Fungi in cheese ripening : Camembert and Roquefort

Historic, archived document

Do not assume content reflects current
scientific knowledge, policies, or practices.

U. S. DEPARTMENT OF AGRICULTURE,

BUREAU OF ANIMAL INDUSTRY. — Bulletin No. 82.

A. D. MELVIN, Chief of Bureau.

FUNGI IN CHEESE RIPENING

CAMEMBERT AND ROQUEFORT.

BY

CHARLES THOM, Ph. D.,

Mycologist in Cheese Investigations, Dairy Division, Bureau of
Animal Industry.

WASHINGTON:
GOVERNMENT PRINTING OFFICE.
1906.

LETTER OF TRANSMITTAL.

U. S. Department of Agriculture,

Bureau of Animal Industry,
Washington, D. C, February 6, 1906.

Sir: I have the honor to transmit herewith the manuscript of an
article entitled "Fungi in Cheese Ripening: Camembert and Roque-
fort," by Charles Thorn, Ph. D., and to recommend its publication as
Bulletin No. 82 of the series of this Bureau. This is the second paper
dealing with the cooperative experiments in soft -cheese making
undertaken by the Dairy Division of this Bureau in conjunction with
the Storrs (Conn.) Agricultural Experiment Station, the first paper
having been published as Bulletin No. 71 of this Bureau.

These experiments have been carried on at the Storrs Station under
the general direction of Prof. L. A. Clinton, the station director, and
under the personal supervision of Dr. H. W. Conn, the station bac-
teriologist, in accordance with the plan outlined in the introduction
to Bulletin No. 71.

While there are many problems yet to be investigated with refer-
ence to the manufacture in this country of soft cheeses of the best
European types, this article indicates that good headway is being
made in that direction, and it is believed that the information here
presented is of considerable scientific and economic value.
Respectfully,

A. D. Melvin,

Chief of Bureau.

Hon. James Wilson,

Secretary of Agriculture.

2

CONTENTS.

Page.

Introduction 5

Camembert cheese 5

Resume of previous paper 5

Culture media and methods 6

Effect of a fungus upon a culture medium 8

Literature of cheese fungi _ 8

Biological analysis of a cheese 9

The flora of Camembert cheese 10

Outline of the work 11

Relation of molds to acidity 12

The breaking down of casein 14

Liquefaction of gelatin 15

Raulin's fluid 16

Casein 16

Sterile milk and curd 17

Does the mycelium penetrate the cheese ? 17

Camembert Penicillium upon cheese 18

Comparative studies of fungous digestion 18

Flavors 21

Temperature 23

Humidity 24

Inoculating material 25

Inoculation with Penicillium 26

Vitality of spores 27

Contaminations 27

Roquefort cheese 28

Cheeses related to Roquefort 29

American Brie and Isigny 30

Molds referred to in this paper 31

The Camembert mold (Penicillium camemberti) 32

Technical characterization of the Camembert mold 33

The Roquefort mold (Penicillium roqueforti) 34

Technical characterization of the Roquefort mold 35

Oidium lactis 36

Summary 38

Camembert cheese 38

Roquefort cheese 38

Other varieties of cheese 39

Bibliography 40

ILLUSTRATIONS

Fig. 1. Camembert Penicillium (P. camemberti) 32

2. Roquefort Penicillium (P. roqueforti) 35

3. Oidium lactis 37

3

FUNGI IN CHEESE RIPENING: CAMEMBERT AND

ROQUEFORT.

INTRODUCTION.

It has been shown in a previous bulletin that certain fungi are the
active agents indispensable to the ripening of Camembert cheese.
The general results and the data upon which they rest are there dis-
cussed, but the more special mycological studies, involving several
lines of work, remained to be brought out in greater detail. These fall
naturally under two heads: (1) The physiological studies of the func-
tions of particular species in the ripening processes of Camembert,
Koquefort, and certain related types of cheese; (2) the classification
and description of these and other forms occurring in dairy work.
This paper includes only the work done under the first head. The
description of the fungi occurring in dairy work is reserved for another
paper.

Aside from such obligations as are mentioned in the discussion of
special topics, the author wishes to acknowledge the assistance of Dr.
B. B. Turner, Prof. W. A. Stocking, Mr. A. W. Bosworth, and Mr.
T. W. Issajeff, members of the experiment station staff, in numerous
cases where the work of each presupposes the results of the other, and
especially to acknowledge the constant assistance of the supervisor of
the investigation, Dr. H. W. Conn, with whom the cheese problems
have been fully discussed at every stage.

CAMEMBERT CHEESE.

RESUME OF PREVIOUS PAPER.

The biological conditions and the physical changes encountered in
the production of a Camembert cheese from market milk may be
restated from our former bulletin 1 "as a basis for defining the special
problems of the mycologist.

Milk as ordinarily received contains bacteria of many species and
the germinating spores of numerous fungi from the stable and from
the food of the cattle. When such milk is curdled for cheese making,

a The figure references are to bibliography at end of bulletin.

5

6

FUNGI IN CHEESE RIPENING.

representatives of all of these species are inclosed in the mass of coagu-
lum. Freshly made cheese from this curd, then, may contain any
species of mold or bacterium found in the locality which is capable of
living in milk or its products. The first step in the, ripening of a Ca-
membert cheese is the production of lactic acid. The lactic bacteria
very soon increase their rate of multiplication so enormously as to be-
come entirely dominant. The acid produced by these forms soon
reaches a percentage sufficiently high to restrict the further growth
of nearly every other species of bacteria, and even to eliminate the
organisms themselves. In a time varying from a few hours to three
or four days, according to the proportional numbers of these antago-
nistic species at the start, further bacterial growth seems to be entirely
stopped. Bacterial development can not begin again until this acidity
is reduced below the critical point for the species involved, and even
then, since the acid is neutralized on the outside first, for most species
it begins at the surface and works slowly inward. The uncertainties
due to the presence of many species of bacteria in the milk are in this
way avoided by the natural, simple, and almost universally successful
process of souring.

The further ripening of a Camembert cheese is attended by a
gradual reduction of this acidity until the ripe cheese is usually alka-
line to litmus. At the same time the mold action in the mass of curd
produces chemical changes which in from three to five weeks reduce
the previously insoluble mass to a high percentage of solubility in
water. In the later stages of this breaking down compounds are
formed which give the characteristic odors and flavors to this type of
cheese. Associated with these chemical changes there is a progressive
physical change from the firm curd to a soft, buttery, or even semi-
liquid texture, characteristic of ripe cheese. The biological problems
then were, in general, the determination of what organisms cause —

(1) The changes in the acidity of the curd.

(2) The breaking down of the casein, with the associated changes in
the physical character of the cheese.

(3) The production of the flavors.

(4) The recognition and control of deleterious species.

CULTURE MEDIA AND METHODS.

The common dairy fungi grow readily upon any of the standard cul-
ture media. Among the media used have been peptone agar, whey
gelatin, sugar gelatin with or without the addition of litmus, milk
agar, gelatin and agar made with Raulin's fluid, potato agar, potato
plugs, and sterilized milk and curd. Special studies have involved
other preparations. The fact that these fungi grow readily upon all
the common media has led to the selection of two preparations for con-
stant use, and the careful study upon these of all species found. For

OAMEMBERT AND ROQUEFORT.

7

this purpose the sugar gelatin, described by Conn 2 for the qualitative
bacteriological analysis of milk, and potato agar have been used.

The sugar-gelatin formula produces an accurately titrated medium
in which every effort is made to secure a uniform composition.
Although absolute uniformity in chemical and physical properties is
never obtained, the reaction of many species of fungi, when grown
upon successive lots of gelatin made after this formula, have been so
reliable as to commend its use for determining physiological charac-
ters. It seems clearly shown, therefore, that slight variations in the
composition of the medium do not produce great differences in the
species studied in this paper. In the discussion of the relation of a
mold to this gelatin it must be borne in mind that the same results
might not follow the use of any other formula.

The other medium, the potato agar, was selected because of its use
in many mycological laboratories. In this medium uniform compo-
sition can hardly be claimed. The following process has been used in
this work: The potatoes are carefully washed, pared, and sliced, then
slowly heated for about two hours in approximately two volumes of
water. At the close of the heating the water is allowed to boil. The
whole is then filtered through cloth, and commonly through cotton
also, water being added to make up the losses of evaporation and fil-
tering. To this is added 1 per cent of shredded agar. It is then
heated for from twenty to thirty minutes in the autoclave to 120° C.
or higher, when it may at once be put into tubes for use, or, if
cloudy, it may be very quickly filtered through absorbent cotton,
after which it should be quite clear. The uncertainties in the com-
position of this medium result from the differences in the potato ex-
tract itself and from the fact that the difficulties in filtering this
extract take out a varying amount, which is replaced with water.
Titration shows that this medium is nearly neutral (4-6 acid on Ful-
ler's scale) in cases tested to phenolphthalein; consequently it is used
without neutralizing. Culture and study of the same species upon
successive lots of this medium show that these differences in compo-
sition have little if any effect upon the morphology of the species
studied.

Petri-dish cultures have been used continually because they admit
of direct study under the microscope. Slanted test tubes were found
useful for stock cultures and for gross studies of physiological effects,
but they are of little value for comparative work. It is useless to
attempt to get a correct idea of the normal gross structure of these
molds from fluid mounts. The extremely delicate hyphse are so tan-
gled in such preparations as to give but very little idea of their ordi-
nary appearance, while the chains of conidia break up immediately
when placed in any fluid. Such mounts are useful and necessary to
get at details of cell structure and cell relations, but in comparative

8

FUNGI IN CHEE8E RIPENING.

studies of species of such a genus as Penicillium their value is only that
of a useful accessory. The primary source of comparative data must
be direct study of the growing colony, undisturbed upon the culture
medium, with the best lenses that admit of such use.

This method of study recognizes that morphology is the basis of
fungus determination, but takes into consideration —

(1) That morphology must not only include the minutest details
of cell structure and cell relations such as are undisturbed in fluid
mounts, but also the appearance and character of the colony.

(2) That the morphology of the colony — i. e., the size of conidio-
phore and fructification, relation of these to substratum, appearance,
and relations of aerial and submerged mycelium — is different upon
various substrata, but has been found to be characteristic for each
particular substratum.

(3) That a description of morphology to be of value must, there-
fore, specify the formula of the medium used and the conditions.

Dilution cultures have been necessary usually to obtain the colonies
pure, but the direct transfer of large numbers of spores upon a plati-
num needle to the surface of gelatine or agar plates which have been
allowed to cool has been found to give equally reliable results, and to
have many advantages for the study of species once obtained in pure
culture. This is often spoken of as inoculation of cold-poured plates.
Litmus solution may be used with either gelatin or agar, and gives
striking evidence of differences in species and the rate of their physio-
logical action. Bacterial contamination has been usually restrained
by the addition of from 2 to 4 drops of normal lactic acid to 8 or
10 c. c. of medium.

