SOME OBSERVATIONS ON THE FREEZING POINT OF

CREAM AND ITS USE IN DETECTING ADDED

WATER*
F. J. DOAN~
Dairy Husbandry Department, Pennsylvania State College, State College,
Pennsylvania
The freezing point of milk is a subject that has received a great
deal of attention from investigators in widely scattered fields.
At present the value is generally accepted as a physiological
constant averaging -0.550°C., with a normal variation of from
-0.534°C. to -0.562°C. according to Hortvet (1)~, and from
-0.530°C. to -0.566°C. according to Bailey and others (2).
The freezing point is now used as a means of detecting added
water in milk and a carefully standardized technique has
been incorporated, in the methods sponsored by the Assoei~ion
of Official Agricultural Chemists, as an "official method" (3).
The freezing point of erea~ has not been studied to any great
extent. In fact the author has been unable to find any published
data dealing with cream as such. However, from cryoseopie
data on related substances and a knowledge of the theory of
freezing point depression it is not impossible to predict what
might be expected.
W. R. G. Atkins (4) found that the addition or removal of fat
from milk has no effect on the freezing point.
Parfitt and Taylor (5) show that variations in the fat content
of ice cream mixes do not alter the temperature at which they
freeze.
Colloidal dispersions, even comparatively concentrated, ex-
hibit very small osmotic pressures and consequently practically
negligible depressions of the freezing points of the continuous
* Received for publication December 16, 1927.
t Published by permission of the Director of the Agricultural Experiment
Station. Contribution from the Department of Dairy Husbandry. The Penn-
sylvania State College, No. 4~5.
353
354 F.J. nOAN

phases, (6) (7). The fat in milk or cream existing as a coarse

emulsion, consequently, would not be expected to influence the
freezing point of the plasma.
If the fat be removed from milk, skimm]lk and cream, either
by extraction or centrifugation, the remaining plasmas are
identical. Exception to this statement might be taken since
Clayton (8) gives data showing that 2 per cent of the casein in
ordinary milk is adsorbed or concentrated on the fat globule-
plasma interface. Milk in creaming, therefore, might be ex-
pected to lose a portion of its colloids to the cream, in which case
the respective plasmas might differ. This point, however, would
not noticeably affect the freezing point since, the colloids, as a
class, do not show a noticeable depression phenomenon. The
freezing points of milk~ skimmilk and cream, then, must be
dependent on the concentration of soluble substances in the
serum and it can hardly be argued that the serums in question
would differ. All of which leads to the conclusion that cream
and skimmilk have the same freezing point as the original
whole milk from which they are derived.
F R E E Z I N G P O I N T S OF M I L K , SKIMMILK AND CREAM

Twelve samples of authentic herd milk were separated and

observations made on the resultant skimmilk and cream which
were compared with similar observations made on the original
whole milk. Cryoscopie readings were obtained following the
exact procedure as given in the A.O.A.C. "Official Methods."
Fat tests and acidity tests were made in accordance with methods
obtained from the s~me source, with the exception that acidity
samples of cream were weighed out instead of measured. The
percentages of solids-not-fat represent the difference between
the total solids content, obtained by means of the Mojonnier
test~ and the fat tests.
Elaborate precautions were taken in obtaining the various
samples and in holding them prior to analysis and the determina-
tion of the freezing points in order that no uncontrolled factors
would make a direct comparison impossible. Table 1 gives the
 OBSERVATIONS ON F R E E Z I N G POINT OF CREAM 355

data taken on these samples and is chiefly of interest from the

standpoint of the cryoscopic results. Samples No. 11 and No. 12
were diluted before separation with 11.5 per cent and 20 per
cent of water respectively. The data seems to confirm the
theoretical conclusion made above, namely that the freezing point
of skimmilk and cream are identical with that of the original
whole milk irrespective of their compositions. An elevation of
the freezing point due to added water in milk is equally detectable
TABLE 1
Freezing points of milk, cream and skimmilk
WHOLE MILK CRE~ SKI~IMILK
SAM~E
NUM-
BER SNT Acid Freezing Fat SNF Avid Freezing
Fat ]----I point point
per ~r °C.
~n$ ~Pn~t cen~Per °O. cp~r$ ~er ~ er I oC
cent nt

1 -- O. 558 -0.559 -0.558

2
3 j
-- 0.537
-0.546
--0.544
/
-0.540
--0.546
-0.538
--0.548
-0.545
4 I--0.545
5 ~--0.55C -0.548
O. 14{ -0.549
6 4.3( 8.47 --0.54~ 22.( 0.0~
6.85 0. 121[-- 0. 54~ 8.91 157 - 0 . 5 4 6
7 3.9,~ 8.51 0.10~ --0.53~ 19.~ 7.91 0.1501-0.539 0.~ 9.0~ 16~ - 0 . 5 3 6
8 4.0~ 8.5~ 0.14~ - 0 . 5 4 ( 27.,~ 6.32 0.1151- 0. 53~ 0.04 9.02 15~ - 0.537
9 4.0( 8.5~ 0.15{ --0.54~ 24.( 7.02 0.121[-0.54~ 0.01 9.41 16] - 0 . 5 4 2
10 4.0~ 8.3{10.141 --0.541 20.{ 7.2~ 0.1181-0.54( 0.0~ 8.8£ 14~ - 0 . 5 4 0
11 3.8q 7.7(i0.13~ -0.46~ 25.( 5.9710. 106l--0.46] 0.0~ 8.17 13~ - 0 . 4 6 0
12 3.3! 6.8~ i0.12~ --0.42~ 21.~ 5.9~1o.o851-o. 43c 0.0,' 7.2~ 13~ - 0 . 4 2 9

