in clean
sea water constantly changed than in those kept out of the
water is to be expected, since in the former the ordinary pro-
cessesof the tissueswould go onunabatedand innormal fashion,
which could not be the case in oysters kept out of the water.

All the oysters of this experiment were, on opening, found
in all respects normal, plump and juicy, their shell well closed.

EXPEEIMENT III.

This experiment is in reality a continuation of Experi-
ment II in this sense, that several oysters of the same batches
left over from Experiment II were subjected to reinfection
and analysis ; at the same time the sea water into which
after infection they were transferred, and which was frequently

31

(every 24 hours) changed, was analysed for B. typhosus, in
order to obtain an insight into the problem whether and to
what extent the decrease of B. typhosus is referable to a
" washing-out " process. The experiment was made in the
following manner : —

Of Experiment II six oysters of the wet lot and four
oysters of the dry lot were left over ; they were transferred
to a fresh sterile tub into 2000 c.c. of sterile sea water ; the
tub and the 2000 c.c. sterile sea water were changed every
day for three days. Seeing that the previously wet oysters
were already free of B. typhosus at the termination of Ex-
periment II, and seeing that in the previously dry oysters
the number of B. typhosus had, by the end of Experiment
II, come down to 90 per oyster, it was quite in accordance
with fact to suppose that if these previously dry oysters
are placed for further three days in clean sea water they
would be free of the microbe. I am referring to the ascer-
tained fact that in clean sea water the previously wet oysters
had in two days from 82 B. typhosus per oyster come down
to 0 ; three days in clean sea water would therefore, in all
probability, bring down the number 90 B. typhosus (oyster
14) to 0 B. typhosus. At any rate, whether or no some stray
B. typhosus are left in the oyster, it would not fundamen-
tally alter the nature of the Experiment III, in which the
remaining oysters were placed in sea water reinfected with a
large number of B. typhosus.

To 2000 c.c. of sterile sea water in a flask, emulsion of B.
typhosus, made by distributing in sterile sea water a 48 hours
old pure agar culture of B. typhosus, was added. 1 c.c. of
the infected sea water of the flask was added to 99 c.c. sterile
water; of this dilution a Drigalski plate was made with
2 J-Q- c.c. The infected sea water was then poured over the
above 10 oysters (two batches) in a clean sterile tub. These
two batches were kept well separated in the tub, and will be
described as "previously wet" and "previously dry " oysters,
both batches, however, being kept, now and afterwards, in the
sea water in the tub, as will be presently described.

The Drigalski plate yielded on incubation 118 colonies of

32

B. typhosus, that is, 118 x 200 X 100 = 2,360,000 B.
typhosus per 1 c.c., or a little over 2J millions.

Of tlie oysters having been kept for 24 hours in the
infected sea water, one of the " previously wet " lot and one
of the " previously dry " lot were taken for analysis, as also a
certain amount — same manner of dilution as above — of the
sea water in the tub ; the remaining oysters were all taken
out of the infected water — keeping the two batches separate
— well rinsed under the tap on the outside, and transferred to
a fresh clean tub with fresh sterile sea water.

Of the sea water in tub infected with B. typhosus 24 hours
previously, 1 c.c. was added to 99 c.c. of sterile water ; of
this dilution a Drigalski plate was made with ^ c.c. This
plate on incubation yielded 63 colonies of B. typhosus ; this
would mean that the infected sea water contained 126,000
B. typhosus per 1 c.c.

Oyster 15 (previously wet), after having been in the
infected sea water for 24 hours — total amount of fluid of the
minced oyster, 3 c.c. ; -^ c.c. of this added to 10 c.c. sterile
sea water, and of this dilution took ^ c.c. for one Drigalski
plate. This plate yielded a pure culture of B. typhosus — 28
colonies; this would mean that oyster 15 contained 28 X 3000
= 84,000 B. typhosus per oyster.

Oyster 16 (previously dry), after having been in infected
sea water for 24 hours, was opened ; total amount of
fluid of the minced oyster, 2 c.c. Of this added -fe c.c. to
10 c.c. sterile water ; with -fa c.c. of this dilution made one
Drigalski plate, which yielded a pure culture of B. typhosus
—659 colonies ; this would mean 659 x 2000 = 1,318,000
B. typhosus per whole oyster. If we suppose that each of
the " previously wet " and the " previously dry " oysters took
out of the infected sea water the same number of B. typhosus-
as oyster 15 and 16 respectively, we would get 84,000 x 6
for the "previously wet" lot and 1,318,000 X 4 for the
" previously dry " lot, that is —

504,000 B. typhosus for the six of the former,
5,272,000 „ „ four of the latter.

5,776,000 Total.

33

The sea water had been infected with B. typhosus to the
amount of 2,360,000 B. typhosus per 1 c.c., and as there
were 2000 c.c. of the sea water, we have then 2,360,000 X
2000.

4,720,000,000 B. typhosus had been originally present in
the 2000 c.c. of the sea water in the tub in which the oysters
were placed ; the ten oysters had therefore removed in 24 hours
from this total only 5,776,000 B. typhosus, so that there should
have remained in the tub after 24 hours — nothing else happen-
ing— a little over 4660 millions of B. typhosus. But accord-
ing to our analysis the sea water in the tub after 24 hours
contained only 126,000 B. typhosus per 1 c.c., that is,
126,000 X 2000 for the total; that is to say, the total sea
water now — 24 hours after infection — contained only 252
millions of B. typhosus. The number of B. typhosus in the
sea water in the tub must have suffered a decrease from
4660 millions to 252 millions, that is, not more than -fa part
of the original number were left. There are no data as to
an active destruction of B. typhosus going on in the oysters
during the same 24 hours, but we may, without much danger of
error, assume that the chief destruction of the microbes was
going on in the sea water itself. The sea water was, before
being infected, sterile sea water; the oysters yielded no
microbes except B. typhosus taken in from the infected water.
It follows from this that the just named rapid destruction in
the sea water in the tub could not have been caused by
the presence of other microbes, but must be referred to an
inimical action of the sea water itself. This was proved
directly by experiment ; 100 c.c. of sterile sea water were
infected with a given number of the same B. typhosus and
kept for 24 hours. The analysis showed that the diminution
amounted to -f-r, against the figure ^ found in the above
experiment. To show that this inimical effect of our sterile
sea water was not due to the sterility of the sea water, but to
the sea water as such, the experiment was made by comparing
in a parallel series the effect of non- sterile sea water exactly
as it had been received from the same portion of the sea
(Lowestoft) from which all our sea water was obtained — that

D

34

is to say, at the same time that the 100 c.c. of the sterile sea
water were infected with a given number of B. typhosus,
100 c.c. of the non-sterile sea water were infected with the same
amount of the same culture of B. typhosus ; 24 hours after,
a Drigalski plate was made with a definite amount and the
number of typhoid colonies ascertained. It turned out that
in the non-sterile sea water the decrease of the number of the
B. typhosus was practically the same, viz., between -fa and -^7.
It is not necessary to enter here into the details of these
experiments, since they were not strictly within the scope of
the shellfish inquiry, but the result is clear, viz., that the sea
water per se had a powerful destructive action on the B.
typhosus. By saying this I do not intend to omit another
important fact, viz., that although sea water is capable ot
materially reducing already in 24 hours the number of
B. typhosus, the reduction does not go on at the same great
rate every subsequent 24 hours, for it has been experimentally
shown by myself, Professor Herdman, Boyce, and others
that in sea water infected with B. typhosus some living
individuals can be recovered from large amounts of the water
even after many days and weeks.

We proceed now with our original analyses.

As mentioned above, the oysters after having been
removed from the infected water were placed for 24 hours
in sterile sea water in fresh tub ; and this change, both of
tub and sterile sea water, was effected every 24 hours.
Analyses of the sea water and of one oyster of the "previously
wet " and one oyster of the " previously dry " lot were made
every 24 hours after change of the sea water, with the
following results : —

Sea water of tub one day after change — -fa c.c. was used
directly for one Drigalski plate ; this yielded a pure culture
of B. typhosus, 25 colonies ; that is to say, this sea water
contained 250 B. typhosus per 1 c.c., or for the total
amount (2000 c.c.) 500,000 B. typhosus. This half-million
of B. typhosus in the water in the tub could have been
derived solely from the eight infected oysters in it, the sea
water having been sterile and the tub having been well

35

brushed and steamed before use. This would indicate that
of the eight oysters a number of living B. typhosus had
actually passed out of their interior.

Oyster 17, "previously wet/' opened 24 hours after
change into sterile sea water — total amount of fluid, 2 • 75 c.c. ;
with Jo c-c- °f tnis ^d made one Drigalski plate; this
yielded 17 colonies of B. typhosus (no other colonies) — this
amounts to 935 B. typhosus per whole oyster.

Oyster 18, " previously dry," 24 hours after change into
sterile sea water — total amount of fluid, 2 • 75 c.c. ; ^ c.c. of
this fluid was added to 0'9 c.c. of sterile sea water; of
this took T\y c.c. for one Drigalski plate. Plate yielded a
number of B. typhosus amounting for the whole oyster to
1900.

Assuming all the previously wet oysters and all the
previously dry oysters contained at the same date the same
number of B. typhosus respectively, we had before the change
into the sterile sea water a total stock of B. typhosus in the
eight oysters of 4,374,000 B. typhosus. At the end of 24
hours in sterile water they would represent a stock of 10,375
B. typhosus only, so that a large margin is here offered for
discharge of B. typhosus by the oysters into the sea water.
The presence of the 500,000 B. typhosus in this sea water
(2000 c.c.) as above found at that stage would, therefore, be
readily explained, although it must be evident that the
difference between 4,374,000 and 10,375 is too large to
permit of ascribing to the discharge of the half-million in the
2000 c.c. of the sea water, the entire cause of this great
reduction of the B. typhosus in the oysters, and we are
justified in concluding that besides a comparatively small
discharge of living B. typhosus from the infected oysters into
the surrounding sea water — even accepting the destruction of
B. typhosus going on in the water — the chief cause of the
great reduction in the number of B. typhosus within the
oysters is due to inimical action by the oysters themselves.
This would be in harmony with what we found in the " dry "
oysters of Experiment II.

Sea water after second change was analysed, -^ c.c. direct

D 2

36

being used for one Drigalski plate. The plate remained free
of any colonies.

Oyster 19, previously wet, after two days' changes — total
amount of fluid, 3 • 5 c.c. ; -j^ c.c. yielded three colonies of
B. typhosus — this amounts to 105 B. typhosus per oyster.

Oyster 20, previously dry, after two days' changes — total
amount of fluid, 3 • 5 c.c. ; ^ c.c. yielded 19 colonies of
B. typhosus — this amounts to 646 B. typhosus for the whole
oyster.

Sea water after third change yielded likewise no B.
typhosus per -fa c.c.

Oyster 21, previously wet, after three days' changes — total
amount, 2*6 c.c. ; fa part of oyster yielded no colonies of B.
typhosus.

