peared below the occiput, which communicated with the intravenous
abscess ; for when this swelling was pressed upon, pus oozed from the
lower part of the sinus groove. The abscess in the bulb must have
perforated and burrowed under the deep nniscles of the neck, until
500 MOUNT SINAI HOSPITAL REPORTS.
it appeared in the upper part of the posterior cervical triangle,
Macewen gives another explanation of these deep abscesses. He
states that they owe their origin to phlebitis of the condyloid emissary
veins. Whatever the cause was, recovery could not ensue before the
second abscess, which was still deeper and which was evidently the
original periphlebitic abscess, had been opened and drained.
There are a number of unusual symptoms connected with the case,
of which occasional mention is made in literature. Wryneck is some-
times found in cases of thrombosis of the jugular, and the explanation
given by Koerner and others is that, the movements of the head to-
ward the other side being painful, the head is instinctively held in
wryneck position to ward olf pain. It is hard to understand why this
symptom should occur only in a comparatively small number of
these cases. At different times in the history of our case we found
for a few days difficulty in swallowing. After the last operation the
symptoms of disturbed sensation of taste and motility of the tongue
made their appearance, proving paralysis of the glossopharyngeus
and hypoglossal nerves on the left side. It is not impossible that this
may have been due to injury, although I feel pretty sure that such
was not the case. Considering that the function of all the three
nerves, the ninth, tenth, and the eleventh, leaving the skull through
the foramen lacerum in close proximity to the jugular bulb, was dis-
turbed, and that the twelfth, leaving through the condyloid foramen,
was paralyzed, it may not be too far-fetched to assume that the peri-
phlebitic abscess was responsible for it, just in the same way as we
sometimes see facial paralysis in cases of mastoiditis.
[Contributions from, the Pathological Department —
F. 8. Mandlehaum, M.D., Pathologist.]
ON CERTAIN FEATURES OF THE GROWTH OF BACTERIA
ON MEDIA CONTAINING SUGARS AND SERUM ; WITH
REMARKS UPON THE ACID PRODUCTION.^
By E. Libman, M.D.,
ASSISTANT PATHOLOGIST TO THE MOUNT SINAI HOSPITAL, NEW YORK.
In the Neiv York Medical Record for May 19, 1900, and in the
Ccntralhlatt fiir Bakteriologie, vol. xxviii., Nos. 10 and 11, I published
a short communication based upon the observation of certain features
of the growth of bacteria on media containing sugars and serum. The
main fact developed at that time was that many bacteria can throw
down a proteid precipitate from human blood serum in the presence
of glucose, this being due to the acid production. Many bacteria
were also found to precipitate ordinary Qg^ albumen. The amount
of glucose used varied from 0.1 per cent, to 0.5 per cent., and pre-
cipitation could be observed with other sugars. On solid media this
precipitation took the form of a diffuse whitening of the media. The
saprophytic bacteria examined at that time did not possess the
property.
In the following paper I shall use the terms "whitening" and
"precipitation" interchangeably, although it will be noted later that
the whitening of solid media sometimes occurs without there being
necessarily precipitation in the corresponding fluid media, and vice
versa.
Since the publication of the preliminary notices I have found a
number of other interesting and important facts bearing upon this
question. On the present occasion I can only hope to give an abstract
of some of my investigations. I shall discuss the question under the
following heads :
1. The Avhitening of agar media and the precipitation in bouillon
media containing sugars and serum.
'Reprinted from the Journal of Medical Research, vol. viii.. No. 1, 1902.
502 MOUNT SINAI HOSPITAL REPORTS.
2. The explanation of the phenomena.
3. Observations on gelatin media.
4. Observations on media containing blood; and some special
phenomena which were occasionally seen.
5. Reactions at which precipitation can occur.
6. The use of sugar-serum media ; and, finally,
7. Some general observations regarding the acid production.
I. THE WHITENING OP AGAR MEDIA AND THE PRECIPITATION IN BOUILLON
MEDIA CONTAINING SUGARS AND SERUM.
To obtain a general idea of the precipitating powers possessed by
various bacteria when grown on various sugar media, the following
bacteria were tested on a large number of media; saccharomyces albus,
bacterium megatherium, bacillus acidi lactici, bacillus prodigiosus,
micrococcus candicans, staphylococcus aureus and albus, bacillus coli
communis, meningococcus,^ micrococcus gonorrhefe, bacillus mesen-
tericus fuscus, bacillus pestis (obtained from Dr. Park), strepto-
coccus, bacillus proteus vulgaris, bacillus diphtherias, bacillus an-
tliracis, bacillus mallei, pneumocbceus, vibrio choleras, and a hog-
cholera bacillus (obtained from Dr. Lartigau). Of some of these,
only one variety was tested ; of others, a large number.