EFFECT OF A FUNGUS UPON A CULTURE MEDIUM.

In studying the relation of a fungus to a culture medium we find (1)
that the fungus absorbs food from the surrounding medium; (2) that
it may secrete or excrete substances into the medium which may
transform its chemical composition and its appearance. The amount
of food absorded by the fungus is small, and for our purposes may be
practically ignored, but the changes induced by indirect action —
secretions from the mycelium — are great and far-reaching. To this
latter group belong the changes in acidity, digestive effects, and fla-
vors produced by fungi.

LITERATURE OF CHEESE FUNGI.

A review of the literature at the outset showed that no work on the
fungous flora of the various types of soft cheese had been published in
English. Epstein, 3 at Prague, studied the ripening of Camembert and
Brie cheeses. He attributes the breaking down of the curd in French
Brie to the action of Penicillium album, but denies the participation

CAMEMBERT AND ROQUEFORT.

9

of molds in the ripening of Camembert. Johan-Olsen, 4 in Sweden, has
published a brief review of the fungi related to the ripening of Gamme-
lost, barely mentioning work done upon Camembert. Constantin and
Ray, 5 in France, have described the appearance upon the cheese of the
species of Penicillium involved in the ripening of the French Brie.
Roger, 6 also in France, has attributed a single phase of Camembert
cheese ripening to the activity of Penicillium candidum, for which he
gives no description. Of these references, that of Epstein and that of
Constantin and Ray describe the mold found upon the French Brie
sufficiently clearly to aid in its recognition. A popular article, signed
Margaret, 7 in the Creamery Journal of October, 1904, gives in entirely
untechnical language a very satisfactory description of the appearance
upon cheese of the penicillium concerned in the ripening of Camem-
bert. The general insufficiency of the literature available made a
first-hand study of the types of cheese found in American markets the
only source from which definite information could be secured.

BIOLOGICAL ANALYSIS OF A CHEESE.

In the biological analysis of a market cheese it is carefully un-
wrapped to avoid contamination as far as possible. Series of dilution
cultures on neutral and acid media are made at once from each part
of its surface which shows any variation in appearance. In this way
all the surface molds and bacteria are secured in one set of plates.
Afterwards this surface is examined in detail, usually with a lens, the
appearance of the different areas being noted, and direct transfers
from each area made to cold agar or gelatin plates. The cheese is
then cut with a sterilized scalpel and cultures are made from various
portions of the interior. Usually the transfers were made from the
center and from the area just inside the rind. Any part showing spe-
cial appearances is reserved for a separate series of cultures.

Most of the brands of Camembert cheese found in our markets, as
well as some sent by Roger, have been examined in this way. For
comparison, similar studies have been made from several specimens of
Roquefort cheese bought in different markets, and from individual
specimens of Gorgonzola and Stilton. Single studies for molds have
been made from Limburger, Port du Salut, Brinse, and from several
brands of prepared cheese found in the market. From these cultures
all species of bacteria found have been isolated and handed over to the
bacteriologists. Each variety of mold occurring upon these cheeses
has been isolated and studied. It has been possible in this way to
show that a comparatively small number of species characteristically
occur upon soft cheese. Although this list may be greatly extended
by including forms which are occasionally found, it is rather surpris-
ing that a restricted group of species occurs with much regularity in
studies of cheese from so widely different countries.
21156— No. 82—06 2

10

FUNGI IN CHEESE RIPENING.

To study the origin and distribution of these molds several labora-
tories and cheese factories have been visited and cultures taken. Cor-
respondents in distant States have kindly sent cultures of molds
occurring in their work. Among those who have sent material are
Dr. C. E. Marshall, Agricultural College, Mich. ; Mr. E. G. Hastings,
Madison, Wis.; Prof. F. C. Harrison, Guelph, Ontario; Dr. H. A.
Harding, Geneva, N. Y., and Prof. P. H. Rolfs, Miami, Fla. Thus, in
addition to a large number of cultures from the dairy laboratories of
the stations at Storrs and at Middletown, we have accumulated from
various sources a considerable number of species representing the
characteristic molds occurring in dairy work, as well as many forms
collected in the field and from laboratories not associated with dairy
investigation.

THE FLORA OF CAMEMBERT CHEESE.

Although a considerable variety of molds appeared in cultures from
Camembert cheeses, a list of possibly twenty species would include
those which were often found. Among these there are perhaps six
species of Penicillium, two or three of Aspergillus, Oidium lactis, Clado-
sporium herbarum, one or two of Mucor, one or more of Fusarium,
Monilia Candida, and two species perhaps related to it, with the inci-
dental occurrence of Acrostalagmus cinnabarinus, a Cephalosporium,
various species of Alternaria, and Stysanus. Besides these, yeasts in
large numbers and considerable variety are found in many cases.

The comparison of the results of culture with comparative studies
of the surfaces of different brands of cheese showed that a single spe-
cies of Penicillium was present upon every Camembert cheese exam-
ined. In partially ripened cheeses this mold often covered the larger
part of the surface. We shall call this the "Camembert Penicillium"
or the "Camembert mold." This species develops a large and charac-
teristic growth of aerial mycelium in addition to a densely felted mass
of threads which penetrate the surface of the cheese for 1 or 2 mm. and
largely constitute the rind. In all except a few very old cheeses
which were almost covered with red slime of bacterial origin it was
readily seen to be the dominant species upon the surface.

Similarly, cultural data showed Oidium (Oospora) lactis to be abun-
dant upon every brand of Camembert. This mold is practically in-
distinguishable upon the surface by its characters, except under very
favorable conditions, and at best its recognition, even with a hand lens,
is not often certain. Mycelium of this fungus develops only in very
moist substrata, and is usually entirely submerged. Only part of its
chains of conidia even rise above the surface. In old and very ripe
cheese, when the rind is covered with yeasts and bacteria, it is often
difficult under the microscope to find the spores of Oidium. In such
cases, unless one is familiar with the peculiar smell associated with its

CAMEMBERT AND ROQUEFORT.

11

action, he must depend entirely upon the culture for evidence of its
presence.

No other species of mold has been found upon every cheese exam-
ined, although no market cheese has failed to show contamination
with at least one or two of the other fungi listed above. In other
words, comparative biological examination of imported Camembert
cheeses established the fact that these two species of mold were pres-
ent upon them all, however abundantly they might be contaminated
with other forms. The examination of hundreds of cheeses in the city
markets has shown the presence of the same two molds upon all the
brands of Camembert offered for sale. Such analyses clearly estab-
lished the presence of these molds upon the ripe cheese, but gave no
information either as to whether they were necessary or what func-
tion, if any, they might have. Experiments were therefore devised
to test the relationship of these molds to the ripening processes out-
lined above. The constant occurrence of other molds upon the cheese
brings up the question, How and to what extent do the latter affect
the ripening process ? The experiments, therefore, have been made to
include as many species as possible. Where detailed chemical analy-
ses had to be made the work has necessarily been restricted to a few
forms.

For this purpose, in addition to the Camembert Penicillium and
Oidium lactis, the Penicillium found in Roquefort cheese ("der Edel-
pilz" of German authors) has been generally used. For convenience
it is called the " Roquefort Penicillium " or " Roquefort mold." One of
the Mucors, probably Mucor or Chlamydomucor racemosus, is so com-
monly found that it has often been included. A pure white mold
closely related to the Camembert Penicillium has given some inter-
esting contrasts. When reference is made to any of the numerous
undetermined green species of Penicillium, they will be indicated by
the letter or number under which they appear in the record book of
cultures, and under which the origin and subsequent cultural history
of all species studied has been kept.

OUTLINE OF THE WORK.

These studies involve two classes of data, first, those experiments
requiring quantitative analyses, which have been conducted in
cooperation with Mr. A. W. Bosworth, chemist to this investigation,
the results of which series of analyses will appear in his report; second,
experiments which show the physiological characters of the fungi by
physical changes in the appearance, texture, or color of the medium
used, or by the production of flavors.

The results may be anticipated here by noting that these two classes
of data did not prove mutually interdependent, but that analysis may
show in general the right stage of chemical changes called for in a ripe

12

FUNGI IN CHEESE RIPENING.

cheese without the necessary texture and flavor; and, conversely, the
practically necessary texture and flavor may be obtained in a cheese
differing considerably in its chemical characteristics from the standard
market article. In our practical experiments we sought first for
proper appearance, texture, and flavor of the cheeses; then, without
disturbing these, endeavored so to control the processes of ripening as
to satisfy the standard of chemical composition established from the
study of market cheeses.

RELATION OF MOLDS TO ACIDITY.

The development of lactic acid has been shown to be of primary im-
portance in the control of deleterious bacteria. In our previous paper
it has also been seen that after doing its work this acidity gradually
disappears in the ripening process. The disappearance of the acid
has been attributed by Soger, 6 by Epstein, 8 and by Maze 9 to the activ-
ity of molds, and interpreted as preparing the way for the action of
peptonizing bacteria. This view of the relation of molds to cheese
ripening has been widely quoted as their only function in the process.

The acid exerts practically no selective action upon any of the molds
studied. Stoll has recently shown that species of Penicillium grow
readily in media containing a much higher percentage of acid than
ever occurs in cheese work. The use of acid in fungous cultures to re-
strain bacteria is practically universal, but the action of the different
species of mold upon the acid is very different. This is strikingly
shown by the introduction of a solution of litmus into the culture
media used. Litmus gelatin or litmus agar may be a deep blue if used
at 15 acid on Fuller's scale, as is usual for bacterial studies, or a clear
bright red if 2 to 4 drops of normal lactic or other acid are added to
10 c. c. of medium. No mold cultivated in this work has failed to
show some definite relation to acidity indicated by litmus reaction.
Some fungi, as soon as they develop visible colonies, begin to change
red (acid) media to blue (alkaline), and consistently maintain this
character. Many others, when grown in blue gelatin (designating by
blue gelatin 15 points acid to phenolphthalein=10 points alkaline to
litmus on Fuller's scale), begin by changing the blue to red. This
change may vary from the faintest tinge of red in only that part of the
medium directly in contact with the threads of the young colony to
deep red over large areas. Oidium lactis and Eoquefort Penicillium
produce at times a very slight pink, which barely traces the outer limits
of the young colonies before the blue reaction begins to appear. At
other times the red, if appearing at all, has been so evanescent as to be
overlooked. It has been suggested that this slight appearance of
acidity might be due to the excretion of carbon dioxide in respiration,
which, although continuous, is afterwards masked by many times
larger changes in other substances.

CAMEMBERT AND ROQUEFORT.