Oyster 22, previously dry, after three days' changes —
total amount, 2*3 c.c.; fa c.c. yielded 31 colonies of B.
typhosus — this amounts to 713 B. typhosus for the whole
oyster.

The sea water was analysed after fourth and sixth
changes ; -fa c.c. direct yielded no colonies.

Of the oysters only two were left of the previously wet
lot, viz., oyster 23 and 25 ; neither of them yielded any colonies
of B. typhosus in fa part of oyster.

It will be noticed in this series that two days after change
the sea water per fa c.c. did not contain any B. typhosus ; at
this stage there were six oysters — four previously wet, two
previously dry — in the water ; assuming that all previously
wet and dry oysters contained, when placed in this sea water
24 hours previously, the same number of B. typhosus, viz.,
935 and 1900 respectively, the total number of B. typhosus
assumed to be in these six oysters would only have amounted
to 7540, so that in 2000 c.c. of the surrounding water, even
assuming that the whole of B. typhosus were passed out into
the water, it would have only amounted to between three
and four B. typhosus per 1 c.c. ; in fa c.c., therefore, none
would have been detected.

Tabulating the results of this Experiment III :

37

TABLE III.
SEA WATER.

Immediately after infection . 2,360,000 B. typhosus per 1 c.c.
Iday „ „ • 126,000

1 day after change . .•*•'. 250 „ „

2 days „ „ . . » 0 „ per TV c.c.

3 0

*> 5) J) » ... 55 55

4 55 55 5) * , • 5J JJ

6 „ „ ,, » ... 0 „

PREVIOUSLY WET OYSTERS.

Oyster 15 — 1 day after infection 84,000 B. typhosus per oyster.

„ 17 — 1 day after change 935 „ „

„ 19—2 days „ „ 105

„ 21—3 „ „ „ 0 „ per TV c.c.

„ 23—6 „ „ „ 0

» 25 7 55 55 55 0 55 „

PREVIOUSLY DRY OYSTERS.

Oyster 16 — 1 day after infection 1,318,000 B. typhosus per oyster.
„ 18 — 1 day after change 1900 „ „

„ 20— 2 days „ „ 646

j) 22 — 3 j, ,, ,, 713 3, ,,

From this Table III it will be seen that the previously
wet oysters cleared themselves of the B. typhosus of the re-
infection in a remarkably short period, from 84,000 24 hours
after infection to 105 after two days' (i.e., twice) change of
the water, and no colonies of B. typhosus could be obtained
from ^G c.c. — that is, if any, they must have been less than
10 — after three days' changes of the sea water ; in other words,
the oysters were even more successful in dealing with the B.
typhosus now than they were after the first infection (Ex-
periment II), for on looking back to Table II it will be
seen that from oysters originally infected with 40,000 B.
typhosus, even after six days in clean sea water, changed

38

every 24 hours, there were still B. typhosus recovered,
whereas in Experiment III we started with 84,000 per
oyster — i.e., more than twice the number — after reinfection,
and could discover no B. typhosus three days after change
to clean water.

The previously dry oysters, on the other hand, do not seem
to have acquired this power of dealing rapidly with the in-
gested B. typhosus during the first 24 hours (1,318,000 after
24 hours in infected water), and it took them an appreciably
longer time to clean themselves although kept in clean water ;
this can be easily understood if we remember that these
oysters had, before reinfection, been kept out of the water
for nine days, that is, under abnormal conditions. This may
well have detracted from the power of their tissues to regain
their full activity when replaced in clean water.

All the oysters of this Experiment III, like those of Ex-
periment II, on opening, looked quite normal, plump and
juicy, and their shells well and tightly closed.

EXPEEIMENT IV.

By this experiment it was sought to ascertain whether
oysters at starting sewage-polluted, that is, coming from
distinctly sewage-polluted beds, behaved in the same or
different way in regard to B. typhosus. For this purpose
oysters were taken from the foreshore of Southend, which,
as also other shellfish of the same locality, Dr. Nash, the
Medical Officer of Health for Southend, had distinctly
declared as sewage-polluted and dangerous, and against the
consumption of which he gave emphatic warning by public
placard.

Most of these Southend oysters (natives) were very small
— some not bigger than the size of a penny — and on the
outside extremely dirty. They were well scraped and
brushed under the tap till from all parts all mud had been
removed as carefully as possible. They were then placed in
a clean tub and covered with sterile sea water (2000 c.c.), to

39

which just previously of an emulsion of B. typhosus so much
had been added that each cubic centimetre contained
2,470,000 B. typhosus (for method, see previous experiment).

The oysters were kept in this typhoid-infected water for
24 hours, and after retaining for analysis one of the smallest
oysters (No. 1), the rest were taken out, well rinsed on outer
surface, and separated into two lots, each lot containing
about the same proportion of " small " and " full-sized "
oysters. Lot 1 (" wet oysters ") was then transferred to fresh
clean tub and covered with 2000 c.c. sterile sea water. The
other lot (" dry oysters ") was laid out on a plate and placed
in cool chamber. The small oyster No. 1 yielded on analysis
95,800 B. typhosus and 900 B. coli communis. It has to be
remembered that this oyster was a very small one, the shell
not larger than the size of a penny. In this case one-
hundredth part of the minced body of the oyster yielded on
a Drigalski plate 958 colonies of B. typhosus and 9 colonies
of B. coli communis.

It is not necessary to detail all the procedure in the analysis
of the sea water and the two lots of these oysters, since they
were the same as were described in the previous experiments,
and we can at .once proceed to give the summary of the
results :

The sterile sea water as also the tub for the wet oysters
were changed every day.

Sea water immediately after infection 2,470,000 B.typh.per Ic.c.

of tub, 24 hours,, „ 1,530,000 „

„ ,, 1 day after change 13,180 ,, „

„ 2 days „ „ 10,580 „

A Of)

35 * 33 33 35 AV 5) 33

„ 6 „ „ „ 0 „ per TV c.c.

33 33 * 35 53 53 ^ 3J 53

It will be seen from this that one day after having been
changed the sea water still contained 13,180 B. typhosus per
1 c.c., which could only have been derived from the interior
of the oysters, since the sea water had been sterile when
added to the infected oysters, and the tub well brushed and

40

cleaned; and the same applies also to the sea water two
days after change (10,580 B. typhosus per 1 c.c.), i.e., the
second lot of originally sterile sea water. It will presently
appear that the oysters of this lot still contained at this
period an enormous number of B. typhosus in the interior,
and therefore the conclusion is obvious, viz., that the above
B. typhosus in the sea water after change had been passed
out by the infected oysters. This is confirmed by the further
fact that as soon as the number of B. typhosus in the oysters
markedly decreased (see below), no B. typhosus could be
discovered in -^ c.c. of the surrounding sea water.
The analysis of the wet oysters showed : —

Oyster 1 (very small) — after 1 day in infected water contained

95,800 B. typhosus per oyster ; 900 B. coli com.
„ 3 (small) — after 2 days in clean water, 752,800 B. typhosus

per oyster ; no B. coli com.
„ 5 (very small) — after 4 days in clean water, 1200 B.

typhosus per oyster ; no B. coli com.
,, 7 (very small) — after 6 days in clean water, 200 B. typhosus

per oyster ; no B. coli com.
„ 9 (medium sized) — after 7 days in clean water, 378 B.

typhosus per oyster ; no B. coli com.
„ 11 (full sized) — after 8 days in clean water, 56 B. typhosus

per oyster ; no B. coli com.
„ 13 (small size) — after 9 days in clean water, 390 B. typhosus

per oyster ; no B. coli com.
„ 15 (full sized) — after 11 days in clean water, 0 B. typhosus

per £ part of oyster.

It appears, therefore, from these analyses that the oysters
cleared themselves of the B. typhosus decidedly less rapidly
than previously clean oysters (Experiments I, II, and III),
which under similar conditions in the course of four to six
days had practically cleaned themselves of this microbe.
It will be, however, noticed that the polluted oysters cleaned
themselves very rapidly of the B. coli communis, for thus
must be interpreted the fact that the very small oysters, con-
taining originally at least 900 B. coli communis, were free of

41

this microbe after two days (two changes) in clean sea
water.

The analysis of the oysters of the dry lot showed : —

Oyster 2 — after 2 days dry contained 58,700 B. typhosus per

oyster ; no B. coli com.

4 „ „ „ 17,400 B. typhosus per

oyster ; no B. coli com.

„ „ „ 37,900 B. typhosus per

oyster ; no B. coli com.

„ „ „ 1300 B. typhosus per

oyster ; no B. coli com.

,, 10 — 5, 11 ,, 3, 3, innumerable B. typhosus,

very large number of
B. coli com.

This oyster always looked weak, it did not close its shell
promptly ; when opened, eleven days dry, it had no liquor
in the shell, and it looked abnormal, brownish. This, there-
fore, must be considered as an abnormal case, in which
the activity of the oyster tissues was unhealthy and in
abeyance, and this would explain the inability of the fish
to deal with either the B. typhosus or the B. coli, both these
microbes having been capable of multiplying in the
oyster.

Omitting this abnormal oyster, we see, then, that also in
this experiment the dry oysters did not clean themselves in
anything like the same extent as did their wet companions :
this is in agreement with the results of the previous experi-
ments, in which clean oysters were used. It is noteworthy
that also the dry oysters were able to effectually deal with
the B. coli communis in two days, which clearly points to
the conclusion that the tissues and activities of the normal
oyster per se are as inimical to the B. coli communis as to the
B. typhosus, both being as regards the oyster aliens, and there-
fore when found in the oyster must have been derived from
the surroundings.

Table IV gives the summary of this Experiment IV.

42

TABLE IV.
SEA WATER.

Immediately after infection . 2,470,000 B. typhosus per 1 c.c.

24 hours „ „ . 1,530,000

1 day after change . . „>. 13,180

2 days „ „ yj. . . 10,580

^ }) » » ".:• -"-• • ,, ,,

6 » >, „ ',•{. • ;. 0 „ per T^ c.c.

* J) J> 53 • *!.; 0 ,, ,,

WET OYSTERS.

Oyster 1 — after 1 day in infected water, 95,800 B. typh. per oyster.
„ 3 — after 2 days in clean water, 752,800 „ ,,

» 5 — 5> 4 » )) )3 1200 ,, ,,

» 7— » 6 „ „ „ 200 „

JJ ** '

„ 13 — „ 9 ,, „ „ 390 ,, „

„ 15— ,,11 „ „ „ 0 per i part of oyster.

DRY OYSTERS.

Oyster 2 — after 2 days dry . . 58,700 B. typhosus per oyster.

,, 4— „ 4 „ „ . . 17,400

„ 6— „ 6 „ „ . . 37,900

„ 8 — . „ 7 „ ,, . . 1,300 „ „

„ 10 — „ 11 „ „ . . Innumerable; abnormal.

The foregoing experiments, confirmatory as they have
been to one another, all point in the same direction, and
are, I think, without further repetition, sufficient for drawing
some general conclusions.

