CIHM
Microfiche
Series
(■Monographs)

ICIMH

Collection de
microfiches
(monographies)

Canadian Inttituta for Historical IMicroroproductions / Institut Canadian da microraproductions historiquas

Technical and Bibliographic Notes / Notes techniques et bibliographiques

The Institute has attempted to obtain the best original
copy available for filming. Features of this copy which
may be bibliographically unique, which may alter any of
the images in the reproduction, or which may
significantly change the usual method of filming are
checked below.

I A Coloured covers /
I — I Couverture de couleur

□ Covers damaged /
Couverture endommagde

□ Covers restored and/or laminated/ •
Couverture restaur^e et/ou pellicula

Cover title missing / Le titre de couverture manque

Coloured maps / Cartes g^ographiques en couleur

□ Coloured ink (i.e. other than blue or black) /
Encre de couleur (i.e. autre que bleue ou noire)

□ Cotoured plates and/or illustrattons /
Planches et/ou illustrations en couleur

□ Bound with other material /
Relid avec d'autres documents

□ Only edition available /
Seule Mition disponible

I I Tight binding nay cause shadows or distortion along
' — I interior margin / La reliure senie peut causer de

I'ombre ou de la distorsion le long de la marge

intdrieure.

I I Blank leaves added during restorations may appear
— within the text. Whenever possible, these have been
omitted from filming / Use peut que certaines pages
blanches ajout^es lors d'une restauration
apparaissent dans le texte. mais, lorsque cela i\aH
possible, ces pages n'ont pas ^t^ film^es.

Additional comments / Various pagfngs.

Commentaires suppl6mentaires:

Thl> item Is filmed at the reduction ratio checlced below /

Ce document est film< au taux de riduction indiquc ci-dessout.

L'Institut a microfilnr)* le meilleur exemplaire qu'il lui a
6\6 possible de se procurer. Les details de cet exem-
plaire qui sont peut-*tre uniques du point de vue bibli-
ographique, qui peuvent modifier une image reproduHe,
ou qui peuvent exiger une modiffcation dans la metho-
ds normale de filmage sont indk]ute ci-dessous.

I I Coloured pages / Pages de couleur

I I Pages damaged / Pages endommagies

Pages restored and/or laminated /
Pages restaurtes et/ou pellteul^es

Pages discoloured, stained or foxed /
Pages dteotortes. tachetdes ou piqudes

I Pages detached/ Pages ddtach^es

\7 Showthrough/ Transparence

I I Quality of print varies/

Quality inhale de rimpresston

Includes supplementary material /
Comprend du nrat^riel suppl^mentaire

Pages wholly or partially obscured by errata slips,
tissues, etc., have been refilmed to ensure the best
possible image / Les pages totalement ou
partiellement obscurcies par un feuiilet d'errata. une
pelure, etc., ont 6t6 film^ k nouveau de fa^on k
obtenir la meilleure image possible.

Opposing pages with varying colouration or
discotourations are filmed twk:e to ensure the best
possible image / Les pages s'opposant ayant des
colorations variables ou des decolorations sont
film^es deux fois afin d'obtenir la meilleure image
possible.

lOx

14x

18x

22x

26x

30x

1
1

12x

16x

20x

24x

28x

32x

Th« copy filmed h«r« has bean raproductd thanks
to tha ganarosity of:

Library
Agriculturt Canada

Tho imasaa appearing hara ara tha bast ruallty
poaaibia eonaidaring tha condition and iagibility
of tha original copy and in Icaaping with tha
filming contract apacif icationa.

L'axamplaira film4 fut raproduit grlca k la
g4n*rositA da:

Blblioth<qus
AgrieuHurt CwiMb

Laa imagaa suivantaa ont 4t* raproduitaa avac la
plus grand soin, compta tanu da la condition at
da la nattat* da l'axamplaira film«, at m
conformity avac laa condHiona du contrat da
filmaga.

Oririnal copiaa in printad papar eovara ara filmad
beginning with tha front covar and ending on
the laat page with a printed or illustrated imprea-
aion. or the back cover when appropriate. All
other origiftal copiaa are filmed beginning on the
firat page with a printed or iiluatratad imprea-
alon, and ending on the laat page with a printad
or illustrated Impreaalon.

The ieat recorded frame on eech microfiche
ahaN contain the aymbol ^^> (meening "CON-
TINUED"), or the aymbol ▼ (meaning "END"),
whichever appNee.

Las axemplairaa ortwinaux dont la couvarture en
papier eat imprimte sent film«s en commenfant
par la premier plat at enterminant soit par la
derniire page qui comporte une emprelnte
dimpreaaion ou dlllustration, soit par la second
plat, salon le caa. Toua laa autraa axempleirea
originaux aont filmte an eommenf ant par la
pramiire pege qui comporte une emprelnte
d'impreaaion ou dlllustration at en terminant par
le dernlAre page qui comporte une telle
emprelnte.

Un dee symbolea suivanta apparaftra sur la
darnlAre imoge de cheque microfiche, aalon le
ces: le symbole — »- signifle "A SUIVRE", le
symbole ▼ signifle "HN".

Maps, platea, charta, etc.. may be filmed at
different reduction ratloe. Thoae too lerge to be
entirely included In one expoaure ara filmad
beginning In the upper left hand corner, left to
right end top to bottom, ea many framea as
required. The following diagrama illustrate the
method:

Lea cartea. planchea. tableaux, etc.. peuvent ttre
fllmte i dee taux de rMuction diffirenta.
Lorsque le document est trop grand pour itre
reproduit en un seul clichA. 11 est film* i partlr
de I'angle 8up4rieur gauche, de gauche k droite.
et de haut an baa. en prenant le nombre
d'imagea n^caaaaira. Laa diagrammaa auivanta
illustrant la mtthoda. -

1 2 3

mxocan mouiTioN mr cnart

(ANSI ond ISO TEST CMAUT No. 2)

^:^PL lED ItVHGE I

1653 East Main Street

nccfmtt. Nn Yon, t4eo9 usa

("8) «2 - 0300 - Phon.
(716) 2U - JW9 - ra«

%.^-,

DXPABlllENT OF AOBIOULTDBI

DAIRY AND COLD STORAGE COMMISSIONER'S BRANCH

OTTA-WA.. OA.VJLI>Jk.

The Testing of Milk, Cream and Dairy By-ProducU
by Means of the Babcock Test

■T

J. F. SINGLETON

Ohuf Intpteior of Dmry PndueU

Bulletin No. 45— Dairy and Gold Storage Seriea

Determination of the Specific Gravity of Milk
The Percentage of Acid and Casein in Milk
Adulteration of Milk by Skimming and Watering
The Percentage of Water and Salt in Butter
The Percentage of Fat and Water in Cheese

■T

J. F. SINGLETON

Chief Inspector of Dairy Prodbcte

Bulletin No. 46— Dairy and Cold St.. age Series

Published by dirMtion of Hon. MARTIN BURRELL, Minister of Agrieulturo, Ottawa, Ont.

OOTOBKB. 181S

DIPAXnilllT or AQBIOULTUltl
DAIEY AND COI.D STORAGE C0MMI8S i:.R'8 BRANCH

OTTA'WA. CA1VAI>A

THE TESTING OF

MILK, CREAM AND DAIRY BY-PRODUCTS

BY MBANS OF

THE BABOOCK TEST

J. F. SINGLETON,

Chief Itupeetor of Dairy ProdueU.

BULLETIN^ No. 45

Dairy uid Cold Storage Seriet.

Published by direction of the Hon. Martin Burrell, Minister of
Agriculture, Ottawa, Ont.

