VOL. Xr. — NEW SEB. I

323

L6

The Origin and Distribution of Microzymes (Bacteria)
in Water, and tlie CIRCUMSTA^CliS which determine their
Existence in the Tissues and Liquids of the Living
Body. By Dr. Burdon-Sanderson, F.R.S.

{Reprinted, bi/ permission, from the author's second "Report of Researches
concerning the Intimate Vathology of Contagion" in the Appendix to the
' 13M Report of the Medical Officer of the Priny Council')

In my previous report on the intimate pathology of con-
tagion, microzymes were defined as living particles which in
their earliest state are spheroids, and do not exceed -a-o,^^ of
an inch in diameter, but subsequently elongate into rods.
As regards the conditions of their development, their ex-
istence was said to be associated with the commencement of
putrefactive decomposition of nitrogenous compounds. The
question of their origin and destiny was left unanswered. It
was left undecided whether on the one hand " they constitute
a race of more or less similar individuals, each of which
springs from and reproduces its like," or, on the other, are
" germs in which a specific form is Avrapped up/' capable of
developing to the higher organisms from which they spring.

It is to this question principally that the experiments we
have now to bring before the reader relate. Our purpose is
to exauiine into the origin, growth, and develoi)ment of
microzymes, to investigate the conditions which are fatal or
favourable to their existence in the liquid and gaseous fluids
by which we are surrounded, in the hope that by doing so
we may be enabled to approach one degree nearer to an
understanding of their influence on the processes Avhich go on
in the living body.

In dealing with the question of origin, we again encounter
the more general question of what is called " spontaneous
generation." I have no intention, however, of entering upon
it. I shall be able to prove in the most decisive manner that
as regards the animal liquids and tissues, and the liquids
which will be used as tests for the presence of microzyme
germs, no spontaneous evolution of any organic form ever
takes places but it will be quite unnecessary either to deny
or to assert its possibility under other and different conditions.

1

324

Before proceeding to state the results of our experiments,
a more complete account must be given of niicrozymos, and
something must be said as to the views entertained by-
naturalists of their nature, origin, and relation to other
organic forms. The methods of investigation which have
been employed must also be explained.

Bacteria or microzymes are placed by most naturalists in
the animal kingdom, and have a position assigned to them
next to the monads. Hallier, on the other hand, believing
that they originate by the cleavage and multiplication of
nuclei in the cells of fungi, and that they develop to the same
forms from which they spring, regards them as plants. Their
claim to be considered animals is founded partly on their
motions, partly on the fact that their chemical reaction on
air, when alive, resembles rather the respiration of animals
than that which is associated with vegetation. The question
is of importance only in so far as it involves that of origin
and development. If it can be shown that they neither
spring from higher forms nor grow to them, the discussion of
their animal or plant nature may be left to those interested
in verbal definitions.

Microzymes grow either in liquids or moist air. In liquids
they present different appearances, as they are observed in
the depth or on the surface. In the former case they show
no tendency to assume any special arrangement to each other
if they are motionless ; nor if they are active are their motions
governed by any mutual relation. At the surface of the
liquid, on the other hand, although the individual bacteria
show no definite arrangement when they first appear, they
soon place themselves in such a manner as to form a mem-
brane, the beginning of the bacterium scum, to which we
shall have frequent occasion to refer. In this membrane,
when it first appears, each rod stands vertically, one end
forming part of the free surface, the other part of the deep
surface of the membrane. The rods adhere together by their
sides after the manner of the elements of columnar epithelium,
but there is, I think, strong reason to believe that this adhe-
sion is not direct, i. e. that they are not in actual contact, but
glued together by a viscous intermediary substance. Con-
sequently on this arrangement the " scum," when first formed,
presents under the microscope the aspect of an evenly dotted
surface, the distance between each dot and its neighbour
corresponding approximately to the diameter of a rod. This
appearance, indeed, is so deceptive, that for a long time I
supposed, as others have done, that the constituent particles
were round ; nor was it until it was discovered that the mem-

325

braiie could be resolved by mechanical means into rods that
I understood the real nature of the membrane. As the struc-
ture (if one may call it so) becomes thick enough to form a
visible scum, the arrangement of the particles can no longer
be made out, for it is not possible to subject it to examination
Avithout dislocating it to such a degree as to render their
relative positions indistinguishable.

When common microzymes grow on moist surfaces, they
with their intervening jelly sometimes form viscous masses of
sufficient size to be cognoscible by the unaided senses, these
consisting of a material similar to that of the " scum " which
forms on the surface of liquids in which microzymes are
growing. This fact is expressed by the term Zooglma applied
to such masses or colonies of microzymes by Cohn.

It is on observations made as to the growth of microzymes
in colonies that tlie little Avhich can be stated as to the /brw
in which they originate is based. In the spheroidal masses
above referred to, and indeed whenever microzymes occur in
a gelatinous matrix which can be distinguished, we observe
foci of growth at which the particles are indefinitely minute
and spheroidal; around these foci there are zones of matrix,
already obsolete and, disintegrating, which are inhabited by
staff-shaped microzymes of larger size, which eventually
become free and display their proper movements. Here
therefore it seems probable that bacteria come into dis-
tinguishable existence not as rods but as spheroids. Sub-
sequently they multiply, as is well known, by division.

As to the conditions of tlieir^ origin there is even less
knowledge and more difference of opinion. There being an
immense preponderance of evidence that they do not spring
into existence of themselves in the media in which they grow,
most observers have looked for germs in the atmosphere, but
with no success. Nor has anyone excepting Professor Hallier
even suggested a plausible theory on the subject. Liqtiids
which contain no particle distinguishable under the highest
powers of the microscope can often (as will be hereafter
shown) be proved to possess the property of evolving
microzymes without contact with external media, and must
therefore contain the germinal substance from which these
organisms spring. In interpreting this fact it may be sup-
posed either that the germinal substance is universally and
equally distributed, i.e. dissolved in such liquids, or that^it
is unequally distributed or particulate. That any living sub-
stance is soluble in water is not at present admissible ; we
must therefore accept the other alternative, and believe that
we have to do with particles so minute that they do not

326

interfere with the homogeneity of the liquid. In so far as
relates to the ultra-microscopical origin of the hacteria, this
inference harmonises entirely with what lias heen stated
ahove as to their development in gelatinous masses from foci.
Here, as in the other case, it would surely he an error to
suppose that in these proliferous foci the apparently hyaline
material is really homogeneous. It appears to he so, merely
because the particles are so extremely minute. Hence when
we apply the term matrix to this subject, we must guard
against the word being understood to imply that in the
present instance bacteria arise out of an amorphous jelly.
What is meant is, that the jelly is itself so organised through-
out, that the smallest conceivable bit of it, if separated from
the rest, would still possess structure, and consequently the
power of reproduction.

Chemical composition of Microzymes, and their relation to
the media in which they groio. — Of the chemical composition
of microzymes we kno^y very little. It is assumed that the
particles are albuminous, because they are readily stained,
with carmine and browned by iodine ; but of the matrix little
can be said, excepting that it is probably also albuminous.
Chemistry can as yet give no account of the difference
betweeen them. As regards their action on the liquids in
which they live the most important facts are : (1) That their
growth is attended with absorption of oxygen and discharge
of carbonic acid. (2) That they are remarkably independent
of the chemical constitution of the medium, provided that
they are supplied with oxygen ; and (3) That they take
nitrogen from almost any source which contains it, and use
it for the building up their own protoplasm.

It is this last power which specially indicates what may be
called their place in nature as the universal destroyers of
nitrogenous substances, acting as the pioneers if not the pro-
ducers of putrefaction. They exercise this function not by
virtue of any special relation of their own nutritive processes
to putrefaction as such, but simply by their extraordinary
power of seizing on the elements which they require for the
construction of their own bodies.