In the first place, it is clearly shown that oysters during
the period these experiments were carried out, viz., September,
October, and November — that is, when oysters are in a fit
state for consumption — and there is no reason why the
same should not be applied to the oysters during the rest

43

of the season (December till April) — are perfectly capable
of living in sterile sea water and to retain their normal
character and aspect in perfect condition. In the second
place, the oysters after infection with even large numbers
of the B. typhosus remain to the eye indistinguishable in
all respects from non-infected normal oysters. This latter
point is of course important from a practical point of
view, inasmuch as oysters which are so infected would
in the ordinary course of things remain undetected. So long
as the oyster shell is well closed and the oyster on opening
would present the normal appearance of colour, juiciness
and plumpness, it would naturally pass as " of good quality."
As has been pointed out on a former page, the longer
persistence of the B. typhosus in oysters out of the water
makes such oysters dangerous to a higher degree than when
they are kept in the water. Now, it is common knowledge
that on many occasions oysters when taken from an in-
fected laying — or, at any rate, from a polluted locality — are
packed and kept in barrels, tubs, or the like, sometimes for
short, sometimes for long periods. This applies, of course,
in a conspicuous degree to oysters coming into England
from distant countries — America, France, Holland — but it
applies also to many oysters coming from distant localities
in England into London or other large towns, viz., they are
kept out of the water, i.e., in " dry " state, sometimes for
several days before they reach the consumer. From the
experiments we have described it must be obvious that this
practice should be done away with, for there is no difficulty
whatever in any part of England or Holland to keep oysters
in clean sea water, which can be frequently changed ; if we
can do so at a very small cost indeed in a laboratory in
London, I do not see that the same thing should not be
possible in seaside and other places ; all that seems required
is a sufficiently large receptacle, which can be thoroughly
brushed out and scalded with boiling water, and a sufficient
supply of clean sea water. We get here delivered in the
laboratory five gallons of sea water (23 litres) at the price of
sixpence, that is to say, sufficient water to give to each four

44

dozen oysters at least eight changes, or a change of nearly
three litres of fresh sea water for eight or ten consecutive days —
surely more than enough for the purpose. All that is there-
fore required is a small primary outlay, insignificant as com-
pared with the price charged for oysters to the consumer.

Our experiments have further shown that even when
oysters are infected with large numbers of B. typhosus,
incomparably larger than would be the case under ordinary
natural conditions, they clean themselves in a comparatively
short time if kept in clean sea water ; under laboratory
conditions even the at first polluted oysters, having been
infected each with between 95,000 and 800,000 B. typhosus,
had done so in less than twelve days.

Although, as pointed out on a former page, the typical
(fluid) typhoid stool during the third week and the typhoid
urine during convalescence contain enormous numbers of
B. typhosus — amounting to many millions per each cubic
centimetre — sewage as it flows out of the sewers, and as even
in the worst places it might directly bathe oyster layings or
oyster ponds, would in no case contain such great numbers
of B. typhosus as were used in our experiments. It will be
remembered that ordinary domestic sewage contains human
dejecta in a highly-diluted state, and therefore unless typhoid
stool or typhoid urine as such are directly allowed to bathe
the oysters, the number of typhoid bacilli in the sewer outfalls
would be under the worst conditions comparatively small.
So much more advantageous that the remedy against the con-
sumption of typhoid-infected oysters, being simple, would be
capable of readier application. The remedy would be this :
Place the oysters after removal from the polluted layings
in tanks or ponds receiving no other than clean sea
water. As far as I can see, to obtain the necessary amount
of clean sea water from outside the range of the polluted area,
and to have this frequently changed in the tanks or ponds,
is a simple matter of arrangement, which after a first outlay
would not involve more than a trifling expenditure, ludicrously
small if compared with the large interests at stake, the high
prices paid for good and safe oysters, and the big profits that

45

would and do accrue from the sale of oysters which would
rightly be considered as perfectly safe.

Not that I would recommend any relaxation in insisting
that oyster beds should be as far removed as possible from
sewage and other pollution ; but in those instances in which
oyster layings are unfortunately for one reason or another un-
controlled and established in localities accessible to pollution
or actually polluted, the remedy for rendering these oysters
clean and safe seems to me simple and well worth trying,
in the interest of the owners whose property at present is
greatly depreciated, unless surreptitiously made active, and
above all in the interest of the public, who in the majority
of instances have to rely on the mere statement of interested
parties to the effect that particular oysters are supposed to
be derived from clean beds.

SERIES B.

EXPERIMENTS WITH THE B. TYPHOSUS IN COCKLES AND

MUSSELS.

Infection with typhoid fever through cockles or mussels
is, in the nature of things, of less extensive occurrence than
through oysters, since cockles and mussels are incomparably
less frequently eaten in a raw state than oysters. Although
the methods generally employed of preparing either cockles or
mussels for consumption are open to criticism in respect of de-
stroying by those methods the infective agent, if present, the
general method is nevertheless capable in some degree of
achieving this. As is well known, both cockles and mussels
are in bulk subjected to a process that is designated as
" cooking," consisting in either plunging a mass of these shell-
fish in boiling water, and taking them out as soon as the
water again commences to bubble, generally sooner, or in
heating the water till it commences to bubble. By either
process the end in view is to expose the shellfish to heat for a
sufficient time till their shell opens, so as to separate the fish
from the shell by simple agitation ; the fish, although

4(5

coagulated on the outside, nevertheless retains its juicy soft
quality, is not tough or too much shrunk, the latter condition
making them unsaleable. Now, I have shown experiment-
ally (see Keport of the Medical Officer of the Local Govern-
ment Board, 1900-1901, p. 570) that pouring boiling water
over a heap of cockles, these at once all open their shells,
although the temperature in the course of very few minutes
falls below 65° C. ; and, as a matter of fact, I have shown that
if cockles previously infected with the B. typhosus are thus
treated in a heap, the B. typhosus can readily be recovered
from the interior of the fish from the middle of the heap,
although of such cockles the shell is widely open and
the outside of the fish is coagulated. It must be obvious
that, if in dealing in practice with these shellfish the object
in view is what it generally is, viz., merely to get the fish
readily out of the shell, and to obtain the former in a
juicy, not shrivelled, condition, then we must expect that
many a consignment of the so-called "cooked" shellfish
cannot be considered safe if they happen to be previously
contaminated, since the amount of "cooking" as gene-
rally practised does not ensure destruction of infective
germs. Some time ago, at the request of the Fishmongers'
Company, experiments were carried out by me on the
premises of the Fishmongers' Company, as also at the
instance of Dr. Collingridge at Leadenhall Market, and we
have shown that mussels and cockles en masse can be
safely "steamed under pressure" without injuring in the
slightest degree the proper aspect and condition of the
fish, and that, treated in this way, few minutes (three
to five minutes) suffice to make them sterile of all infec-
tive germs. At Leigh, I understand this method is followed
with success, and at no greater cost than formerly by
the haphazard methods. Unfortunately, both mussels and
cockles are occasionally eaten in a raw state, as, for
instance, by tourists, excursionists, and children, and as
both mussels and cockles are "dirty feeders," and being
found at or near the foreshore, which in some places is well
exposed to sewage pollution, it is readily understood that

47

infection with typhoid fever through eating raw mussels,
and particularly raw cockles, is an occurrence of which
several instances are on record. I would refer amongst
others to the reports of Dr. Nash, Dr. Thresh, and Dr. Allen.
It is, therefore, important to see in what way the cockle
and mussel are capable of dealing with the B. typhosus
which happen to have access to them.

EXPERIMENT V.

A batch of fine cockles was received direct from Leigh-
on-Sea; they arrived in sand. After well rinsing several
dozen of the cockles on the outside they were transferred to a
clean tub, and 2000 c.c. of sea water infected with B. typhosus
to the amount of four millions per 1 c.c. were poured over
them ; in this infected sea water they were left for 24 hours.
They were then taken out, well rinsed on the outside with
clean sea water, and two cockles being retained for immediate
analysis, the rest were placed in clean sand wetted with
sterile sea water. This procedure, namely, rinsing them with
clean sea water and then keeping them in clean sand wetted
with a little sterile sea water, was found to be the safest way
of keeping them alive. Keeping them in sea water alone —
which was tried with a portion of the infected cockles —
failed, because they soon died; but keeping them in wet
sand succeeded, as it is the nearest approach to the way in
which cockles naturally live.

The fresh sand wetted with sterile sea water was again
changed after two, four, six, nine, and eleven days. As stated
just now, two of the cockles (1 and la) were analysed after
having been 24 hours in the typhoid infected sea water. In
this, as also in all the subsequent analyses, both of cockles and
mussels, exactly the same methods were followed as have
been described in the experiments with the oysters, and it
is not necessary to repeat the details of the entire procedure
again, except to state that after opening the cockles or
mussels the liquor within the shell was drained off as much

48

and as carefully as possible, the fish finely minced and well
mixed, the amount of the resulting fluid carefully measured,
and of this ^ c.c. was used for cultivation by Drigalski plate
after definite dilution. The description of the analysis of the
first two test cockles will suffice.

Cockle 1. — Total amount of fluid, 0 • 8 c.c. ; T\y c.c. of this
was diluted with 10 c.c. sterile sea water, and of this dilution
•fa c.c. was used for one Drigalski plate.

Cockle la. — Total amount of fluid, 0*85 c.c.; same pro-
cedure as above.

The result was : Cockle 1 contained 474,560 B. typhosus ;
cockle la contained a little over 520,000 B. typhosus. This
would average about half a million B. typhosus per
cockle.

In order to ascertain the number of B. coli communis,
Jy part of the total fluid, i.e., -^ c.c. of cockle 1, was used
direct for one Drigalski plate, and in it were found three
colonies of B. coli communis, readily recognised by their
bright red colour and bright red halo. This would mean
that in the whole cockle something like 120 B. coli communis
had been present.

Cockle 2 was taken out and analysed 1 day after change into

clean sand and sterile water; it contained 153,000

B. typhosus, no B. coli communis.
4* was taken out and analysed 2 days after change ; it

contained 382,000 B. typhosus, no B. coli.
„ 6, after 5 days' change, contained 358,000 B. typhosus.
„ 8, „ 6 „ „ „ 1,541,000

„ 10, „ 7 „ „ „ 138,600

„ 12, „ 9 „ „ „ 69,000

,,12a, „ 9 „ „ „ 111,000

„ H, „ 10 „ „ „ 1,600

„ 14a, „ 10 „ „ „ 69,000

„ 14 was not quite normal, for its shell was not closed and

the body was somewhat discoloured and shrunk.

* In this experiment the cockles of analysis are all marked by even
numbers ; the reason is that I meant to keep some (uneven numbers) in
sea water only without sand, but after a day or two they all had died.

49

It appears then, from this series, that the cockles embody
readily a large number of the B. typhosus from the infected
sea water, larger in proportion than oysters. In Experiment
II, after 24 hours in infected water, the proportion of B.
typhosus in the oyster and 1 c.c. of sea water was as 40 to
744 (or about 1 : 18) ; in Experiment III it was as 84 to
2250 (or about 1 : 26) ; in Experiment IV it was as 95 to
2470 (or 1 : 28) whereas in the case of the cockles of Ex-
periment V it was as 500 to 4000 or 1 : 8.