89182—1

OCTOBER, 1915

LETTER OF TRANSMITTAL

Ottawa. November 2T, 1015.
To the Honourable

The Minister of Agriculturp.

8lH, — I have the honour to submit the raanu«cript for a bulletin on the tenlinir
of milk and its product! by meansi of the Babcock milk tciter, which hua been prepaml
under my din-ction by Mr. J. F. Singleton, Thief Ini«pp<"tor of Dairy I'ro<luct».

The growing interest in the testing of cots and the increasing use of thi>
Bttbcoclc tester in this connection call* for authoritative itiHtructinu in it«> niniiipiiln-
tion, and the need of a bulletin of this description for distribution by this branch
hu been felt for some time.

Mr. Singleton is an authority on milk t.'stiuir. his training, and his experience
•I instructor in milk testing at the Kingston Dairy School for several years before
joining the staff of this branch, having q'lalificd him to deal with the subject in n
■cientific as well os a practical manner. It is not claimed that anything fundamentally
new is presented in these poges, but the best practice* in the operation of the test
arc set forth with such clearness and detail, that the bulletin should make a useful
and reliable handbook for those who require instruction or information respecting the
testing of milk, crsam, skim-milk, etc.

I have the honour to recommend thot the ninniiscript be published ns Bulleti!i 45
of the Dairy and Coli "'♦orage series.

. -lave the honour to be, sir,

Your obedient servant,

J. A. RUDDICK,
Dairy and Cold Storage Comr'.issioner.

yl82— U

THE TESTING OF MILK, CREAM AND DAIRY BY-PRODUCTS

BY MBANH OF THK BABCOCK TK8T

IXTROnUCTORV.

The Biiboook tP't flrrivi-*! its niinw from the oriiriniitor, Dr. S. M. Babcock, of
the University o{ VViiconiin. Thia wiik the tint •utiufni'tnry Rhurt methcxl fur
detprtniniiiR tho perccntHgo of fnt in milk, and ninot! it wa* mmie public in 189<l it
hoR be<>n widely ndoptod, parti<Miliirly in the Unitc<l Stntt's, C'unadn, Aiittrnlin nnd
New Zealand. It hni tinrp, with uliitht modlfirntion, b(M>n ftuivpMftilly applied in the
tMtinff of <Ti'iini, •kim-inilk, htittormilk, whey and rheeoe. This te*t hat been of
imnwmw value to t'.ie dairy indutitry. ''nrp it ha» provide*! a prnctioiil mean* .if:—

(a) determining tho fut \< oduction of individual oowi;

<h) mukinit a more o<|uitublo divinion of checte factory nnd rrenmery
pro<<eed( ;

(e) detecting abnormal Iosm's of fat in such by-product* at tkim-milk,
buttermilk and whey;

((i) detecting ndulterutiun* #uch at watering uud skimming.

The te«t it quite cnay to o|)erate nt no extensive training on the part of the
operator it nece»«-iry, The simplicity of the teitt hns proliably boon uvor-fmphusizcd.
producing a corresponding carelessness nn the part I't' sumo operators, resulting in
inaccurate tests and tubsot|ucnt citioism of the niftii.Hl. Wiiih' tiie tcs* is simple to
operate, great care and accuracy muni be ixerrised in uil details of th irk, or the
results will be inaccurate nnd misleading. Tho necessity of euro an ■curm-y in
operating tho test cannot be over-emphasized. It liii" lieeii truly ?!il tliiit o|HT;iting

tho Babcui-k test there is more to leurn in cure than in |irin< ;i>li

THK TKSTINT, OF ., ILK.

Till! apparatus employed in ninl<ina' ii t>st ot ii Miniple of milk •■insists of tour
pi«>i-os : —

(«) pipett*;
(ft) tost iiottle;

(f) ucid measure or aeid burette;
(d) centrifuge.

The Ph'ktte.

The pipette is a glass instrument used to meisure tlie saiuple of iiiillv rcMuirod
for testing. Two different forms of pipette are in use wlileh are usually designated a-!
the "ordinary pipette" nnd the "automatic pipette."'

The ordinary pipette consists of a glass tube enlarged into a bulb about midway
between the two ends (Fig. 1). The lower stem of the pipette should be small enough
in diameter to pass readily into the neck of the test bottle. It should also be drawn
in to a small opening at the point, since if the opening is too large air bubbles will
pass up into the bulb when measuring the sample and thus make accurate measurement
impossible.

The upper stem of the pipette is marked and the pipette filled to this mark contains
seventeen and six-tenths cubic centimetres (17-6 c.c). The pipette should be so con-
structed that the graduation mark is low down on the stem and quite close to the bulb.

The construction of the automatic pipette is shown in Fig. 2. The novice will
measure the sample more quickly and possibly more accurately with the automatic
pipette than with the ordinary pipette. The automatic pipette is more likely to be
broken and is more expensive.

The Milk Test Bottle.

The construction of the milk test bottle is illustrated by Fig. 3. The bulb of the
bottle is about one and one-quarter inches in diameter, and should have a capacity of,
at least, forty-five cubic centimetres (45 c.c). The neck of the bottle is about four
inches in length and is graduated to read the percentage of milk fat when the test is
completed. The graduation consists of ten main divisions numbered from zero at the
bottom to ten at the top. Each division represents one per cent of fat and is sub-
divided into five equal subdivisions, consequently each subdivision represents one-fifth
or two-tenths of one per cent of fat.

The Acid Measure and Acid Burette.

The acid measure (Fig. 4) is a small cylinder graduated to contain seventeen and
five-tenths cubic centimetres (17-5 c.c). Frequently an acid burette (Fig. 6) is used
instead of the acid graduate. Each division on the scale of the burette represents
17-5 c.c. A stand with a clamp attached is used to hold the burette.

The Centrifuge.

The centrifuge (Fig. 6) is a machine for whirling the bottles in making the test
It is fitted with swinging pockets to receive the test bottles and when the machine
is in motion the pockets assume a horizontal position. The machine is usually driven
by either hand power, or by a steam jet or turbine. Hand-driven testers are usually
made to receive either two, four, eight, ten or twelve bottles, while the turbine testers
will usually receive either twenty-four or thirty-six bottles. The two-bottle and four-
bottle hand testers are not made with a covering frame. The larger capacity hand
testers are constructed with a frame (usually cast metal) which closes in the bottles
when the machine is ruiming, which aids in keeping the samples warm. Where
steam is available and the amount of testing to be done will warrant the additional
expense, the turbine tester is preferable as it keeps the samples hot while being revolved.

The Required Speed of the Centrifuge.

The speed at which the tester should be revolved is usually stated on the machine
pnd varies with the diameter of the circle described by the bottom of the bottle in
revolving.

The following table of speeds for machines of different diameters is given by
Farrington and Woll in "Testing Milk and its Products": —

No. of
Revolutions

DiaiTjeter of of bottle Diameter of

Circle. per Mln. Circle.

10 Inches 1.074 18 inches..

12 •■ 980 20 " ..

14 " 909 22

Ifl " S48 24 ■• ..

No. of
R-^- olutlons
of bottle
per Min.
800
759

TO J

693

It'. I.

Fig. 2— Up-to-
dnte Pipette.

Fig. 3— Milk
Test Bottle.

Fig. 4.

Acid

Measure.

Fig. 5.

When operating a turbine tester the speed must be ascertained by means of a
speed indicator, which is applied to the spindle of the machine. The speed is
regulated by varying the steam pressure used, which is indicated by a steam gauge
attached to the machine. The operator must determine what steam pressure on the
gauge will give the desired speed to the tester.