The necessity of oxygen to bacteria is so great that they
cannot grow even for a short time without it. Thus if liquid
containing living bacteria be placed under a cover glass for
Microscopical examination, it is seen that toAvards the centre
of the cover glass their movements become sluggish and
eventually cease, although towards the edges they are still
lively. If bacteria are confined in a tube without air they
soon die. If the supply of air is limited they continue to

3.27

live only so long as the air to which the liquid is exposed
still contains sufficient oxygen. If microzymes exist in great
numbers in a liquid, air which has remained for a length
of time in contact with it has a large excess of carbonic acid,
and occupies less volume than it did originally under the
same conditions of pressure and temperature. We have
found that in such air a taper is immediately extinguished,
whence it would seem that microzymes are able to use up
nearly the whole of the oxygen which is supplied to them.

When microzymes grow at the expense of disintegrating
organic substance, it cannot be supposed that they avail
themselves of the albuminates already existing in it to build
up the material of their own bodies. If this were the case
it would be impossible to understand the fact that they grow
quite as luxuriantly when the nitrogen they require is sup-
plied to them in the form of salts of ammonia, as when it is
in the form of ready-made albumen ; for clearly it must re-
quire a much greater expenditure of plastic energy to build
up protoplasm of elements derived from such sources, than
merely to convert one albuminous compound into another.
It therefore seems probable that bacteria do not use the
material on which they feed until it has already been con-
verted by oxidation or by splitting into lower chemical
combinations.

The question how far microzymes are the cause of putre-
faction will, I think, be elucidated by the results of the fol-
lowing experiments. It wall be shown that so long as the
germinal matter of microzymes is excluded, animal fluids or
tissues withstand decomposition for very long periods, while
the slightest contact with media containing this material at
once determines septic changes. Consequently it can be
asserted positively that under certain circumstances the
presence of microzymes excites putrefaction ; but the facts
do not afford grounds for stating that they a^-e the cause of
putrefaction, or that if it were not for them the process would
be postponed indefinitely. It is indeed asserted by chemists,
and we do not propose to deny, that organic matter may,
under the influence of heat and moisture alone, undergo
decompositions which present all the chemical characters of
putrefaction, even though no microzymes be present.

Method. — As regards the questions which form the prin-
cipal subject of this report, we at present possess no exatt
information. As has been already stated, there is a general
belief that microzymes exist potentially in the air, and it is
also admitted that they may be met with in the blood in
certain septic diseases. Ilallicr, on the other hand, finds

328

thetn not only in septic, but in all contagious fluids, while
Bcchamp imagines that they form part of healthy structures.
To determine these questions it was necessary (1) to subject
the media to the action of some qualitative test by which
the presence of the germinal matter of microzymes could be
detected ; and (2) to make experiments in which their action
on the animal liquids and tissues would be observed under
conditions similar to those which exist in the living body.
As a test for the presence of microzyme germs we have used
first, Pasteur's solution, and secondly animal fluids, either
diluted with pure water or undiluted. These liquids were
selected on the ground first that they contain nitrogen, in
the one case in the form of an amuionia salt, in the other in
that of an albuminous compound, and secondly that although
transparent and free from visible particles when fresh, they
become in a short time peopled with microzymes when kept
under ordinary circumstances and at ordinary temperatures.
Before using them, however, for the purpose intended, it was
necessary to determine that they do not in themselves con-
tain the conditions of evolution; in other words, that they
can be prepared and kept in a state of absolute barrenness
without prejudice to those qualities by which they are fitted
to be employed as tests. These requirements could only be
satisfied by a preliminary series of experiments having for
their purpose to determine the question of so-called " spon-
taneous generation," not in general, but with respect to the
particular liquids to be used. In approaching a question of
such difficulty, even with the limitation above stated, there
are two methods of inquiry Avhich suggest themselves ; one
consists in the comparison of results obtained when the cause
to be investigated is present, with those which are produced
Avhen it is absent, all other conditions remaining unaltered
(method of crucial experiment) ; the other in the comparison
of variations iif the results wath variations in the circum-
stances that lead to them (method of concomitant variations).
We shall find that the first of these methods, which is clearly
the most conclusive, is as applicable as the other to the
particular question before us, that of the spontaneous evolu-
tion of organic forms in any given medium. But even if it
had not been so, the other method would still have been open
to us, for if it could be shown that the appearance of micro-
zymes in a given liquid is either delayed or diminished, in a
degree proportionate to the degree of exposure to external
influences, it might be safely inferred that exposure to the
air is the efficient cause of their development. In the present
instance it is possible to exclude all conceivable sources of

329

contamination, and so to obtain a positive answer to the ques-
tion ; but this does not render it the less advantageous to
compare the varying effects of contamination with the con-
ditions to which they correspond, for by so doing we acquire
a better knowledge of the nature of these causes, and of the
means of obviating them.

The experimental results are stated under three headings,
according as they relate to the conditions w^hich limit the
evolution of organic forms, and particularly microzymes, in
test liquids; to their distribution in ordinary water and in
most substances ; and lastly to their occurrence in the tissues,
and liquids of the animal body.

While considering myself exclusively answerable for the
accuracy of every statement contained in the report, I am
anxious that in so far as the investigation has been a suc-
cessful one, my assistant. Dr. Ferrier, by whom many of the
experiments were both planned and carried out, should
participate in whatever credit may be accorded to me.

Section 1. — Experimental Determination of the Conditions
which govern the Development of Microzymes in certain
Organic Liquids to he used as Tests.

I. — July 22nd, 1870. — A large number of capillary tubes
prepared for the purpose were filled with serum of blood
obtained from a guinea pig a few hours before. According
to the mode of filling, and the conditions under which they
were subsequently placed, the tubes were divided into five
batches, designated respectively a, h, c, d, and e. The tubes
a were exposed, unsealed, to the air of the laboratory ; h
were hermetically sealed ; c were sealed, and thereafter placed
in the incubator, in which a temperature of about 40° C. was
maintained during the whole period of the investigation, with
the aid of a Geissler's regulator ; d were sealed and heated
in the oven to 180° C, and thereafter exposed to the air of
the laboratory by breaking off one end ; e were sealed and
heated in the same manner as d, and then placed in the in-
cubator. The tubes were examined at various periods within
a month after they were prepared. Bacteria were found in
numbers in a, b, and c, but no organic forms could be detected
in d and e. The remainder of these two batches were there-
fore preserved for further examination, until the beginning
of March, 1871. They then exhibited the appearances always
observed under the microscope in superheated serous liquids,^

' The most remarkable peculiarity of such liquids is that they contain

330

but on the most scrupulous examination no organic forms
could be discovered either in the tubes which had been kept
at the ordinary temperature, or in those which had remained
in the incubator.

On August 11 the experiments were repeated under similar
conditions, Avith the exception that the serum employed to
fill the tubes was first rendered alkaline by the addition of
0-5 per cent, of soda. In this case no organic forms had
appeared in any of the tubes at the end of a month, nor
could any be discovered afterwards. The superheated tubes
were examined with the rest in March, 1871, with the same
result. Two quantities of the same serum were kept in
glasses side by side in the laboratory, to one of which only,
soda had been added in the proportion already mentioned.
In the one containing no soda a luxuriant growth of bacteria
and leptothrix appeared in a few days, but nothing whatever
could be found in the other. Soda, therefore, in the propor-
tion of half per cent., appears to prevent the development of
microzymes.