The media which we used were made up of agar to which various
sugars and sera were added, and also of bouillon media prepared in
the same way, and of glucose-agar to which a solution of egg albumen
had been added. The sugar percentages for the glucose media varied
from 0.1 per cent, to 2.0 ; the other sugars, that is, lactose, maltose,
and saccharose, were tested in strengths of 0.5 per cent, and 2.0 per
cent. The 0.1 per cent, glucose was used as a matter of interest, be-
cause that is the amount generally considered to be present in the
normal blood. The 0.5 per cent, was used because, as Theobald Smith
has shown, it represents the most favorable percentage for the growth
cf most bacteria, and because it is the amount which is often present
in the blood in diabetics. The higher percentages of the sugars gave
the larger number of positive results.
There were used human sera (ascitic and hydrocele), cow's and
sheep's sera. Although results can be obtained by adding a very
small amount of serum, we, as a rule, added one-sixth to one-quarter.
^All the meningococcus cultures corresponded accurately to Weichselbaum's
description.
I.IBMAN : GROWTH OF BAC^TERIA. 503
The tubes were all observed live clays: as a rule, if whitening or pre-
cipitation did not occur within two days it did not appear. In some,
results could be noticed within a few hours. The tubes were kept at
37° C. Whitening can also frequently be observed at room tempera-
ture. No notes were made of the variable intensity of the reaction,
but slight results were noted as such.
In the tests, the results of which are soon to 'be stated, ascites serum
was used unless otherwise stated. When a negative result was ob-
tained the experiment was repeated with hydrocele serum or with
one-quarter ascitic fluid if one-sixth had been used. Cow's and
sheep's sera often give a precipitate when human blood serum does
not. Our analyses showed that such sera contained sugar in larger
amount than the human sera we were using. They were also richer
in albumin, and it was noted that precipitates could be obtained more
readily the more albuminous a serum was. The egg-albumen medium
was made by adding one-quarter of a 25 per cent, solution of egg
albumen in water to 2 per cent, glucose-agar. As a rule Witte's
peptone was used in making np the media ; when slight results were
obtained in the way of whitening, the experiments were repeated,
using media made np with Merck's peptone. The results, however,
remained the same. This precaution was taken because whitening
can sometimes be seen on ordinary sugar-agar media made up with
AVitte's peptone— a subject to be discussed in a later paper. It was
noted that some bacteria, whose colonics on agar do not whiten the
medium, may whiten the medium when they grow in the condensation
fluid. No note is specially made of this variation in the phenomenon.
Although all the experiments were made with great care and in
duplicate, I do not claim that the results show exactly what all or
any of the bacteria will always do in these media, as a bacterium
which fails to precipitate in the presence of a certain sugar may very
v^ell be found to precipitate when a more active variety of the same
organism is tested, provided the organism can split up that sugar.
I believe, however, that I can draw some general inferences as to the
occurrence of the precipitation, and as to its variation according to
the amount and vai'iety of sugars present.
The cultures used were for the most part obtained in routine hos-
pital investigations. I am indebted to Dr. Hiss, of the (.'ollege of
Physicians and Surgeons, for many cultures, and to Drs. Lartigau
and Park for numerous cultures an<l !iiaii\' other favors.
504 MOUNT SINAI HOSPITAI; REPORTS.
Inferences from Experiments.
Nearly all of the bacteria which were tested whitened the 2 per
cent, glucose-agar medium, the exceptions being the bacillus mesen-
tericus fuscus, the micrococcus candicans, and the bacillus mallei.
Anaerobically most of the bacteria also whitened the 2 per cent,
glucose medium. In the egg-albumen medium precipitation was
found to occur with all the bacteria that precipitate in the 2.0 per
cent, glucose-serum-agar, except the bacterium megatherium, the
meningococcus, two varieties of bacillus proteus vulgaris, the bacillus
anthracis, and the vibrio cholerae. As a rule the growth was less
active on this medium than on the corresponding serum medium, but
often more active than on ordinary glucose-agar. In the bouillon
media it is often difficult to tell whether the precipitate is present or
not, if it be slight in amount.