in either the skimmilk or the cream after separation. None of

the results on a given sample vary to a greater extent than the
possible experimental error. It might be mentioned that the
freezing points represent the average of three trials all of which
checked within 0.003°C. Results varying outside this range were
discarded and a new trial made. The cyroscopic method as
given seemed to be equally applicable to cream of a richness and
viscosity similar to the samples used. The only difference noted
in making cream readings was the relative sluggishness of the
mercury thread as compared to the movements with milk.
356 F. J. DOAN

TABLE 2
Watered samples of cream
Freezing points °C.

lO PER CENT 15 PER CEN~ 2 0 P E R CENT 25 PER C E ~ FAT IN CREA]

SAMPLE UNDILUTED ADDED ADDED ADDED ADDED I BEFORE
WATER WATER WATER WATER I DILUTION

6 -0.545 -0.465 -0.439 -0.386 -0.355 22.0

7 -0.539 -0.464 -0.436 --0.386 19.5
--0.357 [ 27.5
8 --0.538 --0.384
9 -0.545 --0.461 - O. 417 --0.386 --0.363 24.0
10 I -0.540 -0.465 --0. 414 -0.379 --0.353 20.0

Added water calculated from above depressions

100 ( T - T')
W=
T

6 0 14.7 19.4 29.1 34.8 22.0

7 0 13.9 19.1 28. 4 33.8 19.5
8 0 28.6 27.5
9 0 15.4 23.5 29.1 33.4 24.0
I0 0 13.8 23.3 29.9 34.6 20.0
Average . . . . . 0 14.4 21.4 29.0 34.1 22.6

Added water calculated from above depressions

A ( T - T')
Wffi
T
6 0 10.8 14.7 22.4 27.3 22.0
7 0 10.4 14.8 22.1 26.8 19.5
8 0 21.4 27.5
9 0 11.0 16.9 21.4 25.0 24.0
10 0 10.4 17.7 23.3 27.4 20.0
Average . . . . . 0 10.6 16.0 22.1 26.6 22.6

DILUTED SAMPLES OF CREAM

Winter's formula given in the "official methods" of the A.O.

A. C. for calculating the amount of added water in a sample of
milk is as follows:

100 (T - T')
Wffi
T
 OBSERVATIONS ON FREEZING POINT OF CREAM 357

W, is the percentage of added water; T, the average freezing

point depression of normal n~ilk (0.550°0.) or the original freez-
ing point depression, where this is known; and T', the freezing
point depression of the sample in question. This formula was
used in calculating the added water in a series of cream samples
which were diluted at the rate of 10, 15, 20 and 25 per cent re-
spectively. These samples are the same as numbers 6 to 10 in
Table 1. The freezing points of these samples are listed in the
first portion of Table 2, while results obtained using Winter's
formula are given in the second portion of the table. It will
be noted that the values obtained here are higher than the actual
amount of dilution. This is not surprising when the formula is
considered in the fight of the theoretical conceptions previously
mentioned. Winter's formula is accurate only when the volume
occupied by the substances in solution is the same as the total
volume of the material under observation. Actually then this
formula really gives the percentage of added water based on the
serum of the sample. Thus, the greater the difference in volume
between the serum and the material as a whole, the more in-
accurate will the results be when this forumla is used. In samples
of whole milk or skimmilk the discrepancy is not so large that
results are very far from accurate, especially where the average
normal value 0.055°C. is used for the value of T, but in cream
the differences are considerable as indicated in the table. In
using Winter's formula for these results T was taken as the freez=
ing point depression of the original undiluted sample.
In the lower portion of the table is shown calculated amounts of
water for the samples as in the middle portion, with the exception
that a modified formula correcting for the differences in volnme
between the serum and the cream was used. The formula is
as follows:
A(T - T') i
W-~
T
A, is the per cent of serum (by volume) in the sample having
a freezing point depression of T'. For all practical purposes
The C~joscopic method for cream has been accepted tentatively by the
A. O. A. C. and this formula is to be used.
358 F.z. I)o.~

the value for A may be taken as the per cent of water in the
sample. This is not exactly correct due to a disregarding of
specific gravity but is so nearly so that the difference is negli-
gible, especially where the value of T has to be taken as 0.550°C.
In this work each diluted sample was not analyzed for water
to determine the value of A but the figure was calculated from
the analysis of the sample before dilution according to the follow-
ing formula:
bOoo - c)
A~ ÷C.
lOO