13

The Camembert Penicillium, and several of the very common
green species of Penicillium, when grown upon blue gelatin, at first
turn all the substratum in contact with the growing colonies to a
bright red. Some species produce areas of red beyond the limits of
the mycelium. These effects are most clearly seen by examining the
colony from the under side. Later a spot of blue appears in the cen-
ter of the colony below and gradually extends outward until com-
monly the entire mass of culture medium has become blue. This
often involves a change of reaction in agar or gelatin 2 to 3 cm.
beyond the colony. It is thus clear that there must be either the
secretion or the excretion by the mycelium into the medium of a
substance capable of changing this reaction or the absorption from
the medium of some substance, thus changing its reaction. The
exact nature of this change has not been determined. Increase in the
percentage of acidity or of alkalinity retards the change of reaction.
In certain experiments phenolphthalein was introduced into red
litmus media and several species of Penicillium and Oidium lactis
were grown upon it. With the Camembert Penicillium the entire
mass of agar became blue in a few days, and remained so for nearly
three weeks. Then the characteristic pink color for the alkaline
reaction of phenolphthalein appeared on the under side of the
colony. This was tested by opening the colony with a platinum
needle and introducing a very small drop of normal acid, when the
pink area was changed first to blue and then to red. As the acid
diffused outward from the center the wave of blue traveled outward,
being replaced constantly by red until all tface of the phenolphthalein
reaction was gone. The other species used did not give this reaction.
There are forms including some species of Penicillium, Aspergillus
niger, Moniliafructigena, and others, which produce the acid reaction
in litmus media without any change to blue. Several species of Peni-
cillium rapidly produce the purplish color which is characteristic of
the turning point of litmus at which their further development
occurs. Apparently these bring acid or alkaline, media to that point
without further change. It would appear, then, that the relations of
these molds to acidity, as indicated by the litmus reaction, is reason-
ably uniform. To determine whether the litmus reaction would be
reliable upon a medium closely allied to cheese, test tubes of sepa-
rated milk were prepared, blue litmus added, and the tubes sterilized.
Eleven species of Penicillium were inoculated into these tubes and
observations made every day. Of the eleven species, four, including
the Camembert Penicillium, produced a layer of red milk for a few
millimeters below the colonies, which later was changed back to blue.
The other species either intensified the blue or produced no change.

The suggestion has been made that neutralization of acid is due
to the production of ammonia. A series of cultures were made in

14

FUNGI IN CHEESE RIPENING.

cooperation with Mr. A. W. Bosworth to test the production of ammo-
nia compounds by mold action. The species used were the Roquefort
Penicillium, the Camembert Penicillium, Penicillium sp. (record No.
310), Oidium lactis, Oidium sp. (record B), and Aspergillus niger.
These were grown upon potato ager, to which litmus and lactic acid
were added. The Aspergillus culture remained bright red; all the
others became deep blue. Upon analysis the Aspergillus niger was
found to have produced the largest amount of ammonia. Study of
the figures showed that the ammonia alone was not sufficient to neu-
tralize the acid used in any case. It is clear, then, that the lactic acid
must have been neutralized by some other basic products of digestion
rather than by ammonia. If the acid were absorbed and dissociated
after absorption the area of blue would be restricted to the neighbor-
hood of the hyphae, or the diffusion of the acid for considerable dis-
tances would produce purple tones instead of sharply marked areas
of red and blue. The data seem to indicate that chemical decompo-
sition or neutralization of acid must be the action of some product
excreted by the fungus, probably an enzyme.

It has thus been shown by many experiments that the Camembert
Penicilhum and Oidium lactis are two of many species capable of reduc-
ing the acidity of the media upon which they grow. Many other species
of the same genus produce this effect more quickly than the Camembert
Penicilhum and some act at about the same rate. The reduction of
the acidity of the cheese may clearly be attributed to these molds;
but the study of the relations of many other molds to acid indicates
that any of a large number of species might be equally or more useful
for the accomplishment of this step in cheese ripening. If, there-
fore, these particular molds are essential to Camembert cheese ripen-
ing, their special function must be sought in other steps of the
process.

THE BREAKING DOWN OF CASEIN.

The changes in firm sour curd which result in the production of the
soft, buttery, or semiliquid texture of the Camembert cheese present
some very complex problems. These may be grouped as (1) the
purely chemical questions, which involve qualitative and quantitative
analyses of the material at every stage; (2) The biological and
physical questions, which deal with the agents and conditions which
produce these results and with the gross appearances of the final
products, whose descriptions do not depend upon detailed chemical
analysis.

(1) The chemist describes the general course and extent of these
processes 1 as a change in which the insoluble or but slightly soluble
compounds of casein found in sour curd are rendered almost com-
pletely soluble in water. The details of the process and the data will
appear later in the report of the chemist.

CAMEMBEET AND KOQUEFORT.

15

(2) To determine what relation the molds might have to this change
involved a great many cultures on different media. In some experi-
ments the number of species used was large and the results acquired
in that way a comparative value, but in the more complicated trials
the work was limited to those mentioned above.

It is practically impossible to produce a normal cheese in such a
way as to avoid contamination with bacteria or molds. It is difficult,
therefore, to study directly upon cheese the relations of organisms to
thfe steps of cheese ripening. Even were this possible, the complexity
of the changes encountered would make the interpretation of the phe-
nomena difficult. The activities of these molds have, therefore, been
studied in pure culture upon a series of media which would give infor-
mation as to steps of the process. While these cultural studies were
proceeding, many cheeses were made and inoculated with the Camem-
bert and Koquefort Penicillia. The measure of success obtained from
cheese inoculated with the Camembert Penicillium gave good, practi-
cal ground for its continued study. These detail studies may be dis-
cussed best separately.

LIQUEFACTION OF GELATIN.

The liquefaction of gelatin media has been much used as an index of
digestive activity. All species obtained have been grown upon neu-
tral and acid sugar gelatin and the effects noted carefully.

The difference in action between the molds important in this inves-
tigation are striking. The Mucor produces a slow but rather com-
plete liquefaction; Oidium lactis will gradually soften the gelatin so
that the center of the colony is liquefied; a pigment-producing Peni-
cillium (recorded simply as O) will liquefy all the gelatin in contact
with it so quickly that it becomes in a week a floating colony in a
watery pool twice its own diameter. Several other species of Peni-
cillium have the same effect. The Roquefort Penicillium softens gela-
tin somewhat, but never produces a watery liquefaction. The
Camembert Penicillium often produces a slight liquefaction under the
center of the colony, but never extends that liquid area to half the
total size of the colony. This seems to indicate that the Penicillium O
and its allies would produce a rapid digestion, that the Mucor would
be somewhat slower, that the Camembert mold might have some diges-
tive effect and the Roquefort mold very little, if any, value. The test
of the ability to liquefy the gelatin used gives, therefore, only indefi-
nite or negative results as to any advantageous relation of these par-
ticular species to cheese ripening.

Comparative study of numerous cultures of many species of fufigi
upon gelatin gives, however, some very intefesting suggestions. In
many species which liquefy litmus gelatin rapidly, the area of liquefac-
tion is surrounded by a blue (alkaline) band. For example, in one

16

FUNGI IN CHEESE RIPENING.

experiment with Penicillium 392 at its most active period of growth a
colony 15 mm. in diameter was surrounded by a liquefied area 4 to 8
mm. wide. This area was in turn surrounded by a band of intense
blue shading gradually in a width of perhaps 10 mm. into unchanged
red litmus gelatin. The medium which had been liquefied was almost
colorless.

Several suggestions may be drawn from many such observations.
The change in acidity of the medium, as has been noted above, may
be effected at a distance of 2 to 3 cm. from the colony. This change
of litmus reaction advances faster than the area of liquefaction of
the gelatin. The breadth of the area of liquefaction shows that the
action of the fungus is not a digestion by contact, but the secretion
into the medium of diffusible agents, that is, enzymes. In most of
these species liquefaction occurs only in areas having alkaline reac-
tion. No general relation between acidity and digestion is estab-
lished. The substantial uniformity of the results of repeated cultures
of the same species of fungi upon gelatin made after the formula used
established its usefulness as a test of the ability of an organism to per-
form this particular digestion. It will be shown later that the ability
to liquefy this variety of gelatin is not to be regarded as a general test
of the ability of a species to produce active proteolytic enzymes.

raulin's fluid.

To test the ability of these species to grow in a medium entirely lack-
ing in proteid, Raulin's fluid was used as given by Smith and Swingle, 10
but modified by leaving out the potassium silicate and zinc sulphate.
Sterilized flasks of this solution were inoculated with Mucor, Oidium
lactis, Camembert Penicillium, and Eoquefort Penicillium. All four
grew. The Oidium lactis and Mucor did not appear to develop in an
entirely normal way. Both species of Penicillium grew richly and
fruited normally. The culture of the Camembert mold, after growing
several weeks, was examined chemically and digestive experiments
conducted by Mr. Bosworth demonstrated the presence of a proteolytic
enzyme. In this way it was shown that this fungus could not only
construct proteid from inorganic compounds of nitrogen, but would
produce proteolytic enzymes in such a solution. Enzyme studies
were not made for the other species used in this experiment.

CASEIN.

For a medium at the opposite extreme, the chemists prepared pure
casein. This was weighed into 2-gram lots, moistened, sterilized in
the autoclave, and inoculated with five species of mold. All grew and
fruited luxuriantly. This experiment showed only that the species
used were able to break down casein and to grow normally upon the
products of this digestion without the addition of other nutrients..

CAMEMBERT AND ROQUEFORT.

17

STERILE MILK AND CURD.

' Sterilized milk and sterilized curd offer a substratum related to
cheese. Sterilized milk in quantities varying from 40 c. c. to 150 c. c.
in test tubes and Erlenmeyer flasks has often been used. Nearly all
species of Penicillium grow luxuriantly, forming a felted mass of
mycelium often 2 to 4 mm. in thickness upon the surface of the milk.
With the absorption of the milk in such cultures of the Camembert
and Roquefort species the mass of mycelium buckles and bends,
tubercles of mycelium arise on the under side of the mass and grow
downward, keeping the mold in connection with the fluid. In this
way a culture may continue to grow for several months until it forms
tough, irregular masses of felted hyphse, filling the test tube for an
inch or more downward from the original surface of the milk. The
milk below the colony soon becomes transparent, giving reactions for
digestion, with a residue of curd at the bottom, which in the course of
time may be almost completely dissolved. With the Oidium lactis,
on the contrary, the colonies largely sink below the surface, so that the
milk may be quite well filled with mycelium upon which chains of
spores are only produced in quantity at or just below the surface.
Similar experiments with 100 grams of sterilized curd in flasks, inocu-
lated with the Camembert and Roquefort molds, have shown that
either species is able to change the chemical composition until the
derivatives of casein are amost completely water soluble. Such cul-
tures were plated to show their freedom from contamination by bac-
teria before analysis. The resulting products give the standard reac-
tions for digestion. These experiments show that either of these
molds is capable of producing digestive changes comparable in their
completeness, rapidity, and general nature to those shown by analysis
to have occurred in the ripening of Camembert cheeses.