To convince one's self that a precipitate has been formed, it is
wise to add an alkali and see if w'hat looks like a precipitate can be
dissolved out. Official potassium hydrate does not dissolve the bacteria
and does dissolve the precipitate. It was therefore generally used
for this purpose, so as to avoid the dilution incident to adding weak
alkalies. It was found that many bacteria can throw down a decided
precipitate in the bouillon media, and that positive results were much
more frequently present with cow's serum than with human serum.
The results on agar media containing 1.0 per cent, of glucose were
almost the same as those for the 2.0 per cent, medium, except that
the following bacteria failed to whiten this medium : meningococcus,
bacillus pyocyaneus, bacillus pneumoniae, and the bacillus anthracis.
On the 0.5 per cent, glucose-agar medium, precipitation was to be
seen with saccharomyces albus, bacillus prodigiosus, micrococcus can-
dicans, staphylococcus aureus, meningococcus, gonococcus, bacillus
pneumoniae, bacillus mesentericus fuscus, bacillus proteus vulgaris,
bacillus anthracis, and the bacillus coli communis. On the 0.1 per
cent, glucose-agar medium positive results were obtained only with
streptococci, the plague bacillus, some varieties of typhoid bacilli,
and with the vibrio choleras. Many of the bacteria that precipitate on
glucose media did not precipitate in the presence of other sugars.
Some that whitened lactose, saccharose, and maltose media of 2.0 per
cent, strength did not whiten those of 0.5 per cent, strength. No
positive results were obtained on the 0.5 per cent, sugar media when
they were absent on the 2.0 per cent, media. Bacteria which are
IJBMAN: GROWTH OF BACTERIA. 505
known not to make acids from certain sugars were found not to pre-
cipitate on media which contained only those sugars. As a rule,
precipitation did not occur in the bouillon media when not present
in the agar media. The agar media were found to be much more
delicate for determining the absence or presence of the phenomena
than the bouillon media.
Besides the systematic experiments from which the previous con-
clusions have been drawn, a very large number of other experiments
have been made, mostly with the glucose media, and from the com-
bined observations some general remarks may be made.
Some bacteria did not precipitate in the presence of any of the
sugars or strengths of sugars used ; such are the bacillus mesentericus
fuscus, the micrococcus candicans, and the gonococcus. Staphylo-
cocci were found not to precipitate on the 0.1 per cent, glucose me-
dium. The meningococcus never whitened the 0.5 per cent, glucose
medium, occasionally whitened the 2.0 per cent, slightly. The bacillus
pyocyaneus did not whiten 0.5 per cent, glucose media and grew
white on the 2.0 per cent, glucose media made up with serum or egg
albumen and whitened the medium. It is interesting to note that the
plague bacillus can precipitate on all the glucose media and on the
2.0 per cent, maltose serum-agar. Streptococci were nearly always
found to precipitate on all the strengths of glucose used, but often
did not affect saccharose media. The streptococcus colonies usually
precipitated readily ; the only exception was an attenuated streptococ-
cus isolated from the blood intra vitam from a case of erythema nodo-
sum with joint manifestations, in Dr. Brill's service. This whitened
the media only by means of the condensation fluid. Typhoid bacilli
never affected the lactose media, but ahvays precipitated in the pres-
ence of 2.0 per cent, glucose. The bacillus proteus vulgaris always
precipitated in the presence of 2.0 per cent, glucose. The bacillus of
glanders (two varieties) affected none of the media. The bacillus
diphtheria (five varieties), the bacillus of anthrax (three varieties)
did not affect the 0.1 per cent, glucose medium. The bacillus coli
communis also precipitated in the presence of 0.5 per cent., 1.0 per
cent., and 2.0 per cent glucose, but varied in the other media. Pneu-
mococci do not generally precipitate the 0.5 per cent, media, ex-
cept one variety obtained from Dr. Park, which had been carried
through a large number of animals. The pathogenic bacteria more
frequently precipitate on the media containing low sugar percentages
than the saprophytes.
506 MOUNT SINAI HOSPITAL REPORTS.
II. EXPLANATION OF THE PHENOMENA.
Tests were made which proved that the precipitation is independent
of the variety of peptone used. The results were identical in media
made up with meat infusion and meat extract. Precipitation oc-
curred even if no salt were added to the media, although less fre-
quently. Whitening could be produced on media made vip of agar
plus glucose and egg albumen or serum ; in fluid media precipitation
could be obtained when only serum and glucose were present. Re-
sults could not be obtained in tubes containing only egg albumen and
glucose, as there was no growth. These experiments made it appear
probable that the precipitation was due to the action of acids split off
from the sugars, in the proteid present. This theory was amply con-
firmed by the following facts :
1. Precipitation in a bouillon medium or whitening in an agar
medium was never found to occur without acid production.