II. — August 24, 1870. — The albumen of a fresh egg was
collected in a clean dry test glass, and several tubes of
tolerably large size were filled in the ordinary way and her-
metically sealed. Some of them which were not heated
•were kept at the ordinary temperature, others were subjected
in the hot-air oven to a temperature of 200° C. All of these
tubes were kept until March, 1871, when it was found that
the unheated tubes were still perfectly clear, with the excep-
tion that on the side which was undermost as the tube Liy
on the shelf, its internal surface was lined with whitish
granular deposit. The liquid showed no other change, and
on a microscopical examination no organic forms could be
found. The superheated tubes were in this respect in the
same condition. From the negative results in the tubes
which had not been heated, it might be inferred that white
of egg is incapable of maintaining the life of microzymes,
but we shall see hereafter that the fact admits of a totally
different interpretation.

III. — August 30. — A large number of capillary and other
tubes were filled with a solution of sugar, tartrate of ammonia,
and yeast ash, according to M. Pasteur's formula, and divided
into two batches, designated respectively a and h. Some of
the tubes a after having been sealed, were kept either at the
ordinary temperature or in the incubator. The rest were left

masses of apparently semi-fluid material resembling oil drops. These masses
are of a distinctly yellow colour, and vary indefinitely in size. They are
found in superheated liquids immediately after they are prepared.

331

open and kept in the laboratory, b were sealed and raised
to a temperature of 200° C. Some of them were afterwards
placed in the incubator, others remained at the ordinary
temperature. Specimens of a and h were examined at various
periods up to March, 1871. All of the tubes a became turbid
sooner or later, and were then found to be crowded in different
degrees with bacteria and fungi (torula cells and mycelium).
When the remaining tubes were finally opened it was found
that in many of them gas had been disengaged in such
quantity, that when the end of the tube was broken off the
liquid was expelled with violence. In others this evidence
of increased tension was wanting. On comparative micro-
scopical examination it was found that the liquid in these
last, contained no torula cells, h were kept till March, 1871,
at which time they were found to exhibit no trace of organic
life, whether they had been kept in the incubator or at the
ordinary temperature.

IV. — August 18. — It has been imagined that the so-called
spontaneous evolution of organic forms is materially increased
when the air to which the liquid is exposed has a tension
much inferior to that of the atmosphere, and conversely that
in liquids subjected to pressures greater than that of the
atmosphere the development of such forms is arrested. The
following experiments were made to test this supposition.
Several capillary tubes were filled with fresh serum of blood
of a rabbit kept in an ordinary clean glass. These were
sealed and placed in a larger glass tube closed at one end,
which after having been drawn out at a short distance from
its open end was attached thereby to one branch of a T tube
by means of a vulcanite junction. The stem of the T tube
was then connected with an air pump, and the other branch
with a long barometer tube standing vertically in a cup of
mercury. The air was then exhausted, and as soon as the
mercury had risen in the barometer tube 15 inches, the flame
of a blow-pipe was directed against the narrow drawn-out
part of the experimental tube, which was thus sealed while
the air which it contained had a tension not more than half
that of atmospheric air. The tube was then shaken so as to
break all the capillary tubes, so that the whole of the liquid
which they contained was exposed to the pressure above
indicated. It was then kept at the ordinary temperature.
Another tube was filled with capillary tubes containino-
serum, exhausted to 15 inches, and closed hermetically in
the same way. It was then placed in the oven and raised to
a temperature of 200° C, after which the capillary tubes
inside were broken as before, so as to expose the liquid to

332

superheated air at 15 inches pressure. Both of the tubes
were kept until March, 1871. On opening tlie one which
had not been heated, air rushed into it with great force. Its
contents had a putrid smell, and the liquid on microscopical
examination was found to contain numerous bacteria. When
the superheated tube was opened, the ingress of air was
equally forcible, but on microscopical examination no trace
of organic forms could be discovered.

From this experiment it would appear that diminished
tension has no very considerable effect on the process we are
studying. It is further evident tliat the non-appearance of
organic forms in superheated liquids cannot be accounted for
by supposing that it is attributable either to the relatively
large proportion of the liquid, as compared with the volume
of the air which is enclosed with it, or to any other circum-
stance arising from its being contained in so small a receptacle.
A third experiment of the same kind was made on August
30th. A number of capillary tubes were filled with Pasteur's
solution, and then sealed and introduced into a large tube,
closed in the same manner as in the previous experiment.
The whole was then subjected to a temperature of 170° C,
after which the contained tubes were broken by shaking.
On examining the liquid contained in the broken capillary
tubes after several months no organic form could be detected.

The above observations (I to IV) show conclusively that no
evolution of organisms took place in the superheated liquids,
provided that the air with which they were in contact had also
been superheated, whether they were kept at an ordinary
temperature or at that of the body ; and that the effect was not
modified, either by the tension of the air or by its quantity
as compai-ed with that of the liquid ; and it is further shown
that in all the experiments, organisms appeared in the same
liquids kept under precisely similar conditions, which had
not been superheated. Before, however, drawing any further
conclusions from these facts it may be inquired, in how far
the cause of the non-appearance of organic forms is depen-
dent on the liquids having undergone chemical changes of
such a nature as to render them incapable of supporting life,
in which case the negative results obtained could not be
attributed exclusively to the non-exposure of the liquids to
external media. It will be shown in the sequel that this is
true as regards microzymes, that is to say, that superheated
organic liquids are incapable of supporting the life of these
organisms. It is therefore clear that such liquids do not
furnish a suitable soil for studying the question we have in
view. With respect to fungi, however, the case appears to

333

be difFevcnt; for numerous experiments show that super-
heated liquids, and particularly Pasteur's solution, when
freely exposed to the air in a watch glass or
an open tube, become eventually covered with
tufts of penicilliura.

In the further progress of the inquiry it was
found entirely unnecessary to employ so high a ^
temperature as 170° to 200° C. in order to pre- |
vent the evolution of organic forms, provided |
that the liquids were protected from contamina- t
tion by external media. The experiments which S
led to this result were as follows : —

V. — August 10. — A number of tubes were %
filled with serum of rabbits' or sheep's blood. I
They were then sealed and boiled for an hour |
or two in a water bath, in consequence of which
the liquid contained in several of the tubes ^
became gelatinous, still, however, remaining ^
perfectly transparent. From time to time during J
the next few months a tube was broken for .2
microscopical examination of its contents, the •§
result being always negative. In March, 1871, ^
the remaining tubes were finally examined. No
organic forms could be traced either in those |
which were gelatinous or in those which re- I
mained liquid. S*

VI. — October 5. — Pasteur's solution was pre- ^
pared according to formula (the distilled water I
employed for the purpose being obtained from "2
Messrs. Hopkin and Williams), and placed in ^
a clean capsule. A number of tubes of various §
sizes were then filled, in the manner already |
described, with the solution (which had not 3
been heated), and sealed. The solution was g
then boiled in the capsule for a few minutes, ^
and another batch of tubes were filled in the ^
same manner by breaking their points under-
neath the surface of the liquid while it was ^
still in a state of ebullition. Each tube was \
sealed the moment it was withdrawn from the
boiling liquid. The two sets of tubes were f|
placed side by side in the laboratory under

precisely similar conditions. Some of them were examined
microscopically on the 17th. In those of the first batch
microzymes in immense numbers, and torula cells, were
found along with several filaments of sporotricIm7n. No

334

organisms whatever existed in any of the tubes containing
the boiled liquid. Single tubes of both batches were exainined
from time to time until March, 1871, the results being always
the same. Hence it was concluded that thoroughly boiled
liquids, preserved in tubes first prepared and sealed,* remain
perfectly free from organic forms.

VII. — October 5. — Four tubes, each a quarter of an inch
in diameter, were prepared in the usual manner and filled
with Pasteur's solution which had not been boiled. Tube
a was placed vertically in a cork support, its end being
truncated so as to expose the upper surface of the liquid to
the air. Tube b was also placed upright, its upper end
having been previously drawn out to a long capillary beak,
the tip of which was broken off, so that the interior of the
tube communicated with the atmosphere by a small aperture.
Tube c, of the same form as b, was also placed vertically, but
its open point was bent downwards at a very acute angle.
Tube d was sealed at both ends.