DOES THE MYCELIUM PENETRATE THE CHEESE?

It must be noted carefully that this action of the Camembert mold
goes on without the complete penetration of the substratum by the
mycelium of the mold. That this is true is readily seen in milk cul-
tures, where the limits of the development of the mycelium are sharp
and clear. The same fact has been demonstrated for cheese by hun-
dreds of sections and careful cultural studies many times repeated.
The mycelium forms a dense mat upon the surface of the fluid or the
mass of curd, or the newly made cheese. It follows the irregularities
of the surface and is not found to enter well-packed curd to any extent.
It is very difficult to prove that hyphse of this mold actually appear in
curd of uniform texture below 1 or 2 mm. When found deeper, careful
search usually shows a cracking of the surface, so that the mycelium
may follow the opening already made. In no case of many hundreds
21156— No. 82—06 3

18

FUNGI IN CHEESE RIPENING.

of cheeses studied and experiments performed has the mold been
found to fruit in cavities not opening broadly upon the surface. This
is in marked contrast to the habit of the Penicillium instrumental in
the ripening of Roquefort cheese, which penetrates the channels of the
substratum and fruits in every cavity large enough to accommodate a
conidiophore. The Roquefort mold will make every cavity in a
cracker or piece of bread green with spores, while the Camembert
mold will fruit upon the surface of the bread or cracker with only
vegetative mycelium inside the bread.

Definite experiments to prove that this digestive power on the part
of the Penicillium is due to the secretion of one or more enzymes have
given characteristic reactions for digestion many times. Without
here discussing these chemical reactions, it has been shown that the
chemical action of the fungus is carried on at distances from the
mycelium which preclude direct action. The enzyme must therefore
be secreted and diffuse outward from the mycelium into the sub-
stratum. This explains why the Camembert cheese begins to ripen
just under the surface and the process progresses inward from all
sides until the cheese is entirely ripe. Before this process is complete
the center is simply sour curd. A good illustration of this action is
seen in cheeses which are ripened without turning. In such cases the
development of mold and enzyme on the lower surface is prevented,
and as a consequence ripening is delayed on that surface.

CAMEMBERT PENICILLIUM UPON CHEESE.

Many cheeses have been made and inoculated with this mold in con-
junction with pure cultures of lactic starter. Little difficulty is found'
in this, since, if an abundance of spores are put upon the cheese when
made, this mold seems capable of taking and maintaining the lead of
all others. A cheese made in this way and ripened for from three to
four weeks will finally be rendered creamy, or, under some conditions,
waxy throughout, in color white within, in flavor almost neutral,
having no particular character — good or bad — and hence, to one fond
of Camembert cheese, tasteless and insipid. The important fea-
tures of this ripening process are, then, the completeness of its action
and the entire absence of any objectionable character in its flavor.
Biological analysis has shown that the center of such a ripened cheese
may be practically a pure culture of lactic organisms. The texture is,
therefore, obtainable by the use of the Penicillium alone.

COMPARATIVE STUDIES OF FUNGOUS DIGESTION.

Comparative tests of digestive action have been made for a number
of molds. The Roquefort Penicillium has been used in parallel cul-
tures with the Camembert Penicillium in many determinations. It

CAM EM BERT AND ROQUEFORT.

19

has shown equal or greater ability to digest milk and curd. A typ-
ical example of several series consisted of the cultivation of 1 1 species
of Penicillium upon sterilized milk in large test tubes. Observation
of results after seven days showed digestion by 7 of these species. In
5 of them the amount of action exceeded that of the Camembert Peni-
cillium, and some of them appeared to digest milk at least twice as
rapidly as did that species in the first week.

In another series milk agar was made by dissolving 1 to 2 per cent
of the agar in water at 130° C. and pouring together equal quantities of
the hot agar and hot sterilized milk. If poured into Petri dishes at.
once this medium was smooth and clear, but if acidified or sterilized
after mixing, flakes of precipitate appeared. The flaky precipitate in
the acidified cultures was found very useful as an indication of diges-
tion. In cultures upon the surface of such plates where digestive
action was strong the flakes would entirely disappear. Twenty-
three species of mold were tested upon milk agar in this way. Of
these, 8 produced a distinctly stronger digestion than the Camembert
Penicillium; 5 produced digestion approximately equaling that spe-
cies, and 10 produced less digestion. These cultures were mostly
made in duplicate, and both results in all but two cases agreed fully.
Oidium lactis produced comparatively little effect upon this medium.

Table I. — Reaction of certain species of molds.

Species.

Camembert P . . .

Roquefort P

Oidium

Mucor 12

Mueor 191

O

300

132

310

68

Monilia Candida
198.

P. brevicaule

392

240

A 8pergillus niger

135

136

Litmus.

Red, then blue

Blue,

Blue

Blue

Blue

Blue

Blue

Red, then slowly

blue.
Red, then blue

Red, then blue

Blue

Blue

Blue

Blue

Red

Blue

Red to purple blue

Liquefaction
of gelatin.

Partial. . .
Softening.

Incomplete. .
Incomplete . .
Incomplete . .

Rapid

Partial

Slight..

Slight..

Partial .
Rapid . .

Rapid

Rapid

Rapid

Rapid

Rapid

Partial soft-
ening.

Rate of
digestion
of curd.

Medium.. .
Rapid

Slow

Slow

Slow

Rapid

Medium
to rapid.
Medium...

Slow to
medium.

Slow

Rapid

Rapid

Slight

Rapid

Medium..

Rate of diges-
tion of milk.

Medium .
Rapid . . .
Slow

Rapid .
Rapid .

Slight...

Medium .
Rapid . . .

Rapid

Rather slow.

Slow...
Rapid .

o° to 10° C.

Grow, slow fruit-
ing.

Characteristic

growth.
Characteristic.
Poor growth.

Retarded.

Slow growth.
Slow fruiting.

Characteristic.

Two species of Penicillia, 68 and 310, found closely associated upon
cheese with the Camembert Penicillium, produced little digestion.
The Roquefort Penicillium and several other molds often found upon
Camembert cheese appeared to act much more rapidly than the Ca-
membert mold itself.

20

FUNGI IN CHEESE RIPENING.

All of these series of cultures under different conditions have many
times shown the same results and prove that the ability to digest curd
is common to many species of fungi. The species we have been led to
call the Camembert Penicillium possesses this character in common
with numerous other molds, many of which act more rapidly than
this one.

After the ability of several molds to digest curd is established, the
relation of any particular mold to cheese ripening must be determined
by the character of the products of that digestion and the flavors asso-
ciated with it. No pure culture upon a medium previously sterilized
by heat has given a taste resembling that of Camembert cheese.
Cheese made and kept in an atmosphere of chloroform, which pre-
vented mold and bacterial development, refused to ripen. Numerous
cheeses made and not inoculated with molds have uniformly failed to
develop the texture and flavor of Camembert cheese, although such
cheeses have usually become covered with molds of various species.
The type of cheese made and sold in this country as Isigny and Brie,
and sometimes labeled Camembert, which always shows Oidium lactis
associated with bacteria, differs entirely in appearance, texture, odor,
and flavor from Camembert; yet Oidium lactis is capable of neutraliz-
ing the acid of the cheese much more rapidly than the Camembert
Penicillium. Nevertheless the center of such a cheese remains acid
for a longer time than is required to ripen a Camembert cheese, while
the texture of Camembert is not produced. The necessity for the pres-
ence of another agent in this ripening is clearly established.

More than 2,000 cheeses have been made and ripened at this station
with the Camembert mold under varying conditions. Hundreds of
these cheeses have shown repeatedly that cheese so made will assume
in ripening the texture of the best imported article. The Camembert
Penicillium, therefore, is seen to be able to neutralize the acid of the
freshly made cheese and to produce the texture desired, but not the
flavor. It remains to determine whether other molds may not be
equally useful in this process. For comparison cheeses have been
made and inoculated with the Roquefort Penicillium with undeter-
mined species of Penicillium appearing on the record as O, 300, 310, 68,
132. Of these species one, 310, when cultivated upon every medium
used except the cheese duplicated the reactions of the Camembert
mold completely. Its morphology is scarcely distinguishable. It
differs only in that it remains pure white during its entire cycle of de-
velopment, while the Camembert species turns gray-green in age. The
close relationship apparent, together with a promising test, led to its
use upon over 100 cheeses. The breaking down resulting from its
action was widely different. These cheeses were drier, waxy, with a
mealy crumbling layer just under the rind. The physical character
of the results and the flavor produced were so different that the

OAMEMBERT AND ROQUEFORT.

21

cheeses were entirely worthless. This mold was originally isolated
from a market Camembert cheese, where it was found mixed with
others.

The presence of the Roquefort Penicillium may be seen by the spots
of green it produces and may be detected by a sharp, bitter, perhaps
astringent, taste. The texture of the cheese produced is different, and
the flavor when it is present in any large amount is so strong as to be
very objectionable to many. When present in small amounts upon a
cheese it gives a certain sharpness or piquancy to it, such as has been
found often in certain brands of imported cheese, and is sought for by
some buyers.

The species marked O and 300 secrete a bright yellow pigment into
the cheese, which colors every area with which it comes in contact. A
cheese was inoculated with No. 300 and examined when 8 weeks old.
It had produced no trace of the texture of Camembert. The center of
the cheese remained practically sour curd, while the portion for per-
haps one-fourth of an inch under the colony was decomposed.

The species marked 68 has been obtained from cheese from widely
different sources. In cultures upon milk and milk agar it produced
little change. A cheese inoculated with it remained largely sour curd
for two months. The species marked 132 is a very common green
form, appearing in dairy and other cultures. It has given no satisfac-
tory results when grown upon cheese. In this way related species
found in cheese work have been tested in their effects upon cheese and
shown not to produce digestion comparable in physical character to
that demanded in a Camembert cheese and constantly obtained by
the use of the Camembert Penicillium. There seems to be no further
question that this species of Penicillium, among all the molds so far
studied, is the only agent capable of producing the characteristic
texture of the best type of Camembert cheese, with no objectionable
flavors or colors.

FLAVORS.

All attempts to produce the flavor of Camembert cheese in pure cul-
tures upon milk and curd with particular organisms have failed. Here
again we have had to depend upon the use of cheeses so that direct, posi-
tive proofs have not been possible. The value of the indirect or cir-
cumstantial evidence offered must depend upon the completeness with
which all factors have been considered. It has been previously
shown that a cheese may be ripened to the texture of the best Camem-
bert by the action of lactic bacteria and the Camembert Penicillium,
but that it will lack flavor. A series of difficulties are met here.
The typical flavor does not begin to appear until ripening is well
along. This would indicate that the flavor-producing agent or agents
must act upon already partially ripened cheese to produce the par-

22

FUNGI IN CHEESE RIPENING.

ticular end products which give this flavor. But coincident with this
change the acidity of the curd has become so far reduced that bac-
terial development may now occur on the surface at least, and as a
matter of observation few cheeses begin to show flavor until cultures
from their surface show swarms of bacteria of various species. It has
not been practically possible to change these conditions sufficiently to
make cheeses bearing only pure cultures upon the surface. The prob-
lem becomes, then, one of comparative study and the elimination of
the unnecessary factors one by one, rather than the direct produc-
tion of the flavor sought in a single conclusive experiment.