Another striking fact is the persistence of B. typhosus in
large numbers in the cockles even after ten days' change;
24 hours after infection the cockle examined had half a
million ; after ten days the normal cockle 14a still contained
69,000 B. typhosus.

The abnormal cockle, on the other hand, had a greatly
reduced number, viz., 1600 B. typhosus, so that as compared
with the abnormal oyster 10 of Experiment IV the reverse
condition obtained, for we found that in this abnormal oyster
the B. typhosus had considerably increased, whereas in the
abnormal cockle 14 we found the smallest number of B.
typhosus in the whole series. This would suggest that the
B. typhosus thrives in the cockle well so long as this animal
is in a normal state — a suggestion which is borne out in a
decided manner by the analysis mentioned in the foregoing
Experiment V, for we find that in the cockles after two, five,
and particularly after six days' change the number of B.
typhosus had gradually risen. After one day's change their
number had fallen from 500,000 to 153,000 ; then it rose, till
after six days' change it had surpassed by more than three-
fold the number of B. typhosus in the first cockle, having
risen from 500,000 to 1,540,000. In a former report (I.e.) I
had already observed this phenomenon of increase of B.
typhosus in the cockle as time went on; here we have
definite proof by numerical evidence.

The cockle, then, differs in a dangerous way from the
oyster, inasmuch as not only is the cockle not capable of
dealing so well with the ingested B. typhosus as the oyster,
but it appears to offer to the B. typhosus even facilities for

E

50

increase ; and for these reasons alone, cockles coming from a
polluted locality require, for their being rendered safe, a
thorough and careful disinfection by heat such as was
indicated above, viz., steaming under pressure for at least
three minutes.

Table V gives the summary of Experiment V.

TABLE V.

Cockles kept in sea water infected with B. typhosus to
the amount of four millions per 1 c.c.

Cockle 1 — 24 hours in infected sea water, contained about

500,000 B. typhosus.
„ 2 — after 1 day's change in clean wet sand, contained

about 153,000 B. typhosus.
,, 4 — after 2 days' change in clean wet sand, contained

about 382,000 B. typhosus.
„ 6 — after 5 days' change in clean wet sand, contained

about 358,000 B. typhosus.
?) 8 — after 6 days' change in clean wet sand, contained

about 1,541,000 B. typhosus.
„ 10 — after 7 days' change in clean wet sand, contained

about 138,000 B. typhosus.
„ 12 — after 9 days' change in clean wet sand, contained

about 69,300 B. typhosus.
„ 12a — after 9 days' change in clean wet sand, contained

about 111,000 B. typhosus.
„" 14 — after 10 days' change in clean wet sand, contained

about 1600 B. typhosus; abnormal.
5) 14-a — after 10 days' change in clean wet sand, contained

about 69,000 B. typhosus.

EXPEEIMENT VI.

Several dozen fresh mussels were well cleaned under the
tap and were then placed in sterile sea water in a clean
tub, to which an emulsion of pure culture B. typhosus
was added to the amount of 5,170,000 B. typhosus per

51

1 c.c. After having been kept herein for 24 hours they
were taken out, well washed on the outside, and, except
one mussel which was used immediately for analysis, the
others were transferred to a clean tub, were covered with
sterile sea water and clean well washed seaweed. After six
hours the water was poured off except as was sufficient to
keep the mussels in a wet condition. Under natural con-
ditions the mussels are covered with water at the flood tide
and uncovered with the ebb. The above procedure, viz.,
clean tub, sterile sea water and weeds, was repeated every
24 hours, and each time, i.e., each 24 hours, the excess of the
water was only allowed to cover the mussels for six hours.
In this way the majority of the mussels could be kept alive
and in normal condition for seven days after they were taken
out of the infected water. All those that were examined after
opening looked perfectly normal in every respect, and contained
a considerable amount of liquor. As with the oysters and
cockles after opening, the fluid within the shell was drained
off as completely and as carefully as possible, then the whole
fish was taken out, finely minced and well mixed, and of the
fluid a definite amount was used for cultivation in a Drigalski
plate. The mussels were chosen just as they came to hand,
some being large, some medium sized.

Mussel 1 (large), kept in infected sea water for 24 hours —
total amount of fluid a little over 2 c.c. : added -^ of this to
10 c.c. sterile sea water ; of this dilution used TJ^ c.c. for one
Drigalski plate. This plate yielded 300 colonies ; this would
mean a little over 6 millions B. typhosus per mussel, viz.,
300 x 100 X 100 x 2 = 6,000,000.

Mussel 2 (medium size), total amount of fluid 0 • 4 c.c., after 1 day's

change contained 74,800 B. typhosus per mussel.
„ 3 (large), total amount of fluid 2-4 c.c., after 2 days'

change contained 628,660 B. typhosus per mussel.
„ 4 (medium size), total amount of fluid 1 • 8 c.c., after 3 days'

change contained 36,000 B. typhosus per mussel.
„ 5 (medium size), total amount of fluid 2 • 5 c.c., after 5 days'
change contained 58,000 B. typhosus per mussel.

E 2

52

Mussel 6 (medium size), total amount of fluid 2 • 5 c.c., after 6 days'

change contained 6,250 B. typhosus per mussel.
„ 7 (medium size), total amount of fluid 0 • 5 c.c., after 7 days'
change contained 14,200 B. typhosus per mussel.

This experiment could not, unfortunately, be continued,
because the remaining mussels could not be kept alive. But
as far as it goes it shows that mussels take up the B. typhosus
from the surrounding water with great ease, the first mussel
having taken up in 24 hours from the infected water the
B. typhosus to the enormous amount of over six millions —
greater in proportion than what was observed with oysters or
cockles. Notwithstanding the daily change of sterile sea
water during five days, there were still discovered in the
mussels B. typhosus in considerable numbers. But on the
whole there may be said to have been going on a distinct
decrease of the B. typhosus in the animal — slower than was
the case with the oysters, quicker than with the cockles. No
increase of the microbe, as in the case of the cockles, was
noticed; and therefore it seems justifiable to say that in
respect of dealing with the ingested B. typhosus, the mussel
stands between the oyster and the cockle — that is, it is capable
of ingesting the B. typhosus from the surrounding water in
greater proportion than either the oyster or the cockle,
and the B. typhosus does not undergo increase within the
mussel.

Table YI gives the summary.

Mussels in sea water infected with B. typhosus to the
amount of over 5 millions per 1 c.c.

Mussel 1 (large), kept 24 hours in infected sea water, contained

over 6,000,000 B. typhosus.
„ 2 (medium), 1 day after change, contained over 74,000

B. typhosus.
„ 3 (large), 2 days after change, contained over 628,660

B. typhosus.
,, 4 (medium), 3 days after change, contained over 36,000

B. typhosus.

53

Mussel 5 (medium), 5 days after change, contained over 58,000

B. typhosus.
,, 6 (medium), 6 days after change, contained over 6000

B. typhosus.
,, 7 (medium), 7 days after change, contained 14,200 B.

typhosus.

SEEIES C.

EXPERIMENTS WITH OYSTERS KEPT IN STERILE SEA. WATER
INFECTED WITH HUMAN F^CAL MATTER.

In the foregoing experiments (Experiments IV and V),
we had the opportunity of showing that, just like the
B. typhosus, so also the B. coli communis when originally
present rapidly disappears from the oysters and cockles if
these be kept in clean surroundings. As this question of
the presence of B. coli has recently received a great deal of
attention, and caused a radical divergence of opinion, we
propose to discuss it somewhat in detail, but first wish to
record some experiments made expressly to determine in exact
manner how oysters are capable of dealing with the B. coli
communis, derived from human faecal matter. This deter-
mination by means of Drigalski plates is extremely simple,
since the colonies of this microbe are already after 24 hours
at 37° C. conspicuous by their size, by their red colour and
marked red halo.

EXPERIMENT VII.

One gram of faecal matter of a healthy man was shaken
up in 2000 c.c. sterile sea water, a determination was made
by means of a Drigalski plate with ioo~oo Par^ °f a cubic
centimetre (1 c.c. of the sea water emulsion was added to
99 c.c. sterile water : dilution 1 ; of this 1 c.c. was added to
9 c.c. sterile water : dilution 2 ; -j-1^ c.c. of this dilution 2
was used for the Drigalski plate). This plate showed, after
incubation at 37° C., 23 colonies of B. coli communis. These

54

were visible next day as good-sized red colonies with distinct
red halo ; they were all of the same kind and aspect, viz.,
typical B. coli communis ; this would mean 230,000 B. coli
communis per 1 c.c. of the infected sea water, or 460 millions
of this microbe per 1 gram of ftecal matter.

Thirteen Whitstable oysters were kept for 48 hours in
sterile sea water, and two of them on analysis by Drigalski
plates having been proved to contain no B. coli communis or
any other B. coli, the remaining 11 oysters were placed in a
clean tub into the above foecally infected sea water. Here they
remained for 24 hours. They were now taken out, well rinsed
under the tap, and divided into lots — six oysters were trans-
ferred to a fresh clean tub and supplied with sterile sea water,
which was repeated every 24 hours as long as any of these
" wet " oysters remained ; four oysters were placed " dry " in
cool chamber, and the remaining eleventh oyster was used for
analysis. At the same time of the above infected sea water,
i.e., kept for the 24 hours, analysis by Drigalski plate was made.
Of the sea water, 24 hours after infection with faecal matter,
ToVo Part °f a cukic centimetre yielded 177 colonies of
B. coli communis, or 177,000 B. coli communis per 1 c.c.

Oyster 1 — after 24 hours in infected sea water — treated
in the usual manner, yielded 77 colonies of B. coli communis
per gJo part of the body — that is, 46,200 B. coli communis
for the whole oyster.

Sea water which had been changed and in which five of
the infected oysters had been placed was analysed 24 hours
after the change, using ^- c.c. direct for one Drigalski plate.
This yielded 65 colonies of B. coli communis ; this means
650 B. coli communis per 1 c.c.

Oyster 2, taken from the " dry" lot, one day after removal
from infected sea water yielded 8575 B. coli communis —
343 colonies per ^ part of body.

Oyster 3, taken from the " wet " lot, one day after change
yielded 2325 B. coli communis per whole oyster — 31 colonies
per ^ part of body.

The sea water in the tub was analysed — having been
changed 24 hours previously, as well as the tub into which

55

the five remaining " wet " oysters had been transferred — it
yielded 50 B. coli eommunis per 1 c.c.

Oyster 4, " dry "— having been kept two days dry, yielded 7970

B. coli eommunis.

„ 5, " wet " — having been two days in (twice) changed
sterile sea water and fresh tub, yielded 1305
B. coli eommunis.

The sea water, which contained the remaining four wet
oysters, was analysed after three changes ; it yielded 0 B coli
eommunis per T% c.c.

Oyster 6, " dry " — kept three days dry, yielded 5790 B. coli

eommunis.
„ 7, " wet " — kept three days in (thrice) changed sterile sea

water, yielded 216 B. coli eommunis.