Four similar tubes marked respectively a', b', c' , df , were
then filled with boiling solution and placed side by side with
the others, three of them having openings of the characters
already described, the other being closed. On October 12
the only change which could be distinguished without the
microscope was a very remarkable one. A tuft of penicillium
had appeared on the surface of the liquid in tube b' , the
interior of which communicated with the air only by a capil-
lary aperture. Nothing was visible in the others ; but a few
days later it was observed that all the open tubes {a, b, a' , b'),
excepting those of which the ends had been bent down, had
similar tufts. In the course of the following six weeks the
tufts increased considerably in size. On the 24th of November
the liquid in tubes c and d, in which no penicillium existed,
was observed to be hazy and had a slight scum on the surface.
Tubes c' and d' remained perfectly unaltered. The liquid in
the open tubes was examined microscopically from time to
time during the period of observation, the drop required for
this purpose being on each occasion transferred to the object
glass of the microscope either by means of a glass rod, the end
of which had been first passed through the flame of a Bunsen's
burner, or the capillary tube which had been drawn out immedi-
ately before, so as to avoid all risk of contaminating the liquid.

In all of the open tubes containing unboiled solution torn] a
cells and microzymes began to appear after the first week.
On November 24 they existed in great numbers, in addition
to mycelium and filaments and spores of sporotrichum. In
the closed tube d there were bacteria but no torula or peni-

335

cilliuni. At the same date the open tubes a! and h' , contain-
ing- boiled sohition, were free from microzymes, but contained
numerous toruLa cells and mycelium, c' and d' were not
examined until the 4th of January, at which time both
liquids were perfectly clear and contained no organic forms
of any description.

VIII. — October 5. — Two test glasses were placed side by
side on a shelf under a glass shade, one of which, marked a,
contained unboiled Pasteur's solution, the other, marked h,
boiled solution. On October 10 glass a was turbid, and was
found on microscopical examination to be teeming with
bacteria ; a thick whitish scum had formed on its surface.
Glass h was perfectly clear ; there were, however, great
numbers of torula cells on its surface, but no bacteria. On
October 12 6 exhibited numerous tufts of penicillium, but
the liquiil still remained limpid and free from bacteria ; five
days later similar tufts appeared on the surface of a.

In the last two experiments it is seen that fungi (torula
and penicillium) appeared in unboiled solutions whether they
were exposed or not, but much more abundantly when they
were exposed than when they were protected; and that in
boiled solutions the growth of penicillium was somewhat
more luxuriant than in unboiled under similar circumstances
of exposure. Microzymes did not appear in the boiled liquids
under any circumstances, but were quite as numerous in the
tube d (Obs. VII), which remained closed for many months,
as in any other of the same series. From these facts it seemed
clear, not merely that the conditions of origin and growth of
microzymes and fungi are considerably different, but that as
regards the former the germinal matter from which they
spring does not exist in ordinary air. The experiments to be
next related, however, showed that it would have been wrong
to have inferred from these facts that the boiling of a liquid
is of itself sufficient to prevent the development in it of these
organisms, or that their complete absence in the tubes of the
second series of Observation VII h' , c' , d') was exclusively
attributable to this condition.

IX. — October 25. — A solution (A) according to Pasteur's
formula, was prepared in the same manner as before, with
the exception that water distilled on the same day in the
laboratory was used instead of the ordinary distilled water,
great care being taken to prevent its contamination. At the
same time another solution (B) was made with the same
water, of materials which had been previously heated in the
hot-air bath to 110° C. Eight glasses were at the same time
prepared, of which four, marked severally with the odd num-

336

bcrs 1, 3, 1', and 3', were washed and dried with a towel.
The remainder, numbered 2, 4, 2', and 4', were immersed for
some time in a vessel of boiling water and then dried as
before. The two solutions were then distributed in these
glasses as follows : — In 1 and 2 solution A unboiled ; in S and
4 the same solution after previous boiling ; in V and 2' solu-
tion B unboiled ; in 3' and 4' the same after boiling. Glasses
1, 2, r, and 2' were placed under one shade, and the other
four glasses under another. On November 1, tufts of ])eni-
cillium were obvious on 1, 2, 1', and 2', and were beginning
to appear on the rest. The liquids were examined micro-
scopically at this date and again on November 8, when the
tufts had increased in size. All contained torula cells -and
mycelium, but microzymes were found only in 1, 3, 1', and
2'. Thus it appeared that neither the boiling of the liquids,
nor of the glasses, nor the superheating of the materials, had
exercised any appreciable influence in preventing the develop-
ment of microzymes. It was still more remarkable that in
glass 2, which contained unboiled solution, none of these
organisms could be discovered.

These facts, apparently so contradictory, were explained
by subsequent experiments.

X. — November 11. — Pasteur's solution was prepared with
ordinary distilled water obtained from Messrs. Hopkin and
Williams, and distributed in five glasses designated by num-
bers, the conditions being as folloAvs : — 1, a clean test glass,
taken from the shelf, was filled without further cleansing
with solution which had not been subjected to heat ; 2, a
similar glass, previously rinsed with distilled water, was filled
with the same liquid ; 3, a glass just before heated to 200° C.
was also filled in like manner. The other two glasses (4 and
5) were charged with boiled liquid, the method used being to
boil the solution in a test tube for a few minutes, then to cool
it rapidly by dipping it in a stream of cold water, and transfer
it at once to the experimental glass. Glass 4 was merely
rinsed with distilled water ; 5 was previously heated to 200° C.
The results were as follows : — On November 20, torula cells
were found on the surface of all the liquids. On the 26th,
bacteria had appeared in immense numbers in 1, 2, and 4, so
that the liquid was milky. In 3 it was apparently clear, but
was found on microscopical examination to contain bacteria.
Subsequently it also became opalescent. At the same date
all the glasses showed tufts of penicillium ; those on 3 and 5
were more advanced than the rest, and had become greenish
from the development of heads of spores. At this time, and
on all subsequent occasions, the liquid in 5 was found to be

337

perfectly limpid and free from microzymes. The conditions
under which the liquid in glass 5 was placed differed from
those to which that in glass 4 was subjected in one particular
only, viz., in the fact that the former, instead of being rinsed
with distilled water and dried, had been superheated. The
teaching, therefore, of the experiment was, that the germinal
particles from which the microzymes sprung must have been
contained either in matter adherent to the surface of the
glass, or in the distilled water used to cleanse it, or in both.
That the former was not without its influence is rendered
probable by the circumstance that in glass 3, which differed
frOm 2 only in having been superheated, bacteria appeared
latest. To determine this question was the purpose of the
next experiments.

XI. — December 1. — Pasteur's solution was prepared with
water obtained from a well at Stevington, in Bedfordshire,
which was sent to the laboratory for microscopical examina-
tion. The water in question was perfectly limpid ; but after
it was allowed to stand, a few microzymes could be discovered
in the surface layer. None could be detected in the rest of
the liquid. It contained a scanty deposit in which one or
two monera occurred. The solution was distributed in five
test glasses, the conditions being as follows: — (I) The solu-
tion was boiled in a tube, cooled rapidly, and then poured
into a test glass which had just been heated to 200° C. ; (2)
solution boiled in the same manner was transferred to a glass
which had been rinsed and dried ; (3) the boiled solution was
received in a superheated glass, but just before pouring it in,
the glass was rinsed loith ordinary distilled water ; (4) the
conditions were exactly the same, excepting that the distilled
water used for rinsing had first been h oiled ; (5) the solution
was not boiled, but the glass in which it was placed had been
previously superheated. The five glasses were numbered in
the order in which they have been referred to, and placed
under one shade. On December 7, 5 was already milky, the
turbidity being due to torula cells and bacteria ; 2 and 3 also
contained bacteria. On December 8, the turbidity of 5 had
increased, and 3 was opalescent. There were no microzymes
either in 1 or 4. On the 13th, there were tufts of penicil-
lium on all the liquids ; the tufts were more advanced in
fructification in 1 and 4 than the rest, but these liquids were
still entirely free from microzymes. The last examinations
were made on the 21st of December, when 1 and 4 were still
in the same condition.