Some organism or organisms must be sought for to produce the
flavor. The appearance of the flavor of the imported article in cer-
tain experimental cheeses at this stage of the investigation led to
their immediate study. This showed that Oidium lactis was abundant
upon these cheeses and emphasized the fact that it had always ap-
peared in cultures from market cheeses. Oidium had been excluded
from many experiments in cheese making because it had been found to
be associated with odors that seemed undesirable, as well as because of
the conclusion of Epstein from his researches, that the presence of
Oidium is uniformly deleterious. The inoculation with spores of
Oidium of a half-ripened cheese entirely lacking flavor produced the
flavor distinctly in a single week, but since bacterial action seemed
always associated with this, further evidence was necessary. Roger
and Epstein have attributed the ripening of Camembert to the action
of certain bacteria without distinguishing that the production of the
texture of the cheese is accomplished by a different agent from the
production of flavor. In their descriptions ripened Camembert is
always referred to as slightly reddish in color, and the appearance of
this color is regarded as an indication of the progress of ripening. In
cheeses selected and forwarded by M. Roger this red color was very
prominent and the red layer was found to consist of myriads of bac-
teria of a few species. Cultures from these cheeses showed that
Oidium lactis was also present in abundance. Numerous tests have
been made with the bacteria found associated with the various
brands of Camembert cheese hitherto without producing the flavor in
any case independently of the molds. The comparative study of
many cheeses from the market and from our own cellars seems to
show that cheeses may have the typical Camembert flavor without
the development of any specific surface growth of bacteria. The
character of the bacterial growth upon the surface appears, therefore,
to be incidental or accidental, though its presence may be necessary
to exclude air, as maintained by Maze 9 in a recent paper.

Cheeses of good flavor have been produced here and also purchased
in the market, which indicate that particular surface appearances are
not essential to the typical flavor. Similarly the introduction into

CAMEMBEET AND ROQUEFORT.

23

new cheeses of species of bacteria found in cultures from the interior
of good cheeses has produced either no effect whatever or disagreeable
flavors. Thus far, therefore, no species of bacterium has been found
capable of producing the Camembert flavor. Although the flavor
question is manifestly still unsettled, we may offer the following sum-
mary of the data at hand upon relation of molds to flavor in Camem-
bert cheese :

(1) Oidium lactis has been found in every brand of Camembert
cheese studied.

(2) It has never been found upon a ripened Camembert cheese which
lacked the flavor.

(3) The flavor has never been found in a cheese without the Oidium.

(4) Every other species with which the flavor seemed obtainable has
been eliminated from one or more experiments without loss of flavor.

(5) Bacteria or other molds do in many cases modify the flavor of
Camembert cheese, but do not seem to be able to produce it inde-
pendently of the mold. There thus arise characteristic secondary
flavors which are associated with the output of certain factories and
which command special markets. These varieties are usually more
highly flavored than what we have regarded as typical.

The essential relation of the Camembert Penicillium and Oidium
lactis to the production of Camembert cheese is, therefore, well estab-
lished. Several mycological questions remain: What are the opti-
mum conditions of temperature and moisture for the use of these
molds in cheese ripening? What are the most practicable means of
cultivating material for inoculation? How can the proper inocula-
tion with these molds be most effectually secured ? What other fungi
occur as contaminating species and how can they be controlled ?

TEMPERATURE.

Since the higher temperatures of the ripening cellar lead more rap-
idly to the development of bacteria, it is necessary to determine the
lowest temperature which will permit mold growth and also enzyme
action. The different species respond quite differently to tempera-
ture. In one experiment eight species were inoculated into slanted
tubes of gelatin and put in a refrigerator where the temperature
varied from 5° to 10° C. Of these the Camembert Penicillium and
two nearly related species, Nos. 68 and 310, grew, but fruited very
slowly, showing an inhibiting effect. The Roquefort Penicillium
grew and fruited normally, as also did Oidium lactis. The species of
Mucor used developed very slowly and fruited only slightly. Two of
the very common green species of Penicillium grew richly. Oidium
lactis grows abundantly in the Brie and Isigny cellars visited. In
these the temperature was 50° to 55° F. (11° to 12° C). Numerous
experiments in the ripening cellar show that the Camembert Penicil-

24

FUNGI IN CHEESE RIPENING.

lium does not grow its best in a room cooler than 60° F. (15° C), and
that to obtain rapid development the room should be slightly warmer.
Until this mold is well established, therefore, it is distinctly an advan-
tage to grow it at a temperature of 65° to 70° F. Repeated experi-
ments have shown that lowering the temperature to 52° to 55° F.
checks the rate of ripening very materially. A difference of less than
10 degrees between two rooms will often make as much as two weeks'
difference in the ripening period of cheeses from the same lot in the
two rooms. A temperature as low as 54° to 55° F., as given in an
article in the Creamery Journal previously referred to, appears to pro-
long the ripening period without contributing any compensating
advantages. A half-ripened cheese was cut, the progress of the
softening of the curd was noted, and the cheese put in a refrigerator,
where it was held for four weeks at 48° F. It was then found to be
completely ripened and perhaps . a little old in one place, but the
changes noted at the end of this period would have been produced
within a single week at 60° F. The cold-storage possibilities sug-
gested by this experiment will be further studied.

Some experiments were made to show the resistance of spores to
heat. The spores of the Camembert and Roquefort Penicillia were
inoculated into gelatin and placed in an incubator. Heating for an
hour and fifteen minutes at 56° C. killed all spores of the Camembert
species. Only a few spores of this mold grew after one hour at the
same temperature, while some spores of the Roquefort Penicillium
grew after two and one-half hours.

HUMIDITY.

The use of very moist cellars and caves in the ripening of this class
of cheeses is practically universal. The richest development of mold
is seen in rooms where the atmosphere is saturated or nearly so. This
appears to be exceptionally true for species like the Camembert Peni-
cillium, which is peculiarly a milk fungus, and in which there is a
large development of thin- walled aerial mycelium. So dependent is
the Camembert mold upon abundance of moisture that it has been
found difficult to secure a rich growth upon the surface of a cheese
which has been drained for two or three days before inoculation. Con-
trary to directions commonly given for ripening these cheeses, which
call for a particular degree of humidity, cheeses have been ripened
successfully in our cellars at the saturation point, as well as at various
degrees of humidity below that. A good illustration of a mold which
has adapted itself to changes of moisture is found in mold No. 198
Upon a fresh cheese in a moist room this mold forms a circular,
ringlike colony of floccose hyphse standing often 8 mm. high upon
the surface of the cheese. In a drier situation, or when the cheese
is nearly ripe and the rind becomes harder and dried, the same mold

CAMEMBERT AND ROQUEFORT.

25

produces conidiophores which barely rise above the substratum, so
that the surface of the cheese is covered by a white, powdery layer
which is practically pure spores. The Mucors are so sensitive to mois-
ture that they scarcely develop upon the cheese, except sometimes
during the first few days, when the surfaces are very wet. They
appear to be unable to withstand the rate at which surface evapora-
tion proceeds in the ripening cellars.

INOCULATING MATERIAL.

The problem of propagation of the Camembert Penicillium for inocu-
lation purposes presents some difficulties. This species bears spores
only upon the surface of the culture medium used, in contrast to the
Roquefort species, which, when grown upon bread, develops spores in
every air space, as we'll as on the surface. To produce spores in quan-
tity, therefore, material must be capable of sterilization and must pre-
sent the largest possible amount of free surface in proportion to the
space occupied. For the preparation of such material, quart fruit jars
have been used. Various styles of crackers have been tried. Most of
these were not successful. The most suitable appears to be the hard,
dry " water cracker." The jar is filled with crackers and dry sterilized
at 140° to 160° C. for an hour or more, better twice on successive days.
The spores may be added directly, or first inoculated into about 100
c. c. of sterile water (acidified with 1.5 per cent of lactic acid usually)
and this poured into the jar and shaken until all the crackers are wet.
Various types of "milk cracker" soften to a pasty mass in this mois-
tening process. The best water crackers are not very satisfactory,
because the mycelium tends to transform bread or cracker into a soft,
gummy mass. The crackers become matted together until they pre-
sent much less actual surface than might be expected. The substi-
tutes tried have been excelsior, hay, and sheets of cardboard wetted
with milk or whey. Although some of these have advantages, they
were on the whole less satisfactory than the water crackers. So far,
therefore, on account of the very different habit of our mold, no mate-
rial has been found so easily prepared and so satisfactory as the
"Schimmelbrot " of the Roquefort cheese makers.

From the point of view of the use of pure cultures the Oidium lactis
is even more troublesome. This mold produces a large proportion,
and in some strains all of its spores as chains below the surface of the
substratum. For pure-culture work Petri-dish cultures have been the
only satisfactory vessels used. Its exceedingly rapid development,
however, makes possible the propagation of a culture from day to day
from the draining boards upon which the cheese is made. These
become heavily coated with a slimy mass of mycelium and spores upon
standing overnight. Direct transfers from them have been used with
apparently no serious trouble from contamination. In fact, so capa-

26

FUNGI IN CHEESE RIPENING.

ble is the Oidium of self-propagation in dairy work that Epstein
declares it to be present in all dairy work. Although Roger in his
published statement does not mention it at all, it was found abun-
dantly upon the cheese forwarded by him to this station. We have
succeeded by careful work in making many cheeses entirely free from
Oidium, but with the ordinary treatment of dairy utensils it appears
constantly in factory practice. It is practically possible to rely to a
considerable extent upon the ability of the Oidium to propagate itself,
as has hitherto been done in the factories.

INOCULATION WITH PENICTLLIUM.

With the Penicillium, however, numerous experiments indicate that
there is much advantage in early and effective inoculation from cul-
tures of known purity. Whether such inoculation must be always
made from specially grown laboratory cultures is questionable. In
factory practice, the making room and the ripening cellar are usually
adjacent. If precautions are taken always to have on hand some
cheeses bearing pure cultures (and the cheese maker must know his
mold so well that there will be no question about it), one or two such
cheeses will furnish enough inoculation material for much newly made
product. This would be indicated by the rough calculation that from
the abundance of the chains of fruit and the size of the spores (0.005
mm. in diameter) probably about enough spores are produced to
cover evenly the surface upon which they grow — perhaps 25,000,000
to the square inch. Very successful inoculation in 75 pounds of milk
has commonly been secured by tapping a Petri-dish culture over the
vat, or by breaking a piece of cracker about an inch square or less and
stirring it into the milk.