„ 8, " dry " — kept four days dry, yielded 2625 B. coli corn-
munis.
„ 9, "wet" — kept four days in (four times) changed sterile

sea water, yielded 1 1 B. coli eommunis.
„ 11, " wet" — kept seven days in (five times) changed sterile

sea water, yielded 18 B. coli eommunis.
„ 13, "wet" — kept eight days in (six times) changed sea

water, contained no B. coli eommunis per f

part of oyster.

Summarising the above results in tabular form, we
have : —

TABLE VII.

SEA WATER INFECTED WITH NORMAL FAECAL MATTER.
SEA WATER.

Immediately after infection . . 230,000 B coli per 1 c.c.

After 24 hours 177,000

„ 1 day change .... 650 „ „

„ 2 days' „ .... 50 „ „

„ 3 „ .... 0 per TV c.c.

56

WET OYSTERS.

Oyster 1 — after 24 hrs. in infected water, 46,200 B. coliper oyster.
,,3 ,,1 day change . . . 2325 „
5 ,,2 days' „ ... 1305

*•*•/•* )) 5J

jj 7 „ 3 ,, ,, ... 216 ,, ,,

9 A 11

JJ * )J )J ... A i. ,, ,,

117 1 ft

)) -1 A )> ' 5J 3} ... IO ,, ,j

» 13 „ 8 „ „ ... 0 „ per | oyster.

DRY OYSTERS.

Oyster 2 — 1 day dry . . 8575 B. coli communis per oyster.

„ 4— 2 days dry . . 7940 „

„ 6—3 „ „ . . 5796 „ „ „

„ 8—4 „ „ . . 2625 „

Starting, then, with 230,000 B. coli communis per 1 c.c.,
the sea water had diminished them in 24 hours down to
177,000; but then we should not forget that there have been
11 oysters in the water ; taking all these 11 oysters as having
contained, or rather having withdrawn from the water,
the same number of B. coli communis, i.e., 11 times 46,200,
there would have been on account of this withdrawal only
an insignificant number (half a million) removed from the
460,000,000 originally added to the water. From this it
follows that the number of B. coli communis had actually
been reduced by the sea water from the original 460,000,000
to 353,500,000 (177,000 X 2000 c.c., minus 500,000 for the
eleven oysters). This reduction does not, therefore, compare
with that observed in connection with the B. typhosus, and
this coincides with what is known as to the greater hardiness
and greater resistance of the B. coli communis over the
B. typhosus. We notice, however, that in the oysters kept
in clean sea water, frequently changed, the reduction of the
B. coli communis was rapid and marked. By three days'
changes the number of B. coli communis had been brought
down to 216 (from the original 46,200), and after eight days'
changes no B. coli communis could be discovered any longer.

57

Whereas in the oysters not kept in sea water the decrease,
though progressing, was distinctly slower ; after four days
dry the oyster still contained more B. coli communis (2625)
than the wet oyster after one day change in clean water.

There can, then, be no question about the fact that the
oyster per se is capable of dealing with the B. coli communis
in the same manner as with the B. typhosus, mz.t that
also the B. coli communis does not multiply in the oyster ;
when taken in from the surroundings by oysters clean at
starting the number of B. coli communis decreases, and if
the oysters are kept in clean water the microbe rapidly
disappears.

I consider this additional definite proof that the B. coli
communis is as foreign to the oyster as the B. typhosus, and
that therefore when B. coli communis is found in oysters it
is derived from the surroundings, and must be of fairly
recent importation.

EXPERIMENT VIII.

In order to obtain further confirmation, the next experi-
ment was made in the same manner, but this time with
native oysters of a different kind, viz., Colchester oysters.

One of these oysters (1), immediately on receiving them,
was well brushed and cleaned on the outside and then used
in the usual manner for analysis, -j1- part of the body of the
oyster being used for a Drigalski plate. The others were
placed in sterile sea water in a fresh tub. No colonies of B.
coli came up in the plate of oyster (1).

After 24 hours in clean sea water we took out a further
oyster (2), ^ part of the oyster body being used for a Drig-
alski plate. No colonies of B. coli came up in this plate.

We now infected 2000 c.c. of sterile sea water with one
gram of fecal matter of a healthy person. By making a
Drigalski plate on the same plan of dilution as in Experi-
ment VII, it was ascertained that each c.c. of this water
contained 22,000 colonies of B. coli communis — that is to

58

say, 44 millions per 1 gram of faeces — a considerably smaller
number than in Experiment VII ; in fact, barely the tenth
part of the number used in the latter experiment.

Into this infected sea water were placed the remaining
ten oysters, and they remained herein for 48 hours. After
this period the infected sea water as also one oyster (3) were
analysed, the remainder were well rinsed under the tap and
then transferred to a fresh tub and fresh sterile water, and
this procedure was repeated after a further day, also after
two, four, six and seven days ; at each of these periods of
change, oyster or oysters were taken out and analysed.

The sea water, 48 hours after infection, contained 4110
B. coli communis per 1 c.c. The sea water in tub one day
after change contained 40 B. coli communis per 1 c.c. —
that is, the originally sterile sea water in the tub to which
the infected oysters had been transferred. The sea water
twice changed contained 0 B. coli per ~£-ff c.c.

Oyster 3 — having been kept 48 hours in the fsecally infected sea

water, contained in its body 650 B. coli communis.
,, 4 — having been kept 1 day in sterile sea water, contained

84 B. coli communis.
„ 5 — having been kept 2 days in sterile sea water, contained

600 B. coli communis.
„ 6 — having been kept 3 days in sterile sea water, contained

484 B. coli communis.
„ 7 — having been kept 5 days in sterile sea water, contained

52 B. coli communis.
„ 8 — having been kept 7 days in sterile sea water, contained

48 B. coli communis.
„ 9 — having been kept 7 days in sterile sea water, contained

0 B. coli communis per | oyster.
„ 10 — having been kept 8 days in sterile sea water, contained

18 B. coli communis.
„ 11 — having been kept 8 days in sterile sea water, contained

0 B. coli communis per -J- oyster.

One of the oysters, not analysed, died two days after the
commencement of the experiment.

59

From this we learn that by using for infection of the
sea water a considerably smaller number of B. coli communis,
and therefore the number initially ingested by the oysters
being relatively small, it made no appreciable difference
in the manner and time in which the oysters became clear
of the B. coli communis, as compared with the oysters of
the preceding experiment in which the initial number in
the sea water was ten times greater.

Table VIII gives a summary of the results of Experi-
ment VIII, and for comparison we repeat Table VII as
Table IX, giving the analysis both of the wet oysters of
Experiment VIII and of the oysters of previous Experi-
ment VII.

TABLE VIII.

SEA WATER INFECTED WITH NORMAL FAECAL MATTER.
SEA WATER.

Immediately after infection. 22,000 B. coli communis per 1 c.c.

After 48 hours .... 440 „

After 1 day's change . 40 „ „ „

After 2 days' change . . 0 „ „ per T^ c.c.

OYSTERS.

Oyster 3 — 48 hours in infected sea water, contained 650 B. coli

communis per oyster.

5, 4 — after 1 day's change, contained 84 B. coli com. per oyster.
„ 5— „ 2 days' „ „ 600

„ 6— „ 3 „ „ „ 484

" ' » "

» " » • » » j> 48 ,, „ ,,

» " » ' )> » » 0 ,, ,, per i part

of oyster.

» 10 — „ 8 „ „ „ 18 „ „ per oyster.

» 11 — • » 8 „ „ „ 0 „ „ per | part

of oyster.

„

„ „

» » 5) 52 ,, ,, „

60
TABLE IX.

PRECEDING EXPERIMENT VII.— SEA WATER INFECTED WITH
230,000 B. COLI COMMUNIS PER 1 c.c.

Oyster. 24 hours in infected sea water 46,200 B. coli communis.

„ 1 day's change .... 2325 „ „

„ 2 days' „ . . . . 1305 „

,, 3 ,, „ ... . 216 ,, ,,

4. 11

}) » .... J. 1 J, ,,

» ' 5J 5> * 1° }J )J

5> *" 3J 5> .... U ,, ,,

This comparison emphasises, therefore, the fact that even
when the initial number of the ingested B. coli communis is
very large, the oyster is capable of dealing with it success-
fully, and it cannot be a question of mere " washing out " by
the water of the ingested B. coli, but must depend on the
activity of the tissues of the oysters in dealing with the
foreign intruder.

The difference observed here may, it is true, be due to
the two samples of oysters coming from different localities,
but I hardly think this a satisfactory explanation. If there
be an extraneous cause it is more likely to be due to the
sample of the oysters used in Experiment VIII not having been
quite so good or fresh as those of Experiment VII ; one oyster
dying early in the experiment would point in that direction.
At any rate, about the fact that even when the initial number
of the B. coli communis is very great, the normal oyster is
capable in clean water of rapidly clearing itself of this microbe.
This fact would, then, once for all set at rest the implied
suggestion by the Eoyal Commission on Sewage Disposal,
viz., that the B. coli communis is a normal inhabitant of the
body of the oyster. If there is one thing clear, it certainly
is the fact contrary to that suggestion, viz., it is a fact con-
clusively proved that oysters from clean places and oysters
kept in clean water are free of B. coli communis ; and
further, that if they should happen to have imbibed them
from the surrounding water, they, by being again placed in
clean water, rapidly clean themselves of the intruder.

61

EXPEEIMENT IX.

This experiment was undertaken to test the behaviour of
oysters towards the B. coli communis of ordinary domestic
sewage. One dozen " small Dutch oysters " were obtained
from a first-class shop in the City — the oysters had just
arrived direct from Holland. The oysters were carefully
brushed under the tap and placed in a clean tank in 2000 c.c.
of sterile sea water. Twenty-four hours after, two were taken
out for analysis by Drigalski plates — two plates for each
oyster. The result was that neither of them contained in
-j^ and £ part of the body of the fish any B. coli communis —
that is to say, in none of the four plates were any red
colonies with red halo to be found.

The test by Drigalski plates, as I have already explained,
is undoubtedly the best that we have, and for this reason :
that only B. coli communis produces on these plates already
after 24 hours at 37° C. round distinctly red colonies,
several millimetres in breadth with red halo ; and if no such
colonies, viz., largish red with red halo, make their appear-
ance in 48 hours, according to my experience — extending
now over a considerable number of analyses — the conclusion
is justified, and is confirmed by other culture tests, that no
B. coli communis is present in the material analysed. I
must insist on this, for the reason that all other methods
used for preliminary diagnosis are decidedly inferior, because
colonies of the above kind invariably respond to all tests
of B. coli communis, whereas the colonies of other microbic
species, unlike the above red colonies with red halo, on the
complete series of tests being made do not answer to
the true B. coli communis. Although in several of the tests
they may simulate the B. coli, if the tests are completed, they
can be proved to be not the true B. coli communis, but to
be different from this, the typical and constant microbe of ex-
cremental matters. I have for years past — now more than
12 years — insisted on a distinction being drawn between
the typical B. coli communis of excremental matters and
what, owing to one or the other similar character, may be

62

designated as a coli-like microbe. I have repeatedly drawn
attention to this, that whereas the B. coli communis is the
typical microbe of faecal matter, the derivation and distribu-
tion of many coli-like microbes is at present not sufficiently
known and cannot therefore be used for diagnostic purposes.
It is therefore satisfactory to find that the American ob-
servers* draw the same sharp distinction ; they find that
oysters coming from clean, not sewage-polluted, layings have
no B. coli communis, although they may on first tests show
microbes which are coli-like,f and that in proportion to the
pollution of the layings by sewage they contain the B. coli
communis.