II ere the two liquids in which no development of micro-
zymes took place differed from each other in the circumstance

338

that the glass in which one of them was contained (4) was
rinsed with boiled distilled water just before it was charged,
both glasses having been superheated. On the other hand^
3, in which microzymes appeared, differed from 4 only in the
omission of the boiling of the water used for cleansing. By
the comparison of these two results we were enabled to con-
clude that ordinary distilled water may contain the germinal
particles of microzymes in such profusion that even so small
a quantity as is introduced into a glass in rinsing is sufficient
to render a relatively enormous volume of liquid fruitful.
The following is one of a series of experiments which were
made to confirm this result: —

XII, — December 3. — Pasteur's solution prepared with the
same (Stevington) water was distributed in six glasses, all of
which were superheated. Of these three, marked c, were
filled with solution which had not been boiled, the remainder,
b, with boiled solution. They were placed in pair's, one of
each series in each pair, in different rooms. On December
8, the glasses c were all hazy, and found to contain in-
numerable bacteria, b were perfectly transparent ; as time
went on the contrast became more and more striking in con-
sequence of the increased turbidity of c. Subsequently tufts
of penicillium appeared on the surfaces of all the glasses,
which in this as in the previous experiments progressed more
rapidly in the clear solutions than in the others.

This experiment was repeated several times with corre-
sponding results.

In many preceding experiments it has been shown that
although torula cells and penicillium appear invariably and
without exception on all nutritive liquids of which the sur-
faces are exposed to the air, without reference to their mode
of preparation, no amount of exposure has any effect in deter-
mining the evolution of microzymes. This conclusion although
it is in complete accordance with what we have already learnt
as to their relations both in the visible and invisible state to
moisture, is of such importance that it seemed necessary to
establish it by special experiments.

XIII. — January 7. — The bent glass tube for the absorp-
tion of carbonic acid by potash, known as I.iebig's bulbs,
was heated to 200° C. and filled with boiling test solution.
It was then attached by a vulcanite connecter which had
been previously boiled, to an aspirator. During the following
week air was drawn through it for a few hours daily. On
the 23rd there were numerous torula cells with submerged
tufts of mycelium in the liquid, especially in those bulbs to
which the air had access first, but no trace of microzymes.

339

On March 18 the surface of the liquid in the first bulb was
crowded with a dense crust of pcnicilliuni ; in the last bulb
there were no tufts, and the liquid was still entirely free from
microzymes. The result shows in a most striking manner
not only that ordinary air is entirely free from living micro-
zymes, but that the activity of the development of penicil-
lium is in proportion to the degree of exposure.

XIV. — March 2. — A test tube, containing Pasteur's solu-
tion, in which there were immense numbers of microzymes
and torula cells (penicillium), was plugged with cotton wool,
boiled for a few minutes, and placed, still plugged, in a rack,
where it remained for some time. The liquid which, at the
time of boiling, was quite opalescent, gradually became clear,
from the subsidence of the organisms it had contained. It
remained perfectly clear and free from organic forms until
the 18th of March. The plug of cotton wool was then re-
moved, soon after which tufts of penicillium appeared on its
surface; but up to the present time (March 31) the liquid is
entirely free from microzymes.

Section II. — Distribution of the Germial Matter of
Microzymes in ordinary Water.

Having thus found in a number of cases that either contact
with surfaces which had not been superheated, or the admix-
ture of water which had not been boiled, was the exclusive
cause of the growth of microzymes in the experimental
liquid, it was not necessary to go far in order to arrive at the
inference that water is the primary source from which the
germinal particles of bacteria are derived, whenever they
seem to originate spontaneously in organic solutions. To
prove this is a number of experiments were made with
different varieties of water in ordinary use, in order in the
first place to confirm the observations already made, and to
ascertain whether all waters possess the properties in ques-
tion in a like degree.

XV. — January 2, 1871. — A number of eprouvettes of the
form shown in the margin were placed in the hot-air oven
and heated to ;300°C. They were then filled with Pasteur's
solution made Avith ordinary distilled water, under the fol-
lowing conditions : — a. Solution not subject to heat, h. Solu-
tion introduced boiling, which was then allowed to cool;
immediately after, a single drop of cold distilled water was
added to it. c. The same as h, with the exception that water
from the tap was used instead of distilled water, d. The
eprouvette was filled with boiled solution, as in h and c, but

2

340

ci o m u

M -t> Pj w J3

nothing was added to it. The glasses were then carefully

plugged with cotton wool. On Jan.
12, glass a was quite milky in ap-
pearance, and had a gelatinous scum
on the surface. It contained myriads
of bacteria and a few torula cells.
Glasses b and c were also turbid,
the former more than the latter.
The microscopical appearances were
the same in all. In d no change
could be detected either by the
naked eye or with the microscope.
Some of the liquid which remained
in a glass exposed to the air was
covered with tufts of penicil-
lium.

In this experiment, which was
confirmatory of the preceding, it is
worthy of note that the two waters
used to impregnate the test solution
the most decided effects were produced by the distilled water.

XVI. — January 17. — Pasteur's solution was made with
ordinary distilled water. A sufficient quantity was then
boiled, and immediately distributed in four eprouvettes, all
of which had been heated immediately before to a tempera-
ture of 200° C, the quantity of liquid in each being about
equal. The conditions of experiment were as follows : — 1.
The liquid was allowed to cool, and then three drops of
freshly distilled water were added to it with the aid of a
small pipette which had just been heated in the flame of a
Bunsen's burner. This water was collected in a superheated
glass from the glass distilling apparatus, which had been
previously thoroughly steamed out. 2. The same, excepting
that three drops of ordinary distilled water were used. S.
Three drops of water from the tap were added. To a fourth
eprouvette no addition was made.

On January 24 the liquids in 1 and 4 were perfectly limpid,
and showed no trace of organic forms ; 2 and 3 were milky,
especially the former. Both contained bacteria. Up to the
preseiit time the eprouvettes 1 and 4 remain perfectly barren.

This experiment shows that if due precautions are taken,
distilled water may be obtained in such complete purity that
it is free from germinal particles, whether of microzymes or
fungi ; and that the zymotic power (if I may be permitted to
use lliis term to express the faculty to determine the develop-
ment of organic forms in a test solution to which it is added)

341

of ordinary distilled water is acquired after distillation, either
by mixture with extremely small quantities of other waters,
or by contact Avith the surface of the vessels in which it is
contained. It was also evident that between waters of dif-
ferent kinds and of different sources there are corresponding
differences in the degree of zymotic effect they produce,
whence it seemed probable that a practical method of
judging of the amount of zymotic impurity contained in
any two waters might be founded on the comparison of the
degree of opalescence produced by each in the same time
and at the same temperature. In how far this surmise was
justified may he judged of by the results of experiments to
be hereafter referred to, relating to the zymotic powers of
the waters supplied to the metropolis.