The most economical and successful method of inoculation so far
devised has been the use of a sprinkling jar or can. For this purpose
holes 1 mm. or less in diameter in the jar lid are demanded. A small
amount of water is put into the jar, a piece of cracker or cheese covered
with mold is broken into the water, the top is then screwed on, and
the jar thoroughly shaken. The water is then sprinkled upon the
newly made cheese at the time of first turning, so that both sides of
each cheese receive a few drops of water. Excellent results have
been obtained in this way with the smallest amount of inoculating
material and the least requirement of labor and skill. Such a jar
should be emptied and washed immediately after using. The mix-
ture is made fresh each time. Milk may be used instead of water, as
was first suggested and tried by Doctor Conn; but the water has been
found the more easily managed. The practical method for factory
use will probably vary with the conditions and skill of the maker.

©5MEMBERT AND ROQUEFORT. 3JT

VITALITY OF SPORES.

Studies have been made upon the vitality of the spores of the spe-
cies used. This varies greatly in different species. In some of the
most common forms spores have been reputed to remain viable for
several years. Recent studies by Wehmer showed that five species
of Penicillium used in his experiment were entirely dead in labora-
tory cultures at the end of two and one-half years. Cultures of the
Camembert Penicillium grown upon potato in test tubes plugged with
cotton have refused entirely to germinate at the age . of one year.
Other cultures have seemed entirely dead inside of six months. In
fact, the spores of this mold are very thin walled and die very rapidly
when stored. Under such conditions they lose turgidity and become
crenulated or indented. Spores of Monilia Candida and several
others have grown after more than a year in laboratory cultures, but
their germination was much retarded. Oidium lactis seems to be
very easily killed by drying, as would be expected from a species with
such thin-walled spores. The Roquefort Penicillium under some con-
ditions is more resistant, but loses vitality quite rapidly. It is cer-
tain, therefore, that to give the best results material for inoculation
should be fresh and vigorous. Under ordinary circumstances it
would not be desirable to use material more than a few weeks old.

CONTAMINATIONS .

The number of molds found upon market Camembert cheese shows
the need of care in guarding against contamination of cultures. Ex-
traneous molds may come from the milk or from the utensils used or
from the clothes and hands of the workmen. Although the milk is
the primary source of most infections, practical experiments have
shown that if the proper molds are put upon the cheese at the time of
making the troubles arising in this way may be minimized. In fact,
sufficient contamination from this source directly to ruin a cheese is
very uncommon. '

The very habit in some countries of washing or rinsing cheese-
making utensils in whey will account readily for the universal presence
of Oidium lactis and perhaps for many of the bacterial infections that
result in loss. But the source of the most trouble in a cheese cellar
is found to be the cheese maker himself. The cheeses are commonly
exposed upon curing boards and turned and examined in the hands.
In this way spores from molds or bacteria occurring accidentally as
single colonies upon single cheeses are distributed by thousands to
hundreds of cheeses. The product of a factory may almost be identi-
fied in the markets by the contaminations upon the surface of its
cheeses. Certain brands of the cheese always bear Monilia Candida
and commonly one op two other Monilias. A species of Fusarium is
distinctive of another brand, with Acrostalagmus cinnabarinus occa-

28

FUNGI IN CHEESE RIPENING.

sionally present. After numerous experiences with all sorts of con-
tamination this trouble has been practically eliminated from our
experimental work by putting the fresh cheeses, as soon as they are
drained, salted, and comparatively dry upon the surface, into boxes
which are slightly larger than the cheeses, leaving air space and room
for mold to develop normally. In this way fingering is done away
with, the cheese is turned by turning the box, and examined by
removing the lid without touching the surface, so that a colony of
mold appearing upon one cheese is no longer distributed throughout
the cellar.

It is therefore possible to produce cheeses practically free from
molds other than those inoculated upon their surface. Although
such boxing upon a large scale may be practically undesirable on
account of expense, it remains certain that it may be useful in elimi-
nating certain troubles without so large a loss as would come from dis-
carding all infected cheeses, many of which would ripen satisfactorily
but for the danger of spreading obnoxious fungi over great numbers of
cheeses.

ROQUEFORT CHEESE.

The well-known Roquefort cheese is another highly flavored cheese
in which mold has long been known to play a part. In manufacture
this cheese approaches the hard type, but the ripened cheese bears a
closer relation to the soft cheeses. Many complete descriptions give
the details of its making and curing. These need not be repeated
here. Roquefort is by description a goat's or sheep's milk cheese,
made in France principally, though cheese of nearly the same quality
is said to be made in other parts of Europe from mixed cow's and
sheep's milk or from cow's milk alone.

The great popularity of Roquefort cheese makes information as to
the biology of its ripening processes very desirable. To this end nu-
merous specimens of Roquefort have been purchased and analyzed.
The results of this work have been very much simpler than the stud-
ies of Camembert. The ordinary Roquefort cheese before it is sent to
the market is carefully cleaned and covered with tin foil. Its surface
would, therefore, tell very little. When cut it is seen to be traversed
by channels or luxes made by the prickle machine (Stechmaschine)
and by cracks. Every air space is lined with green Penicillium, so
that the cut surface is said to be marbled with green. The texture of
the cheese is reasonably uniform, with every indication that ripening
is simultaneous throughout the cheese or at least approximately so.
Its texture is rather crumbling than waxy, with a tendency to dissolve
readily in the mouth. The taste is a characteristic sharp flavor, in
which a rather high salt content is noticeable. Its odor is strong,
cheesy rather than offensive in any way, except as pronounced

CAMEMBERT AND ROQUEFORT.

29

putrefactive odors are sometimes developed in the rind. Cultures from
the surface often show various species of fungi. There is no regu-
larity about the surface, however, while uniformity of texture and ap-
pearance is universal on the inside. Cultures from the interior show
a remarkable uniformity. In many cheeses examined a pure culture
of a single species of Penicillium has been found. The extremely rare
appearance of any other mold in the cultures has been remarkab e.
Similarly the bacterial content is usually limited to typical lactic forms.
Sufficient analyses have been made to establish clearly that a first-
class Roquefort cheese should conta n only lact-c bacteria and the
Roquefort Penicillium. This Penicillium is often referred to by writ-
ers as P. glaucum and regarded as the common green species, but as it
has very characteristic morphological and physiological characters it
seems best to designate it as the Roquefort Penicillium, even though
it quite often occurs upon other substrata.

The cultures which have been conducted in connection with the
study of Camembert cheese have shown that the Roquefort Peni-
cillium is capable of digesting curd very completely. Here, as in
Camembert cheese, chemical analyses have shown that the derivatives
of casein become almost completely water soluble. Further pure-
culture experiments upon sterile curd have shown that this mold dur-
ing the process of digestion produces bitter flavors during the first few
weeks, but that its continued action changes these to typical flavors
of the Roquefort cheese. Here, then, we have a definite, positive re-
sult. It is thus shown that the Roquefort Penicillium, acting with the
lactic bacteria, is capable of ripening Roquefort cheese without the in-
troduction of other enzyme-producing or flavor-producing organisms.
The investigations of the chemical nature of these changes have barely
been touched upon at this time. In a recent experiment a cheese of
the Roquefort type was made of cow's milk inoculated with the
Roquefort Penicillium and kept in a room at a temperature of about
60° F. At the end of five weeks this cheese was found to have ac-
quired both the texture and the flavor of genuine Roquefort. There
seems to be no doubt that it will be possible to develop methods of
making and ripening that will produce the Roquefort type of cheese
successfully in the United States. Details of making and handling
will then be offered.

CHEESES BELATED TO ROQUEFORT.

Single studies have been made from the Italian Gorgonzola, Eng-
lish Stilton, and Hungarian Brinse (Brindze or Brimse). Gorgonzola
and Stilton are made from cow's milk. Brinse is described as made
from sheep's milk, mixed sometimes with goat's milk. These three
varieties of cheese are found marbled with green Penicillia in pure
cultures, which are unquestionably one or more strains of the Ro'que-

30

FUNGI IN CHEESE RIPENING.

fort Penicillium. In the Gorgonzola and Stilton cheeses examined
lactic species were the only bacteria found. Comparison of the flavors
in these cheeses shows that the differences lie in the qualities of the
materials used in the making and the handling of the cheeses rather
than in the qualities attributable to ripening organisms. It is pecul-
iarly interesting to find the same species of mold in the interior of
ripened cheese in four countries so widely separated, where no efforts
at the use of pure cultures are known to be made. Experiments show
that in every locality so far studied there are many green species of
Penicillium. It is evident, then, that the food material or the condi-
tions, or both, presented by these types of cheese must exert a selective
influence upon the molds, which results in the dominance of the one
species so universally found. This species has been introduced into
experimental cheeses at this station.

AMERICAN BRIE AND ISIGNY.

Cheeses of the type referred to in our previous bulletin as the Ameri-
can Brie have been studied for comparison. This was a collective
term suggested to cover cheese sold under various labels as Brie,
Isigny, Wiener, Miniature, and others, designated commonly by the
retailer simply as Brie. The name "Brie" seems to be applied in the
French dairy literature to a cheese which differs from the Camembert
in the process of making, but ripened by the same fungi and approxi-
mately in the same way as Camembert. The domestic product so far
as examined is quite different, with the exception of the output of one
factory, which is conducted by imported cheese makers. The cheese
met in the eastern markets under these names shows no trace of the
Camembert Penicillium. Numerous brands have been examined in
the market and many hundreds of cheeses have been seen in the cel-
lars of two of the largest cheese companies. Oidium lactis is univer-
sally present upon these cheeses, but its presence goes practically un-
noticed by the makers, since it produces neither color nor aerial
mycelium. All noticeable molds are washed or scraped from the sur-
face of the cheese. The washing produces exactly the best conditions
for the growth of bacteria and Oidium. This treatment results in a
cheese without a very definite fungous rind and with a strong flavor
and smell.

Cultures from this type of cheese, indicate that there is an asso-
ciative action between the Oidium lactis and various species of bacte-
ria. Several species of Penicillium occur as contaminations in these
cellars and sometimes are found upon the cheeses in the market.
Every effort is made to eliminate mold action other than that of
Oidium lactis, which usually passes unrecognized. Cheeses of this
type usually bear rich growths of yeasts, giving a characteristic greasy
feeling to the surface. Exactly what parts these various organisms
play in the production of Brie is as yet undetermined.

CAMEMBERT AND ROQUEFORT.

31

Single studies have shown that Oidium lactis is the dominant mold
upon the surface of some brands of Limburger, brick, and Port du
Salut. There is, then, good reason to believe that this fungus is asso-
ciated with nearly every type of highly flavored, ripened soft cheese
met in the American market.