Messrs. Clark and Gage give the following table, based on
analyses of shellfish and water during three years : —

SHELLFISH AND SHELL WATER CONTAINING THE TRUE

B. COLI COMMUNIS.

Character of
Source.

Number
of
Sources.

Percentage of Samples Positive.

Shellfish.

Sea Water.

Shell Water.

Intestine.

Not polluted . .

15

0

0

0

Doubtful . . .

22

8

6

17

Polluted . . .

8

41

35

44

Statements such as have been repeatedly made (see the
Keport of the Bacteriologist of the Sewage Commission ;
Dr. Foulerton's paper read at the Folkestone Meeting 1904
of the Koyal Institute of Public Health, in the Bacteriology

* Thirty-fourth Annual Keport of the State Board of Health of Massa-
chusetts for 1902, p. 18.

f Of 58 species which gave the " presumptive " tests, only 12 were
found to be B. coli communis (I.e. p. 20).

63

section), viz., that oysters derived from " clean " layings, or
sea water taken many miles away from the shore, contained
large numbers of the true B. coli communis, are to me
perfectly unintelligible. I must confess I have not suc-
ceeded in verifying this. I have not found, for instance,
that oysters coming from layings which are miles away from
any source of sewage or manure pollution, e.g., some layings
in Halford Eiver, some layings in Hayling Island, " contain
1000 B. coli communis per oyster" (see Eeport of Sewage
Commission) ; in fact, I have not found anything approaching
such a condition even in oysters directly from sewage-bathed
ponds.* I have quite recently had the opportunity to examine
oysters which came from the mouth of Langston Harbour, only
600 yards distant from the principal sewer outfall of Ports-
mouth, and what I found was that of nine oysters examined
all contained B. coli communis. But in what numbers ?
Three were specially tested by Drigalski plates each made
with one-fiftieth part of the body, and they were found to
contain : one 200 but not 300 B. coli communis per
oyster, a second one 150 but not 200 B. coli communis
per oyster, and a third 50 only. A sample of oysters
derived from layings several miles away from the above
showed B. coli communis only in one out of nine oysters,

* The culture tests for B. coli communis of faecal matter of man and of
ordinary domestic sewage are these : —

1. B. coli communis forms on Drigalski medium at 37° C., after 24-36
hours, colonies several millimetres in size, distinctly red, with distinct
red halo when viewed in transmitted light.

2. In ordinary nutrient gelatine shake culture it forms colonies all
through the medium with numerous gas bubbles already in 24 hours
at 20° C. ; in gelatine streak it forms at 20° C. a rapidly spreading
dry band with irregular margin ; no liquefaction of gelatine at any
time.

3. Neutral red broth at 37° C. is changed in 24-36 hours from cherry red
to greenish fluorescent.

4. It turns MacConkey fluid (litmus, glucose, taurocholate of soda, pep-
tone) red, forming acid with numerous gas bubbles.

5. It grows well in phenol broth at 37° C., making it turbid in 24 hours
with copious gas formation.

6. It produces indol in nutrient broth at 37° C. in 3-5 days.

7. It turns lactose peptone litmus in 24-36 hours at 37° G. red (acid
production) with copious gas formation.

8. Litmus milk at 37° C. becomes red in 24 hours, due to acid production,
the milk becoming clotted in 1-3 days.

64

and in that one there was only one colony of B. coli com-
munis per one-tenth part of the body of the oyster. And it
is precisely on account of the ready and reliable manner in
which B. coli communis can be identified by the method of
Drigalski plates that the greatest importance attaches itself
to this method.

Having ascertained, then, that our small Dutch natives
contain no B. coli communis, they were divided in two lots :
one (five oysters) was kept in sterile sea water (2000 c.c.)
without any addition, the other lot (five oysters) was trans-
ferred to a fresh tub with sterile sea water (2000 c.c.) to
which 5 c.c. of crude sewage (of St. Bartholomew's Hospital)
were added. Analysis of this sewage made by Drigalski
plate at the same time showed that it contained 220,000
B. coli communis per 1 c.c. — that is to say, each cubic centi-
metre of the infected sea water contained 550 B. coli com-
munis. The first lot of oysters, viz., in clean tub with clean
sea water, will be mentioned here as " clean lot," the second
lot, mz.t in sewage polluted sea water, will be mentioned as
"polluted lot."

Oyster 1 (clean lot) — kept 1 day in sea water, contained no
B. coli communis.

„ 2 (polluted lot) — kept 1 day in sewage polluted water,
contained 800 B. coli communis.

„ 3 (clean lot) — kept 2 days in clean sea water, contained
no B. coli communis.

„ 4 (polluted lot) — kept 2 days in polluted sea water, con-
tained 150 B. coli communis.

The polluted lot were taken out after having been kept
48 hours in the polluted water, well rinsed under the tap
and transferred to clean tub and 2000 c.c. clean (sterile)
sea water. The clean lot received 2000 c.c. fresh sterile sea
water, and this procedure was repeated on each of the fol-
lowing two days.

Oyster 5 (clean lot) — kept 3 days in clean sea water, contained
no B. coli communis.

65

Oyster 6 (polluted lot) — kept 1 day in clean water, contained

100 B. coli communis.
„ 7 (clean lot) — kept 4 days in clean water, contained no

B. coli communis.
„ 8 (polluted lot) — kept 2 days in clean water, contained

no B. coli communis per TL part of body.
,, 9 (clean lot) — kept 5 days in clean water, contained no

B. coli communis.
,, 10 (polluted lot) — kept 3 days in clean water, contained

no B. coli communis.

Tabulating the results of this experiment :

TABLE X.

Sea water infected with crude sewage from St. Bartholomew's
Hospital to the amount of 550 B. coli communis per
1 c.c. sea water (220,000 B. coli communia per 1 c.c.
crude sewage).

CLEAN LOT.

Oyster 1 — after 1 day,* contained no B. coli communis.
„ 3— ,, 2 days, „ „ „

K O

" ° 11 ° » 5> 5) >»

7 4.

•>•> »

POLLUTED LOT.

Oyster 2 — kept 1 day in polluted sea water, 800 B. coli communis.
„ 4— „ 2 days „ „ 150

„ 6 — kept 1 day in clean sea water, 100 ,, ,,

„ 8— „ 2 days „ „ 0 „ „ per

•fa part of oyster.

» 10 — „ 3 ,, ,, „ 0 B. coli communis

per -j^ part of oyster.

Starting, then, with clean oysters and placing them in
sewage-polluted sea water — polluted with sewage B. coli
communis to the amount of 550 per 1 c.c. — we find that after

* This will be understood as the time when the comparison between
the two lots commenced.

66

24 hours in this polluted water the oyster had taken in 800
B. coli communis. After a further day only 150 were found
in the next oyster, and after having been transferred to clean
water they practically cleared themselves in two further
days. The previous experiment has shown us that the sea
water per se is capable of materially reducing in 48 hours the
number of B. coli communis, and, therefore, the reduction
of this microbe in the oysters after 48 hours from 800 to
150 is what we might expect. This experiment is in so far
interesting, as it appears to point to this, viz., that clean
oysters are capable of dealing promptly with the B. coli
communis of sewage, seemingly more promptly than with
the B. coli communis directly derived from human faecal
matter.

SEEIES D,

In the following series of observations, an attempt was
made to differentiate those microbes of sewage and of fecal
matter which in Drigalski plates are capable of forming
" blue " colonies — that is, colonies that might interfere with
and aggravate the diagnosis and recognition of the colonies
of B. typhosus and similar microbes of a pathogenic character
like the B. enteritidis Gaertner, both of which would indicate
specific pollution ; the former being derived from the typhoid
patient (bowel discharge, urine), the latter from the bowel
discharges of a person affected with certain forms of acute
gastro-enteritis, and most probably also with the acute ailment
called paratyphoid, recognised now as different from typhoid
fever. The microbe of this disease, viz., the Bacillus para-
typhosus appears from all accounts to closely resemble the
B. Gaertner, being possibly a variety of this latter.

The B. Gaertner forms blue-violet colonies on Drigalski
medium at 37° C., but they grow slower, coming up slower and
remaining smaller than those of B. typhosus, and are marked
also from the latter by showing a central opaque spot. By
microscopic examination in the hanging drop, by subculture
in neutral red broth andin litmus milk, in MacConkey fluid, and
by the agglutination test with typhoid serum, the differential

67

diagnosis is readily established. For B. Gaertner is shorter
than B. typhosus ; B. Gaertner gives positive neutral red broth
test, it turns the litmus milk at first slightly acid (red), but
after two or three days gradually alkaline (blue and slate
colour), and it produces acid and gas in MacConkey fluid. It
does not agglutinate with typhoid in anything like the high
dilution that B. typhosus does — that is to say, it is by all the
above tests easily distinguished from the B. typhosus (see
a former page). There would be, therefore, none but a
preliminary difficulty in differentiating between the two
microbes. B. Gaertner in small doses is highly virulent to
rodents, both after subcutaneous injection, as also by feeding,
and in these respects differs from the B. typhosus. We have
already indicated that all colonies of B. coli communis can
by means of the Drigalski plates at 37° C. be recognised
already in 24 hours, they being several millimetres in dia-
meter, being red and surrounded by a distinct red halo ; other
acid-forming (red) coli-like microbes are slower in develop-
ing, and are much smaller. These are always numerous in
sewage, but being small and slow in coming, although red,
and even some with indication of red halo,, can at once be
neglected, as far as the search for specific microbes is
concerned.

The B. dysenteric forms neither red nor blue colonies,
being neutral, like many other microbes not belonging to the
ooli-typhoid group. These, therefore, do not offer any basis
for further inquiry by means of the Drigalski medium..

But there are a number of species of microbes of sewage
and of faecal matters which, on account of their forming blue
or bluish colonies in the Drigalski medium, require special
considerations in reference to shellfish, in order to differentiate
them from B. typhosus and from B. Gaertner.