When operating a ' hand-driven tester the number oi revolutions which the
bottle makes, to each revolution of the handle, should be determined by counting.
The diameter of the circle described by the bottom of the bottle when in the horizontal
running position should be measured. By consulting the above table, the required
speed of the bottle is obtained. The number of revolutions per minute required as
indicated by the table is divided by the number of revolutions the bottle goes to each
revolution of the handle. The result will be the number of revolutions of the handle
per minute.

Fig. C-

Fip. 6.

Kij.'. 1).

For cxnmplo, if the bottle is foiind to complete twelve revolutions for each
revolution of the liandle, ami the dinmotor of the circle described by the bottom of the
bottle is fourteen inehes, wo find liy cnnsultini; the above tiihlo that for a fourteen-
ineh diameter, the bottle must revolve imO times per minute: 909 divided by 12 is 7C
(almost), wliieh is the number of revolutions required of the handle caoh minute.

The tester should be placed perfectly livel on a firm bench and l)e kept well oiled.

The Milk-Test Act.

The Milk-Test Act which came into force January 1, 1911, provides that all test
bottles and pipettes used, or sold to be used, for the testing of milk and cream in
connection with tlie Babcock test must bn verified by the Standards Brnnch of the
Inland Revenue Department, Ottawa. Glassware, when verified and found to bo
correct within a specified liTiiit of error, must "be ineffaceably marked with the outline
of a crown having within it the initial letter of the reigning sovereign." All milk
and cream test bottles and pipettes, now in use or offered for sale, must be so marked.

Making the Test.

Before starting to make a test of whole milk see that nil glassware to be used is
clean and bears the verification mark.

Secliiing a Representative Sample of Milk.

In testing milk it is necessary that tlie sample taken for testing represents the
average quality of the quantity of milk to be tested. If such is not the case, the result
of the test will, of course, be inaccurate and misleading. When milk is allowed to
stand for evena short time, cream rises to the surface and in order to thoroughly mix
the milk before taking a sample for testing, it is best to pour the entire quantity of
milk from one vessel to another several times. If the quantity is too great to permit
of pouring, it fchould be well stirred. After being thoroughly mixed a smaller qiinntity
(three to six ounces) should be taken out and pat into a separate vessel.

Samplino and Addikg Milk to the Test Bottle.

The sample should be brought to a temperature of 60° F. to 70° F. and then
poured from one vessel to another several times. Care must be taken that c'l the cream
mixes back with the milk and that none adheres to the sides of the vessel. In pouring
allow the milk to follow down the side of the vessel into which it is being poured.
By so doing there is less tendency to partially churn the sample which would render
the results less accurate. After the sample is thoroughly mixed, if using the ordinary
pipette, insert the lower stem of the pipette into the milk and by suction of the mouth
raise the milk above the graduation mark on the pipette. Quickly cover the top of the
pipette with the index finger, taking care to keep the top of the pipette and the finger
dry. By slightly removing the finger allow the milk to drop slowly from the pipette
until it comes exactly to the 17-6 c.c. graduation mark on the stem of the pipette.
The lower stem of the pipette is now inserted into the neck of the test bottle and the
sample allowed to run into the bottle. The last few drops should be expelled from the
pipette into the test bottle by blowing through the pipette.

If using the " up-to-date " automatic pipette, have the glass petcock of the pipette
open and insert the lower stem of the pipette into the milk By suction on the tube lead-
ing from the large upper bulb, draw the milk up into i e pipette until the lower bulb is
filled and the milk is overflowing from the upper stem into the large bulb. While the
milk is still overflowing into the large bulb, quickly close the petcock. The lower stem
of the pipette is inserted into the neck of the bottle and the petcock is opened to allow
the milk to flow into the test bottle. As with the ordinary pipette the last few drops
should be expelled from the pipette into the test bottle by blowing through the pipette.

It is a rather common practice for operators of the test to blow through the pipette
into the milk before drawing the sample up into the pipette. This should never be
done as air is incorporated in the sample wliieh will affeet the result of the test.

While the sample of milk is measured into the test bottle, the test is based on
weight. The 17-6 c.c. pipette will deliver, of average milk, a definite weight — eighteen
grams — into the bottle. Since tho weight of a given volume of milks of different rich-
ness is fairly constant, measuring with the pipette is quite accurate and does not
introduce any appreciable error.

89182—2

If several samples are to be tested, each sample should be given a serial number
and a test bottle mariced with a corresponding serial number for each sample. A part
of the bulb of the test bottle is usually frosted so that such a number may be written
on with a lead penciL

Aonixo THR Acid.

The next step is the addition to the test bottle of the acid which is used in
making the test. Commercial sulphuric acid with a specific gravity (Sp. Gr.) of
1-82 to 1-83 is used and should bo at a temiwrnturo of 60° F. to 70° F. The acid is
measured in the cylinder provided for the purpose which is graduated to contain
17-5 c.c. After measuring the correct quantity of acid into the graduate it is slowly
pr ured into the test bottle. When pouring the acid into the test bottle the bottle should
'lo held in a slanting position and if any drops of milk are adhering in the neck of the
bottle, the bottle should be slowly revolved so that the acid will carry the milk down
into the bottle. By holding the bottle in a slanting position the acid fl.nws down the
walls of the bottle nnd lies under the milk. The aoid should never be allowed to drop
directly on the milk, and after adding the acid to the bottle there should bo a distinct
line between the milk and ocid. If the bottle is placed on the table after adding the
ocid. care should be taken not to jar the bottle which will tend to partially mix the
milk and acid.

In case the acid burette is being used it is filled with a(!id to the top graduation
mark. The test bottle is held in a slanting position underneath the burette with the
top of the burette inserted into the neck of the test bottle. By opening the petcock
the acid is olio wed to flow into the bottle. When the surface of the acid in the
burette is lowered to the second groduation mark, the petcock is closed and 17-5 c.c.
of acid will have been delivered into the test bottle.

The sulphuric acid used in making the test is extremely corrosive. It is advisable
to have, at all times, a bottle of liquid ammonia nt hand, and in case of any acid
coming in contact with the clothing, a liberal application of the ammonia will
neutralize the acid and prevent the drstniction of the cloth. Should the acid come
in contact with the face or hands, wash immediately with cold water.

Mixing s-ae Milk wn Acid.

After adding the acid to the test bottle the milk and acid should be thoroiifihly
mixed, shaking the bottle with a rotary motion. The neck of the bottle must not
bo covered with the finger while mixing the milk and acid and care should be taken
to avoid splashing particles of milk or curd into the neck of the bottle. As the
milk and acid are mixed the milk is first curdled, then the clots of curd are dissolved
nnd disappear, the mixture turns a dark chocolate colour, and becomes hot, due to the
action of the acid on the milk.

Placing the Bottles in the Tester.

The bottle or bottles should, at once, be placed in the tester in such a way that
the machine will be balanced. If an odd number of samples are being tested an
extra test bottle may be filled with water and placed in the machine to properly balance
it. If the machine is not properly balanced it will not run smoothly and the bottles
are likely to be broken.

Keeping the Bottles Warm.

If the bottles are allowed to become cool, previous to, or while whirling, an
ini'oniplete separation of fat will result and the reading will be too low. There is,
of course, no difficulty in keeping the samples sufficiently warm when lining a turbine
tester. When using a hand machine in a cool room it is advisable to partially fill
the frame of the machine with boiling water before commencing the whirling.
When using the two, or four-bottle hand tester, which has no covering frame, the
j.oi'kets should be filled with boiling water surrounding the bottles.

WiiiKUNo THE Buttles.

