Historic, archived document

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

), BULLETIN OF THE yy
1 Bec USDEDARIVENT OF AGRICULTURE
eo N Le ES

oO. 202

Contribution from the Bureau of Animal Industry, A. D. Melvin, Chief.
May 12, 1915.

(PROFESSIONAL PAPER.)

THE ALCOHOL TEST IN RELATION TO MILK.

By 8.. Henry Ayers, Bacteriologist, and Witu1AM T. JoHNnson, Jr., Scientific Assistant,
Dairy Division.

INTRODUCTION.

The alcohol test as generally used consists in the mixing of equal
volumes of alcohol and milk. Usually 2 cubic centimeters of 68 per
cent alcohol are added to 2 cubic centimeters of milk and shaken
gently in a test tube. The test is considered positive when a precipi-
tate is formed, or in other terms, when a coagulum is produced.
When a positive test is obtained with fresh milk from a single cow or
small herd, it is generally believed that it indicates an abnormal milk,
due to physiological or pathological conditions in the cow. A positive
test with market milk is supposed to indicate that changes have been
produced in the milk as a result of bacterial fermentations.

According to Fleischmann (11)! the first account of the alcohol
test was published by Martinn in 1890 in the Deutsche (Berliner)
Molkerei Zeitung. It is stated that Martinn used 68 per cent alcohol
with equal parts of milk. Héft (13) in 1898 used the alcohol test to
give anideaof theacidity of milk. He found thatthe higher the acidity
the greater the amount of coagulation by alcohol. In the same year
Petri and Maaszen (24) made use of the alcohol test to determine the
quality of pasteurized milk, and Weber (31) in 1900 studied the alcohol
test in relation to the so-called sterilized milk.

Since 1900 numerous investigators, mostly in Europe, have studied
the alcohol test. Of those who have worked with this test Morres is
probably its most ardent supporter. He strongly advocates the
alcohol test in combination with the alizarin test, which he calls the

- alizarol test. This test will be described later. Morres and the

other advocates of the alcohol test claim that it is of great value,
since it affords a simple and quick means of determining the condition
and keeping quality of milk.

In this country the alcohol test is used by only one large company
which manufactures milk powder. Any milk which shows a precipi-

1See list of citations to literature at end of bulletin,
$2832°—Bull. 202—15——-1

2 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

tate when mixed with equal volumes of 75 per cent alcohol is rejected
by thiscompany. We are not aware that any practical use of the test
is made by any one else in America. In Europe the alcohol test is
more generally used, but we are unable to state to what extent the
test is employed at present, although Farrington and Woll (9) say
that in European creameries and city milk depots the alcohol test is
often applied to every can of ntilk received; milk that is sufficiently
sour to be noticed by the taste will coagulate when mixed with an
equal volume of 70 per cent alcohol.

The Berlin police regulation of 1902 (82) regarding the sale of milk
and cream required that cow’s milk coming from a distance must,
at the time of delivery to the consumer, stand without coagulation
the cooking or alcohol test (mixture of 70 per cent alcohol by volume
with equal parts of milk). According to Devarda and Weich (6),
only fresh milk, which shows no precipitate or only a very fine coagu-
lation with the alcohol test, is accepted in the Vienna market.

OBJECT OF THIS WORK.

The principal object of this work was to determine the practical
value of the alcohol test as a test for the quality of market milk. As
incidental to our primary object, it was our purpose to determine
some of the causes for the precipitation or coagulation of milk by
aroatcls METHOD OF MAKING THE ALCOHOL TEST.

In our work we have used the single alcohol test; that is to say, a
mixture of equal volumes of alcohol and milk. A few investigators
have used the double alcohol test, in which two parts by volume of
alcohol are mixed with one part of milk. In general equal volumes
of 68 per cent alcohol and milk are mixed for the test, but in our
work 75 per cent, 68 per cent, and 44 per cent alcohol were used.
Three tests were made on each sample of milk, 2 ¢. c. of alcohol
being mixed with 2 c. c. of milk in a test tube. The milk was always
at a temperature of from 15° to 20°C. After adding the milk to the
alcohol the tube was shaken and examined for the appearance of a
precipitate. The precipitate appears as flakes the size of which were
recorded as follows: VS for very small, S for small, M for medium-
sized, and L for large.

The different percentages of alcohol were obtained by dilatiaes a
high grade of absolute alcohol with distilled water. Reiss (27) has
shown that alcohol should always be tested for acid before using in
the alcohol test, as acetic acid sometimes found in the alcohol may ~
make the milk sufficiently acid to cause a coagulation with alcohol.
The acidity was determined by titrating 10 c. c. of milk with N/10
NaOH, and is expressed throughout this paper as per cent of normalacid.

Any special methods employed in this work will be discussed when
mentioned in the text,

7 ee

~ So, *

Pi Ee

THE ALCOHOL TEST IN RELATION TO MILK, 3

THE ALCOHOL TEST IN RELATION TO FRESH MILK FROM A SINGLE COW
OR HERD.

While reviewing the literature on the alcohol test it became evi-
dent that the value of the test must be considered from two stand-
points: First, its relation to fresh milk from a single cow or small
herd, and, second, its relation to mixed market milk. Although
our work on this subject deals principally with the relation of the
alcohol test to mixed market milk, we feel justified, after a careful
survey of the literature, first in briefly discussing the test in itsrela-
tion to fresh milk from a single cow or herd.

In the consideration of fresh milk from a single, normal cow we
must omit the changes in milk due to bacterial growth and the in-
fluences of the changes on the alcohol test. The changes as a result
of bacterial activities are of greater importance in the relation of the
alcohol test to the mixed market milk and will be discussed later.

It is evident from the results of other investigators and from our
own tests on milk from a few cows that fresh, normal milk occasion-
ally coagulates with 68 or 70 per cent alcohol when mixed in equal
volumes. Henkel (12) found, after an examination of more than
1,600 samples of milk from a single cow, that 6 showed a coagulation
with 68 or 70 per cent alcohol. This is a very low percentage of
positive results and he concluded that, generally speaking, the milk
of a single animal does not coagulate with 68 or 70 per cent alcohol.
After an extensive study of the alcohol test Auzinger (2) concluded
that the alcohol coagulation of fresh single milk is not so rare as
Henkel had observed. Auzinger (2) also found great fluctuations in
the alcohol test (70 per cent) with milk from single cows. Occa-
sionally milk from the same cow gave a positive test in the morning
and not in the evening, or vice versa. The test might be positive
one day and not the next, but might reappear on the third day.
Sometimes he found that the first and last milk from a single cow
showed fluctuations in the alcohol test. Auzinger also found that
milk from single quarters may coagulate with alcohol independently
of the other quarters, although these cases were rare. He concludes
that the alcohol test in normal milk from a single cow is independent
of the acidity and when the test is positive it is caused by a change
in the milk salts, especially the calcium, in their relation to the milk
proteids. His opinion as to the reason for the occasional coagula-
tion of fresh, normal milk is strengthened by one of his experiments,
in which calcium phosphate was fed toa cow. It was found that the
milk from this cow coagulated with a smaller volume of alcohol or
with a lower percentage of alcohol than did the normal milk.

When fresh, normal milk from a single cow coagulates with 68 per
cent alcohol it is evidently due to some slight change in the com-
position of the milk. What the exact changes are it is impossible at
present to state.

4 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

When we speak of fresh, normal milk we mean fresh milk from a
healthy cow in the middle of the period of lactation. Mulk in the early
period of lactation, that is, colostrum milk, or milk taken late in the
lactation period—‘‘old’’ milk, as it is sometimes called—usually
coagulates with the alcohol test. Henkel (12), Metzger (17), and
also Auzinger (2), found that the milk from a cow in the first of the
lactation period, while apparently normal, may show a positive alcohol
test at irregular intervals. Auzinger (2) believes that the high albu-
men and globulin content of colostrum milk and the calcium salts are
responsible for the positive alcohol reaction.

EFFECT OF COLOSTRUM AND OF “OLD” MILK ON THE ALCOHOL TEST.

In Table 1 are shown the results of the alcohol tests which we have
made on colostrum milk from two cows. Three tests were made,
using 75, 68, and 44 per cent alcohol. The results show clearly that
colostrum milk gives a positive alcohol test and that the stronger the
alcohol the longer the test will be positive. It will be noticed that the
milk from cow 16 gave a positive test with 68 per cent alcohol for 24
days, although the acidity was low after the fourth day. It is evi-
dent from these results and from those obtained by other investi-
gators that the coagulation of milk in the first of the lactation period
by alcohol is largely independent of acidity.

TABLE 1.—Alcohol tests with colostrum milk.

Alcohol test.

Cow | Days after Ncidit
No. | calving. ia ee
75 per cent. | 68 per cent. | 44 per cent.

—— SS ESS ES SS ESS

4 2, 2. 61 14, +L +VS
3 2. 45 +L +L +VS
4 2529 +L + —
5 1. 87 +L +S _
6 1. 80 +M _ _
8 1.50 +M — ~
9 1.70 +M - os

10 1.55 — _
11 1. 52 +S —_ —
12 1.50 +M — _
13 1.31 +8 _ —
19 1.35 — —_ —
21 2.10 _ _ —
22 1512 = _ _

16 1 2. 40 +L +L =
2 2. 20 +L +S —
3 2.70 +M +M a
4 2. 26 +L +8 _
5 1.60 +M +5 -
6 1. 84 +M +S _

13 1.36 +L +M _
15 1.65 +M +8 —_
16 1.53 +M +M ~
19 1.57 +M +8 —
22 1.70 +M +58 —
23 1.60 +M +58 _
24 1.50 +M +S =
25 1.45 +M — =

1In this and succeeding tables the initial letters denoting the degree of the positive (+) tests signify:
L, large flakes; M, medium flakes; S, small flakes; and VS, very small flakes. Minus sign (—) signifies
negative test. ;

a.

THE ALCOHOL TEST IN RELATION TO MILK, 5

In order to determine whether or not the alcohol test would be
positive in a mixed colostrum and normal milk, one experiment was
performed. Colostrum milk from two cows 24 hours after calving
was mixed in various proportions with fresh, normal milk which gave
a negative alcohol test. The results of this experiment, in Table 2,
show that from 80 to 90 percent of colostrum milk had to be mixed with
normal milk in order to cause a positive test with 68 per cent alcohol.
When 75 per cent alcohol was used the test was positive with as low
as 25 per cent of. colostrum milk from cow 5, but when colostrum
milk from cow 16 was used, a mixture of 80 per cent was required to
give a positive reaction with 75 per cent alcohol. It seems evident
from these results that the mixing of colostrum and normal milk
would not cause a positive alcohol test unless a very large percentage
of the milk were colostrum milk.

TABLE 2.—The alcohol test with a mixture of normal and colostrum milk.

Coustrich Hereoutage Rerdentage Alcohol test.
0 0 i
ae at normal | colostrum
; milk. milk. 75 per cent. | 68 per cent. | 44 per cent.
5 10 90 1+1 hole ns
20 80 ai +V8 ak
25 75 4M 4 az
50 50 48 tL x
75 25 +8 we a
90 10 a ne ies
16 10 90 +M +M ZB
20 80 +8 am ty
25 75 Be ez i

1 See footnote under Table 1.

Having discussed the relation of the alcohol test to colostrum milk,
let us consider its relation to milk drawn at the last of the lactation
period, or what is known as “old” milk. Several investigators have
shown that “‘old”’ milk gives a positive alcohol test. It is well known
that milk changes in composition toward the end of the lactation
period, and it is undoubtedly these changes which cause the coagu-
lation with alcohol. While no definite changes have been attributed
to the positive alcohol reaction, it is believed by some to be due to
the high content of solids (not fat). Henkel (12), however, found
that this could not explain in all cases the coagulation by alcohol.
Auzinger (2) believes that on account of the variation of solids (not
fat) the alcohol test has no significance in milk from ‘ old” milk
cows.