If the apparently inevitable contamination of originally
pure water, when kept, is due not merely to admixture with
other water, but also to contact with surfaces impregnated
with living matter, it becomes of interest to inquire by what
conditions the action of such surfaces is limited or determined.
In the course of one of the observations already related it
was observed that a boiled liquid contained in a superheated
test glass, which had long remained perfectly linipid, and
entirely free from organic forms, became turbid after a pipette
employed in order to procure a specimen for microscopical
examination had been dipped in it ; and that the time which
intervened corresponded with that which usually elapses after
impregnation before the effect manifests itself. This occur-
rence suggested the following experiments, which were under-
taken in order to ascertain how far it is necessary that a
surface should be moist in order to its acting zymotically.

XVII. — January 30. — A glass rod was chaiged with
bacteria by dipping it into a solution on the surface of which
there was a viscous scum, consisting entirely of these bodies
imbedded in gelatinous matrix. The rod was allowed to dry
in the air for a few days ; it was then introduced into boiled
test solution contained in a superheated glass. On February
6 the liquid was already milky and teemed wdth microzymes.
On the same day a portion of the same scum was introduced
into a test glass and dried with a gentle heat. The glass
was then filled with test solution which had just before been
boiled and cooled in the usual way. The result was the same
as in the previous experiment.

In these instances it may be readily understood that the
drying was very imperfect. To determine the effect of more
complete desiccation, an ei)rouvette containing one cubic
centimetre of cold water previously ascertained to be zymotic,

312

was evaporated to dryness in the incubator and kept for sorne
days at a temperature of 40° C. On February 20 the dried
glass was charged with boiled and cooled solution, and
plugged with cotton wool in the usual way. The liquid
Avas examined microscopically on March 2, wlien it contained
numerous torula cells, but no trace of microzyrnes. It there-
fore appeared that the germinal particles of microzyrnes are
rendered inactive by thorough drying without the a])plication
of heat. As, however, it could not be concluded therefrom
that drying acted in a similar manner on the microzymes
themselves, an experiment was made on this point also.

XVIII. — March 4. — As it appeared probable that in the
previous experiments with bacteria scum, desiccation might
be prevented by the gelatinous matrix, a portion of the same
scum was thoroughly washed with water, collected in an
eprouvette, and dried for some days in the incubator. The
eprouvette was then (March 4) charged Avith boiled and
cooled Pasteur's solution, and plugged with cotton wool.
On March 11, the liquid was slightly hazy, but on micro-
scopical examination was found to contain no trace of
microzymes. The haziness was due to the presence of torula
cells in great numbers. On the 18th the appearances were
similar, but mycelium now existed in addition to torula. It
thus appeared that fully formed bacteria are deprived of their
power of further development by thorough desiccation ; so
that we may conclude that the contamination of water by
apparently dry surfaces happens only in those cases in lohich
desiccation is incomplete. It Avill be seen that this conclusion
is quite consistent with the previous observations.

Method of testing the zymotic property of water. — As a test
of the faculty possessed by all water which is not absolutely
pure, of determining the growth of microzymes, Pasteur's
solution gives results so constant and satisfactory that it
appears scarcely necessary to seek for better, although there
is no doubt that many other liquids would answer the same
purpose, and that some would react with greater delicacy.
The method consists, as already indicated, in the addition
of a small quantity of the suspected water to a relatively large
volume of the solution. As it is very desirable that the con-
ditions of experiment should be subject to as little variation
as possible, in our test experiments we add one drop of water
to a centimetre of solution, ahvays using the same drop])ing
pijjette. As the eprouvettes commonly employed contain
live centimetres half full, this quantity is preferred,

so that in the following ])aragraphs the term "charged
ejuouvette" is un(ferstoo(l to uuniu an eprouvette which has

313

been first superheated and then tilled to five cubic centi-
metres with boiling solution. After each testing the pipette
is immersed for several minutes in boiling distilled water.
In six days after impregnation with any zymotic water such
an eprouvette becomes hazy. It need scarcely be added that
in each experiment a second charged eprouvette must be
placed beside the impregnated one for comparison. Both
must be protected from the air by plugs of cotton wool.

From the most careful and repeated examinations of waters
known to be zymotic, Ave have learnt that such waters often
contain no elements or particles whatever which can be
detected by the microscope ; so that it may be concluded
that the elements of which the germinal substance of micro-
zymes consists are of extreme minuteness. It therefore
appeared to be of great importance to extend our inquiries
to water which is optically pure, not merely in the sense that
it contains nothing which can be detected by the microscope,
but in the much higher sense that when viewed in the electric
beam by the method employed by Professor Tyndall it shows
no haze. Unfortunately it is not as yet possible to procure
such Avater. Professor Tyndall has, however, been good
enough to give us the opportunity of testing specimens
obtained by the fusion of ice which approach the standard
of optical purity so nearly that the electric beam in passing-
through them displays a blue colour. Of the results of our
examination of these specimens it is sufficient to state that
they are as zymotic as many other varieties of water which
in the beam are seen to be full of light-scattering particles.

To determine in how far the zymotic properties of water
are affected by chemical compounds which are believed to
have the power of arresting the evolution of living forms in
organic liquids, a series of experiments were made in which
the zymotic power of water was tested before and after the
addition of such compounds, the supposed anti-zymotic being
contained sometimes in the water to be tested, sometimes in
the test solution.

XIX. — March 2-10. — 1. A quantity of water previously
ascertained to be zymotic was ozonised by subjecting it to
the action of air which has passed over fresh and moist phos-
phorus, the apparatus for this purpose consisting of (a) two
Woulff's bottles containing sticks of phosphorus ; (b) a wash-
ing bottle containing solution of caustic potash; (c) a flask
containing the Avater to be ozonised. Air Avas made to pass
slowly through a, b, c in succession, by means of an aspirator,
for several hour, after Avhich the liquid in c, and the air in
contact with it, reacted strongly on iodide of potassium and

314

starch paper. Charged eprouvettes were then (March 2)
prepared in the usual way, to each of which a few drops of
the ozonised water was added. On March 21 no organic
forms whatever coukl ho discovered in the Hquid. The phigs
Avere then removed. On the 27th the first tufts of penicil-
lium appeared, Avhich have increased up to the present time.
There are no microzymes in the liquid.

2. Water known to be zymotic was treated with Condy's
liquid in quantity sufficient to colour it slightly, A few drops
were then (March 2) added to a charged eprouvette. Up to
the present time the liquid remains free from microzymes,
but contains torula cells. On the same day a second charged
eprouvette was treated with two drops of undiluted Condy's
liquid and plugged. It remained absolutely barren till March
21, when the plug was removed. In a few days torula cells
appeared, and on the 27th there were tufls of penicillium.

3. A charged eprouvette was impregnated (March 2) with
ordinary distilled water containing O'l per cent, of carbolic
acid. At the end of a week the liquid was hazy and teemed
with bacteria and torula cells. Ultimately penicillium tufts
appeared on the surface. On March 8 the experiment was
repeated with water containing 0*5 per cent, of carbolic acid.
It remains up to the present moment free from microzymes,
but contains torula cells and mycelium.

4. A charged eprouvette was impregnated with ordinary
distilled water (March 2) containing 0*1 per cent, of sul-
phate of quinia, and plugged. At the end of a week it was
opalescent and full of bacteria ; it also contained torula cells
and mycelium. On March 8 the experiment was repeated
with water containing 0*5 per cent, of the salt. At the
end of a week it was hazy, but on microscopical examination
this was found to be due exclusively to torula cells.

5. March 11. — A charged eprouvette was impregnated
with distilled water containing 10 per cent, of the solution
of peroxide of hydrogen. The liquid remained free from
microzymes until March 21, when the plug was removed.
Tufts of penicillium had already appeared on the 27th.

6. March 11. — A charged eprouvette was impregnated with
distilled water containing five per cent, of liquor chlori. The
liquid remained barren until March 21, when the plug was
removed. It is now crusted with penicillium.