The bottles nro whirled lor live minutes at the proi»r speed. As pointed out above
the bottle when in niotiuu uHsumcK a liorizoiitid pcjsitiou The bottle, rapidly revolv-
ing, is subject to a force which tends to throw it awiiy from tho centre. This is known
as " centrifugal force." This force is exerted most strongly on tlie heaviest parts of
the mixture which work to the outside, thereby forcing the fat, which is lightest, to
the centre. When the whirling ceases the bottle assumes a vertical position with the
fat on the surface. If the machine is not run long enough or sufficiently fast, the
separation of fat will not be complete.

ADDixa THE Hot Watei; to the Test Bottles.

Hot water is now added to the bottle to float the separated fat up into tho neck of
the bottle so that the percentage may be read.

Tlie turhine tester usually hiis a sninll pail attached for heatinp anil iiddiii;.' tlie
water to tlic test bottles. The pail is fitted with a small rubber tube leading from tlie
bottom of the pail, into the end of which is fitted a piece of glass tubing drawn to a
point similar to the glass of an eye-dropper. There is a spring pinch cock on the rubber
tubing to shut off tlie water. If using a hand tester and only a few samples are being
tested at one time, the pipette will answer very well for adding tlie water to the bottles.
If many samples are being tested it is advisable to provide such a pail as described
above and attached to the turbine tester as shown in Fig. 6.

IJain watei- or condensed steam is preferabh; to hard water as it will give a clearer
fat eolunni in the finished test. If hard water must bo used, a few drops of sulphuric
acid should be added to the water bei'ore it is heated. Adding acid to hot water is
dangerous as it is liable to be splashed on tho faoe. hands or clothes of the operiitor.

If a turbine tester is being used a temperature of 140° F. to 160° F. will be high
enough at which to have the water for adding to the bottle. If a hand machine is
being used th(> water should be made much hotter, in fact there will be no disadvantage
in usitig boiling water. As far as possible the work should be done in such a manner
as to have the t' iniier:>tiire of the siimi>les 1 ctwfen ^"l)'-' Y. nrnl 1+0° T". when the test i-
completed, aii<l the temiieraturc of the water used may bi: varied somewhat with this
end in view.

Snflicicnt water is added t) fill tlie brittle to the bottom of the neck and the bottle
whirh^d for one minute at the proper speed, after which water is again added until the
top of the fiit column in the neck is raised to about the eight per pent niark on the
scale. In adding water the second time it is advisalile to allow tho water to drop
directly on the fat in tlie neck of the buttle. The water passing through the fat column
tends to wash the fat and carry any inipu'-ilies. which may be in the fat, down into the
body of the bottle. The bottles siiould be ag.iin wliirleil for one minute.

^fany operators add all the water at one time but e!;\Trer readings will usually bo
olifained by adiling the water iwiee as outlined abo\e. If the water is all added at once
the bottles sliould he whirled for two minutes nfter adding the water.

Ti'Mi'iRATLitK 01" Fat wiikn Rkaiunt. tur Test.

The Third Dominion Conference of Dairy Experts held au Ottawa, December 6
r.nd T. I'.'ll, a'lopti'd a tempiratnre of l;ii)" F. to 140° F. .ns tho projior one at wl.ii'h fi.
have the fat when the percentage is read. If the fat is too hot when read, tl,e result
will be too high owing to expansion and if the fat is too cold the result will be low
owing to contraction.

Thk Fse of a Watei? Bath.

In order to get aeenrate residts and uniformity from time to time, the bottles
should be placed in a water bath at a temperature of 130° F. to 140° F. for, nt least, two
minutes before the per cent of fat is rciid. The water surrounding the bottles should

S9182— 2i

extend up on the neck nf the bottle n« high ai the top of the fat column. If inuiiy
■amples are to be tested, it is well to have a rectangrular tin diah (Fig. 7) made to be

Fig. 7.

used as a water bath for the bottles. This dish should be about one-half inch less in
depth than the length of the bottles, so that the bottles cannot be overflowed with
water. A false bottom in the bath, with round holes in it, or wire rack to receive the
bottles, will prevent them from being overturned in the water.

Reading the Test.

If the work of making the test has been properly performed, the fat column will
be a bright amber colour, free from any dark or curdy specks. The top of the fat
oolumn will appear slightly hollow or concave and the bottom of the column will
iippear slightly rounded or convex (Fig. 8). The bottle should be held level with

B ^2

-1

-0

Fi(f. S.

the eye to be read, and the reading taken from the extreme points of the fat column,
that is from A to B and not from C to B. By reading from the extreme points,
allowance is made for a slight amount of fat which is not raised into the neck of the
bottle, and the results will correspond more closely with chemical analyses.

The reading of the fat h most conveniently taken by means of a pair of dividers
(.Flu. 8)i with fine, ihurp points. The hinf^e nf the dividers must bo stiti enough that
the dividers will not work too freely. Th(> dividers are »pr<«(l iiiitii the pnintft are
farther apart than the fat column is long, then one point is placed at the extreme lower
end of the frti column, and the divider^* slowly <'liised until the other point is ut the
extreme upper eid of the fnt coliinin. One point of the dividers is now placed at the
zero mnrk on the scale and the other point will indicate on the scale the correct
readincr. For instance, if the length of the fat column, ns shown by the dividers, is
three main divisions ond four subdivisions, the reading is 3% per cent. This may also

Kilf. !l.

be expressed as 3^jn per cent or 3-8 per rent. This would mean that iu each one
hundred pounds of such milk there are 3-S pounds of milk f:it.

Cleaxisc the Bottlks.

As soon as the readings of fat are tnkon tlio bottles should, while still hot, be
emptied and rinsed with warm water. Tlii-.v should next ho thoroughly washed with
hot water to which has been added n little of some good washing compound, using the
brush provided to clean out the neek. The bottles should be again rinsed with hot
water to thoroughly remove the washing compound from the bottle. Even with such
washing the bottles will in time become coated on the inside. To prevent this an
occasional cleaning, using a very strong sohition of washing compound, with some
shot in the bottle, is advisable. Sulph\irie acid, to which luis been added as much
bichromate of potash as the acid will dissolve, makes an excellent material for
cleaning any dirty glassware, and the same solution msiy be used many times.

DAnK-C0L0UI!ED OR BfliNT KkADINOS.

If, when the test is completed, the fat column contains black specks or is too dark
in colour, the test is not satisfactory and a duplicate test should he made.

Dark-coloured or burnt readings may be due to one or more of the following
causes :—

(a) the milk, the acid, or botli. iii'itig at tno liish :i temperature (over 70°) ;
(6) too much or too strong acid;

(d) allowing the bottles to stcnd too long after adding the acid before
mixing.

LiGi(T-Cou)ti;i:n or Cuhdv Ukadixos.

If ttlion the test is completed the fat column is too liglit in colour or shows curdy,
the results of the test may be too high and a second test should be made.

Light-coloured or curdy readings arc due to one or more of the following
causes :—

(a) the milk, the acid, or Loth, being at too low a temnerature (under 60°) ;

(b) too little or too weak acid;

Acid supplied for testing is not always of the proper strength. If tho test is
carefully and properly made, a dark-coloured or burned reading indieait* ikat the aeid
M too lirong, whiln a light-coloured or curdy reading indicatei that the acid it too
weak. If the auid is only sliithtly too strong, satisfactory results may be obtained by
using somewhat less than 17S c.c, ond if the acid is only slightly weak, using a little
more than 17-5 c.c. will give satisfactory results. Acid that is much too strong or
much too weak cannot be usod satisfactorily. Acid, if left exposed to the air, becomes
weaker by absorbing moisture from the air, consequently acid of the correct strength
should be kept stoppered when not in use. Acid, which is too strong, will in time
weaken to the correct strength if left uncorked. If a cork stopper is used, the acid
will char the cork and the acid becomes dark. A glass-stoppered bottle is prefenible.