SUMMARY OF CAUSES FOR POSITIVE TESTS IN MILK OF SINGLE COWS.
It is apparent that fresh milk from a single cow may occasionally

give a positive alcohol reaction with 68 or 70 per cent of alcohol.
Colostrum milk gives a positive reaction, and the same is true usu-

6 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

ally of ‘old’? milk—that is, milk from a cow in the last of its lac-
tation period. ;

The causes for a positive alcohol test may be summarized by the
opinion of Ernst (8) who states that a positive alcohol test of fresh
milk from a single cow indicates a physiological or severe patho-
logical condition of irritation of the milk glands. There is, how-
ever, a difference in the opinions of various investigators as to the
reaction of the alcohol test to pathological conditions of the udder.
Ruhm (28) noticed the alcohol test in milk from cows with in-
fected udders. In some cases he found the test was positive during
the infection and frequently a positive test was observed for three
or four weeks later when the milk had a normal appearance and
taste. He points out that in udder infection the milk may vary in
many ways, and in consequence the alcohol test varies. Auzinger
found that there was no relation between streptococci in infected
udders and the alcohol test and that a positive test is produced
through chemical changes in the secretions. Rullmann and Tromms-
dorff (29) also observed a positive alcohol reaction in milk from cows
with infected udders, but according to these authors the alcohol test
shows no definite relation to the leucocyte count. They point out
that the variation in ash salts and high albumin content probably
influences the alcohol test. Campbell (5) also believes that the alcohol
test is of value in determining the diseased condition of the udder.
Besides udder infection Auzinger (2) states that the general infections
and infections of the vaginal canal may cause a positive alcohol test;
also that milk from cows which have aborted may coagulate with
alcohol. Metzger (17), however, after a study of the alcohol test
with milk from sick cows concludes that the milk from them shows
no relation between the acidity and alcohol test. According to this
author fever had no influence on the acid and alcohol tests. There
was no relation between tuberculosis of the animal and the alcohol
test. When animals were very lean from disease the milk inclined
toward coagulation with alcohol. Infectious inflammation of the
vagina was without influence on the test. Infection of the uterus
shows almost regularly with the alcohol test, but not without ex-
ception. Metzger also found that there was no relation between the
alcohol test and various forms of indigestion. He points out that
the chief value of the test lies in its use for the freshness of milk.

We have not had an opportunity to study the alcohol test in its
relation to the milk from sick cows, but from a study of the literature
on this subject we are inclined to believe that the alcohol test would
be of but little value as a routine test of the milk from a single cow
or from a small herd. If the alcohol test were used regularly to test
fresh milk of single cows a positive reaction would indicate some
change in the milk from normal. Subsequent examination of the

f

eee

THE ALCOHOL TEST IN RELATION TO MILK, 7

cow might reveal some pathological condition, or there might be
some physiological reason for a slight variation in the composition
of the milk. If the test were performed on the milk from a few cows
a positive reaction might be caused, as Auzinger (3) believes, by the
mixing of milk which is changed by physiological or pathological
conditions with milk from normal cows. If there were a large per-
centage of abnormal milk which gave an alcohol test with a coagula-
tion with large flakes, the mixed milk might show a positive alcohol
test in which the coagulation would be in the form of small flakes.
When mixed milk from a large number of sources gives a positive
alcohol test it must be interpreted in an entirely different manner,
and this leads us to another phase of the subject.

THE ALCOHOL TEST IN RELATION TO MARKET MILK.

Since 1900 a considerable number of papers have appeared on the
use of the alcohol test in its relation to market milk. According to
Kirchner (15), Morres in 1905 showed that the alcohol test was of
value for determining the keeping quality of milk and indicating its
acidity. Reiss (26) in 1906 pointed out the practical value of the
test, and Morres (18) again in 1909 showed the value of the alcohol
test as a means of determining the keeping quality of milk. He added
2 c.c. of milk to 2 c.c. of 68 per cent (by volume) alcohol, and states
that if the milk coagulates with alcohol then decomposition has
already started and the extent is shown by the size of the flakes.
If the precipitate is in fine flakes then the acidity corresponds to
4 degrees Soxhlet; however, the coagulation may not be due to an
increase in acidity, but may be due to the action of rennet-forming
bacteria. In later work Morres has combined the alcohol and
alizarin tests. This will be discussed later. Morres considers that
the coagulation of mixed market milk is due largely to the formation
of acid or the action of rennet-forming bacteria or to a combination
of both. Henkel (12) concludes from his work that the alcohol test
does not afford a proper means for determining acidity, but that the
value of the test lies in the fact that it gives a knowledge of the souring
and other changes in the properties of milk or in variations from the
normal properties which the acid test does not show. Other investi-
gators believe that the alcohol test is of value only as a preliminary
test. Fendler and Borkel (10) after a large number of tests to

_ determine the relation of the acidity of milk to the alcohol test con-

cluded that the double test with 70 per cent alcohol was not a proper
criterion for the freshness of market milk, including infants’ milk
and superior grades of milk. They state that the double test using
50 per cent alcohol is suitable as a preliminary test for food in-
spectors, but the milk should be submitted to further tests. These
authors also found that no consistent relation existed between the

8 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

alcohol test and the acidity of milk. Rammstedt (25) also agrees
with Fendler and Borkel, so far as he found, that no consistent
relation existed between the alcohol test and the acidity of milk.
He considers that the test gives preliminary knowledge of the hygienic
quality of a milk.

It is evident from the literature that in a mixed market milk the
acidity plays a part in connection with the alcohol test, so that in
considering the factors which influence the test we may first take up
the question of acidity.

THE INFLUENCE OF ACIDITY ON THE ALCOHOL TEST.

In. our first experiments the acidity of milk was raised by the
addition of N/10 lactic acid. The results of two experiments recorded
in Table 3 show that a very slight increase in the acidity of milk may
cause a positive alcohol test with 75 per cent and 68 per cent alcohol,
but a considerably higher acidity is required to cause a positive test
with 44 per cent alcohol.

These results show clearly that the alcohol test is sensitive to slight
changes in acidity when these changes are produced by the addition
of lactic acid. Since an increase in acidity will cause a positive
alcohol test it is evident that the growth of acid-forming bacteria in
milk will cause a positive test.

TABLE 3.—Influence of acidity on the alcohol test.

N/10 lactic Alcohol test.

acid adHse
Pane Acidity.
of mili 75 per cent. | 68 per cent. | 44 per cent.
C5C-
0 1.81 — - —
0.5 1.88 = —
1.0 1.94 11M +M =
3.0 Zook +L +L a
3:5 2.38 +L +L _
4.0 2.47 +L +L +M
0 1.70 - _ a
oD iy +M — —
1.0 1. 84 +M +8 —
2.0 2.00 +L +L —
3.0 2.20 +L +L —
3. 4 25 +L +L _
oe) 2326 +L +L +few VS
4.0 Deal +L +L +few L

1 See footnote under Table 1.

In order to determine the relation between the number of acid-
forming bacteria, the acidity, and the alcohol test, two experiments
were performed, using a pure culture of a lactic-acid-producing organ- .
ism. The culture was inoculated into sterile skim milk and incubated
at 37° C. <A bacterial count was made while the acidity and the
alcohol test were determined at the same time. From the results
shown in Table 4 it may be seen that in Experiment I the alcohol
test was negative even after seven hours of incubation. At that time

THE ALCOHOL TEST IN RELATION TO MILK, 9

the acidity had increased from 1.98 to 2.14, and the bacteria from
from 82,000 to 15,100,000 per cubic centimeter. It is interesting to
note that an extensive multiplication of lactic-acid-forming bacteria
may occur without causing a positive alcohol test. In the second
experiment, also shown in Table 4, a heavier inoculation was used, and
it will be seen that the milk at the beginning of the incubation period
contained 480,000 bacteria per cubic centimeter. The 68 per cent
alcohol test was not positive until the bacteria had increased to
31,400,000 per cubic centimeter.

These figures show that when a pure culture of lactic-acid-forming
bacteria is grown in skim milk there must be a very great increase in
order to produce acidity enough to cause a positive alcohol test. In
these experiments there were no positive alcohol tests until the bac-
teria had increased from less than 500,000 to over 16,000,000 per cubic
centimeter. From these results it is apparent that the growth of
acid-forming bacteria in milk may, through the formation of acid,
cause a positive alcohol test. However, when there is sufficient acid
produced to cause a coagulation with 68 per cent alcohol the number
of acid-forming bacteria would be very high.

TasiLe 4.—IJnfluence on the alcohol test of acid produced by the growth of a pure culture of
lactic-acid bacteria.

Alcohol test.
Age of Bacteria co es
PaO pOECHbIC. eA ciditya! | ee

RE LEO oy AGUS 75 per cent. | 68 per cent. | 44 per cent.

' Experi-
ment No.

I 0 2, 000 1.98 = a me
2 dt 3000 3 ete ss tcc = = ci

4 1,510, 000 2. 06 = = vst

5 ,300, 000 2.08 = = z

6 | 11,700,000 2.09 = = Ms

7 | 15,100,000 2.14 fs = as

II 0 480, 000 1.94 = ES a
2 15060)000. 15.222... 3 = = =

4 7,500, 000 2. 08 = ae a

5 | 16,100,000 2.08 148 = Ls

6 | 31; 400,000 2.30 +1 +M i

7 | 46,000,000 2. 47 ie +L =

1 See footnote under Table 1.
EFFECT OF PHOSPHATES.

We have so far discussed in a general way the effect of increasing
the acidity of milk by the addition of lactic acid and by the generation
of the acid in milk. Since the acidity of milk when titrated with
phenolphthalein is due partly to acid phosphates, it will be of interest
to show the effect on the alcohol test of the increase in acidity by
acid phosphates. In Table 5 are shown the results of a few tests,
using sodium and potassium acid phosphate. Various amounts of
a 5 per cent solution of these salts were added to 50 c. c. of milk. It
will be seen from the table that when the acidity was increased by
sodium acid phosphate from 2.15 to 3.33 the alcohol test with 75

82832°—Bull. 202—15——2

10 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

per cent alcohol was positive. At an acidity of 4.27 the milk coagu-
lated with 68 per cent alcohol but the flakes were very small. In
order to cause a coagulation with 68 per cent alcohol with medium-
sized flakes 1t was necessary to increase the acidity to 6.16. When
potassium acid phosphate was used the results were about the same.

These results show that it is possible by increasing the acidity of
milk with acid phosphates to cause a coagulation with the alcohol
test, but the acidity has to be increased to a high degree and there
would never be enough acid phosphate in a mixed market milk for
it to be entirely responsible for a positive alcohol test.

TABLE 5.—Influence on the alcohol test of the addition of acid phosphates to milk.

Sodium-acid phosphate. Potassium-acid phosphate.
Amount Alcohol test. Amount Alcohol test.
of 5 of 5
per cent per cent
solution ; solution ;
ofacid | Acid- ofacid | Acid-
phos. ity: | 75 per | 68 per | 44 per nes ity. | 75 per | 68 per | 44 per
sadn 46 cent. | cent. | cent. |. Pada to cent. | cent. | cent.
50 c. c. of 50 c. c. of
milk. milk.
Gic: Cre.
es ee a — — 0: (eee = _
1 2.75 o = 1 = = =e
2 3.33 |1+M - - 2 3.13 |1+M - =
3 | 427 | +M |4+VS] — 3 | 4.00 | +M | +VS | —
5 | 5.50 | +M vs} — 5 5.20 | +M | +VS }] —
| 6 | 6.16 +M | +VS 6 +M | +M +VS

1 See footnote under Table 1.

In some cases where we increased the acidity of milk by adding
lactic acid it was noticed that a very slight increase in acidity caused
a positive alcohol test. At other times the acidity had to be in-
creased to a considerable extent before the milk coagulated with
alcohol. It occurred to us that the explanation for these differences
might be that there were different amounts of dibasic phosphates
present in milk and that the acid converted the dibasic phosphate into
acid phosphate, which increased the acidity but did not. cause a posi-
tive alcohol test. In order to test this theory one experiment was
performed, the results of which are shown in Table 6. Two flasks of
milk were used, each containing 50 c. c. of milk. One flask was left
normal and 0.5 per cent dibasic sodium phosphate was added to the
other. Various amounts of N/10 lactic acid were then added to each
flask. As may be seen from the table, when 3 c. c. of N/10 lactic
acid was added to the normal milk, the acidity-was 2.37 and the
alcohol test was positive with both 75 per cent and 68 per cent alco-
hol. The flakes were large and medium, respectively. The same
amount of acid added to the milk with dibasic phosphate increased

THE ALCOHOL TEST IN RELATION TO MILK, ia

the acidity to 2.55 and yet the alcohol test was negative. When 7
c. c. of N/10 lactic acid was added to the normal milk, the acidity was
increased to 3.00 and the milk coagulated with large flakes with all
the different percentages of alcohol. When 7 c. c. of N/10 lactic
acid was added to milk with dibasic phosphate the acidity was
increased to 3.05 and only the 75 per cent alcohol test was positive,
and the coagulation was in the form of small flakes. When 8 e. e.
of acid was added to the milk with dibasic phosphate the acidity
was 3.19 and the alcohol test with both 75 per cent and 68 per cent
alcohol was positive. It was found by titration that 10 c. c. of a
0.5 solution of dibasic phosphate required 1.56 c. c. of N/10 lactic
acid to convert the dibasic into the monobasic phosphate; therefore
50 c. c. of milk containing 0.5 per cent of dibasic sodium phosphate
would require 7.8 c. c. of N/10 lactic acid to convert the dibasic
into the monobasic phosphate. It will be seen from Table 6 that
when from 7 to 8 c. ec. of N/10 lactic acid was added to the milk
with dibasic phosphate, the alcohol test became positive; that is, when
the dibasic phosphate had been converted into monobasic phosphate
then further increase in acidity caused a positive alcohol test.