I. — MICEOBES OF SEWAGE FORMING "BLUE" COLONIES ON
DRIGALSKI MEDIUM.

(a.) Amongst the many clean oysters which I have ex-
amined, I have not found any as yet which contain bacteria

F 2

68

which, on Drigalski medium incubated at 37° C., are capable of
yielding blue colonies like B. typhosus. I have examined a
considerable number of Burnham oysters, of Colchester oysters,
of clean Dutch oysters, of Whitstable oysters, and others ; and
I have not found in them any bacteria which, incubated on
Drigalski medium at 37° C. for 24-72 hours, form such blue
colonies. On three occasions when examining clean oysters, I
have been, however, greatly puzzled by the appearance in the
Drigalski plates of blue colonies with violet margin, which, by
their size, by their slightly irregular contour, and their more or
less conical form, and uniformly granular appearance, looked
very like those of B. typhosus. And the difficulty became:
considerably enhanced* by noticing that an emulsion of the
bacilli — which were very motile — of these colonies, subjected
to agglutination test with typhoid serum, gave a strikingly
positive test. When, after the method of Koch-Drigalski, a
loop of typhoid serum was added to several big drops of the
emulsion — made by distributing a trace of such a blue
colony in sterile beef broth — arrest of movement and agglo-
meration into large conspicuous clumps took place within a
a few minutes, the process of agglutination being quite
complete within five or six minutes.

This occurred with quite a series of clean Whitstable
oysters, derived from Whitstable and from Langstone
channel. The difficulty and puzzle was, however, soon
solved, viz., these blue colonies appeared only after the
plates had been taken out of the hot incubator, i.e., at
37° C., and were then kept for several days either at 20° C.,
or at the ordinary temperature of the laboratory. So long
as the plates had been kept in the hot incubator, i.e., at
37° C., there was no trace of these blue colonies, but they
gradually made their appearance after they had been kept
for several days at the lower temperature.

Another fact noticed about these typhoid-like blue
colonies was this, that the bacilli composing them were thinner
than the B. typhosus and decidedly more cylindrical, and
even filamentous, but they showed active mobility just like
the B. typhosus. Sub-cultures in the different media estab-

69

lished the differences in a striking manner. (1) They failed
to grow in all media, if these are incubated at 37° C. ; there
was no perceptible growth at this temperature either on agar,
or in broth, or in milk. (2) They completely failed to alter
MacConkey fluid at 37° C. ; B. typhosus turns it red, but
forms no gas. (3) They failed to cause any change in litmus
milk ; B. typhosus forms gradually acid (red) without altering
the fluid character of the milk. (4) They produced no
growth either in neutral red broth or in phenol broth.

In addition to these differences the microbe grew very
slowly on gelatine : it took two to three days before any
distinct growth could be noticed, and then it was very trans-
parent, unlike that produced by B. typhosus ; and, lastly, in
shake gelatine it did not form its colonies in the depth of the
medium like the B. typhosus, but, besides being much slower
in its growth, it formed colonies only or principally near the
free surface of the medium.

It is therefore clear that this microbe need cause no
difficulty in the search for the B. typhosus on Drigalski
plates kept at 37° C., being a microbe not capable of growing
at that temperature.

(b.) Of the most frequent "blue" or " violet" colonies found
in Drigalski plates inoculated with sewage or faecal matter,
with sewage or fsecal matter polluted shellfish — oysters, mussels,
and cockles — are those of streptococci. These appear already
after 24 hours at 37° C. ; better and more conspicuously later
as violet-blue small round dots, uniformly raised and moist
looking. A particle of a colony emulsified and looked at
under the microscope is at once recognised as a compound of
diplococci and short streptococci. They possess the additional
character of staining with gram. Owing to their small size,
their violet-blue colour, and their appearance under the
microscope, viz., being cocci, they need not further offer any
difficulty in respect of being mistaken for anything else.

(c.) Eepeatedly I have come across in Drigalski. plates at
37° C., that had been inoculated with sewage, bright Hue small
colonies, more or less conical in shape. They are composed
of vibrios or comma bacilli ; they resemble in size, shape,

70

and staining more or less the well-known vibrio of cholera,
but they differ from this latter in this essential respect, that
they do not liquefy gelatine at any period of their growth,
forming on this medium non-liquefying translucent, more or
less angular colonies. In peptone salt they grow feebly, do
not form any marked pellicle on the surface of this fluid, and
do not produce nitroso-indol : sufficiently distinct differences
from the vibrio of cholera. Besides, this non-liquefying
sewage vibrio, when injected, even in large doses, intra-
peritoneally into a guinea-pig, causes no disease. This same
vibrio was found on Drigalski plates inoculated with the
fluid of a mussel derived from a polluted locality, and
twice in oysters which had been distinctly polluted with
sewage.

I attribute, therefore, to the presence of this vibrio in
Drigalski plates, inoculated from oysters or other shellfish, an
important diagnostic value, because this vibrio appears
present in sewage in small numbers only, and when, there-
fore, present in shellfish is a fortiori strong presumptive
evidence of sewage pollution.

All vibrios form on Drigalski plates bright blue, small,
moist, round colonies, and in respect of this marked colour
and small size can readily be recognised ; a simple microscopic
specimen in the hanging drop shows the actively motile
comma-shaped, S-shaped, and shorter or longer spirillar forms.
From a cockle derived from a particularly polluted
locality — black mud on the Upper Orwell — I have isolated
by the Drigalski plate at 37° C. a motile vibrio which differs
from the above non-liquefying sewage vibrio in the following
respects: its colonies are bright blue, moist, round, small,
raised in the centre — that is, conical. After two to three
days the colonies reach the diameter of several millimetres.
This vibrio liquefies gelatine a little faster than the cholera
vibrio, but slower than the vibrio finkler. It grows well in
peptone salt water, and forms thereon in several days a
pellicle composed of matted masses of wavy or spiral
threads ; it does not produce nitroso-indol ; it grows well in
litmus milk, which becomes distinctly reddened (acid pro-

71

duction) by the growth; the milk remains fluid for
about a week, after that date it becomes firmly clotted. The
vibrio when injected intraperitoneally into the guinea-pig in
very minute doses — a loopful of a recent agar surface growth
— causes acute peritonitis and death, the turbid peritoneal
exudation being crowded with the vibrio, and the intestines
much inflamed.

As stated above, the vibrio is motile ; it is distinctly
shorter than the cholera vibrio, and is possessed of one or
two short terminal flagella. It does not become agglutinated
with blood serum of an animal protected by cholera vibrio.
Owing to its having been obtained from a cockle, I have named
it vibrio cardii* (cardium edule — the common cockle). As I
have not had opportunity of making further analysis in this
direction of cockles of other localities, I am at present
unable to attribute to this vibrio any particular diagnostic
value.

(d.) A third group of microbes forming " blue " or
"bluish" colonies on Drigalski medium, which can be
isolated from ordinary crude sewage and from human faecal
matter, and which are not to be met with in clean oysters,
comprises various species, but none of them are to be mis-
taken for either B. typhosus or B. Gaertner, as will presently
be described. Some of these species are such as occur in
filth of various kinds, and, therefore, cannot be considered
as diagnostic for sewage or fsecals, but we will consider
them here nevertheless, because they are not microbes of
shellfish per se.

(1.) First and foremost are bright blue colonies, which
already after 24 hours at 37° C. can be noticed with the unaided
eye as distinctly blue, flat, roundish patches ; after another 24
hours their margin becomes fringed, filmy, and rapidly spread-
ing. Examined in the hanging drop, they appear as rapidly
motile cylindrical bacilli, but they give no sign of agglutina-
tion with typhoid blood serum. Making a sub-culture on
gelatine they are seen to rapidly liquefy the gelatine ; they are,
in fact, the common Proteus vulgaris.

* " Centralblatt f. Bakteriologie," 1905.

72

(2.) Another microbe of common and copious occurrence in
sewage, and forming pale blue colonies on Drigalski medium >
is one of which the colonies appear rounded, uniformly raised,
and in two or three days show in reflected light a distinctly
greenish peripheral portion contrasting markedly with the
violet centre. The uniformly raised condition, the pale blue
colour, and the greenish margin are sufficiently distinct to
differentiate them from the typhoid or Gaertner colonies.
They are composed of short motile bacilli, and show no sign
of agglutination with typhoid blood serum.

(3.) A further cylindrical motile microbe, forming deep blue
round colonies, is one which can be readily distinguished by
an ordinary gelatine sub-culture, for in this it forms here a
distinct bluish-green fluorescence, without liquefying the
gelatine, being, in fact, the common Bacillus fluorescens
putidus, viz., a microbe common in filth of all kinds.*

(4.) One of the most frequent microbes botli of sewage and
of faecal matter, which forms round, bluish or blue-violet colonies,
raised in centre, flat at the periphery, and on inspection of the
Drigalski plates in some respects resembles the colonies of B.
typhosus or Gaertner, is one which is represented by short
cylindrical motile bacilli. On account of the blue colour of
the colonies with violet marginal part, on account of their
raised condition, and on account of their being made up of
cylindrical motile bacilli, they could be easily mistaken for B.
typhosus, the more so since they show distinct agglutination
with a highly potent typhoid serum, although in high dilution
no agglutination takes place. But the above characters pos-
sessed by the colonies as they appear on the Drigalski plate
might be presumptive of B. typhosus or B. Gaertner. More
careful observation and sub-cultures, however, soon show that
they are altogether different. In the first place, these blue
colonies are slower in their development, i.e., smaller than
those of B. typhosus or B. Gaertner. This difference is, how-

* A microbe frequently met with in Drigalski plates infected with
eewage, forming bright blue colonies, is conspicuous by its forming rounded,
very flat, dry, scaly colonies ; it need not trouble us here in our diagnosis of
faecal microbes specific or non-specific.

73

ever, obviously only of secondary value, since in a plate in
which colonies of B. typhosus or B. Gaertner are numerous
those that are in more crowded position are always much
smaller, although not slower in coming up, than where the
colonies are more isolated.

This last fact applies not only to Drigalski plates, or to the microbes at
present under consideration, but it applies to all kinds of microbes, and to all
kinds of cultures, viz., where isolated and sufficiently apart from one
another the colonies are much larger than in places where they are more
crowded.

A more important difference is the fact that in the hanging
drop made with a small particle of a colony many of the
bacilli are seen to be arranged in longer or shorter chains, a
fact not observed in similar colonies (from Drigalski medium)
of B. typhosus or B. Gaertner.

Sub-cultures made in the different media soon decide the
separate position of the microbe under consideration, for on
the gelatine surface it forms a slow-growing translucent
film ; in gelatine shake culture colonies do not appear in the
deeper parts, but are all crowded on and near the surface of
the gelatine, and it does not form gas in glucose gelatine ; in
litmus milk (at 37° C.) it forms strong alkalescence (blue)
after some days ; neutral red broth (at 37° C.) after some
days is turned orange, with slight fluorescence ; MacConkey
fluid (at 37° C.) remains unaltered ; it grows only very feebly
in phenol broth at 37° C. ; no agglutination is observed with
typhoid blood serum or with Gaertner blood serum even in
moderately high dilutions (e.g., 1 in 100) if a particle of a
recent gelatine culture is submitted to the test.