Composite Samples or Milk.

A composite sample of milk is u quantity of milk composed of several smaller
simples taken from different sourct-s and should represent the average quality of the
different quantitieg from which the samples are taken.

Cheese factories, which divide the proceeds on a basin of the test, use the com-
posite sample and the great mojority test only monthly.

A tightly stoppered bottle is provided for each patron, and some means provided
to identify each patron's bottle. A convenient way is to gum a label bearing the
patron's nome, or a number to designate the patron, to each bottle. If the label is
covered with two coats of shellac, the bottle may be washed without injury to the label.

The Use of a Preservative.

Some chemical is used as a preservative to prevent souring and other fermenta-
tions. Since nearly all strong preservatives suitable for this purpose are very poison-
ous, some colouring matter is mixed with the preservative which gives the sample a
distinctive colour and thereby indicates that the sample is unfit for use as a food.
Preservative may be purchased in tablet form from the dairy supply houses. These
commercial tablets are usually composed largely of corrosive sublimate and are very
satisfactory. Powdered corrosive sublimate is also very efficient. If this is used a
small proportion of magenta should be mixed with the corrosive sublimate to colour
the sample.

In case one is troubled with mould growing on the walls of the bottle a few drop«
of formaldehyde may bo added to the sample. This will prevent the growth of mould.

The Quantity or Preservative to Use.

The preservative is added to the bottle before any sample is put in. No definite
quantity of preservative con be said to be the correct quantity. The correct quantity
to use is the least that will preserve the sample efficiently, and this depends on :

(a) the quantity of milk that will be in the sample bottle;

(b) the length of time over which a sample extends;

(d) the degree of ripeness of the milk composing the sample.

An excess of corrosive sublimate affects the casein in such a manner that it spcma
more difficult to dissolve and more shaking is required in mixing the milk and acid in
the test bottle.

Sampling Milk for the Composite .,r.

The sample of milk added each day to the composite jar should not only represent
tlie average quality of the quantity from which it is taken, but should also be propor-
tionate to the quantity.

Two methods of taking the sample are in common use:

(a) the small conc-shapotl dipper (Fig. 10) ;

(b) the sampling tube (Fig. 11).

Kijc. 10. Fig. 11.

The small cune-sliupvd or " ounce dipper " as it is commonly called does uol take
a Bani|ilo proportionate to the (|uantity of the milk being sampled. Since milk
deliven»d at ii ohceso fui-tory is well mixed from the iiKitiition received on the wiigon
and by i)ouring into the weighing cun the sample taken by the ounce dtp|)er will
represent the <iuality of the milk <iuite accurately.

The ounce dipper is much more commonly used in cheese factory work than the
sampling tube owing to its greater convenience. Since the quantity and quality of
milk delivered by a patron to a cheene factory is fairly uniform from day to day, the
use of the ounce dipper in taking samples for the composite test will not introduce
any serious error.

Sampling Frozen Milk.

lUilk should not be allowed to freeze, but in severe weather partially frozen milk
is occasionally received ot cheese factories. In freezing, the fat and other milk solids
not in solution are to a great e-xtent forced out of the ice. The ice of frozen milk will
frequently contain less than one per cent of fat, the fat l)eing largely in the unfrozen
portion. Consequently partially frozen milk should not be sampled until the ice has
been melted and the whole quantity thoroughly mixed. It is very diiRcult to secure a
uniform distribution of fut in a quantity of milk which has been allowed to freeze.

Mixing the Composite Samples.

After adding the sample to the composite jar each day, the sample should be
mixed by shaking the bottle in a rotary motion, care being taken not to splash any
creari up on the walls of the bottle. If any clots of cream should be splashed on the
walls of the bottle, the agitation should be continued until it is completely washed
dawn.

Care of Composite Samples.

Composite samples should be kept in a cool place nnd not exposed to sunlight.
Neither should they be exposed to frost as it is extremely difficult to get a correct
sample from the jar if the samples have been partially frozen. They should be kept
tightly corked at all times since if not tightly stopper.^ evaporation of water takes
place which will result in the test being too high.

Preparing Composite Samples for Testing.

The composite sample should be prepared for testing by warming to a tempera-
ture of 100° V. to 110" F. in order to soften anv dots of orcam nnd to rf move all cream
from the wi\lls of the composite jar. The temperature of the sample should not be
allowed to become high enough to melt the fat into oil as it is then difficult to get a
representative sample from the jar since the oil will quickly rise to the surface. The
sample should be carefully poured from one vessel to another several times and the
sample immediately taken for testing.

Tnma Thick ur Curdled Milk.

TmtiitB thick or curdled milk if not to b« recommended ai it it mure diiRcult to
cot • correct sample. However in hot weather a namplo may curdle due to inaufficient
pr«M>rrutive, or to over-ripe milk having [m'u .uhM to tho jar. In mcli ni«is a vorv
imall quantity of powdered lyo may be added to the sample and the tamplo poured
from one veaai'l to another several times. The lye n-'utrnlifcs the acid and when suffi-
cient acid ii neutralized the milk airain becomet liquid. The lyo ohoiild be added in
small quantities, pouring the sample several tinien after each adilitioii of lye. In this
way the use of an excessive quantity of lye may \>e avoided. When the lumps of curd
disappear and the sample becomes fluid it is samn1e<l in the us\ml way. Samples
treated in this manner require less acid than a normal sample.

THE TESTDIO OF CREAM.

DrrnutNATioN or tub per cent ok Fat m Cream.

Cream is that ortion of mi'k, rich in fnt, which rises to the surface of milk on
ttandinir, or is sej rated from it by centrifugal force.

The Babcock test is used to determine the per cent of fat in cream as well as in
milk. In determining the per cent of fnt in cream certain modifications of the method
already outlined for determining the per cent of fnt in milk are necessary.

Cream Test Bottles.

In testing cream specially graduated test bottles i.ro used. Several different styles
of buttles (fig. 12) are on the market, b'lt those most commonly used are: —

(a) the six and one-half (6i) inch bottle graduated to read fifty (50)
per cent of fat, using a nine (1)) gramme sample;

(b) the .six and oie-iialf (.I'i) Im-'h bott.e giuduatcd to read either forty
(40) or fifty (60) per cent of fnt, using an eighteen (IS) griimmo .--.iTr.ple;

Each main division on the graduation scale of these bottles, representing uno
per cent of fnt, should be subdivided into two equal subdivisions, each of which
represents one-half of one per cent of fat. This is not always done and on many
bottles the smallest divisi' a of the scale represents one per cent of fat.

Whether the bottle i.. constructed for a nine (1>) gramme or an eighteen (18)
gramme sample, the bulb should have a volume, at least, eiiual to that of the ordinary
ten (10) per cent iri"t bottle.

Both tho six aua one-half-ineh nine-gramme bottle nnd the nine-inch cightcen-
grnmnio bottle have necks of smaller bore than tho six and onc-lialf-incli eighteen-
grnmmo bottle. In tliis respect citiior of tho fnrmcr is prefiTaMe to the latter, since
the smaller the bore of the neck tho less error tlioro is likely to Ih? in reading tho per
cent of fat. In using tho nine-grannno buttle, however, ouy error in wei^jhing tlie
sample produces twice .ns great an error in tho tost as tho same error in woighiug
wi'l produce in tho eii,'liteen-grammo bottle. This is roaiHly soon from the following
calculation : —

If a nino-gramme sample in r. niiie-granimi^ hottle rends W por cent fat, each onf"
gramme of the sample roads r5C^!)=4 por oout fat, nnd onoh ono-half gramme roads 2
per cent fat. If an eighteen gramme sample in an oifthtoon-grnmmc bottle rends Of)
per cent fat, c ono gramme of tho sample roads 36~lS=-2 ikt cent fat, nnd each
one-half gran --nds 1 por rent f.<it.