7. February 13 — A superheated eprouvette was charged
with some of the superheated Pasteur's solution which had
been prepared five months before. The liquid was then im-
pregnated with distilled water known to be zymotic. On
March 4 the liquid was examined and found to be entirely

345

barren. It was then treated with boiled solution of pure
sugar. As on March 21 the liquid was still entirely free
from organic forms it was sown with some fresh spores of
penicillium. Up to the present time there has been no
change,

XX. — February 23. — For the purpose of investigating the
zymotic property of any water, it may be conveniently col-
lected in a tube of the form and size shown in Fig. 1.
It is essential that it should be used in a state of absolute
purity. As "it is seldom possible to draw the tube at the
time that it is to be filled, it must, as a rule, be prepared
beforehand, in which case it is necessary to close it hermeti-
cally at both ends before it is removed from the flame. Such
a tube obviously contains calcined air, of which the tension
is very much less than that of the atmosphere ; consequently
it is very easily filled by breaking off its end under the sur-
face of the water of which a specimen is to be collected. If
the water is flowing from a tap or in a stream it must be
received in a boiled capsule, in the contents of which the
tube may be dipped.

Fifteen specimens of water supplied by the London com-
panies were collected in this way during February last, and
tested with reference to their zymotic power on the 23vd of
the same month. The results of the experiment are exhibited
in the following table, in which the number printed below
the designation of each specimen indicates the order in which
it would have stood if the tubes had been arranged in a linear
series according to the degree of turbidity which each mani-
fested on the ninth day after impregnation.

Letter designating
Water Company.

Water before
filtration (from sub-
sidence reservoir).

Water after filtration
(from pump well).

Water as distributed
(from main).

A

15

3

13

B

14

10

7

C

5

2

6

D

1

12

8

E

4

9

11

The specimens to which the numbers 15, 14, 13, 8, 7, 5,
correspond, became hazy as early as the fifth day. On the
ninth day it is noted that the eprouvettes to which the six
highest numbers correspond were milky, while those in whicli
the turbidity was least marked were merely hazy ; the rest

310

are described as being opalescent. All, therefore, acted zymo-
tically in different degrees.

All of the eprouvettes were plugged with cotton wool. As
in previous experiments, the quantity of water introduced
was in each case measured with the same pipette, which was
immersed in boiling distilled water for a few minutes between
each impregnation. A check eprouvette was then impreg-
nated in the same way as the rest with the water in which
the pipette was washed. It remains to the present moment
perfectly transparent and ban'en.

Excepting in so far as this experiment shows that filtra-
tion exercises no perceptible influence on the zymotic power
of water, no conclusion can be drawn from the comparison
of the results. It happens that the water designated C
stands considerably higher than the rest, and that designated
A considerably lower. It Avould be premature, however, to
attach importance to this fact.

Section III. — Circumstances which Determine the Existence
of Microzymes in Organic Liquids and Tissues.

The experiments to be related in the following paragraphs
Ts^ere undertaken for the purpose of ascertaining whether the
tissues and liquids of the living body participate in the
zymotic property Avhich has been shown to exist in ordinary
water and moist substances : in other words, whether the
living matter with which the body is in constant contact by
its external surface penetrates into its interior.^

XXI. — March 24. — A glass canula of suitable size, which
had just been drawn, was introduced into the carotid artery
of a rabbit, and secured with a ligature. The arterial blood
as it flowed from the canula was received into four ordinary
test glasses (marked a, a, and h, h), and into an eprouvette
(marked c). The quantities of blood collected in a, a were
mixed with boiled and cooled distilled water, and left freely
exposed under a bell jar. In two or three days bacteria
appeared and the liquid became offensive. The quantity in
h, b was left undiluted, and each glass was covered with a
layer of cotton wool. On the 30th they remained unaltered,

1 As to the existence of vmble microzymes in the liquids of persons
affected with contagious diseases, I iiad aU-eady satisfied myself that I could
rot accept Hailic/s observations ; for on examininj^ the blood of patients
affected with scarlatina (in which, according to Hallier's statement, micro-
coccus is constantly observed and very abundant) at all stages of the disease,
I had found that no such bodies existed in it. It does not, however, follow
from this that organisms are not present potentially, i.e. iu the form of
germinal particles not distinguishable by the microscope.

317

and contained no organic forms excepting those proper to the
liquids. They were then carefully mixed with hoiled dis-
tilled water by the aid of a freshly prepared pipette, and
again covered with cotton wool. [On April 3 the liquid was
entirely free from microzymes, and exhibited no sign of
decomposition.]^ The blood contained in the eprouvette was
allowed to coagulate, and yielded a clot and very limpid
serum. Up to March 30 it remained quite unaltered, and
on microscopical examination it was found to be quite free
from microzymes. On that day the serum was transferred
by means of a superheated pipette into another superheated
eprouvette, and diluted with boiled and cooled distilled water:
it w^as then placed in the incubator. To the clot distilled
water was added in quantity corresponding to that of the
serum which had been abstracted : it w^as placed in the incu-
bator. [When these preparations were examined on April 3,
the serum was still limpid and perfectly free from organic
forms, and the clot-preparation showed no change.]

Other experiments were made, consisting in impregnating
charged eprouvettes with drops of blood taken directly from
the finger, great care being taken in each case to cleanse the
surface of the skin where the puncture was made. In each
case the blood-corpuscles subsided to the bottom of the eprou-
vette, leaving a clear liquid in which no development of
microzymes took place, although they were kept under obser-
vation for several weeks.

We have no hesitation in attributing the development of
bacteria in the liquid in the test glasses marked a, a, to an
accidental contamination (e.^. to the falling into the glass of
a hair of the rabbit, or possibly a drop of saliva), and in con-
cluding that normal blood contains no microzymes potentially
or actually.

XXII. — February 24. — A guinea pig was killed, and,
immediately after,- the integument was stripped off the back.
Portions of the muscles and cellular tissue of the rump were
then rapidly cut out Avith scissors which had just been heated
in the flame of a Bunsen's lamp. The pieces were then
seized with the aid of glass hooks which had just been made
for the purpose, and transferred into charged eprouvettes
(marked d). Others were placed in superheated test glasses,
and covered with boiled and cooled distilled water, but by
accident one of them fell from the hook on to the table
(marked h). The skin was then stripped oiSf the thighs,
which were immediately separated from the body with the

^ The Report bears (late March 8]st, 1871. The passages in brackets
were added during the first week of k\\\\\.

348

same precautions as before, and hung up under a bell jar by
wires which had been heated and cooled. The liquid in the
eprouvette a was subjected to repeated examination until
March 3, when it was still perfectly limpid and entirely free
from organic forms. A single drop of common distilled water
was then added to it. In a few days it became milky and
acquired a putrid smell. In the glass h, there were already
signs of bacteria on March 2, and the liquid soon became
offensive. The thighs which were hung up, shortly became
covered with a crust of penicillium. One of them was ex-
amined on March 9. On removing the crust and cutting
into the muscle it was found to be less moist, but otherwise
of natural appearance. There was a musty but no putre-
factive smell. The cut surface was neutral to test paper.
The other thigh was examined March 27, and was in a
similar condition, excepting that the muscular substance was
drier and of darker colour.

XXI II. — February 1. — Five centimetres of urine were
introduced into an eprouvette, which was then plugged with
cotton wool and placed under a glass shade. It retained its
acid reaction and limpidity till February 9, when a drop or
two of ordinary cold distilled water was introduced from a
fine capillary pipette prepared just before. On the 16th the
liquid was hazy and crowded with bacteria. In the course
of a few days more, a sediment subsided to the bottom of
the eprouvette, and the liquid became alkaline and ammoni-
acal. This experiment was subsequently repeated with similar
results.