2. If the organism precipitated in the presence of a high percent-
age of sugar and not a lower, it was found to make less acid in the
latter. The presence of peptonizing power seems to interfere with
precipitation, at times, in the media containing smaller sugar per-
centages. This is made especially probable by some observations on
the symbiosis of certain streptococci and the bacillus mesentericus
fuscus. For, if, in a tube which is being whitened by a streptococcus,
the bacillus is inoculated before the tube becomes too acid to allow the
bacillus to grow, the precipitate formed can be dissolved. This is not
due, as might be suspected, to alkali production, as our figures show
that the bacillus makes acid, although in small amounts, on glucose
media.-
3. When bacteria whitened media aerobically and not anaerobically,
they were found to produce less acid anaerobically.
4. When cultures in sugar bouillons of various strengths were
added to diluted sera or white of egg, a precipitate soluble in alkalies
was often obtained. By adding alkalies to such cultures they could
no longer precipitate sera when the acidity had been reduced to a
certain figure. This figure was reached sooner when stronger solu-
tions of alkalies were used. This was not due to dilution, as the cul-
tures could be diluted so as to reach a very low acidity and still pos-
sess the power of precipitating. The point at which such cultures no
"Peptonization and precipitation can, however, be observed together.
LIBMAN : GROWTH OF BACTERIA. 507
longei' would precipitate seemed to be about the same for all the bac-
teria when added to a given serum. Filtrates of cultures gave the same
results. Boiled cultures or boiled filtrates did likewise, although the
results were less marked. Recently an ascitic serum was used in which
a coagulum could be produced by the filtrates of cultures in glucose-
bouillon. The filtrates of the growths of the same organisms in or-
dinary bouillon failed to produce a similar coagulum.
This serum was obtained from a case of multiple venous thrombosis
apparently of infectious origin. (Service of Dr. Manges.)^
5. Occasionally a filtrate of a certain acidity, which would not
produce a precipitate in a certain serum, could be made to do so by
adding a small amount of one of the dilute mineral acids. The added
amount of acid was too small to have itself produced any precipitate.
Chemical examinations made by Dr. S. Bookman (of the Patho-
logical Institute of the New York State Hospital) of a precipitate
produced by adding the filtrates of cultures of streptococci in sugar-
bouillon to cow serum or egg albumen showed that the precipitate
was a proteid and belonged to the group of the acid albumins.
Dr. Bookman in a personal communication, and later Dr. Levien
at the May meeting of the Pathological Society, suggested that the
precipitation might have been due to the presence of nucleic acid in the
cultures acting in the presence of an acid medium. I append here a
part of Dr. Bookman 's report :
"The precipitates produced from cow serum (diluted with three
parts water) and from 25 per cent, solution of egg albumen (white
of egg) seem to be identical.
"The precipitate produced from the egg albumen contains 14.156
per cent, nitrogen ; that from cow serum 14.168 per cent, nitrogen.
"The phosphorus contents were exceedingly low, being about 0.1
per cent., which precluded the precipitate from being a phosphorus-
rich nuclein or nucleic or paranucleic acid, the phosphorus contents
of which vary from 9.0 per cent, to 11.0 per cent. The dried precipi-
tate is soluble in centinormal or 0.04 per cent, sodium hydrate. It
gives a weak but distinct biuret reaction, and other properties point
to classing it among the group of acid albumins. On account of the
small yield further researches have been impossbile. I hope, however,
'Since writing the above I have had a number of samples of serum in
which, when added to glucose-bouillon, coagula could be produced by the
growth of organisms. They could also be produced by adding filtrates (sugar-
bouillon cultures) to the pure sera.
508 MOUNT SINAI HOSPITAL REPORTS.
in the near future to give its total percentage composition as well as
description of its other properties."
It was attempted during part of this work to obtain precipitates
free from bacteria, and in the course of these endeavors certain varie-
ties of superimposed media were used, to one of which I wish to cal;
your attention, simply because of its interest. If, over a stratum of
agar to which egg albumen or serum has been added, a layer of or-
dinary glucose-agar is added and a stick or surface inoculation made
on the latter, the lower layer can often be whitened, leaving the upper
layer clear. The point of solubility of the whitening in the tubes I
have not absolutely established, but, as a rule, one-quarter decinormal
sodium hydrate will suffice to dissolve them out.