That is. ai. . i. <r of one-half gramme in weighing tho sample produops in ono cn-ie
an error of 2 per cent in the rondin? nnd in the other case an error of 1 per cent.

In utinc th« nins-tnch bottle a ipcciully roiiitructoil mm'hiiu* U roquirm), which
It mow expentiva than th« ord'nary te»tiT. The ninu iiu'h b«)ttU»* uro more expeniivo
than tho lix and one-half-inrh ufittlva. and br«<ukMK<-» nrv more freqiimt. not only
when whirlini the bottlea but al»o in handling, linre tho loniiir bottle i* mure easily
oTitrt limed.

The nine-Rrarome bottle haa an advantasn over the eiBhtwu-Kriimnio Uttlo in thai
nine <Mibi<' wntiniftrw of water aro mixi-d with tho nino urNmnMi) of rn-nm. whirh
ti'nd* to iriTe a clearrr rending. '

Whi.'h bottle one ihall uw it Inwly a mnltiT of prpfrren.c on tho pnrt of the
operator, nt uny one of th<- thrw will Rive »ati»fii(tc.r.v ro«iiU* if tho work it projicrly
done.

To tecure iiniform rt'snltt nil botth-n unod in iiiiv plnnt thoiild b*- mndo of uniform
conttriiftion, nil nmdo to ti-tt the tnnio woitrht of miniplo and wiih uniform grnduntion
and dinmoter of bore.

SoracKa or Eaaon is Mkahiriso Car.AM rou TruTiso.

A* pointed out proviouMy the Bul«'o<-k tint i« haneu on weiyhl and it it tiniply for
convfiiicmv thn' the tuur ' • i* nieanunHl with tin- pi|M'tto in letting milk. Tho pipette
I'.-livert nptiroxinintfly tho tame wt'ittht of milk from time to timn. Severnl factort,
howfvfr, tend to render mcanurinn of cream by nii'nnt of a pii>etto iiinccurnte, Theae
are:—

(u) Variation in richncs* of cniim.—Crenm iiiuy t»'«t at low at 15^^ of fat or aa
Liph lit 50', . / the p«'r cent of fat in cream iiicrraMii, the weight of a given vilurae
(Iccrenscs. Therefore a iiipctte, which will mcnoure a sample of the proper weight from
a low letting cream, will mcnture, from a richer crcnm, n tnniplo that will be too liifht.

(b) Gat and Air in the crMW.— More or Kh gns duo to touring or other fcrmcutn-
tiont i* prcM'nt in crcani. Tho hcav.v body of the crenm tendt to retain theto gates in
the cream niid therefore reduce the uriulit of cream which will ho mcniured by a
pipette. This will not hnvc to great an intluenco if the cream it warmed before 'he
tnmi>lo it measured with a pipetU'. a* the wariniiik' reduces the body of tho cream no
\hat the gat will, to a great extent, escape.

This is illustrated by the following table which also illustrates the difference
between the resuHs obtained when the scale is use<l to weigh tho sample as compared
with nioasurinjf with the pipette.

The different siilli|rlcs of tweet cream were tested u-ing l)oth scnlet and pipette
and were then all.wi'd t.. snur in tijrhtl.v stoppered bottles, after which they were again
teste.1 while cohl, usiiip both scales and pipett4'. The si.niples wore then warmed to
100" F. and tested usinjt the pipette.

TABLE.

s»>.

1 =^'

iir CrMiiii.

Cold Crea
I'ilirtti-. i

(X..i.-.t.)

III.
McbIim.

nriii Cn^aiii .

Scnie*.

I'ilH'tt...

Pipt'tti'.

:i(i

3!)

St) 5

:m ^

.^■« .'•.

:«i 5

.11 .•.

31 .'.

3'-'

31 ft

1)..,

3J 6

3i;

.<« 5

:« .-•

36 3

?<,

:v.r>

:i."i ."i

«i

31 .-,

...

31 .1

:« 5

34 5

.■M

:«;

:«

2!! :>

.to

29 5

Thcie HgurtM »n typionl of iIm< rmulu obUiniHi in Mveral mure luch taat*. Prao-
ticalljr no iliffentniw, bpyoiiil • n>a*oiialtl« limit <*( vrrur, ii iiotin<<l Iwtwrvn Ute t«t • of
lbs Mtno crpam iwcet or auur, when th« wtur >ain|)li> i* warnM>il to IW) doKrco* befor*
•amiilinic. thuuffh in lunie fttiri a ilivhtly lowir n'mlinic i* imlifttl wh> lh« ■mii.tliv
wa« tulii'ii witliuut warniiiitf. ] » i\\titv |K»iiil)U> that in iiM-i-ial ca^i — with very g»»*y
crviim- thi< error would nhow •n-ntcr than in thiiH> fliruri'x, a* ilu>ttfi numitli'* lumnHi
quito i'ifan in Huvour.

(!•) Cri-am mihrrxHO to thf wullt of lh<f pii>illr. Sonm cn-iini mlhi-ri>4 !■> the walU
uf tliii |ii|>i'tti* untl if thin in not thoruuithly riiiM'd olT and uddi-d to thu text bottle tlio
mult uf thu ti'Kt will be inoccuratc.

Thk I'm*; or ScALKM iciii WKHiiiiNo Sami'lkm or Ckkvm.

In onliT to HViiid iUvm- Miurifn of rrrur in nu'uxurinir i'ri>uni «uni|ilt'>, m-alut huvv
IxH'n dfvixi'il f<ir weiKhinir th)> creiini minipli-M of fithrr nino (IM |{riitnnii"< i>r riuhiirii
(IH) (rrninnitM into tln' iNittliM, Crcnni t«'i>tinif wmIim nrc niii-itriii'ti'il i>f iJilTiri'iit
(>ii|iiicitii'», S(inn> Willi'" liiivr i ivipiirily nf Inilvr I. "I liutllr* ( l-'iir. i:t), rhnt \* iti'iimi

samples may be wuigiieU into each of twelve bottles with oi:e balaiifins; t' the scales.
Other M-ales have eapacifies of four bottles, two bottles or one buiile ( i . U). The
fewer the Im'tles the seuleg will earry the more accurate will be 'be v.i'KfhinK ami the
loiiRer will the work require since the scales must be balaiioeil nidro frequently. The
scales should be kept in u dry place to protract the bearinprs frnii: rust which would
soiin render the .scales inacciir :it<'. When in use the scale sbimbl lie p'ac.vl i>m a firm.
Iri-fl ,bi'lf or tabl.'.

In usinjj the twelve-bottle scale the bottles are inimlierix' and placed on the scale.
The weisht on the weiRhinpr Iwam is placed ai the extreme left notch on the beam
and the scale is then balanced by niovintr tl:e ball on the tare beam. The b.nll or weifrht
on the wciRhinfir beam is then moved to the ritrht, to the nine CO) (rramnie or eiirliteen
(IS) (tramnie mark, deiwndiiiK on thr style of bottle beins u.sed. and cream is added to
bottle No. 1 until the scales aprain bnlanco. Fsnally the weight is ngnin moved to
the rijrbt and the weiffhinjr repeated into battle No. 2. This is repeated until the
bottles on tlie left pan of the scale each contain a sample, when the weight on the
beam is moved to the left and samples weijrlicd into the bottles on the right arm of
the scale. A better practice is to wi'igli a sample intf n bottle on the left pan, then

89182—31

move the weight back to the U'ft and weigli u suiiiplu into a bottle on the right pan,
weighing alternately into bottles on the right and left puns until all bottlea contain
samples. The four-bollle scale is used in the same manner as the tweUo-bottle scale,
except that nine (!)) and eighteen (IS) grnmiiie weigiits are used on the pans of the
scale instead uf the weight on the notelicd beam. With the one-bottle scale (fig.
14) after balancing the scale with the bottle placed on the left pan a nine (9) gramme

ur eigliU'on (!>) i:rrainnic \vi iuht is placed nu tlic ri^bt pun of the scale and the sample
is tlu'ii woiiibcHl into the bottle.