As a very general explanation of this result it may be said that when
acid is added to milk it converts the dibasic phosphate into the
monobasic phosphate. It follows that the acid and also the monobasic
phosphate probably affect the casein and thereby change it into a
condition in which it is possible to precipitate the casein by alcohol
and cause a positive test. This action on the dibasic phosphate prob-
ably explains in part the positive alcohol tests with different low
acidities.

TABLE 6.—Influence on the alcohol test of the addition of dibasic phosphate to milk.

Wiormaletitics Normal milk + 0.5 per cent

N/10 dibasic sodium phosphate.
lactic
aci
added Alcohol test. Alcohol test.
to
50 ¢c. c. aie: zy
of Acidity. 75 68 44 Acidity. 75 68 44
milk. per er per per | per per
cent. cent. cent. cent. | cent. cent.
CHC: |
0 1.85 — — 2. 03 — — --
1 2.01 {1485 -- — 2.13 -- — --
3 PEAY | +L +M — 725 (995) _ _- _
5 2. 63 +L +L +5 2.81 -- _- _
6 2. 80 +L +L +M 2. 92 — a —
ie 3. 00 +L +L +L 3.05 | +S _
C2 TN aS eS AU ae a NP 3.19 | +L +M —
CE oe I ea eg ee a ee 3.43 | +L +L —
TE Ae oR ete SW A We 3.97 | +L +L +L

1See footnote under Table 1.

-

12 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.
RESULT OF MIXING SOUR AND NORMAL MILK.

Since a positive alcohol test may~be produced by increasing the
acidity, several investigators have pointed out that a mixture of sour
and normal milks will give a positive test. The amount of sour milk
which can be added to fresh milk without causing a positive alcohol
test will, of course, depend upon the acidity of the sour milk. In
one experiment, the results of which are shown in Table 7, various
percentages of sour, raw, and pasteurized milk were added to fresh
milk. The addition of 1 per cent of sour milk caused a positive test
with 75 per cent alcohol, 2.5 per cent caused a positive test with 68
per cent alcohol, and the addition of 10 per cent of sour milk was
necessary to cause a positive test with 44 per cent alcohol.

It must be taken into consideration in this experiment that the
sour milk had a high acidity. If the acidity had been low a much
higher per cent could undoubtedly have been added to the fresh
milk without increasing the acidity sufficiently to cause a positive
alcohol test.

TasBLE 7.—The alcohol test with a mixture of normal and sour milk.

Addition of sourraw milk. Acidity 10.23. Addition of sour pasteurized milk.

Acidity 9.87.
Per | | Alcohol test. Per Alcohol test.
cent of Acid- }—————__"_—_- cent of 0 ——————————————
sour ite Ls sour ity
milk i 75 per | 68 per | 44 per | milk * | 75 per | 68 per | 44 per
added. cent. | cent. | cent. | added. cent. | cent. | cent.
0.0 1. 84 — = See Sees ae one _ _— _-
1.0 1. 93 145 - —_ 1.0 1. 88 +S _ =
2.5 2. 06 +M +M - 2.5 2. 00 +M +M =_
5.0 2.30 |} +L +L _ 5. 0 2. 29 +L +L _
10.0 2.70 | +L +L +M 10.0 2. 68 +L +L +VS

1 See footnote under Table 1.

In connection with the relation of acidity to the alcohol test the
question arises as to whether or not the acidity of a sour milk can be
neutralized so that the alcohol test will be negative. Some investi-
gators have shown that the neutralization of the acidity does not
cause a positive test to become negative, although the size of the
flakes in the coagulation is somewhat reduced. We have tried one
experiment in which various amounts of normal lactic acid were
added to fresh milk, after which the acidity was reduced to the original
acidity by the addition of sodium hydrate. From the results which —
are shown in Table 8 it will be seen that when the acidity was in-
creased to 4.3, then neutralized to 1.90, the 68 per cent alcohol test
was positive. The positive alcohol tests with 68 per cent alcohol
could be made negative at acidities below 4.30 by reducing to about
the original acidity of the normal milk.

THE ALCOHOL TEST IN RELATION TO MILK, 13

TaBLE 8.—Effect on the alcohol test produced by neutralizing the acidity of milk.

Amount Alcohol test. Alcohol test.
of normal
lactic Acidity
acid Acidity. after neu-
added to 75 per | 68 per | 44 per | tralizing.| 75 per | 68 per | 44 per
50 ¢. ¢. cent. | cent. | cent. cent. | cent. | cent.
of milk
Ciuc)
0.0 1. 84 = = mn Wee cece = |Socace sin nemo elation
0.1 2.08 |1+M +M _ 1.80 _ - —
0.3 2. 42 +M +M — 1.81 _ - —
0.5 2.99 +L +M +VS 1. 64 a _
0.8 3.54 +L +L +M 1. 60 +VS — _
1.0 3.94 +L +L +L 1. 60 +5 — _
15 4.30 +L +L +L 1.90 +L +-M. =

1 See footnote under Table 1.
EFFECT OF HEAT COMBINED WITH ACIDITY.

As a matter of general interest we may mention the effect of heat-
ing milk which gave a positive alcohol test. Auzinger (2) found
that a milk which gave a positive test at 15° C. sometimes did not
give the test when heated for 30 minutes at 60° C. Then again he
found that the test might remain positive in milk heated to boiling.
In Table 9 is shown the result of an experiment showing the effect
of heat on the alcohol test with milk of two different acidities. No
effect of heat was found on the sample of milk with an acidity of
2.30, but there was a marked effect when the acidity was lower.

We have no explanation to offer for this negative result of the test
when the acidity is low. This action of heat might be of importance
when the alcohol test is applied to pasteurized milk.

TABLE 9.—Effect of heat on the alcohol test which is positive on account of acid action.

Alcohol test.

Milk heated Acidity, 2. Acidity, 2.30.
to— is iter) Caer Ok

75 per | 68 per | 75 per | 68 per
cent. | cent. | cent. cent.

°

OE etre a ciate S +M +VS +L +L
SO sees +5 - +L +L
SO recess. _ — +L +L
LOO Sse 5: — - +L +L

1 See footnote under Table 1.

14 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.
INFLUENCE OF THE ACTION OF RENNET.

The relation of the alcohol test to the acidity of milk shows that
acidity is one factor which may cause a positive alcohol reaction,
but from the work of other investigators it is evident that it is not
the sole cause. Morres throughout his papers points out that the
alcohol test may be caused by an acid fermentation or by a rennet
fermentation or by a mixture of both fermentations. :

In order to determine the effect of rennet action in relation to the
test, we first tried the effect of prepared rennet. Four flasks of fresh
milk were used and to each a different percentage of rennet was
added. The milk in each flask was tested by the alcohol test at
intervals of one hour. It will be seen from Table 10 that four
different percentages of rennet were used, ranging from 0.00005 per
cent to 0.0015 per cent. The acidity of the milk increased during
the four hours from 1.64 to 1.70; therefore the influence of acidity
can be neglected, since it is only a slight change.

The results show that the action of rennet in milk may produce
changes which cause a positive alcohol test and that two main factors
are of importance, namely, the amount of rennet and the length of
time the rennet has to act. Undoubtedly a third factor must be
taken into consideration; that is, the temperature at which the milk
is held. In our experiments the milk was held at room temperature.
These results confirm those obtained by other investigators and
indicate that the action of the rennet-forming bacteria might cause
a positive alcohol reaction. |

TABLE 10.—Influence of rennet on the alcohol test.

{

Alcohol test.
Hours. ee t
: 75 per 68 per 44 per
cent. cent. cent.
Per cent.
- 00005 _ _ —
. 00025 — — —
- 0005 — —
. 0015 _ — —
1 - 00005 _ — _
. 00025 _ = _
. 0005 _— = --
. 0015 14+M +M —
2 . 00005 — _ _
. 00025 — — a
- 0005 +M +M —
. 0015 +L +L +M
3 - 00005 _ — -—
. 00025 +M +M —
- 0005 +M +M —
0015 +L +L +M
4 00005 _ = —
00025 +L +L L
0005 +L L +VS
0015 (?) () 2)

1 See footnote under Table 1.
2 Milk curdled by rennet.

ee"

PARE BO

THE ALCOHOL TEST IN RELATION TO MILK, 15

In order to show the effect of rennet of bacterial origin, the action
of a pure culture of a rennet-forming organism was studied. Two
flasks of sterile skim milk were inoculated with different amounts of a
pure culture of a rennet-forming organism. These flasks were
incubated at 37° C., and the bacterial increase was determined at
definite intervals, together with the alcohol test. The results are
shown in Table 11. From a study of the table it is evident that
rennet-forming bacteria will cause a positive alcohol test, but there
must be a large bacterial increase to produce rennet enough to cause
a positive test.

The acidity was also increased during the incubation, but we believe
this acidity played a minor part in causing the positive alcohol test.

TABLE 11.—Jnfluence on the alcohol test of rennet produced in milk by the growth of a pure
culture of a rennet-forming organism.

Alcohol test.
Experi-| Age of | Bacteria :
ment cul- per cubie pea
No. ture. | centimeter. Y- | 75 per | 68 per | 44 per
cent. cent. cent.
Hours.
34, 000 1.98 _ —
2 62.000 |. cee = = =
4 4, 700, 000 2.02 — — =
5 | 9,000, 000 2.06 | 1+L —- _
6 | 21,000, 000 2.10 +L +L —
7 | 31,000, 000 Pili +L +L +L
II 0 147, 000 1.94 — - —
2 200000) Eee see — _— —
4 | 15,000, 000 2.10 14 +L +8

1 See footnote under Table 1.
DIFFERENTIATION BETWEEN ACIDITY AND RENNET ACTION.

The fact that reducing the acidity did not cause a negative alcohol
test, as mentioned above, led us to believe that it might be possible
to differentiate between a positive alcohol test caused by acidity and
one caused by rennet action. In order to determine whether this
was true two flasks of sterile skim milk were prepared. One was
inoculated with a pure culture of a lactic-acid-forming organism and
the other with equal amounts of a pure culture of lactic-acid bacteria
and rennet-forming bacteria. The two flasks were then incubated at
37° C. As may be seen from Table 12, the milk containing the
lactic-acid bacteria had an acidity of 2.23 after 3 hours’ incubation
and the test was positive with both 75 per cent and 68 per cent
alcohol. When the acidity was reduced to 1.49 all the alcohol tests
were negative. The milk containing a mixed culture of lactic-acid
bacteria and rennet-forming bacteria after 3 hours’ incubation had
an acidity of 2.32 and the alcohol test was positive with 75 per cent
and 68 per cent alcohol. In both cases the coagulation was in the
form of large flakes. When the acidity was reduced to 1.70 the

16 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

alcohol test remained positive,’ although the size of the flakes was
reduced. This milk after 54 hours’ incubation had an acidity of 4.38
and the milk coagulated with large flakes with each percentage of alco-
hol. When the acidity was reduced to 1.49 the alcohol test remained
positive, the only change being with the 44 per cent alcohol, in which
case the size of the flakes was reduced. This experiment was re-
peated, as will be seen from Table 12, and the results confirmed
those of the first experiment. ‘These results indicate that it may be
possible to differentiate between an acid and an acid-and-rennet
fermentation in milk, provided the acidity is not high.

TABLE 12.—Differentiation between an acid and a mixed acid-and-rennet fermentation by
neutralizing the acidity and using the alcohol test.

After incubation at 37° C. for 3 hours. | After incubation at 37° C. for 53 hours.

Ex-
peri- | Pure culture Alcohol test. Alcohol test.
ment of—
No. Acidity. Acidity.
75 per | 68 per | 44 per 75 per | 68 per | 44 per
cent. cent. cent. cent. cent. cent.
I. | Lactic - acid 2. 23 14], +L — Boal +L +L +L
bacteria.
Neuuelized to _ _ — Neutralized to — =_ -
1.66. 49,
Mixtureof 2.32 +L +L _ 4.38 +L +L +L
rennet-
forming
and lactic-
acid bacte-
ria.
Sisuizeuized to +S |] +VS — Neutralized to +L +L +S
1.70. 1.49.
BP artic acid hiceee cc each cee eee cc ees locos eeeeeoeeeee 22.70 +L +M —
bacteria.
Mixtureof 32.00 +L +M _ 23.42 +L +L +L
rennet-
forming
and lactic-
acid bacte- ¢
Tia.
Neutralized to — — — Neutralized to +L +M]+VS
60.

| 1.80. 1:
}

Note.—Acidity of ncrmal milk in experiment I, 1.75; in experiment II, 1.78.