It has been long known that the tendency of urine to
undergo decomposition may be obviated by protecting it
against contamination from without. The preceding experi-
ments show that here, as elsewhere, water is the contami-
nating agent.

XXIV. — January 2. — An abundant flow of saliva having
been determined by introducing a few drops of ether into
the mouth, one or two drops were allowed to fall into a
charged eprouvette. The liquid was repeatedly examined
during the next three weeks, but no microzymes could be
detected. The salivary secretion, as it is discharged from
the salivary ducts, is no doubt inactive ; but inasmuch as the
mixed liquid with which the mucous membrane is moistened
is exposed to several sources of contamination, and moreover
can be often shown to contain leptothrix filaments, it would
not have been surprising if the result of the experiment had
been otherwise.

XXV. — It is scarcely possible to obtain milk in a state of

349

purity, for the liquid as it issues is exposed to contamination
both from the hands of the milker and from the surface of
the test itself. It is not therefore surprising that the results
of our experiments with this secretion were not uniform.
Their variations, however, exhibit so complete a correspond-
ence with the varying conditions of the experiments, that
they are scarcely less confirmatory of the general conclusions
we have arrived at than if they had been positive.

February 28. — Milk was received directly from the cow
into two flasks (marked a and b) which had been previously
superheated. The flasks were immediately plugged with
cotton, wool. Another specimen of milk "as delivered to
customers," was brought from the dairy at the same time
in a clean bottle which had not been superheated. All the
specimens were alkaline. On March 4 it was found that the
milk in the bottle was slightly acid and crowded with bacteria.
On the 9th it was curdled and smelt offensively. The flask
a was also acid on March 4, and contained a few groups of
bacteria. In the flask b the acid reaction Avas scarcely appre-
ciable, and no bacteria could be discovered in it. On the 9th
the contrast between a and b was still very striking, the
liquid in a having separated into Avhey and curd, while b
remained apparently homogeneous. Charged eprouvettes
were then impregnated with drops of the liquid in b in
which no bacteria could be detected. After a few days
bacteria appeared in the test liquid, and in the liquid which
still remained in the flask.

The difference between a and b was of course accidental,
for both were exposed to equal chances of impregnation.

XXVI. — Februarv 21. — It has been already stated that
superheated tubes containing egg albumen which had been
kept from August, 1870 to March, 1871, were found abso-
lutely free from organisms, and to all appearance unaltered.
The liquid contained in one of these tubes which was per-
fectly limpid was emptied into a superheated eprouvette and
impregnated with two drops of cold distilled water. On
March 2 the liquid had acquired a yellowish green tint, a
scum had formed on its surface, and the liquid was full of
separate bacteria.

XXVII. — March 20.— P us was collected from a deep-
seated abscess in the thigh of a child by introducing the
capillary end of a collecting tube into the path of the
bistoury which had been used for opening it, the bistoury
having been itself immersed in boiling water. It was then
transferred to a small eprouvette and exposed to the air. On
March 30 there were no bacteria. It was then diluted with

350

boiled and cooled distilled water, [[t was again examined
on April o, when it contained no organic forms whatever.]

February 7.— A pyaemic abscess of the elbow joint was
opened ; a full stream of pus issued from the incision. Several
large but still capillary tubes were then filled by inserting
their open ends into the track of the bistoury. The tubes
were immediately sealed, and the contents used the same day
to impregnate a charged eprouvette. After a few days the
test liquid was teeming with bacteria.^ In this case the
knife was not previously immersed in boiling water, but the
discharge of pus from the Avound Avas so copious that I do
not think there is the slightest doubt that the quantity used
was collected without any contamination, whether arising
from this source or from the surface of the skin.

XXVIII. — The collection of blister fluid is attended with
inuch greater difficulties than that of pus, for it is almost
impossible to abstract it from the vesicles without risk of
contact with the surface of the skin. It can be best obtained
by opening both ends of a collecting tube, and then intro-
ducing the capillary end into a vesicle after first snipping the
epidermis. This done, the liquid must be drawn into the
tube by suction. Liquid thus collected was used as
follows : —

January 10. — Blister fluid was added in the usual propor-
tion (one drop to one cubic centimetre) to a charged eprou-
vette. For a long time the liquid remained clear, but even-
tually bacteria appeared in small numbers.

February 13. — Blister fluid from another source was used
in a similar manner and with a similar result.

March 27. — The same experiment was repeated with dif-
ferent fluid, but in this case the eprouvette was kept in the
incubator. The development of bacteria was much more
rapid. On the same day another quantity of the same liquid
was diluted with boiled and cooled distilled water in a super-
heated eprouvette and also placed in the incubator. [In a
few days it became turbid and swarmed Avith bacteria.]

The equivocal results of these experiments are to be
attributed eiitirely to the difficulty of obtaining blister fluid
pure, that is, to accidental contamination in the process of
collection.

^ This important experiment could not be repealed, for an attendant who
entered the laboratory in my absence carelessly destroyed all the tubes
exceptin£( the one which had been already used. Tlie single result was so
satisfactory that I myself entertained no doubt of its significance.

351

From the consideration of a number of facts Avhich pre-
sented themselves in the course of the experiments related in
the previous pages, it has appeared certain that there is. no
developmental connection between microzymes and torula
cells, and that their apparent " association is one of mere
juxtaposition. The grounds of this conclusion may be
shortly stated thus : — \ ^

1. The prompt appearance of torula cells in Pasteur's
solution whenever it is exposed to the air, and the rapid
development and luxuriant fructification of the higher form
(penicillium), sHow that so far as the chemical composition
of the liquid is concerned, there exist in it all the conditions
favourable to the process.

2. Our experi«ients prove that when precautions are taken
to prevent contamination by impure surfaces or liquids, the
development which ends in penicillium goes on from first to
last without the appearance of microzymes.

3. Whenever it is possible to impregnate the test liquid
with microzymes without at the same time introducing torula
cells or germs, the development of the former begins and con-
tinues by itself without any transformation into the latter.

Thus fungi are not developed, notwithstanding the presence
of microzymes in the same liquid in which, microzymes being
absent, but air having access, they appear with the greatest
readiness.

This being the case we are enabled to eliminate the ques-
tion of the quasi spontaneous evolution of fungi altogether
in the present discussion, as lying beyond the limits of our
inquiry. It can hardly, however, be considered out of place
to state to the reader some of the results to which our obser-
vations have led us with reference to this question, especially
considering that however improbable it may seem to ourselves
that fungi have any important relation with the processes of
disease, there are others who are of a difierent opinion.

To determine the forms in which germs of fungi exist in
the air, the best method is that long ago used by Pouchet —
that of projecting a jet of air on a glass plate moistened with
glycerine or syrup. A few experiments were made, but the
results were mostly negative, for in London the particles of
soot and refuse fragments which are collected by this method
are so numerous that organised particles, even if present,
could scarcely be distinguished. We find it a much more
successful plan simply to expose a glass surface covered
with glycerine to the air. In examining such a surface it
was always possible to discover a certain number of cells

352

which resembled torula cells, and occasionally penicillium
acrospores.

From this result we do not, however, conclude that it is
by these forms that the cosmopolitan fungus (as Hallier calls
it) is usually propagated ; it frequently happens that liquids
which have been once exposed, although they contain no
visible cells whatever, rapidly germinate without further ex-
posure. We are also certain that although air is the main
source of what we may venture to call fungus impregnation,
as distinguished from impregnation with microzymes, yet the
tAvo acts may take place at the same moment — germs of
torula being often contained in the same liquid media as the
germ particles of microzymes. That this is so is proved by
instances already referred to, in which liquids protected from
air filled with torula cells. Here we relinquish this question,
although in a biological point of view it is of the greatest
interest and importance.