SaMPI.INC Cur.AM l""l! Tkstino.

In sampling ereain for tesliiig siiceial prd'aiitioiis must be taken to secure a
sample which represents the average quality ti> lio tested. It is more difficult to get
a representative sample uf a quantity of iTi'ain than it is to <;t'.t a representative
sami (■ of an eiinal • ;antity nf milk, since tbi> rre:im ibu^s not mix as readily. If
circumstances pern)it it is lu'st tci pour tlio quantity of cream from one ves.sel to
another several tiinos. If pouriiiir is nut practicable, tb(> (ream should be thoroughly
stirred by means of a stirrinir rod (Fiir. l.'i') Ci.n>tructod for the purpose.

i"i

i:.

If the quantity of cream is thnrouiihly mixed a small dipper (Fig. 10) will he
(piite satisfactory fur lakiiii.' tlic s:nuple. Owing to the difficulty of thoroughly
mixin.ii a can of cream, several special ileviees have been recommended for taking
the sample. Of tbe-e the " McKay Sami)ler " (Fig. 11) is i)robably the most satis-
factory. This is ('(instructed with two sliitte(l tubes, one inside the other. The tubes
are turned to clo^c the slot and the ':auq)ler inserted in the cream to the bottom of
tlie can. The slot is. then, opened to admit the cream to the tidie, after which the
slot is closed and the tube withdrawn. The sann.do thus taken is a small column of
cream exteiuling from the bottom of the can to the surface of the cream, and is repre-
sentative both of the quaJihf and (luantiti/ of the cream sampled.

Pbepabisg and Weiohino the Sample for Testing.

The sample of cream to be tested should be warmed to remove any lumps from
the cream. Should lumps be pre?c>nt, which will not disappear upon warming, the
sample should be poured through a fine wire strainer and the lumps broken up nnd
forced through the strainer. The sample is then carefully poured from one vessel to
another several times and, by menus of a pipette, cream is transferred to the bottle
on th(> scale until the scale balances exactly.

If usinpr a nine (9) gramme sample in a nine (9) gramme bottle, nine cubic centi-
metres of water should be added to the 8nmi>le in the bottle from a nine oubic
centimetre pipette. The full quantity of acid (17r>ec.) will be required with a
nine-gramme sample, if nine cubic centimetres of water have Seen added to the bottle.
The addition of water in the bottle usually gives a clearer fat column. For the same
reason it is advisable to add n few cubic cpiitimptros of water to the ciarhteen-
gramnip paniplo. The addition of watnr, howrver, ncrrssitntns the use of more than
the usual (|uantity of arid, and care must he ta!:en not to add toi mncli wnt(^r to the
eiglitc'di-prammo sample, as the bulb of the botth^ will not b(^ la'ge rnoiigh to hold
the extra acid required in addition to the water.

Measuring Cream Samples with the Pum:tte.

When strict accuracy is not essential fairly correct results may be obtained by
using a pipette and measuring the cream into the test bottle. Since cream is lighter
than milk, the pipette used for measuring the sample into the eighteen-gramnie bottle
should be larger than that used for testing milk. A pii)ette with a volume of eighteen
cubic centimetres is used irt connection with the eighteen-gramme bottle, and after the
sample has been measured into the bottle a few cubic centimetres of warm water
should be used to rinse the pipette, which rinse water is added to the bottle. For
measuring the sample into the nine-gramme bottle a nine-cubic centimetre pipette ia
used and nine cubic centimetres of warm water is used to rinse the pipette and is added
to the sample in the bottle.

As previously stated, the presence, in a cream sample, of gases due to souring or
other fermentations, or of air incorporated by pouring, while introducing no appre-
ciable error when the scales are used, will produce an appreciable error if the pipette
is used. The presence of air and gas in the cream lessens the weight delivered by the
pipette. Warming the sample reduces the body or thickness of the cream, facilitating
the escape of the gas or air from the sample and to a great extent prevents error from
this source. For this reason especial attention should be given to the warming of the
sample when the pipette is to be used.

Reading Cream Tests.

Especial care should be taken to have the fat at a temperature of 130° F. to 140°
". for reading. Owing to the volume of fat present in the nwk of the cream bottle,
considerable error may be introduced by having the samples too hot when read. Cream
samples also require longer than milk samples to become adjusted to the temperature
of the water bath.

Unlike reading the tests of milk, the reading of cream tests is not taken from the
extreme points of the fat column but from the bottom of the fat column to the
bottom of the meniscus on the surface of the fat column (Fig. 17). The reading
should be taken from A to B, not from C to B.

The Use of Oil in Reading Cream Tests.

Owing to i' difficulty of determining where the bottom of the meniscus is, a few
drops of a lig? <loured oil is frequently added to the top of the fat column. This
oil must be lighK>r than the fat, so that it will not mix with, but float on top of the fat.
The oil is conveniently added to the bottle by means of a pipette, allowing it to flow
down the wall of the neck of the bottle. The miniscus is raised up on the surface of
the oil leaving a sharp distinct line between the fat column and the oil. The reading
should be taken from this line to the bottom of the fat column.

This oil may be secured from any of the leading; dairy mipply lionsos, or may be
prepared by any one requiring it. A mineral oil sufficiently light that it will not mix
with the fat must first be secured. An oil sold, by one of the leading companies, under
the trade name " Albolite " is very satisfactory. Alkanet root, which may be obtained
from any good drug house, is used to colour the oil. One ounce of alkanet root will
colour one gallon of oil. The alkanet root should be rolled in doiiblo ply chctY-o cloth
and soaked for twenty-four hours in the oil. The alkanet root is then removed and
the oil will be of a light reddish colour and ready '.'or nse.

Composite Samples op Cream.

While many creameries, and the number is increasing test each delivery of cream
received from each patron, many still use the composite samples and test either once
or twice each month.

What has been said regarding composite samples of milk will also apply to com-
posite samples of cream.

The ounce dipper is not as satisfactory for sampling cream as for saii4>ling milk.
Since it is more difficult to thoroughly mix cream than milk and since the deliveries of
cream shew greater variation both in quality and quantity than deliveries of milk
^ome such device as the McKay sampler already described should be used.

Evaporation of Water fri Composfte Samples op Cream.

Especial care should be taken to see that the sample jars are tighly corked to prevent
evaporation of water from the jar, which would cause the samples to test too high.

To illustrate this point, some years ago the writer after testing a number of
composite cream samples at the end of a month left the s.implcs uncorked in n cold
room for seventeen days and again tested the samples. The tests were as follows: —

At end of month. Seventten day later.

A 40 48

B 25 88

C 40 4S

D 30 81

E 21 24

F 30 82-6

G 29-5 34

These figures illustrate plainly the necessity of keeping the conipoaite jars
tiphtly corked. The extreme variation in sample A, is accounted for by the fact
that there was very little cream in the jar and the evaporation was higher proportion-
ally than in the other jars.