1 See footnote under Table 1.
2 After 43 hours incubation at 37° C.
3 Acidity after adding pure cultures to milk and before incubation.

EFFECT OF HEAT COMBINED WITH RENNET ACTION.

Earlier in this paper we have shown the effect of heat on the alcohol
test with milk of high and low acidity, and as a matter of general
interest the effect of heat on the alcohol test produced by rennet
action may now be considered. The results of two experiments
shown in Table 13 explain themselves clearly. Sufficient rennet was_
added to two samples of milk to cause a positive alcohol test with 75
and 68 per cent alcohol. The milk was then heated, and it was found
that at 90° C. the milk no longer gave a positive alcohol test. Both
experiments showed the same results.

3
:
j

THE ALCOHOL TEST IN RELATION TO MILK, ley

- TasBie 13.—Effect of heat on the alcohol test with milk in which the positive test is due to
rennet action.

Alcohol test.
Milk heated | 75 per cent. 68 per cent.
to—
| Experi- Experi- Experi- Experi-
ment I. ment II. ment I. ment IT.
Pas Heated: +L +L +L +L
Sa eee eee +L +L +L +L
60 Cen meee ae: +L +L +L +L
FOR sai +L +L L +L
BOL eee si +S +VS +S +VS
OQ 8 Aecey eh = — — =

Novte.—<Acidity of milk in experiment I, 1.82; in experiment II, 1.84.
1 See footnote under Table 1.

The results which we have shown on the effect of rennet action in
relation to the alcohol test confirm the work of other investigators,
and it is evident that the rennet-forming group of bacteria in milk
can play an important part in the production of a positive alcohol test.

INFLUENCE ON THE ALCOHOL TEST OF CARBON DIOXID IN MILE.

There are probably numerous minor’factors which influence the
alcohol test with market milk. While the two principal factors are
probably acidity and the effect of rennet action, it is beheved by some
investigators that carbon dioxid plays a more or less important part.
Auzinger (2) found that milk one hour old which gave a positive
alcohol test gave a negative test after it had been held for 18 hours.
He believes that carbon dioxid might be partly responsible for such a
change. |

We have passed carbon-dioxid gas into milk many times and have
always been able to cause a positive alcohol test. In one experiment
carbon-dioxid gas was passed into milk until the acidity was 2.36
(titration in cold milk with phenolphthalein as an indicator), and a
positive alcohol test was obtained with 75 per cent and 68 per cent
alcohol. As is shown in Table 14, this milk was heated at different
temperatures up to 100° C. With the increase in temperature the
acidity was reduced, due probably to the expelling of the CO,. Barillé
(4) has shown that carbon dioxid forms a very unstable compound,
which he calls carbono phosphate of calcium and is easily broken up
by heat. When the temperature reached 70° C., the alcohol test with
68 per cent alcohol was negative and the acidity had been reduced
from 2.36 to 2.05. At 90° C. the acidity was 1.91 and the alcohol
test was negative with 75 per cent alcohol.

There can be no doubt as to the fact that carbon dioxid may cause
@ positive alcohol test, provided there is a large enough amount in the

82832°—Bull. 202—15——3

18 BULLETIN 202, U. 8S, DEPARTMENT OF AGRICULTURE.

milk. In order to determine how much carbon dioxid was required
to cause a positive test with 68 per cent alcohol, the gas was passed
into a flask of fresh milk until a positive alcohol test was produced.
The amount of CO, in this milk and in the original milk was then
determined. It was found that the normal milk contained 0.76 per
cent of CO, by volume at 32° C., and the milk through which the gas
had been passed contained 13.05 per cent of CO, by volume. [fn this
experiment it was necessary to merease the CO, content to 13.05 per
cent by volume in order to cause a positive alcohol reaction with
68 per cent alcohol. According to Kastle and Roberts (14) carbon
dioxid is present in milk to the extent of 3 to 4 per cent by volume
and partly escapes into the air when the milk is drawn. This being the
case, it is evident that there is not enough carbon-dioxid gas in normal
milk to cause of its own accord a positive alcohol test with 68 per cent
alcohol. Of course, the presence of CO, may assist other factors to
cause a positive alcohol test and in the case of bacterial fermentation
where the gas is produced it might play a small part, but we believe.
that when 68 per cent alcohol is used in the test the influence of CO,
in mixed market milk would be very small, if it has any effect.

TasLe 14.—Effect of heat on alcohol test with milk-made acid to phenolphthalein with
carbon dioxid.

|

Alcohol test.

BEE Beeted | . Acidity.
75 per cent. | 68 per cent.
2G.
Not heated 2.36 1+L +L
40 2.32 +L +L
50 2.30 +L
60 2.19 +L +VS
70 2.05 +M | =
80 2.05 +VS —
90 1.91 — ~
100 tO 2 aa _ _
Original milk 1.90 | -

1 See footnote under Table 1.
THE RELATION OF THE ALCOHOL TEST TO THE BACTERIA IN MIXED MARKET MILK.

Having discussed the effect of acidity and the effect of rennet
action on the alcohol test, let us consider the relation of the test to the
bacteria in market milk. Since an increased acidity and also rennet’
action may cause a positive test, it is natural to suppose that there
may be some definite relation between the alcohol test and the number
of bacteria in milk, as the increase in the acidity and the rennet in
milk is the result of bacterial growth.

It is claimed by some authorities that the alcohol test is of oreat
value for determining the freshness of milk, and as this is a question

1 We are indebted to Dr. Clark, of the Dairy Division laboratory, for this analysis.

ee

Ee eee

al te

THE ALCOHOL TEST IN RELATION TO MILK. 19

of great importance we have examined a number of samples of market
milk which had been held for a number of days.

In Table 15 are shown the results of the examination of four samples
of market milk. One was raw milk and the others were pasteurized.
Each bottle of milk was obtained from a different dairy and was held
in a refrigerator at a temperature of about 9° C. The acidity and
alcohol test were determined daily and bacterial counts were made on
the first day and again when the 68 per cent alcohol gave a positive test.

It is evident from the results obtained tuat the alcohol test (68 per
cent) does not show the freshness of milk. The samples were held
from 8 to 13 days at 9° C. before the alcohol test became positive, and
during that time the bacteria had increased to more than 100,000,000
per cubic centimeter.

TABLE 15.—Fffect on the alcohol test of holding milk at 9° C.

Raw milk. Pasteurized milk A.
Days Alcohol test. Alcohol test.
ee ia. |) Bacteria | cua lo 171 aeteria
ity. | 75 CBee ge ceo ity. |. 7 a ie is
per per per : per per per 5
cent. eent. | cent. cent. cent. | cent.
0 1.72 |1+V.S8. _ - Si200 ni sea ee oe — — —
| bal (aaa Oe +V.S. _ Seer glia eee Senet el a 8 — — me eRe ed a ae cee ts
Dei iee sh 5 +V.S8. _ Sg ee ee aes | shee ees — _ et | Ae ee ere eee
Sos a Soe +V.S. - OF Be eas ee ae Siena _— — Sten ean po ce Aone
ee PE te | ec fare Mot sp ote palm rnin! ote we, = ate lobe MiSI SP alc eioere alors Nieyeke ieee Si]icielaatele!l a slovcevereierajetacaiere
Fil aig Seta +V.S - a a ge ee rene [1 Ae _ _ Sie aati ee
Griese = 7 _ _— co Atl = 44 eset gre banged fas eyan rene — — tN el ie es
7 Sie cae — _ Sr es ee eek a Le Ae eae - — Ae ener es errs
8 1.88 |+V.S — pale eset La eae 1.78 — — SIH eee Ne Aa
9 1.95 _ _ Hie | 5 RR ae Set aaa 1. 88 +8. — ery oie Ue ae eeenany enacc
10 2. 02 _ _— aN aoe Be atl, 3 aa ae 1. 93 +S. — td [ia ene ee et eh
enn I ree es fo Seis) eerie slalom ooo So ciate Sasa elo clos Kasia [are eiuis Mies |b eve mreredie [ise close illo wicterats o bicere ind
12 2.08 |+ M — Hath hs co AE, Be or 2. 30 +L. +M —
13 2.43 /+ L +M. TUNIS alee eee eme ore 2. 48 +L. +L a IP eel
14 2.72/+ L +L — 146, 000, 000 715) +L. +L. |] +L
Pasteurized milk B. Pasteurized milk C.
D ays Aleohol test. Alcohol test.
€ d. * SF ae hk ee Bacteria 5 ae ie aa a es Bacteria
Acid- ‘ Acid- | :
ity 75 6g | 44 | Percubic | “ity. 75 68 | 44 | Pert.cubic
per per per centimeter. per per per centimeter.
cent. cent. | cent. cent. cent. | cent.
0 1.74 +M. _ ~ 182° 0008 |Heoa ss —_ — == | 133, 000
ieee a be +L. - ne esos sre te P+ mn SMR ary — — oe ai bis, eevniisest aea
D1) le pe ES +M. = ay || Sai ae 1.75 — — =F Pil ee SE hae on ee
35 aa a +M. _ aM eee te ciara 5) Re ne —_ — SF fa ab ean yan tig ater
a eae se Fe Re eg oe Aw oi ate)k 5 Sl mci bar <| o sicteiasaseiei] Sn amet teae Scala Sel ewes
Baye oe cher +M. _ oe SiS se ee = — oi | | te a eas
6 1. 88 +M. _ SPE AE Sa i ee hel a — — aaa ee Wc s ae
it ie +M. _ merits Ape ears 22 js, apenas ay — — ae Ay NS Pa eae ea
8) 2.00) +M.| +M./ — 145,000,000} 2.12} +M.| +M.}| — | 221,000,000
9 2.10 +M. +M RARE 2 Ss se 34 +L. Fs) DE | pee | eae
a4 2. 38 +L. +L ee hl ot Rae SM ES Po Cll +1. +L. | +M. | 650,000, 000
12 poe ret 2F Si ee NE cle 700000 OOUR Se oes aloece cle ee secls beacohiee coves ccs
Pune eee ena esheets) Fae ie

1 See footnote under Table 1.

20 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

EFFECT OF HOLDING THE MILK AT DIFFERENT TEMPERATURES.

The temperature at which the milk is held would, of course,
affect the length of time before the alcohol test becomes positive.
To show the effect of temperature we held samples of raw and pas-
teurized milk at 9° C. and also at room temperature, and examined
them in the same manner as in the preceding experiment. The
results are shown in Table 16. In order to have about the same
bacterial content in the milk held at the different temperatures, a
quart of milk was thoroughly mixed and placed in sterilized pint
bottles. |

The results show clearly that the temperature at which milk is held
has a marked influence on the time when the alcohol test will be posi-
tive. Also, as shown by sample of raw milk C, the bacterial content
of the milk is an important factor. In all the samples it will be no-
ticed that the bacterial counts show an exceedingly high number

when the 68 per cent alcohol test was positive.