MOLDS REFERRED TO IN THIS PAPER.

The Camembert and Roquefort molds belong to the hyphomycete
genus Penicillium, which has been characterized by one author — ■

Hyphse broadly effused, creeping; conidiophore branched at the apex in an irregularly
verticillate manner, producing brush or broom-like forms; conidia in chains, hyaline or
bright colored, spherical or elliptical.

This genus of fungi contains a large number of very poorly de-
scribed forms which are everywhere abundant as the "green" or
"blue" mold of the household, the dairy, and the granary. They
form patches upon and just under the surface of the materials upon
which they grow. The patches are composed of delicate threads of
mold, which are matted together, forming more or less cottony sur-
faces, never rising more than a small fraction of an inch above the sub-
stratum. At first these areas are always white, but in most species
the ripening of a crop of spores is indicated by the change to a color
which is usually some shade of green, though this may later give
place to a brown. In a few species other colors appear. These spores
(conidia), or propagating bodies, are minute thin- walled cells averag-
ing possibly one five-thousandth of an inch in diameter, and so light
that they float freely in the air. A breath upon the surface of such a
colony carries away thousands of them, when if held in a proper posi-
tion they may commonly be seen to rise in a cloud. If the colony be
held to the nose and inhaled they give the sensation commonly called
the "smell of mold." They are, then, exceedingly light; they are pro-
duced in immense numbers; they are capable of growing in almost
every conceivable situation, upon anything which is not definitely and
strongly poisonous. Some of these spores are short lived, others cling
tenaciously to their power to germinate. Of the species, probably a
dozen common ones may be expected in any locality, perhaps more.
Our studies have shown that they affect very differently the sub-
stances upon which they grow. It is, then, clearly necessary that by
thorough study of their characters and habits we know the forms we
are to use, and just as important that we know how to get rid and stay
rid, if it be possible, of those we do not want. The discussion of the whole'
group will be reserved for another paper. Here we may describe in
simple terms the two cheese fungi we find important, but it may as
well be acknowledged at the outset that, with the possible exception
of the Camembert species, safe recognition of species without technical
knowledge and cultural study is out of the question.

32

FUNGI IN CHEESE RIPENING.

THE CAMEMBERT MOLD (PENICILLIUM CAMEMBERTl).

The spores of the Camembert mold grow rather slowly in compari-
son with the other molds of the group. They first swell to nearly
double size, and then produce fine threads or hyphse at from one to
three points 6n their surface. Upon a cheese or in laboratory culture
the subsequent growth of these threads forms a colony large enough to
be visible to the naked eye, in ordinary room temperature, in about two
days. Usually in four or five days the colony will have become loosely
white, cottony, about one-half inch or less in diameter, and perhaps

standing one - twentieth of an inch
above the surrounding surface. At
or about this stage the center of this
colony begins to turn a shade of green-
ish gray, which is characteristic of
this species, though one or two other
forms produce colors closely resem-
bling this shade, and difficult to dis-
tinguish from it except to one very
familiar with the colors in question.
This is due to the presence of ripe
spores. Upon the cheese in the cellar
this color often does not appear in less
than a week or even ten days. Micro-
scopic examination shows that the
submerged threads of mycelium of
such a colony do not go deeper into the
solid media than one-sixteenth of an
inch," and that the superficial portion
of the mycelium spreads as fast, or
nearly so, as the part beneath the sur-
face of the substratum. This fungus
grows and fruits for about two weeks —
in some cases this may be prolonged
to three weeks — and at the end of
that period no further growth is to be expected from the primary
colonies, nor, if the medium is undisturbed, is there a secondary
growth from the germination of the spores produced by the first
colony. In case the rind of the cheese is broken so that a fresh
surface is presented, the spores will develop new colonies upon such
areas. A colony, then, produces a single crop of spores and dies,
under ordinary circumstances, and in undisturbed cultures there is
usually no second growth from the spores or from the old mycelium,
although the contrary has been claimed for this fungus by a recent
writer (Maz6 9 ) . A cheese inoculated with this mold will become

Fig. 1. — Camembert Penicilllum (P. ca-
memberti). a, conidiophore showing a
common type of branching and the pro-
duction of basidia and conldia, highly
magnified; 6, a common form showing
much less branching; c, d, /, diagrams
of large fructifications (X 80); g, i, j,
germinating conidia.

CAMEMBEBT AND ROQUEFORT.

33

covered with pure white cottony mycelium in about a week. The
color will then begin to show the gray -green shade characteristic
of the species, which spreads, until at the end of the second week the
entire surface, if left undisturbed, will be colored.

Persistent search has failed to find a single colony in America whose
presence can be attributed to anything but Camembert cheese im-
ported from Europe. The mold may then be regarded as a typical
dairy form which is not well adapted to cosmopolitan conditions and
to the struggle for existence on all sorts of media. In fact, in the
course of laboratory practice involving thousands of cultures, even in
the laboratories of this station, this mold rarely appears as a contam-
ination, although it has been cultivated in quantity and used in the
inoculation of large numbers of cheeses in the same building with the
bacteriological laboratory. Moreover, the spores are easily killed by
heat and retain their vitality for only a few weeks in ordinary cultures
allowed to dry in the air at room temperature.

TECHNICAL CHARACTERIZATION OP THE CAMEMBERT MOLD."

The following technical characterization of Penicillium camemherti
(fig. 1) may be offered, based upon studies made upon the sugar
gelatin and potato agar described in this paper :

Colonies effused, white, slowly changing to gray-green (glaucous); surface of colony floc-
cose, of loosely felted hypha? about 5 fi in diameter; reverse of colony yellowish white;
conidiophores 300 to 800 /i in length, 3 to 4 //in diameter, septate, cells thin-walled, often
collapsing in age, arising as branches of aerial hyphen; fructification sometimes 175 fi in
length, but usually much less, consisting commonly of one main branch and one lateral
sparingly branched to produce rather few basidia, which bear long, loosely divergent chains
of conidia. Basidia 8 to 11 by 2.4 to 3 n; conidia at first cylindrical, then elliptical, and
finally globose when ripe, smooth, bluish-green by transmitted light, thin-walled and com-
monly guttulate, 4.5 to 5.5 /( in diameter, swelling in germination to 8 to 10 f.i. Germ-
tubes one to several. Cells of mycelium about 5 by 20 to 40 /z; liquefies sugar gelatin only
under the center of the colony. Changes blue litmus to red strongly at first, then after four
to six days begins to turn the red back to blue at the center and continues outward concen-
trically until all has become blue. Growing and fruiting period about two weeks. Fruits
only upon exposed surfaces of the substrata — never produces spores in cavities not very
broadly open. Habitat, cheese.

a PeniciUium camemherti (nomen novum). This species is unquestionably the one
referred to by Maz6 in his recent papers as P.. album Epstein. Professor Maz6 was kind
enough to show me the cultures. But the name P. album was already used by Preuss some
fifty years earlier for a species of Penicillium, hence by the rules of nomenclature should not
be used again for a species whose identity with P. album Preuss is not claimed by Epstein.
Upon this ground Lindau, in Rabenhorst's Kryptogamenflora, has changed the name of
Epstein's fungus to P. epsteini Lindau, and extracted from the article written by Epstein a
brief and totally insufficient diagnosis. A careful study of the physiological data given by
Epstein shows that they differ from the data so far found for this species so materially as to
lead to the probability that he was studying another form entirely. I therefore give P.
album Epstein in the list of possible synonymy only, because the name is accepted by Maz6
for what I know to be this species.

84

FUNGI IN CHEESE RIPENING.

THE ROQUEFORT MOLD (PENIOILLIUM ROQUEFORTl) .

The spores of the Koquefort mold grow very rapidly, often produc-
ing new mycelium and ripe spores within thirty-six hours. The colo-
nies are white at the very first, but begin to become green at the cen-
ter within two days in a rapidly growing colony. Such a colony may
become a half inch in diameter in the first two days. The mycelium
is mostly submerged, but very close to the surface, and grows rapidly
outward from the starting point in a radial manner, which is rendered
prominent by certain of the threads lying just under the surface for the
most part, but making loops into the air by rising just above the sub-
stratum for a little way, then reentering the medium again. This
gives a grayish, almost cobwebby (arachnoid) , appearance to the mar-
gin of the young colony. The rate of growth is not uniform in the cir-
cumference of such a colony, which makes the border of a colony
uneven instead of regularly circular, as most species appear. The
superficial portion of the Roquefort mold is almost entirely composed
of the fruiting hyphse or conidiophores, the vast majority of which
arise as branches of submerged hyphse and consequently stand sepa-
rately as short, unbranched threads of approximately equal length,
which gives the surface a velvety appearance. They are usually 0.2
or 0.3 mm. or less in length, say one seventy-fifth of an inch. Such a
colony spreads indefinitely in the substratum, so that the center will
be composed of ripe fruit, while the margin is still actively growing.
In laboratory culture, however, the development is so rapid that the
entire surface is covered within the first few days; then growth ceases.
The mycelium here, as in the Camembert mold, produces but a single
crop of spores, then dies. These spores are a bright green at first, bfft
in a short time become a dirty-brown color in dry culture. The spores
of this fungus are much more resistant than those of the Camembert
mold both to heat and to natural exposures. They will retain their
viability for months in old cultures under the ordinary conditions of
exposure in the laboratory. Upon a cheese this mold produces a
bright green area which extends rapidly. Its action can be detected
in a few days by the bitter taste of the curd near to the mycelium. A
similar taste is, however, produced at least in some measure by other
green forms, so that it is not diagnostic except as between this and the
Camembert species. A colony upon the surface of a cheese becomes
brown in two or three weeks, but colonies growing in the cavities
which are so characteristic of the center of this type of cheese retain
their bright green color for long periods.

This mold is not limited to dairy products, but is widely distributed.
It has been sent to the laboratory from the most distant correspond-
ents. It has been found in silage, and in laboratory cultures from
many substances. It has been found to be the green mold of Stilton,
Gorgonzola, and Brinse, as well as in certain types of prepared cheese

CAMEMBERT AND ROQUEFORT.

35

purchased in the market. Once in a laboratory it stays and seems
to get into everything. In other words, this is one of the cosmopoli-
tan and omnivorous species of the genus. One character seems to
differentiate this mold from most of the others — that is, its power of
growing into and fruiting normally within narrow cavities, such as
appear in cheese. It appears that this character exerts a sort of
automatic (perhaps we may call it a truly "natural") selection which
eliminates all other species from the ripening processes of Roquefort
and related types of cheese.