Here b, represents the per cent of water in the undiluted sample

and C, the per cent dilution.
It is noted that the results calculated by the "serum" formula
a r e considerably closer to the act~ial amounts of water added
than the results obtained using ~ m t e r ' s formula. The average
algebraic variation in the former is but 1.4 per cent, while in the
latter it amounts to 7.3 per cent, roughly five times greater.
It is probably unnecessary to state that, when unknown sam-
ples of cream are examined, it is necessary to get the value of
A by analysis, as it would be rather difficult to calculate it from
the percentage of fat when the amount of possible dilution
would be tmknown.
CREAM S A M P L E S F R O M D I L U T E D M I L K

Sample No. 11 and No. 12 in table 1 were diluted before sepa-

ration. Sample No. 11 showed a freezing point depression of
0.536°C., before dilution and sample no. 12 one of 0.558°C. un-
diluted. Therefore, fairing table 1 as evidence the respective
cream s_~mples would have had like values in the undiluted state.
If the "serum" formula is used and the amount of added water
calculated for the cream samples, and in the same way for the
sklmmilk samples, the results are as follows:
Cream s a m p l e no. 11 . . . . . . . . . . . . . . . . . . . . . . . . 9.4 per cent added water
C r e a m s a m p l e no. 12 . . . . . . . . . . . . . . . . . . . . . . . . 16.8 per cent added water
S k i m m i l k s a m p l e no, 11 . . . . . . . . . . . . . . . . . . . . . 13,0 per cent added water
S k i m m i l k s a m p l e no. 12 . . . . . . . . . . . . . . . . . . . . . 21,5 per cent added water
 OBSERVATIONS ON F R E E Z I N G POINT OF CP~EAM 359

Actually there was added to the whole milk samples prior to

separation 11.5 per cent and 20 per cent of water respectively.
These values are closely approached when calculated from the
data by the "serum" formula being 12.2 per cent and 20.7 per
cent respectively. I t is to be noted, however, that the amounts
of added water calculated for the cream and skimmilk portions
of these samples do not agree with the amount added to the whole
milk, nor do they agree with each other. This of course, is due
to the fact that the amounts of serum in the cream and in the skim-
milk differ from the amount in the original whole milk and differ
more widely as the richness of the cream increases.
In order to determine the amount of water added to a lot of
milk prior to separation, from an examination of the cream or
the skimmilk, the volumes of each and the amount of added water
in each must be known, in which case the following formula may
be used.
(C X W') + (S X W")
~+B

C, is the volume of cream; W' the per cent of added water in

the cream; S, the volume of skimmilk; W ~, the per cent of
added water in the skimmilk and W, the per cent of added
water in the original whole milk. This formula naturally would
have its limitations when commercial samples are being inves-
tigated.
SUMMARY AND CONCLUSIONS

A very brief review of literature, only, is included since no

published data was found bearing directly on the subject of this
paper.
From data presented it is concluded that cream and sMmmilk
have the s~me freezing points as the original sample of milk
from which they are separated. If this is so then the normal
range must be from -0.530°C. to -0.566°C. as for milk, with
an average of -0.550°C.
Winter's formula as given in the "official methods" of the
A.O.A.C. is theoretically incorrect and while this fact does not
360 F. #. V O ~

ma.ke it entirely inapplicable for Whole milk or skimmilk it shows

a great degree of inaccuracy when applied to samples of cream,
the degree increasing as the cream samples become richer in fat.
D a t a is given showing the inaccuracies when it is applied to a
series of diluted cream samples.
A modified formula is given which is theoretically more ac-
curate and when applied to the same ss.mples of cream mentioned
above gives fairly close results as compared with actual amounts
of water known to have been added.
The smount of added water found in a cream sample or a
skimmilk sample does not correspond with the amount added to
the original whole tniltr, in case such diultion was made before
separation; but the latter can be calculated by the formula
shown if the weights or volumes of both the cream and skimmUk
portions are known together with the respective amounts of
water in each, as found by cryoscopic exRmination and the use
of the " s e r u m " formula.
REFERENCES
(1) HORTVET, JULIUS: Jour. Ind. and Eng. Chem., xiii, no. 3, p. 198. (1921).
(2) BAILEY, E. M.: Conn. Agri. Exp. Sta., New Haven Conn., Bul. 236~ p. 251,
(1922).
(3) Official and Tentative Methods of Analysis of the A. O. A. C., (1924).
(4) ATKINS, W. R. G. : Chemical News, xcvii, p. 241, (1908).
(5) PARFITT, E. H., AND TAYI,ER, C. L. : Jour. Dairy Science, viii, no. 3, p. 230,
(1925).
(6) PHILIP, J. C.: Physical Chemistry, 3d ed., Longmans, Green & Co., New
York City, p. 207, (1925).
(7) MATeEWS, A. P. : Physiological Chemistry, William Wood & Co., New York
City, p. 243, (1920).
(8) CLAYTON, WM.: Second Report on Colloid Chemistry, British Assn. Adv.
Science, p. 96, (1921).