TABLE 16.—Comparison of the alcohol tests with milk held at 9° C. and at 24° C.

| | Raw milk A. Pasteurized milk B.
pooner Days Alcohol test. Alcohol test.
which | held. : Bacteria per Bacteria per
held, | pre 75 68 44 | Cubic centi- fies 75 68 44 | cubic centi-
| ° | per | per | per meter. per | per | per meter.
cent. | cent. | cent. cent. | cent. | cent.
fe Se Catena eens) fe G00)| debs ta) ue le 139, 000
| 1 eS Sa te ell ee | pre eee 1.60 | (— 2] Se ee eee
2} 1.90 - _ penal’ | Wodet rere en eterehe rac 1.60 — - se Uae Sac eS
3 | 1.80 _ _ eed tl Fes See Emenee ae 1.70 - ~— +b [Po Gesekt Beene
BN, -- — aS eo are ere STAL - -- ce bp ee EP ede 5)
Wi [la trere ered oy Seen etree stall cca cscs | = Brcvotave wicreyarcnre sare [le pees a at Re pea, Dee | ee ee ee |
6} 1.85 _~ — amy” Jee eE Se 1.60 _ — Sa es eee ee
7 | 1.90 - _ —" | Shs oeeeetaeares 1.80 _ _ eS tea eee
| 8 | 1.80 _ - bc eee epee ae 1.65 - _ Ee Sah ten
OF See | a— - — tL ates eae aia | ae ee - _ piel Pha 2 SE oe
BOW GISO eee cf eee a 179 le Ea eae) San eee
1... ee. | ees 2.00|1+8 | — | — | 197,000,000
Tes cba oe oe aoe te fee cc oo [ak aisle ree pe Bk ee eer ee | Dee ee | ee
| 13 PATA ag ee _ 400,000,000 | 2.3 +L }+M] — 960, 000, 000
14 Se oe eee I ee cieis | Be oe ae bites |b e en ee ees Cee eee
17 1 (Ee aay ner, 440,000,000 | 4.3 = eS |G | odicin sive oie
Room
tem-
pera-
turel2 0 | 1.85 - _ _ 2,900 | 1.65 _ - - 146,000
(9a.m.)} 1.90 _ — ei te See 1.85 | +S — 325 600, 000
(3p.m.)} 3.1 +L/4+L/]4+L 534,000,000 | 3.70} +L rae +L 694, 000, 000

1 See footnote under Table 1.

2 About 24° C.

THE ALCOHOL TEST IN RELATION TO MILK. 21

TABLE 16.—Comparison of the alcohol tests with milk held at 9° C. and at 24° C.—

Continued.
Raw milk C,
Temper-
ature at Days. Alcohol test.
ee held. Neild: | a> SRG | eee tema per
© : ae ; cubic centi-
ity 75 68 44 ti
per per el on
cent cent. cent
heal Oa ) 1.85 — — — 7,870, 000
1 1.80 a _ Ti lel | ete ere Ee
2 1.95 |1+58 _ — 93, 000, 000
3 PA +M +S — 130, 000, 000
4 2.36 +L +L _ 188, 000, 000
£0 (be | |" SRE) ch GS oo) bs MR ee BL te
6 2.88 +L +L +M 430, 000, 000
(KS etek a | aie rere ||P ian es RON Oy Renae et Ok
SN eee esr [eg eee ea perenne eek |) Re oe Ec Ad a mean
£O aH Tye Seat ate | EE ACT® ELS rl oat LT 8 ea
DUCA Si | TRIE a et (Es ea a | Us Re ee a
TUT GE eet i tc aah Ae RI 2 eR) Um ne 1 ea aera ee
TAG IRR Pe is Sela Pee La BA ae [Se 2 TNA a be ae ee A
B35) |e ae ae | Ra eae} " 5 SS Re mle Us <a CE
CAN rahe oa WAS 1b os em (ER ee (oe as hae Ae er eC
SDE Teh| teeete carpe ee oes = ire eps | ees eet OR pata Nga ae
Room
tem
pera
ture? . O:) 31585, | — — — 7,610, 000
1(9a.m.) 2100 00) PAR A) te TI (er at Curdled.
1 See footnote under Table 1. 2 About 24° C.

At various times different investigators have used the alcohol test
on market milk. Aurnhammer (1) in an examination of 250 samples
of market milk during July and August of 1907 found the 68 per cent
alcohol test positive in 82 samples. In a study of market milk in
Philadelphia, Campbell (5) found that 37 of 100 samples of milk gave
a positive test with 68 per cent alcohol. Of these 37 samples 17
contained less than and 20 more than 1,000,000 bacteria per cubic
centimeter. It was found by Nurenberg and Lythgoe (23) during an
examination of 2,600 samples of market milk that only 63 gave a
positive test with 68 per cent alcohol.

We made alcohol tests on 236 samples of Washington market
milk during the period from March 20 to June 4, 1914. These
samples and their bacterial counts were supplied by the Health
Department, District of Columbia.‘ Of the 236 samples we found
that 37 gave an alcohol test with 75 per cent alcohol, 20 with 68 per
cent alcohol, and 5 with 44 per cent alcohol. The samples which
gave a positive test are tabulated in Table 17 with their acidity and
bacterial counts. There were 177 samples of raw milk and 59 samples
of pasteurized milk in the 236 samples examined. As may be seen
from the table, 35 of the raw-milk samples gave a positive test with
75 per cent alcohol and only 2 of the 59 samples of pasteurized milk.

1 We take this occasion to express our thanks to Dr. Kinyoun and Dr. Dieter, of the Health Depart-
ment, for the samples of market milk and their bacterial counts which they so kindly furnished us.

22 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

TABLE 17.—Raw and pasteurized market milk which gave positive alcohol tests.

5 Alcohol tests.
ara Bacteria per
Milk. ae Acidity. cubic
ber centimeter. 75 44
: per cent. | per cent. | per cent
Raweeee eee 1 2.01 2,100 |1 +M -
2 2.30 7,000 | +L —
3 2.00 12,000 | +M _
4 1.60 14,000} +M —
5 1. 97 18,000 | +M _
6 1.92 24,000 | +M +VS
7 1.91 24,000 | +S — oo
8 1.70 29,000 |} +S _ —
9 1.62 51,000 | +M +S _
10 2.30 67,000 |} +L +M —
11 1.79 121,000 | +M +8 +VS
12 lay 156,000 | +S _ _
13 2.00 200,000 | +M _— -
14 1.94 350,000 | +M ~ _
15 1.90 442,000 | +S ~- -
16 2.06 464,000 | +L +M -
17 1.95 1,200,000 | +M +VS _
18 ey 1,300,000 |} +VS — _
19 75 1,500,000 | +S _ -
20 1.80 1,600,000 | +M +58 —
21 Sy 2,100,000 | +M _ _
22 2.03 2,120,000 | +S — _
23 1.90 2,200,000 | +S _ —
24 1.95 2,300,000 | +M +S _
25 2.45 2,600,000 | +L +M +VS
26 1.95 4,100, 000 +M +M _
Pat 1.93 4,700,000 ; +VS aa —
28 2.15 7,200,000 |} +M +8 _
29 2.19 8,600,000 | +M +8 =
30 3. 65 10,200,000 | +L +L +L
31 1.90 10,500,000 | +M - —
32 2.05 20,200,000 |} +M — =
33 1.95 20,400,000 | +M +M -
34 Aas 20,600,000 | +L +L +L
35 2.10 21,200,000 | +L +M —
Pasteurized.. 1 1. 68 2,000 | +S _ _
2 1.90 8,000 | +M +S -

1 See footnote under Table 1.

RESULTS OF TESTS WITH SAMPLES OF KNOWN BACTERIAL CONTENT.

When we consider the alcohol test in relation to the number of
bacteria in milk, a short survey of the results is sufficient to show
that there is no definite relation. Of the 35 samples of raw milk
which showed a positive test with 75 per cent alcohol, 16, or 45.7
per cent, contained less than 500,000 bacteria per cubic centimeter.
Of the 19 samples positive with 68 per cent alcohol, 8, or 42.1 per cent,
contained less, and 11, or 57.9 per cent, more than 500,000 bacteria
per cubic centimeter. Of the 5 samples positive with 44 per cent
alcohol, 2 samples, or 40 per cent, contained less, and 3, or 60 per cent,
more than 500,000 bacteria per cubic centimeter. The number of
bacteria in soar dles which gave a positive alcohol test ranged from
2,100 to 21,200,000 per onlie centimeter.

The eaples of pasteurized milk which showed a positive alcohol —
test had a very low bacterial count.

In order further to show that the alcohol test has no definite
relation to the bacterial count, there are tabulated in Tables 18 and 19
the samples of raw and pasteurized milk which gave negative alcohol

ht ar ewer ee

THE ALCOHOL TEST IN RELATION TO MILK. 23

tests, together with their acidity and bacterial counts. We wish to
call particular attention to the bacterial counts of 142 samples of
raw milk which ranged from 2,000 to 19,600,000 bacteria per cubic
centimeter. Of these 142 samples none gave a positive alcohol test,
yet 86, or 60.6 per cent, contained less than and 39.4 per cent more
than 500,000 bacteria per cubic centimeter.

The bacterial counts of the samples of pasteurized milk which gave
a negative alcohol test ranged from 1,200 to 3,600,000 per cubic
centimeter, as may be seen from Table 19.

From our results we believe that there is no definite relation between
the alcohol test and the bacterial count, except in special cases where
the bacteria have developed to a point where there is sufficient acid
produced or where rennet-forming bacteria have acted sufficiently
to influence the test.

Tas_e 18.—Acidity and bacterial count of samples of raw market milk which gave negative
alcoho! tests with 75 per cent, 68 per cent, and 44 per cent alcohol.

Sam- Neidt Bacteria Sam- | 4 cig- Bacteria Sam-]| 4 cid: Bacteria

ple} per cubic ple per cubic ple F per cubic
No. ity. | centimeter. || No. Le centimeter. || No. aby, centimeter.
1 1.08. 2,000 49 1. 82 82, 000 97 1.95 812, 000
2 2.10 5, 000 50 1.75 86, 000 98 1.85 840, 000
3 1. 85 6, 000 51 1.08 92, 000 99 1.80 860, 000
4 1275 7, 000 52 1.90 93, 000 100 1. 60 880, 000
5 1. 82 §, 000 53 1.85 93, 000 101 1. 62 906, 000
6 2. 05 8, 000 54 1.60 105, 000 102 1. 69 910, 000
7 2. 05 8, 500 55 1.70 109, 000 103 1.90 910, 000
8 1.83 11,000 56 1.93 115, 000 104 1.90 910, 000
9 1.92 13,000 57 2. 00 118, 000 105 1. 80 920, 000
10 1.97 13,000 58 1. 87 120,000 106 1.75 1, 040, 000
il 1.75 14, 000 59 2. 06 120, 000 107 US 74) 1, 100, 000
12 2. 00 14, 000 60 1.75 130, 000 108 1.75 1,170, 000
13 2.15 16, 000 61 1.75 132, 000 t09 1.65 1, 200, 000
14 2.10 18, 000 62 2. 00 147,000 110 1.70 1, 200, 000
15 1.75 21,000 63 1.85 149, 000 111 1.90 1, 210, 000
16 1.95 21,000 64 1.90 150, 000 112 1.70 1, 400, 000
17 1.62 22, 000 65 1.90 157, 000 113 1. 85 1, 400, 000
18 2.15 22, 000 66 1.48 160, 000 114 1. 85 1, 460, 000
19 1. 80 25, 000 67 2. 00 164,000 115 1. 80 1, 600, 000
20 1.52 26, 000 68 1.85 172,000 116 2.08 1, 600, 000
21 1.80 26, 000 69 1.80 206, 000 117 79) 1, 630, 000
22 1.95 26, 000 70 2. 00 210, 000 118 1.98 1,710, 000
23 1.89 27,000 71 2. 03 212, 000 119 2. 20 1, 800, 000
24 1.65 29,000 72 1.90 214, 000 120 2.10 2,120, 000
25 2. 00 32,000 73 1.85 216, 000 121 1.74 2, 210, 000
26 1.60 33, 000 74 2. 00 220, 000 122 1:75 2, 260, 000
27 1. 84 33, 000 75 1. 82 238, 000 123 1.85 2, 340, 000
28 1.65 34,000 76 1. 83 238, 000 124 1.90 2, 580, 000
29 1.76 34, 000 77 2. 00 242,000 125 1.85 2,710, 000
30 2.10 35, 000 78 1.70 266, 000 126 VATGI 2, 840, 000
31 1.90 36, 000 79 2. 02 268, 000 127 1.70 2,900, 000
32 1.97 36, 000 80 1.90 270, 000 128 1. 80 2,920, 000
33 1.74 37,000 81 1. 80 278, 000 129 1.70 3,300, 000
34 1.45 37,000 82 1. 80 310, 000 130 2.07 3, 800, 000
35 1.55 38, 000 83 1. 80 350, 000 131 1.85 4,300, 000
36 1. 80 38, 000 84 TAV5 360, 000 132 1.96 4, 800, 000
37 2. 00 39, 000 Shale nets 422, 000 133 Dale 5, 100, 000
38 1.75 42,000 86 2.10 451,000 134 2.05 5, 300, 000
39 1.81 42,000 87 1. 80 506, 000 135 1. 80 5, 700, 000
40 1.70 43,000 88 2.05 510, 000 136 1. 80 6, 400, 000
41 1. 80 46, 000 89 1.90 560, 000 137 1.85 6, 900, 000
42 1.78 51, 000 90 Tait 610, 000 138 1.90 6, 900, 000
43 1.90 54, 000 91 2.10 620, 000 139 1. 80 8, 800, 000
44 1. 86 56, 000 92 1. 88 624, 000 140 1.90 | 12,600,000
45 1. 60 , 000 93 1.70 640, 000 141 1. 85 12, 700, 000
46 e735 69, 000 94 1.74 740, 000 142 2.10 19, 600, 000
47} 1.90 74, 000 95} 1.75 740, 000
48 LA 79, 000 96 2. 00 800, 000