Fig. 2. — Roquefort Penicillium (P. roqueforti). a, part of conidiophore and of bas of fructification,
highly magnified, showing the production of basidia on the sides as well as at the apex of the
basidiophore; 6, c, other types of branching; d, young conidiophore just branching; e, f, basidia
and the formation of conidia, highly magnified; g, h,j, diagrams of types of fructification as seen
under low power ( x 80) ; k, I, m, n, germination of conidia and new conidia produced directly on
the first hyphse.

TECHNICAL CHARACTERIZATION OF THE ROQUEFORT MOLD."

A technical characterization is offered of Penicillium roqueforti
(fig. 2), as follows:

Colonies quickly turning green, becoming a dirty brown in age, velvety strict, indetermi-
nately spreading by large main radiating, branching hyphse, giving a somewhat uneven or

<• Penicillium roqueforti (nomen novum) . In offering a new specific name for this well-
known fungus, the author is perfectly aware that the mold is often referred to in the litera-
ture as P. glaucum. A careful study of the literature fails to disclose a single description
which indicates that this is identical with the plant described as P. glaucum. As a prelimi-
nary step, therefore, to the proper determination of the green species of Penicillium which
have hitherto been collectively referred to as P. glaucum, this very distinct and easily rec-
ognized form is named from its universal occurrence P. roqueforti.

36

FUNGI IN CHEESE RIPENING.

indefinite margin, which gets a white, fibrous, almost spider-web appearance from its alter-
nation of submerged parts of hypha? with short prostrate aerial loops; reverse of colony yel-
lowish white. Conidiophores arising separately and in acropetal succession from the grow-
ing parts of submerged hyphse (comparatively few from aerial parts, but some), 200 to 300 fi
septate. Fructification 90 to 120 /« or at times 160 n by 30 to 60 /( at broadest place,
usually appearing double by the divergence of the lowest branch ; branchlets (basidiophores)
irregularly verticillate, bearing crowded verticils of appressed basidia 9. to 11 )i by 2.5 /u
with long divergent chains of conidia. Conidia bluish green, cylindrical to globose, smooth,
rather firm-walled, 4 to 5 n in diameter, germinating by a straight tube. Colonies do not
liquefy sugar gelatin, though they soften it somewhat. The fungus changes litmus from
red to blue very rapidly and strongly, almost from the beginning of growth. Fruiting period
short, but one crop of spores upon the mycelium. Cosmopolitan and omnivorous, or nearly
so. Characteristic of Roquefort and related types of cheese.

OIDIUM LAOTIS.

The mold (fig. 3) variously known as Oidium, or Oospera, laetis
is another cosmopolitan organism. This fungus differs widely from
the species previously described. Inoculated into any suitable
medium it grows with enormous rapidity. A single spore (or oidium)
may give rise to several centimeters of mycelium and hundreds of
spores in twenty-four hours. It prefers very moist situations, since
almost the entire mycelium is developed below the surface of the sub-
stratum. It is therefore passed unnoticed many times or produces
changes which are attributed by the observer to bacteria. Descrip-
tion, therefore, must depend upon microscopic characters. The study
of the border of the young colony shows numerous vegetative hyphse
radiating outward. Each of these is found to divide dichotomously
(fig. 3, a, I), so that the border is a crowded series of forking branches.
In the older parts of the mycelium a branch may be produced at each
end of every cell, or several at each end, and these branch indefinitely.
The fruiting branches are mostly produced as outgrowths from the dis-
tal ends of the cells. These extend Upward into the air or remain en-
tirely submerged in many cases. From the ends of these outgrowths
one to several rows of oblong or cylindrical cells begin to be pinched off.
If extending above the surface this gives rise to chains of delicate shim-
mering cells appearing as a powdery covering upon the surface, which
can be seen with a good lens to be arranged in chains. In some strains
of Oidium all of these chains (and some of the chains in all strains) of
spores remain submerged and germinate at once, so that they give
rise to unintelligible mats of hyphae. Oidium produces a very slight
acid reaction to litmus at first, then a strong and continued alkaline
reaction. It liquefies sugar gelatin under the colonies, but does not
extend the area of liquefaction beyond the edge of the colony.
Oidium always and everywhere tested has produced a strong and very
characteristic odor. Once familiar with this odor the worker may
recognize its presence by its spores or oidia, which are hyaline,

CAMEMBEKT AND ROQUEFORT.

37

smooth, cylindrical, 3.5 to 5 }x by 6 to 30 M, varying with the condi-
tions and the substratum and perhaps at times exceeding these limits.
These swell variously and germinate in many ways, so that no germi-
nation characters are definite. Upon some media this mold may be
induced to produce a large growth of aerial mycelium, but the limits
here defined will include the variations to be found upon the usual
culture media.

Oidium lactis is described as universally present on milk and its
products. Epstein even suggests that experiments upon milk and
cheese can not be freed from its presence without sterilizing. The

Fig. 3. — Oidium lactis. a, ft, dichotomous branching of growing hyphre; c, d, g, simple chains of oidia
breaking through substratum at dotted line x-y, dotted portions submerged; e,f, chains of oidia
from a branching outgrowth of a submerged cell; A, branching chain of oidia; k, I, m, n, o, p, s,
types of germination of oidia under varying conditions; (, diagram of a portion of a colony show-
ing habit of Oidium lactis as seen in culture media.

same or almost indistinguishable forms are found upon decaying vege-
tables and fruits, which may give reason for the statement that the
odor produced by Oidium is that of rotten cabbage. There seems to
be good reason for saying that all these forms are but varieties or
strains of the same species. Comparison of several of them shows
that under uniform conditions the morphology of all these forms is
very nearly the same. This is largely true also of their physiological
effects. This mold has been much studied and numerous papers dis-
cuss its nature and physiological effects as well as its relationships.

38

FUNGI IN CHEESE BIPENING.

It will be sufficient to describe here the fungus and to give figures to
assist in its recognition. Its relations to the problems of cheese ripen-
ing have already been indicated.

SUMMARY.

CAMEMBEBT CHEESE.

The acidity of the curd resulting from the action of lactic organ- .
isms reduces where it does not entirely eliminate the growth of objec-
tionable bacteria.

Many species of dairy fungi exert in the course of their development
the power of changing this reaction to alkaline. The Camembert
Penicillium and Oidium lactis possess this power, but not in greater
degree than many other species.

Many species of fungi possess the ability to change curd to a greater
or less extent.

The breaking down of curd by fungi is due in the cases studied to
the production of enzymes.

The texture, appearance, and flavor of curd acted upon by such
fungi are different for different species.

The Camembert Penicillium (P. camemberti) is the only species so
far studied with which the particular appearance and texture sought
in the ripened Camembert can be produced from curd soured by
lactic bacteria without producing any objectionable flavor.

Oidium lactis is always found upon Camembert cheese and so closely
associated with the presence of the flavor as to indicate its agency in
flavor production, though only circumstantial proof of such function
has been possible thus far. The participation of bacteria in flavor
production is not excluded by these results.

Other species of fungi have been shown to produce variations in
this flavor such as have been often found in certain market cheeses.
In this way it is possible to look for the cause of differences in flavor
in contamination of the cultures upon the cheeses. This points
toward the use of pure cultures for inoculation, with the addition of
special organisms if certain variations from what we have regarded as
typical flavor are found to be of value in the market rather than
dependence upon accidental occurrence of the desired species in the
factory.

EOQUEFOBT CHEESE.

In the ripening of Roquefort cheese the only organisms found neces-
sary are lactic bacteria and the Roquefort species of Penicillium.

The Roquefort Penicillium has been shown to possess the power to
reduce the acidity, to digest the curd, and to produce the typical
flavor.

CAMEMBERT AND ROQUEFORT.

39

OTHER VARIETIES OF CHEESE.

The Roquefort species of Penicillium is found in the imported Stil-
ton, Gorgonzola, and Brinse, as well as in Roquefort cheese.

Oidium lactis alone of the forms studied has been found upon the
various brands of Limburger, Brie (American type), Isigny, and
related cheeses found in the market. Other species incidentally
occur, but not uniformly, and such occurrence is avoided as far as
possible by the makers.

BIBLIOGRAPHY.

(1) Conn, Herbert William; Thom, Charles; Bosworth, A. W. ; Stocking, W. A., Jr.,

and Issajeff, T. W. The Camembert type of soft cheese in the United States.
Bull. No. 71, U. S. Department of Agriculture, Bureau of Animal Industry.
Washington, 1905. Also published as Bull. No. 35 of the Storrs Agricultural
Experiment Station, Storrs, Conn., Apr., 1905.

(2) Conn, Herbert William. Bacteria in milk and its products. Illus. 306. pp. Phila-

delphia, Blakiston's Sons & Co., 1903. See p. 268.

(3) Epstein, Stanislaus. Untersuchungen fiber die Reifung von Weichkasen. Arch. f.

Hyg., Bd. 43, Hft. 1, pp. 1-20; Bd. 45, Hft. 4, pp. 354-376. Munich and Leipzig,
1902.

(4) Johan-Olsen, Olav. Die bei der Kasereifung wirksamen Pilze. Cent. f. Bakt., Abt.

2. Bd. 4, No. 5, pp. 162-169. Jena, March 5, 1898.

(5) Constantin, J., and Ray, J. Sur les champignons du fromage de Brie. Compt. rend.

Soc. de biol., Paris, aer. 10, t. 5, No. 16, pp. 504-507. Paris, May 13, 1898.

(6) Roger, Georges. [Article in] Revue hebdomadaire, v. 7, p. 334. Paris.

(7) Margaret, pseudonym. The practice of cheesemaking at home and abroad. The

Creamery Journal, v. 1, No. 11, pp. 313-315. London, July 20, 1905.
{8) Epstein, Stanislaus. See Citation 3, above, p. 373.

(9) Maze, P. Les microbes dans l'industrie fromagere. Ann. de l'Inst. Past., ann. 19,

No. 6, pp. 378-403, June 25; No. 8, pp. 481-493, August 25. Paris, 1905.

(10) Smith, Erwtn F.,and Swingle, Dean B. The dry rot of potatoes, due to Fusarium

oxysporurn. Bull. No. 55, U. S. Department of Agriculture, Bureau of Plant

Industry. Washington, February 16, 1904.
Lang, M., and Freudenreich, Eduard von. Uber Oidium lactis. Landwirthschaftl.

Jahrbuch der Schweiz, Bd. 7, pp. 229-237. Bern, 1893.
Marpmann, G. Beitrage zur Kaseflora. Ztschr. f . angewandte Mikroskopie, Bd. 2, Hft.

3, pp. 68-79. Berlin, June, 1896.

Teichert, Kurt. Beitrage zur Biologie einiger in Molkereiproduction vorkommenden
Schimmelpilzen. Milch-Zeitung, v. 32, No. 50, pp. 786-787. Bremen, December
12, 1903.

Thom, Charles. Some suggestions from the study of dairy fungi. Jrn. of Mycology, v. 2,
No. 77, pp. 117-124. Columbus, Ohio, May, 1905.

o