As stated above, this microbe is of constant and of fairly
numerous occurrence in sewage and in faecal matter; its
colonies on Drigalski medium resemble by their colour and
general aspect those of B. typhosus, the bacilli are short
cylindrical, very motile and forming chains, and, as just
detailed, are in all media in sub-culture distinctly and easily
differentiated from B. typhosus and -from B. Gaertner. It is
further to be remembered that unlike the typhoid-like
colonies not capable of growing at 37° C., mentioned on a

74

former page as having been met with in certain oysters, the
sewage microbe at present under consideration grows well at
37° C., and has not been met with in any of the numerous
Drigalski plates from clean shellfish which I have had before
me ; and therefore, without attributing at present to it any
special derivation, I can with confidence say that it is foreign
to the shellfish per se, and when found in it may with proba-
bility be taken to be derived from sewage or similar filth. In
order to be able to refer to it, I propose to name it Bacillus
streptoides, on account of its tendency to form chains.

(5.) As a last and important microbe which I have met
with, once in Drigalski plates inoculated with a trace of typical
fluid typhoid stool and twice out of five samples of sewage
of St. Bartholomew's Hospital, is a species which in all
morphological and cultural characters coincides with the
Bacillus fsecalis (alkaligenes), first isolated by Petruschki
from typhoid stools. The microbe which we isolated pro-
duced on Drigalski plates distinctly blue colonies, which in
their growth, form, and aspect might be mistaken for those
of B. typhosus or of B. Gaertner ; examined in the hanging
drop the component bacilli were cylindrical, motile, and
multiflagellated, and therefore not different from the B.
typhosus or B. Gaertner ; but they failed to become agglu-
tinated with typhoid serum or with Gaertner serum in
moderate dilution (1 : 20). On gelatine surface and in gela-
tine shake culture the growth quite resembles that of B.
typhosus, but here the similarity ends, for the microbe in
question produces in litmus milk (at 37° C.) forthwith
distinct alkali, the milk retaining its fluid character ; it turns
MacConkey fluid (at 37° C.) blue, and it produces no gas in
glucose media ; injected into guinea-pigs subcutaneously it
exerted no pathogenic action.

Dr. Durham * has given it as his opinion that the
Bacillus fsecalis (alkaligenes) of Petruschki is identical
with B. Gaertner. As stated just now, and as has been
well known to Petruschki, it has certain, characters in
common both with the B. typhosus as also the B. Gaertner.

* Brit. Med. Journal, December 17, 1898.

75

It appears to me that Durham's statement seems to rely,
besides the morphological characters of the microbe in ques-
tion, chiefly on the fact that it produces alkali in litmus
milk. But this will barely be considered a sufficient reason,
since, in the first place, other alkali-producing microbes do
the same, and, in the second place, the manner of the alkali
production is distinctly different for the B. fsecalis and the
B. Gaertner. If the B. fsecalis is planted in litmus milk,
and this is kept at 37° C., it will be noticed that the litmus
milk from the outset becomes more and more blue, whereas
with the B. Gaertner under the same conditions the litmus
milk for the first 24-48 hours shows a tinge of redness, that
is, slight acid production, and this gradually gives way to a
change into slate colour and later into deep blue. It
requires only sub-cultures in glucose gelatine, in MacConkey
fluid and in neutral red broth to establish marked differences,
for, as mentioned on a former page :
B. Gaertner turns neutral red broth greenish fluorescent ;

„ „ MacConkey fluid red, and forms slowly gas

therein ;

„ ferments glucose gelatine, forming gas therein ;
whereas the action of the B. faecalis is in these respects
quite negative.

The presence of blue colonies looking like those of
B. typhosus or B. Gaertner, composed of motile cylindrical
bacilli, and answering to the tests characteristic of
the B. fsecalis (alkaligenes), is therefore of an im-
portant diagnostic value, inasmuch as they have not been
at present found in any but human excremental matter ;
unfortunately, their occurrence in highly diluted 'matter,
such as is employed for making a Drigalski plate, is
rather rare, but if present it is of so much greater im-
portance.

On a former page we mentioned an experiment (IX) in
which one lot of Dutch oysters were placed in sewage
polluted sea water, the other lot being kept in sterile sea
water for control. After 24 hours one oyster of each lot, as
described in Experiment IX, was analysed by Drigalski

76

plates. In the plates of both oysters there came up blue
colonies of the following characters.

Drigalski plate of oyster 1, from clean sea water, con-
tained three blue colonies, that is, 150 per oyster ; these
proved to be colonies of a motile rapidly liquefying vibrio,
in many respects similar to vibrio finkler.

Drigalski plate of oyster 2, from polluted water, contained
crowds of beautifully blue colonies, slightly raised, homo-
geneous and watery. On microscopic examination the
majority proved to be motile vibrios, rapidly liquefying
gelatine like vibrio finkler.

Drigalski plate of oyster 5, kept three days in clean
water, had no blue colonies.

Drigalski plate of oyster 6, two days in polluted water, one
day in clean water, had two bright blue colonies. This would
make 100 per oyster. These were shown to be the same, viz.,
small watery colonies as above, viz., vibrio, but non-liquefying,
and in sub-culture of the same characters as those mentioned
as occurring in the sewage.

Oyster 4, having been kept for 48 hours in sewage
polluted sea water, when examined showed in a Drigalski
plate (prepared with ^ part of the oyster) two watery blue
colonies. This would amount to 100 per oyster. Sub-
cultures of both these blue colonies proved them to be of the
nature of the motile Bacillus streptoides described as isolated
from the sewage.

From this experiment we learn, then, that an oyster not
placed in the sewage polluted water (oyster 1) contained a
liquefying vibrio (probably similar to vibrio finkler), the
same as oyster 2, which had been 24 hours in sewage polluted
water ; further, that in another oyster (6), which had been
48 hours in sewage polluted and one day afterwards in clean
water, we identified the non-liquefying vibrio isolated from
the sewage ; and in an oyster (4) which had been 48 hours in
sewage polluted water we isolated the motile B. streptoides,
the same as was found in the sewage.

From these facts it appears, then, justifiable to conclude
that the oysters placed in sewage polluted water had imbibed

77

from that sewage the non-liquefying vibrio and the motile
B. streptoides, both forming blue colonies but of different
aspect, size, and constitution, and therefore capable of being
at once recognised by means of the Drigalski plates. In the
analysis of oysters for sewage microbes the presence in
Drigalski plates of these two species may therefore be of
diagnostic value.

We append here a number of photographic representa-
tions of the appearances in Drigalski plates made with
typhoid materials ; although the colonies are not coloured in
these photograms, it is not difficult to recognise their
character.

These photograms are reduced to half the actual size ; all
colonies, except when specially mentioned, appeared blue,
and of the uniform character and aspect of the typhoid
colonies, as described in the text; there was no difficulty
when inspecting them under a glass of ascertaining their
nature, which was also confirmed by agglutination test and sub-
cultures in the various media made of various colonies
indiscriminately chosen.

SERIES I.

Fig. 1. — Plate charged with TOO<JO~ part of a cubic centi-
metre of sea water immediately after infection of the sterile
water with culture of B. typhosus (see Experiment IV).
This plate contained 247 colonies of B. typhosus and no
others ; this amounts to 2,470,000 B. typhosus per 1 c.c.

Fig. 2. — Same sea water, 75000 Par^ °f a cubi° centi-
metre 24 hours after infection ; the plate contains 153
colonies of B. typhosus ; in addition one large (opaque) and
one small (dot-like) colony not blue and not B. typhosus;
this would therefore amount to 1,530,000 B. typhosus per
1 c.c.

Fig. 3. — The sea water in which the infected oysters had
been kept had been changed 24 hours previously ; TL c.c. of
it now examined yielded 1318 blue colonies of B. typhosus,

78

and in addition two stray, not typhoid, colonies : one on the
left middle, one near the upper left part of margin. The sea
water therefore contained 13,180 B. typhosus per 1 c.c.

SERIES II.

Figs. 1, 2, 3, and 4 represent plates made from wet oysters
of same lot as Experiment IV.

Fig. 1. — Oyster 1, kept for 24 hours in typhoid-infected
water — the plate was made with TJ-0 part of the oyster ; it
yielded 958 blue colonies of B. typhosus, and 9 colonies of
B. coli communis, seen in the photo as large, white, opaque
colonies. There were in addition three or four other non-
descript colonies ; the above result would amount to 95,800
B. typhosus and 900 B. coli communis per oyster.

Fig. 2. — From same lot, oyster 3, 24 hours in typhoid-
infected water, two days afterwards in sterile water ; TJ -0 part
of the oyster yielded 7520 blue colonies of B. typhosus,
calculated by counting as carefully as possible two separate
-^ sections of the plate ; there were no colonies of B. coli in
the plate. This would mean 752,000 B. typhosus per oyster.

Fig. 3. — From same lot, oyster 5, 24 hours in typhoid-in-
fected water, four days afterwards in sterile water ; TJ-0- part of
the oyster was used for the plate ; it yielded 12 (blue) colonies
of B. typhosus, three large and ten minute neutral nondescript
colonies. This would mean 1200 B. typhosus per oyster.

Fig. 4. — From same lot, oyster 9, 24 hours in typhoid-
infected water, seven days afterwards in sterile sea water;
J part of the oyster yielded 63 (blue) colonies of B. typhosus.
This amounts to 378 B. typhosus per oyster.

SERIES III.

Figs. 1, 2, and 3 represent plates made from oysters of
dry lot, of same Experiment IV.

Fig. 1. — Oyster 2, 24 hours in typhoid-infected water,
and two days afterwards kept dry ; jfa part of the oyster

79

was used for the plate; this yielded 587 (blue) colonies of B.
typhosus. This amounts to 58,700 B. typhosus per oyster.

Fig. 2. — Oyster 4, 24 hours in typhoid-infected water,
four days afterwards kept dry ; TJff part of the oyster was
used for the plate; this yielded 174 (blue) colonies of B.
typhosus. This amounts to 17,400 B. typhosus per oyster.

Fig. 3. — Oyster 8, 24 hours in typhoid-infected water,
seven days afterwards kept dry ; TJD part of the oyster was
used for the plate ; this yielded 13 (blue) colonies of B.
typhosus. This amounts to 1300 B. typhosus per oyster.

•LONDON : PRINTED BY WILLIAM CLOWES AND SONS, LIMITED,
DUKE STREET, STAMFORD STREET, S.E., AND GREAT WINDMILL STREET, W.

SERIES I.

FIG. 1,

FIG. 2.

FIG. 3.

SERIES II.

FIG. 1,

FIG. 2.

FIG. 3.

FIG. 4.

SERIES III.

FIG. 1.

FIG. 2.

FIG. 3.

,t

RETURN BIOLOGY LIBRARY

TO—*- 3503 Life Sciences Bldg. 642-2531

LOAN PERIOD 1

2

3

1 -MONTH

•5-MON<X

&APH

ALL BOOKS MAY BE RECALLED AFTER 7 DAYS
Renewed books are subject to immediate recall

.^DUE AS STAMPED BELOW

APR .134^

9

d

r

%

&

UNIVERSITY OF CALIFORNIA, BERKELEY

FORM NO. DD4, 12m, 12/80 BERKELEY, CA 94720

®$

VUI