24 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

TABLE 19.—Acidity and bacterial count of samples of pasteurized market milk which
gave negative alcohol tests with 75 per cent, 68 per cent, and 44 per cent alcohol. ©

Sam- ed Bacteria Sam- cide Bacteria Sam- New: Bacteria

ple F per cubic ple F per cubic le : per cubic
0. ity centimeter 0. ity centimeter. YO. ity. centimeter.
1 1.85 1, 200 20 1.75 15, 000 39 1. 85 104, 000
i el Ear 1, 200 Pa ees (2 16, 000 40} 2.05 110, 000
3 1. 66 1,900 22 1. 65 16, 000 41 1.65 114, 000
4| 1.75 3, 000 23} 2.05 16, 000 42| 1.76 120, 000
5 1. 66 4,000 24]. 1.69 17,500 43 1.80 133, 000
6 1. 80 5, 000 25 iy! , 000 44 1.75 194, 000
7 1. 85 7, 000 26 1. 66 21,000 46 13 264,
8 1. 85 7, 600: 27 1. 80 24, 000 47 1.70 284, 000
9 1.80 8, 000 28 1.83 32, 000 48 1.90 340, 000
10} 1.80 9,000 29] 1.67 37, 000 49| 1.76 446, 000
11 eal 9,000 30 1.96 41,000 50 1.85 720, 000
12 ie 11,000 31 1. 75 52, 000 51 1.65 740, 000
13 1. 85 11, 000 32 1.90 59, 000 52 1.75 940, 000
14 1.70 11,000 33 1. 85 62, 000 53 1.74 1, 280, 000
15 1. 65 12, 000 34 1. 85 64, 000 54 1. 60 1,660, 000
16 1.85 13, 000 35 1.70 65, 000 55 1.97 2, 460, 000
hg Nigees Oy fi 14, 000 36} 1.75 68, 000 56| 1.60] 3,100,000
18.) 21.75 15, 000 371 1280 71, 000 57 | 2.00] 3,600,000
19 1. 80 15, 000 38 1.70 74, 000
|

In the early stages of the growth of acid-forming bacteria in milk,
when the numbers are low, there is a period in which a rapid increase
in numbers takes place without any increase in acidity which can be
detected by ordinary chemical methods, or it may occur with only a
slight increase in acidity; consequently if the alcohol test were made
during that period there would be a high bacterial count and yet not
high acidity enough to cause a positive alcohol test. The same is
true of the action of the rennet-forming bacteria in their growth and
action, as we have shown earlier in this paper when dealing with the
relation of acidity, and also the effect of rennet on the alcohol test.
Besides these facts there are other groups of bacteria which may
develop in milk and yet have no influence on the alcohol test, as, for
example, the alkali-forming group of bacteria. We have tried cul-
tures of this group of organisms and found that they did not produce
a positive alcohol test. There are other groups of bacteria in the
flora of milk, such as the inert group, which also would probably
develop without influencing the alcohol test in any way. When we
consider all these facts it is not strange that there is no definite
relation between the bacterial flora of milk and the bacterial count.

When the 68 per cent alcohol test is positive with a sample of
market milk, it is evidence that there is some change in the milk from
normal. In some cases it may be due to an increased acidity and in
consequence a change in the casein of the milk, due to bacterial
action. In other cases it may be due to a pure rennet fermentation
or there may be a combination of an acid-and-rennet fermentation.
In such cases the bacterial count would undoubtedly be high. How-
ever, there still remains to be explained the reason for a positive
alcohol test in samples of market milk with a low bacterial count and
low acidity.

S

THE ALCOHOL TEST IN RELATION TO MILK. 25

We can not see that the alcohol test is of any particular value in
the control of a market milk supply except as a means of evidence
that milk from a particular source is abnormal in some way and
should be examined by other tests. It might be of value at a re-
ceiving station as a means of detecting sour milk, but the test would
be expensive compared with the use of alkaline tablets for the rapid
determination of acidity as described by Farrington and Woll (9).

THE TITRATION METHOD OF APPLYING THE ALCOHOL TEST.

From the simple alcohol test in which a definite volume of a definite-
percentage alcohol is added to an equal volume of milk there has
developed a method in which a definite volume of milk is titrated
with certain percentages of alcohol until a coagulation of the milk is
produced.

Loéhnis (16) has found this titration method to be of value as a test
for the quality of market milk. He found that there was quite a
definite relation between the titration with 80 per cent alcohol and
the bacterial content of market milk. He titrated 2 c. c. of milk in a
beaker against a black background with 90, 80, and 70 per cent alco-
hols, the titration being made at a temperature of from 15° to 20° C.
The first appearance of flakes was considered the end point.

We have used this method in the titration of 116 samples of market
-milk furnished with bacterial counts by Dr. Kinyoun and Dr. Dieter,
of the Health Department of the District of Columbia. In our
titrations of 92 samples of raw and 24 samples of pasteurized milk we
have not found any definite relation between the titration with 90
per cent and 80 per cent alcohols and the bacterial count. In Table
20 is shown the acidity, bacterial counts, and alcohol titration of 92
samples of raw milk, and in Table 21 the results of an examination
of 24 samples of pasteurized milk. The bacterial counts of the raw
milk ranged from 2,100 to 20,600,000 per cubic centimeter, and the
pasteurized milk from 1,200 to 3,100,000 bacteria per cubic centimeter.
Consequently we were able to titrate samples having a great variation
in their bacterial content. If a study is made of the bacterial counts
and the alcohol titrations shown in Tables 20 and 21 it will be seen
that there is no definite relation between them. In order to bring
this point out more clearly the titrations of samples containing more
than 500,000 and less than 500,000 bacteria per cubic centimeter
have been averaged, as shown in Table 22. The average titration
with 90 per cent alcohol of 46 samples of raw milk containing more
than 500,000 bacteria per cubic centimeter was 1.95 c. c., while the
average titration of 46 samples containing less than 500,000 per cubic
centimeter was 2.39 c.c. The average titration of 46 samples with
80 per cent alcohol was 4.61 c. c. when the bacterial count was more

°6 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

than 500,000 per cubic centimeter and 5.61 c. c. when the counts
were less than 500,000 per cubic centimeter. The average titrations
of the pasteurized milk samples showed even smaller differences.
The small differences in the average titration of samples with a high
and a low bacterial count show that there is little, if any, relation
between the alcohol titration and the bacterial count. This is shown
even more strikingly in Table 23, where the range in titrations
among samples grouped according to bacterial counts is recorded.
With these extreme ranges among samples of milk with high and low

_ bacterial contents it would be almost impossible to interpret an
alcohol titration in terms of bacteria.

TABLE 20.—Alcohol titrations of raw market milk.

Alcohol titra- Alcohol titra-
Bacteria ion. Bacteria tion.
Sams Acid- | per cubic eas Acid- | per cubic
k Pa ity centi- N a ity centi-

; meter 90 per | 80 per ; meter. | 90 per | 80 per

cent. | cent. cent. | cent.

Cx: CNC: CXC: CHC:
1 2.01 2,100 1.03 1. 86 47 1.80 | 506,000 1.50} 1.80
2 2.30 7, 000 ol 93 48 1.75 | 610,000 2.93 | 5.25
3 1.75 7, 000 Pea ly 5. 30 49 2.10 | 620,000 1.80} 5.10
4 1. 82 8, 000 2. 76 4. 54 50 1.88 | 624,000 2.91 | 9.20
5 2.05 8, 500 2.00 5. 30 51 1.74 | 740,000 2.96 | 6.80
6 175 14, 000 2. 21 4. 67 52 1.75 | 812,000 1.91} 6.76
7 2.15 16, 000 2. 80 7.60 53 1.80 | 860,000 1.62 | 3.00
8 1.95 21, 000 2. 70 7.43 54 1.62 | 906,000 1.68 | 3.30
9 1.91 24, 000 1.59 3.08 55 1.69 | 910,000 2.08 | 6.13
10 1.80 25, 000 3. 29 11. 73 56 1.80} 920,000 1.85 4.50
11 1.52 26, 000 3.12 7. 28 57 1.75 |1, 040, 000 1.92} 4.37
12 1.65 29, 000 2.68 | 10.07 58 1.75 |1, 100, 000 2.36 | 4.00
13 2. 00 32, 000 3. 64 9. 59 59 1.75 |1, 170,000 2.20 | 6.04
14 1. 60 33, 000 2. 56 5. 55 60 1.65 |1, 200, 000 1.90} 5.76
15 1.65 34, 000 2. 48 5. 61 61 1.76 |1,300, 000 2.72 | 6.59
16 2.10 35, 000 1. 24 2n71 62 1.70 |1, 400, 000 1.63 | 3.74
17 1.75 37, 000 2.26} 4.40 63 1.85 |1, 400, 000 2.36 | 4.42
18 1. 80 38, 000 2.71 6. 50 64 1. 85 |1, 460, 000 3.04] 9.61
19 1.55 38, 000 3. 06 6.92 65 1. 80 |1, 600, 000 1.23 | 5.40
20 1375 42,000 3. 00 4. 28 66 2.08 |1, 600, 000 2.51 | 5.94
21 1.80 46, 000 2.78 9. 68 67 1.70 |1, 630, 000 2.40] 6.00
22 2.30 67, 000 . 50 Fal 68 2.20 |1, 800, 000 1.30 | 2.37
23 1.75 69, 000 2. 58 5. 98 69 1.70 |2, 100, 000 1.16 | 1.60
24 1.90 74, 000 3.03 6. 33 70 2.03 |2, 120, 000 1.16} 2.34
25 i Ler 79, 000 3. 50 6.91 71 1.74 |2, 210, 000 2.10 | 3.96
26 1.85 93, 000 15% 4. 36 72 1.75 |2, 260, 000 1.81 | 4.12
27 1.70 | 109,000 3. 04 5. 20 73 1.85 |2,340, 000 2.42} 4.78
28 1.87} 120,000 2. 46 5. 80 74 2.45 |2, 600, 000 . 78 97
29 1.75 | 130,000 1. 60 3. 20 75 1.71 |2, 840,000 2.95 | 3.97
30 1.75 1 132,000 250 10. 45 76 1.70 |2, 900, 000 2.12:|\ 6.36
31 2.00 | 147,000 1.72 3. 80 77 1. 80 |2,920, 000 3.00 | 6.85
32 1.85 | 149,000 3. 54 8.17 78 1.70 |3,300, 900 2.80 | 6.30
33 1.80 | 160,000 1.48 2.49 79 2.00 |3, 600, 000 1.90} 5.40
34 1.80} 206,000 3.16 5. 52 80 1.85 |4, 300, 000 2.05 |. 5.00
35 2.00 | 210,000 2. 68 7.00 81 1.93 |4, 700, G00 1.40 | 3.80
36 2.03 | 212,000 2. 35 5. 70 82 1.96 |4, 800, 000 1.46 | 2.70
37 1.90 | 214,000 1.68 4.52 83 2.05 |5, 300, 000 1.60 | 2.68 2
38 1.82 | 238,000 2.75 7.00 84 1. 80 |5, 760, 000 2.94) 7.25
39 2.00 | 242,000 1. 24 2.90 85 1. 85 |6, 900, 000 LOL |) 42248
40 1.76 | 266,000 1.95 3.94 86 2.19 |8, 620, 000 .92} 1.86
41 2.02 | 268,000 2.57 5. 20 87 1. 80 |8, 800, 000 2.96 | 10. 83
42 1.90 | 270,000 2.74 4.45 88 1.90 |12,600,000 1583045: 10
43 1.80 | 278,000 Bee 8. 70 89 1. 85 {12,700,000 1.40 | 2.48
44 1.80} 310,000 2.34 4.58 90 2.15 |19,600,000 Le 5Galera-al 0
45 1.80 | 350,600 2. 68 4.67 91 1.95 |20,400,000 -86} 1.40
46 1.75 | 360,000 2.35 5. 29 92 2.55 |20,600,000 52 . 60

, ee Fs,

THE ALCOHOL TEST IN RELATION TO MILK. 27

TABLE 21.—Alcohol titrations of pasteurized market milk.