III. GELATIN MEDIA.
These experiments were carried out on a small scale with a 2.0 per
cent, glucose-gelatin to which one-quarter of serum had been added.
A number of facts were developed which can only be summarized
now. Whitening was produced by a number of bacteria ; many bac-
teria grew more rapidly and reached a final more voluminous growth
on this medium than on gelatin or 2.0 per cent, glucose-gelatin. Some
bacteria that liquefied ordinary gelatin would not liquefy this me-
dium, but the same is also true of glucose-gelatin media. The pig-
ments produced by the bacteria vary as to their occurrence, some being
absent on this medium and present on ordinary gelatin, and vice
versa.
IV. BLOOD MEDIA.
On plates made with glucose-serum-agar for purposes of bacterio-
logical examination of the blood, a few interesting observations were
made. Around many of the colonies there was found a dark halo.
It was made evident by using plates made with less blood that the
dark area was often due to precipitation only, although in some flasks
of glucose-bouillon to which blood had been added there often
ensued a diffuse darkening of the medium, which would go to show
that there had been a change in the hemoglobin, due probably to the
acid production. The clear spaces developed around some colonies
(e.g., staphylococci) on the blood plates have been noted by other
observers and need not detain us here.
IJBMAN : GROWTH OP BxVCTERIA. 509
V. REACTIONS AT WHICH PRECIPITATION CAN OCCUR.
The limits of the acidity and alkalinity at which precipitation can
occur were not absolutely determined, but on agar media precipita-
tion could be obtained on media the titre of which varied from 1.8
per cent, acid (1.8 c.c. of decinormal sodium hydrate necessary to
neutralize 10 c.c. of the medium) to 1.0 per cent, alkaline (on adding
the serum such media became 1.6 per cent, acid to 0.4 per cent, alka-
line). In bouillon media precipitation was found in media ranging
from 2.6 per cent, acid to 2.0 per cent, alkaline (on' adding the sera
such media were 2.3 per cent, acid to 0.3 per cent, alkaline). These
titrations were made according to the formula of the American Pub-
lic Health Association. If they were made according to the more com-
mon method of titrating for determining acid production (i.e., add-
ing alkali until the first change in color occurs) the acidity would
amount to about 0.3 per cent, to 0.5 per cent. less.
A point of decided interest noted was that whereas the growth was
greatest in bouillon media of which the titre was 1.0 per cent, acid
when no sugar was used, it was best in a medium 0.2 per cent, alkaline
when sugar was present.
VI. USE OF STJGAR-SERUM MEDIA.
On these media, particularly on the one containing 0.5 per cent, glu-
cose, the growth of most 'bacteria is much more marked and more
rapid than on ordinary serum-agar. This is particularly true of the
streptococcus, pneumococcus, gonococcus, and the meningococcus.
I wish to draw especial attention to the value of adding glucose
to the serum-agar ordinarily used for cultivating the gonococcus, for
by means of this addition I could obtain colonies from urethral dis-
charges when absent on ordinary serum-agar, or if present in the
latter they were often smaller. As a rule we have used a mixture of
two parts of 0.5 per cent, glucose-agar and one part of human serum
(hydrocele fluid preferably). Even the addition of 0.1 per cent,
glucose seems to favor the development of the gonococcus.
I would recommend the media particularly on account of the
rapidity of gro\\i;h upon them. This is especially true in making
cultures from the blood. I have made some use of these media for
510 MOUNT SINAI HOSPITAL REPORTS.
saving time in plate work. The principle involved is that we should
use the medium which contains a sugar which does not allow of pre-
cipitation by a certain bacterium. Thus, for finding the bacillus
typhosus I use a medium consisting of 0.5 per cent, (sometimes 2.0
per cent.) lactose-serum-agar. On such a plate the bacteria which
precipitate are surrounded by a white halo. According to our ob-
servations thus far, the typhoid bacillus does not precipitate in the
presence of lactose ; therefore colonies on the plate with a distinct
halo can at once be excluded as not being those of the typhoid bacillus,
and only the others need to be investigated. I believe this method can
be developed by further study. I must draw attention, however, to
one drawback, which, however, thus far has not been a serious one.
If there are numerous colonies on the plate the entire plate becomes