Aleohol titra- Alcohol titra-
Sam- Bacteria tion. Sam- Bacteria tion.
le | Acid- | per cubic ple | Acid- | per cubic
o. | ity. | centimeter. No. | ity. | centimeter,
90 per | 80 per 90 per | 80 per
cent. cent. cent, cent.
Cute Gat. Cue! cee:
1 Ue 1,200 2.20 5. 74 13 1.70 65, 000 2.88 5. 04
2 1.70 11,000 2. 84 6. 76 14 1.75 68, 000 2. 40 5.50
3 1.65 12,000 3.08 8.80 15 1.80 71,000 2. 50 4.49
4 1.85 13, 000 1.56 4.43 16 1.85 104, 000 2.43 3. 96
5 1.75 14, 000 1. 65 3. 10 17 1.70 120, 000 2.10 5. 96
6 1.75 15, 000 2.22 3. 42 18 1. 75 194, 000 3.05 6. 01
7 1.70 16, 000 2.76 4.10 19 1.73 264, 000 2.24 3. 00
8 1.78 21,000 2.33 5.91 20 1.70 284, 000 2. 86 5. 48
9 1.80 24, 000 1.69 6. 10 All 1.76 446, 000 1,82 5.34
10 1.90 59, 000 2.57 3. 68 22 1.60 | 1,600,000 2.10 4,78
11 1.85 62, 000 2.23 5. 47 23 1.97 | 2,460,000 2.08 3. 66
12 1.60 63, 000 2.53 3.24 24 1.60 | 3,100,000 3.53 7.07

TABLE 22.—Average alcohol titrations of samples of raw and pasteurized market milk in
tables 20 and 21.

Average alcohol

Num! titration.
Milk. MY Oe Bacteria per cubic
samples centimeter.
90 per 80 per
cent. cent
Gace Cues
AVON see = 46 | More than 500,000 - 1.95 4.61
46 Less than 500,000. - 2.39 et
Pasteurized . . 3 | Morethan 500,000-. 2.57 a ale
21 Less than 500,000. - 2.28 5. 02

TABLE 23.—Range in alcohol titrations of market milk shown in detail in tables 20 and 21.

Alcohol titration.

Bacteria per cubic centimeter. 90 per cent. 80 per cent.

Lowest. | Highest. | Lowest. | Highest.

Cac: (EC (3G: calc
26 samples with less than 100,000. ..... 0.51 3. 64 0. 93 11.73
30 samples with from 100,000 to 1,000,000) 1.24 3. 54 1.80 10. 45

36 samples with over 1,000,000....-.--.. | 02 3.04 . 60 10. 83

For the sake of clearness we have plotted in figure 1 the bacterial
counts and the 90 per cent alcohol titration. In this figure the
titrations of 116 samples of milk were plotted as ordinates and the
logarithms of the bacterial counts as abscisse. The numbers 3, 4,
5, 6, 7, and 8 represent the mantissa of the logarithms of the bacterial
counts. Consequently from 3 to 4 was plotted the logarithm of
samples with a bacterial count of from 1,000 to 9,999, from 4 to 5
counts from 10,000 to 99,999, and so on, as may be seen from the
figure. By this method of plotting it is possible to plot bacterial
counts ranging from low to high numbers, which would otherwise be

28 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

impossible in a limited space. A glance at the plot which shows the
90 per cent alcohol titration and the bacterial count of 116 samples
indicates clearly that there is no definite relation between them. In
figure 2 we have plotted in the same way the 80 per cent alcohol

LOGIT? OF ee, COUNT S..

ES : 4000 ~ 9,999 bs 10,000 -99,999 ++ 100,000 999,999 +. 4000, 000 ~9.999,999 Re

hyn Soe

90% ALCOPIOL.

S

CUBIC CENTIMIETERS OF

Fic. 1—Relation o falcohol titration to the bacterial count of milk. Titrations of 116 samples of raw
and pasteurized market milk with 90 per cent alcohol.

titration and the bacterial counts. It may be seen that among the
116 samples plotted there is a wide range in titration of samples
with low and high bacterial counts. Some sampies with a low count
show a low titration and others a high titration. Among samples
with a high count some show a low and others a high titration.

epee OF BAC. TEU COLNTS.
4000-3999 + 12000-93999 + 100.000 ~993,999 ++ ,009000- 9.999999 + 12009. 000-929999984

SACTERSAL b
COUNTS

Fic. 2.—Relation of alcohol titration to the bacterial count of milk. Titrations of 116 samples of raw
and pasteurized market milk with 80 per cent alcohol.

Our results indicate that there is no definite relation between
alcohol titration and acidity unless the acidity is more than about
2.20. This is shown in figure 3, where 116 samples are plotted ac-
cording to their acidity and titration with 90 per cent alcohol, and
also in figure 4, where the 80 per cent titrations and acidities are

THE ALCOHOL TEST IN RELATION TO MILK. 29

plotted. The plots show that there is a wide range in the alcohol
titration at all acidities until they reach about 2.20, after which the
alcohol titration becomes lower in general as the acidities increase.
This fact holds true for the small number of samples at these high

OF

NY & A

<9 “
se
s

2.20 250 2.40 2.50

CUBIC CLV IVWIIE TERS
IOV ALCO/VIOL.

ACAT YT
Fic. 3.—Relation of alcohol titration to the acidity of milk. Titrations of 116 samples of raw and pas-
teurized market milk with 90 per cent alcohol.

acidities and would probably have been brought out more clearly if
we had had a larger number of samples with acidity above 2.20.

If we were dealing with pure cultures of organisms which influence
the alcohol test the titration with alcohol might be of value in giving
an idea of the bacterial numbers, as is shown in Table 24, from the

6

(Z 490 <00 2./0 «20

460 470 4G 2.30 £40 250

CUBIC CENTIMETERS OF 80% ALCOVIOL.

o
LS

0)

ACIOITY.
Fig. 4.—Relation of alcohol titration to acidity of milk. Titrations of 116 samples of raw and pasteur-
ized market milk with 80 per cent alcohol.

results of experiments in which we used pure cultures of lactic-acid
and rennet-forming bacteria. In milk, however, we have a varied
bacterial flora to contend with and we can not see from our results
that the alcohol titration method is of much greater value than the

simple alcohol test.

80 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

TABLE 24,—Alcohol titrations of milk inoculated with pure cultures of bacteria.

Alcohol titra-
tion.
Age of F BGvenane | —_——<————$————————
Experi- : Acid- ‘
ment. | Cuteore | te| fy. | Bez eabls.| 90 per | 80 per

cent to | cent to

2c.c. of | 2¢.c. of
milk
IES ee Lactic-acid bac- | Hours. Cc. cues
CLIA WS. See 0 1.95 64, 000 3.08 | 8.88
2 TIS GSi| aes See Gee 2.86] 9.13
| 3 1.90 131,000 3.14} 9.42
4 1.95 120, 000 2.60 | 9.52
5 1.95 361, 000 3.05 |} 9.40
6 2.05 736, 000 3505) |< 8285
63 2.07 | 1,660,000 1.85 | 7.08
Rennet-forming
bacteria......- 0 AOS a aye ee tee ee O12) |- 902
2 1.90 1,600 SBUBY | Ch Gy
3 1.95 70, 000 2.50] 9.48
4 2.03 230, 000 2.44 | 9.83
5 2.04 | 2,850,000 3.20] 9.90
6 2.04 | 9,300,000 2.80] 8.12
63 2.07 | 11,100, 000 1.87 | 6.18
PARE Lactic-acid bac-
tenia sae eee 0 2. 00 25, 000 Sye8) platy
2 2.00 184, 000 3.58 | 13. 82
3 2.08 475, 000 S024 ALLS
4 2.04 | 1,710,000 3.73 | 9.85
5 2.18 | 4,900,000 2:00)! “adeoe
6 2.26 | 8,400,000 2.00} 5.50
7( 2.33 | 22,500,000 1.06 | 1.79
Rennet-forming
bacteria....... 0 2.00 6, 100 Suse alan 9D
2 2.00 51, 500 3.40 | 13.36
3 2. 00 234, 000 3.98 | 10.52
4 2.00 | 1,325,000 SRE Cas
5 2.00} 1,300,000 1.56 | - 8.98
6 2.04 | 13,000, 000 1.25 | 2.93
if 2.20 | 21,800, 000 0.83 | 1.21

THE ALIZAROL TEST.

When the alcohol has alizarin added to it to act as an indicator
for the acidity the alcohol test is known as the alizarol test. This
name was given to the test by Morres (21). The use of alizarin as
an indicator for the acidity of milk has been known for a long time,
but Morres (19) was probably the first to combine the alcohol and
alizarin test. He pointed out that the alcohol test was of more value
than the litmus test and that the combination with alizarin was
better than the combination of litmus and alcohol. Morres (20)
used a 68 per cent alcohol with 1.2 grams of fresh alizarin paste, or
0.4 gram of dry alizarin to 1,000 c. c. of alcohol. Two cubic centi-
meters of this alizarin-alcohol solution are mixed with 2 ec. c. of milk,
the same as in the alcohol test. This author found that from the
coagulation by alcohol and the color of the alizarin it was possible
to obtain a picture of the condition of the milk. According to
Morres (20) the alizarol test shows the following conditions:

1. Lilac-red color. (Milk titrated 7° acid.)
(a) With no coagulation. The milk should keep sweet more than 6 hours.
(6) With fine flaky coagulation. The beginning of rennet production is shown
here.

(c) With heavy flocculent co2gulation. This indicates advanced rennet forma-
Tony), :

THE ALCOHOL TEST IN RELATION TO MILK, DL

2. Pale-red color. (Milk titrated 8° acid.)

(a) With no coagulation or only very fine coagulation. This shows the beginning
of lactic-acid fermentation.

(6) With flaky coagulation. Acid and rennet fermentation is indicated.

(c) With coagulation with very thick flakes. A mixed fermentation is indicated
with advanced rennet and the beginning of acid fermentation.

3. Brownish-red color. (Milk titrated 9° acid.)

(a) With coagulation with fine flakes. Well-advanced pure acid fermentation is
indicated.

(6) With coagulation with thick flakes. A mixed fermentation with advanced
rennet and strong acid fermentation is indicated.

(c) With coagulation with very thick flakes. A very advanced rennet production
and little less important acid fermentation is indicated.

4. Reddish-brown color. (Milk titrates 10° acid.)

(a) With flaky coagulation. Advanced pure acid fermentation is indicated.

(b) With thick flaky coagulation. Advanced acid fermentation and the begin-
ning of rennet production is indicated.

(c) With very thick flaky coagulation. A proportional mixed fermentation
which is well advanced is indicated.

5. Brown color. (Milk titrates 11° acid.)

(a) With thick flaky coagulation. Pure acid fermentation is indicated. Milk is
sour; to be detected by smell.

(6) With very thick flaky coagulation. Some rernet production and well ad-
vanced acid fermentation is indicated.

6. Yellowish-brown color. (Milk titrates 12° acid.)
(a) With very thick flaky coagulation. Acid fermentation is indicated. Milk
tastes acid.
7. Brownish-yellow color. (Milk titrates 14° acid.)
(a) With very thick flaky coagulation. Sour taste is distinctly noticeable.
8. Yellow color. (Milk titrates 20° acid.)

(a) With very thick flaky coagulation. Pure acid fermentation is indicated.
Milk smells and tastes strongly acid and is near the normal coagulation
point.

9. Violet color. (Milk titrates 7° acid.)
No fermentation is indicated, but the milk is abnormal.

It can not be disputed that a simple test which will picture con-
ditions in milk, as claimed by Morres, would be of considerable value.
But will the alizarol test indicate all that Morres claims? Devarda
and Weich (6) in 1913, after working with this test, decided that it
had no value over the alcohol test. In a later paper Devarda (7)
draws conclusions as follows:

1. For market control the alcohol test is satisfactory for the determination of the
quality of milk.

2. The assertion of Morres that the alizarol test can show a pure rennet and mixed
fermentation is without scientific or practical significance.

3. In a pure lactic fermentation the alizarol test stands close to the acidity in its
color relation, but for the determination of the keeping quality of milk it is of slight
significance.

4. The diagnostic value of the alizarol test is limited to an empirical test for milk,
principally as to its suitability for cheese making which was already employed by
Eugling in 1882. -

Thoni (30), in a study of the milk supply of Berne, found that 12
of 85 samples examined were more or less abnormal, according to

32 BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE.

the alizarol test. Among the other 73 samples of milk, which ac-
cording to the alizarol test were normal, there were samples which
had a high bacterial content and which were abnormal according to
the leucocyte and other tests. From his results Théni believes that
the alizarol test is not sufficiently delicate for use in market-milk
investigations. However, he believes the test is of value as a quick
means for detecting udder infection in animals. .

It is evident that there is a diversity of opinion as to the value of
the alizarol test, and our experiments have not been extensive
enough for us to form a definite opinion in regard to it.

We have tried the test on a number of samples of milk and have
not been able to obtain all the color changes which are described by
Morres. When the acidity was slightly above normal we found a
change from lilac red to pale red and brownish red. In one sample
of milk we increased the acidity by the addition of lactic acid and
obtained the colors named below.

Amount of N/10 lactic aie .
Sadbdtealso Gictot milk Acidity. |Color of alizarol test.

Normalimilk- 25% 332 1.85 | Lilac red.
DONC eee ee see 2.10 | Pale red.
ArOsC) 5 sees owe 6 rd 2.42 | Brownish red.
1a ik Cee a 2 Zo Do.

Sievers Sra 3. 00 Do.

1.5 ¢. ec. normal acid.. O25 Do.

From our results we believe that alizarin will show slight changes
in the acidity when the acidity is low, but that the indicator did not
seem to be very sensitive to high acidities in milk. Morres (22), ina
paper in 1913, also states that alizarin is of greatest value in indi-
cating the first changes in acidity and that the color change is so
gradual at acidities over 16° that the test is of no particular value.

In regard to the value of the alizarol test we believe that wherever
the alcohol test can be considered of value, the addition of an indi-
cator, such as alizarin, may increase the value of the alcohol test by
possibly giving additional information as to acidity.

On account of the complexity of the bacterial fermentations in
market milk we do not believe that the alizarol test gives any very
valuable information as to the conditions existing in the milk.

CONCLUSIONS.

In conclusion, we wish to point out again that the alcohol test
must be considered from two standpoints: First, in its relation to |
the milk from a single cow or small herd, and, second, in its relation
to mixed market milk.

As to the relation of the alcohol test to milk from a single cow, it
seems evident from the work of other investigators, which is con-
firmed to some extent by our results, that a positive 68 per cent

THE ALCOHOL TEST IN RELATION TO MILK. 338

alcohol test indicates some change in the milk from its normal con-
dition. In our opinion the value of the alcohol test with milk from
a single cow or small herd lies in the fact that it would show that the
milk was abnormal, and in consequence a car eful examination should
be made of the herd,

When the relation of the alcohol test to mixed market milk is dis-
cussed, we must consider it on an entirely different basis. In this
case the test with 68 per cent alcohol may be positive as a result of
changes produced in milk through bacterial action. The results of
our work confirm some of the results of other investigators and show
that the alcohol test may be positive as a result of the growth in milk
of lactic-acid and rennet-forming bacteria. When the growth of
these bacteria has reached a point where the acid or rennet is pro-
duced in sufficient quantities to affect the casein, a coagulation is
produced when equal volumes of 68 per cent alcohol and milk are
mixed. Our results, however, do not show that there is any definite
relation between the alcohol test and the number of bacteria in milk.
During an examination of 177 samples of raw milk we found that 20
samples gave a positive test with 68 per cent alcohol. Of these 20
samples 8, or 42.1 per cent, contained less than 500,000, and 11, or
57.9 per cent, more than 500,000 bacteria per cubic centimeter. It
was also found that 39.4 per cent of 142 samples of milk which gave
no positive alcohol tests contained over 500,000 bacteria per cubic
centimeter. That there is no definite relation is probably explained
by the fact that bacteria may increase in large numbers before there
is much acid or rennet produced. Consequently, if an alcohol test
were made during that period there would be a high bacterial content
and yet not enough change produced in the milk by acid or rennet
to cause a positive test. Besides this point it must be remembered
that in market milk there is a bacterial flora representing many dif-
ferent species, many of which may increase without influencing the
alcohol test.

As stated before, generally speaking, when the bacterial fermenta-
tions have advanced to a point where chemical changes are produced,
the alcohol test will be positive as a result of lactic or rennet fermen-
tations, or a mixture of both. In such cases the alizarol test may
be of more value than the plain alcohol test, so far as it may give
additional information as to the kind of fermentation. From our
results it seems evident that the acid-and-rennet fermentations may
be differentiated by means of neutralization of the acidity by sodium
hydrate.

The alcohol titration method according to our tests seems to offer
no particular advantages over the alcohol test. In a study of 116
samples we were not able to find any definite relation between the
alcohol titration and the bacterial count.

34

10.
1a
12.
13.
14.

15.
16.

we

LS.

19.

BULLETIN 202, U. S. DEPARTMENT OF AGRICULTURE,

LITERATURE CITED.

. Aurnhammer, Albert. Milchversorgung der Stadt Miinchen. Inaug.-Dissert.

Munchen, 1907. Jn Centralblatt fiir Bakteriologie, Parasitenkunde und Infek-
tionskrankheiten, Abteilung 2, vol. 21, no. 17/19, p. 529, Jena, 1908.

. Auzinger, August. Studien iiber die Alkoholprobe der Milch, ihre Verwend-

barkeit zum Nachweis abnormer Milchen und ihre Beziehungen zu anderen
Priifungsmethoden pathologischer Milch. Jn Milchwirtschaftliches .Zentral-
blatt, vol. 5, no. 7, p. 293-315; no. 8, p. 352-370; no. 9, p. 393-413; no. 10,
p. 430-446, Leipzig, 1909.

. Auzinger, August. Die Alkoholreaktion der Milch. Jn Molkerei-Zeitung, vol.

28, no. 25, p. 457-459, Hildesheim, March 27, 1914.

. Barillé, A. De l’existence des carbonophosphates dans le lait. Leur précipita-

tion par la pasteurisation. Jn Comptes Rendus des Séances de l’Académie des
Sciences, vol. 149, no. 5, p. 356-358, Paris, Aug. 2, 1909.

. Campbell, H. C. Biochemic reactions and the bacterial count of milk. In

U.8. Department of Agriculture, 28th Annual Report of the Bureau of Animal
Industry, p. 195-224, Washington, D. C., 1911.

. Devarda, A., and Weich, A. Die Morressche Alizarolprobe zur Priifung der

Haltbarkeit der Milch. Jn Archiv fiir Chemie und Mikroskopie, vol. 4, no. 4,
p. 207-212, Wien, 1913. ‘

. Devarda, A. Welchen Wert hat die Alizarolprobe fiir die Untersuchung der

Milch zum Zwecke der Marktkontrolle. Jn Osterreichische Molkerei-Zeitung,
vol. 21, no. 2, p. 17-19, Wien, Feb. 25, 1914.

. Ernst, Wilhelm. Grundriss der Milchhygiene fiir Tierarzte, p. 274, Stuttgart,

1913.

. Farrington, E. H., and Woll, F. W. Testing milk and its products, Madison,

Wis., 1911.

Fendler, G., and Borkel, C. Alkoholprobe und Sauregrad der Milch. Jn Zeit-
schrift fiir Untersuchung der Nahrungs-und Genussmittel, vol. 21, no. 8, p.
477-480, Berlin, Apr. 15, 1911.

Fleischmann, Gustav F. W. Lehrbuch der Milchwirthschaft, Dritte auflage,
p. 120. Leipzig, 1901.

Henkel, Th. Die Aciditaét der Milch, deren Beziehungen zur Gerinnung beim
Kochen und mit Alkohol, die Saiurebestimmungsmethoden, der Verlauf der
Sauerung. Jn Milchwirtschaftliches Zentralblatt, vol. 3, no. 8, p. 340-369;
no. 9, p. 378-405, Leipzig, 1907.

Hoft, H. Zur Priifung und Beurteilung saurer Milch nach der Alkohol und
Alizarinprobe. Jn Molkerei-Zeitung, vol. 12, no. 17, p. 277-278, Hildesheim,
Apr. 23, 1898.

Kastle, Joseph H., and Roberts, Norman. The chemistry of milk. Treasury
Department, Public Health and Marine-Hospital Service of the U. 8., Hygienic
Laboratory bulletin no. 56, p. 313-425, Washington, D. C., 1909.

Kirchner, W. Handbuch der Milchwirtschaft, Fiinfte auflage, p. 180, Berlin,
1907.

Lohnis, F. Die Titration der Milch mit Alkohol von verschiedener Konzentra-
tion. Jn Molkerei—Zeitung, vol. 28, no. 9, p. 153-155, Hildesheim, Jan. 30, 1914.

Metzger, Karl. Untersuchungen iiber die Alkoholprobe bei Milch von kranken
Kiihen. Diss. Stuttgart 1912. Jn Centralblatt fiir Bakteriologie, Parasiten-
kunde und Infektionskrankheiten, Abteilung 2, vol. 39, no. 4/7, p. 181, Jena,
Oct. 11, 1913.

Morres, Wilhelm. Die Haltbarkeitspriifung der Milch. Jn Molkerei-Zeitung,
vol. 19, no. 36, p. 421-422, Berlin, Sept. 4, 1909.

Morres, Wilhelm. Zur Alkoholprobe der Milch. Jn Molkerei-Zeitung, vol. 23,
no. 47, p. 1319-1321, Hildesheim, Nov. 20, 1909.

21.

23.

24.

20.

26.

27.

28.

29.

30.

dl.

THE ALCOHOL TEST IN RELATION TO MILK. 35

. Morres, Wilhelm. Die Zersetzung der Milch und das einfachste Verfahren, um

sowohl ihre Art als auch ihren Grad zuverlissig zu bestimmen. Jn Molkerei-
Zeitung, vol. 24, no. 98, p. 1837-1838, Hildesheim, Dec. 16, 1910.

Morres, Wilhelm. Die hiufigsten Zersetzungsarten der Milch und ihr bestes
Erkennungsmittel in der Hand des Molkereipraktikers. Jn Molkerei-Zeitung,
vol. 22, no. 38, p. 445-446, Berlin, Sept. 21, 1912.

. Morres, Wilhelm. Neue Versuche mit der Alizarolprobe. Jn Osterreichische

Molkerei-Zeitung, vol. 20, no. 21, p. 331-333; no. 22, p. 349-351, Wien, Nov.
1913.

Nurenberg, Lewis I., and Lythgoe, Herman C. Detecting pasteurized milk and
old milk. Jn The Creamery and Milk Plant Monthly, vol. 11, no. 6, p. 4-6,
Chicago, Feb. 1914.

Petri, R. J., and Maaszen, Albert. Zur Beurtheilung der Hochdruch-Pasteur-
isir-Apparate. Jn Arbeiten aus dem Kaiserlichen Gesundheitsamte, vol. 14,
p. 53-70, Berlin, 1898.

Rammstedt, O. Kochprobe, Alkoholprobe und Sauregrad der Milch. Jn Zeit-
schrift fiir Offentliche Chemie, vol. 17, no. 23, p. 441-455, Plauen, Dec. 1911.
Reiss, F. Ueber eine schnellere und billigere Ausfitihrung der Alkoholprobe in
den Milchhandlungen. Jn Molkerei-Zeitung, vol. 20, no. 3, p. 50-51, Hilde-

sheim, Jan. 20, 1906.

Reiss, F. Wie musz der Alkohol zur Priifung der Milch auf Kochfaihigkeit
beschaffen sein? Jn Molkerei-Zeitung, vol. 18, no. 35, p. 831-832, Hildesheim,
Aug. 27, 1904.

Rihm, G. Untersuchungen tiber das Vorkommen und die Haufigkeit der Strep-
tokokkenmastitis bei Kihen. Jn Wochenschrift fiir Tierheilkunde und
Viehzucht, vol. 52, no. 7, p. 125-130; no. 8, p. 147, Munich, Feb. 18, 1908.

Rullmann, W., and Trommsdorff, R. Milchhygienische Untersuchungen. In
Archiv fiir hygiene, vol. 59, no. 3, p. 224-265, Munich und Berlin, 1906.

Thoni, J. Untersuchungen iiber die hygienisch-bakteriologische Beschaffenheit
der Berner Marktmilch mit Beriicksichtigung des Vorkommens von Tuberkel-
bacillen. Jn Centralblatt fiir Bakteriologie, Parasitenkunde und Infektions-
krankheiten, Erste Abteilung, Originale, vol. 74, no. 1/2, p. 11-69, Jena,
May 27, 1914. .

Weber, A. Die Bakterien der sogenannten sterilisirten Milch des Handels, ihre
biologischen Eigenschaften und ihre Beziehungen zu den Magen-Darmkrank-
heiten der Sduglinge, mit besonderer Beriicksichtigung der giftigen peptoni-
sirenden Bakterien Fliigge’s. Jn Arbeiten aus dem Kaiserlichen Gesundheit-
samte, vol. 17, p. 108-155, Berlin, 1900.

32. Die neue Polizeiverordnung, betreffend den Verkehr mit Kuhmilch und Sahne

in Berlin. Jn Molkerei-Zeitung, vol. 16, no. 13, p. 225-226, Hildesheim, Mar.
29, 1902.

ADDITIONAL COPIES

OF THIS PUBLICATION MAY BE PROCURED FROM
THE SUPERINTENDENT OF DOCUMENTS
GOVERNMENT PRINTING OFFICE
WASHINGTON, D. C.

AT

5 CENTS PER COPY
V

WASHINGTON : GOVERNMENT PRINTING OFFICE ; 1315