Testing or By-Phoducts.

DETERMINATION OF THE PER CENT OF FAT IN 8KIM-MILK AND BUTTERMILK.

Under the moat favourable conditions of separation and churning of crenm there
is some fat lost in the skim-milk and the buttermilk. The Babcock test may be used
to determine the extent of these losses.

FiR. 16.

40^^

10-

30—^

c5 — =

20 — =
15-==

5-^

rig. 17

THK DOUBLE-NECK BOTTLE.

A specially constructed double-neck bottle (Fig. 16) is used. The larger neck
is to admit the skim-milk or buttermilk and acid to the bottle and should be conducted
down close to the bottom of the bottle. The smaller neck is graduated to read the
percentage of fat. The older st.yle of double-neck bottle was usually graduated so
that the entire scale consisting of ten divisions represented one-h.alf of one per cent
of fat. Each division would, therefore, represent five one-hundredths of one per
cent of fat.

"Most double-neck bottles now on the market have either five or ten main divisions,
each of which represents five one-hundredths of one per cent of fat and is subdivided
into five equal divisions. Each small division, therefore, represents one one-hundredth
of one per cent of fat.

In testing skim-milk or buttermilk especiul care must be token to hare all glasa-
ware perfectly clean, as a slight amount of fat in the pipette or bottle would seriously
affect the results.

A 17-C cc. pipette is usp«l to measure the saraph and 17-li cc. of arid is urdinarily
used. It is probnble thnt a sliphtly more perfect separation of the fat will be obtiiinod
if the milk and acid are cooled below 60°F. nnd nlimit 2flcc. of ncid used.

Care must be taken in niixiiiR the milk nii<l noid to avoid olosinR the jfradnated
neck of the bottle with small pieces of curd. If tliis occurs the mixture of milk and
acid will be forced out of the other ncrk nnd the work must then be repeated. It is
a good practice to add half the acid, and mix, then add the remaining half of the acid
and mix again. The water must be added plowly to avoid forcing the fat out of the
neck. If nece?9nry when reading the tp<<t, tho f.it i\in be raised by pressing the finger
over the opening of the larger neck.

In testing skim-milk or buttermilk the fat is not all recovered in the neck of the
bottle. Some fat remains in the mixture of milk and acid. It has been recommended
that five onc-hundredths of one per cent be added to tin- reading shown on the bottle
to allow for this error and thus make tho test correspond more cln.Nely with chemical
analysis. However, since the fat remaining in tho mixture in the bottle cannot be
recovered by mechanical means and is, therefore, not avnilnble for the manufacture of
cheese or butter, and since the testing of these by-products is largely to determine the
comparative loss from time to time, such addition to the reading would seem to be
unnecessary.

DETERMINATION OF THE I'ER CEST OF F.VT IN WHEY.

The double-neck bottle and 17-6 cc. pipette are used in testing whey. Care should
Iv taken to have the temperature of the whey well down to 00' F. Owing to the fact
tLat part of the milk solids have been removed in tlio process of cheeseninl;iiig less
acid is required, and usually felightly over half a measure of ncid will be sufTleicnt.

THE AVERAGE COMPOSITION OF MILK, SKIM-MILK, BUTTERMILK AND WHET.

Milk.*

Skiminilk. +

liiitUriiiilk.t

Wh.y.t

Water

Fat..

Per cent.

87-5
30
2 5
07
60
07

Per cent.

90 30
ID
2-75
80
5 25
80

Per cent.

HO
1
2-8
8
4 4
7
C

Per cent.

93 40
,«
010

&':;:::.•;;. .::.:-.;.-.v:;.v:::;-::.

Lactic Acid

480

(ill

'Dean. fVanSlyke.

DEPARTMENT OF A'miOIjLTURE

DAIRY AND COLD STOHAGK rOMMISSlONKkS BRANCH

OTTA'WA, fANAJ)A.

Determination of the Specific Gravity of Milk ;
the Percentage of Acid and Casein in Milk ;
the Adulteration of Milk by Skimming and
Watering ; the Percentage of Water
and Salt in Butter ; the Percent-
age of Fat and Water
in Cheese

J. F. SINGLETON

Chief Inspector of Dairy Products.

Bulletin No. 46--J9airy and Cold Storage Series

Published by direction of Hon. Martin i3urrell, Minister of
Agriculture, Ottawa, Ont.

OCTOBEK, 1915

LETTER OF TRANSAflTTAL.

Ottawa, November 27, 1915.
To the Honourable

The Minister of Agriculture.

Sib, — I have the honour to submit the manuscript for a bulletin dealing with
certain determinations as to the composition and condition of milk, cheese and butter,
which has been prepared by Mr. J. F. Singleton, Chief Inspector of Dairy Product*
in this branch.

The information herein presented should be of great value to chceseniakers,
buttermakers and others who have to deal with milk and its products.

I have the honour to recommend that it be published as Bulletin 46 of the Dairy
And Cold Storage series.

I havo the honour to be, sir.

Your obedient servant,

J. A. RUDDICK,
Dairy and Cold Storage Commissioner.

THE AVERAGE COMPOSITION OF BUTTER AND CANADIAN CHEDDAR

CHEESE.

P«rc«nl— Butter* Cheeie.**

Water 13-n 3;'0«

*■»« »i3-:i 34-43

•'"••In l-(( 2s'ni,

Bait and n«h 2-5 s-ill

• Van Slyke. •• Shuit.

UETEUMINATION OF TlIK SPECIFIC OUAVITV OF MILK.

By sporiKr Bravity (Sp. Or.) is meant tho wi-iKht of a ilefinite volume of any
substaiico as compared witli tlip wciftlit of »ii pinial volume of some other substance
chosen lis a standard, both liriii;» at the same temperature.

In dotermiiiinjr the KiK'cifie jjravify nf liipiids or solids, pure distilled water i(
taken as a stumlard and the specitic! Krlivity of water ii .'preseiiteil by 1.

One (lalli)ii of water weighs ten (ID) pounds. If oiio (fallen of sulphuric iieid
is found to weifih eifilitcen HS) pound-, the specific gravity of thj acid is found by
tho followinf? calculation: —

When a volume of water weislis 10 lbs. an eiiual volume of sulphuric acid woiglia
18 lbs.

'When ft V(dumc of wafer wi Ir'hs 1 lb. an e(pial volume i,f sulphuric acid weighs
18h-1<)=^--1.M lbs.

This means tiiat the acid is 1-8 times as hcavv as water or has a specific pravitv
of 1.8.

TlIK I'sr. OF nviinoMKTKIiS.

The specific Rrnvity of a litpiid is not usually olitaincd liy w(i^'hinK a definite
volume of the liipiid, but it is obtained by means of an instrument lamuii ;is an
hydrometer. The hydrometer is a >rlass instrument which floats upii.-ht in the rHpaiil.
constructed with a small bulb on the lower end, wliicji is biiuied with either nicrcnry
or fine sliot, and a larger liellovv bulb above, to the top of wliich is attached a slender
stem with a trnuluatcd paper scale insiilc from which to ri>.id the specific pravity.
Tho use of such ,in instrument for determinin;,' the specific irravities of li(inids is
based on the natural law tliat "A ho<\y lloiitiiur in a li(iuid will displace a volume of
that liipjid cipial in weight to the wcicht of the body floatluLr." The liydronu'lei-
sinks in the lii|uid until it h:;s displaceil a vnbime of the li(juid eiual in woiirlit to the
weisht of the hydrometer. In constructing hydrometers for deti'rniininsr the specific
gravities of different Tuiuids, the weiirht of shot or mercur;\' in the lower bulb, and tho
size of the hollow bulb above, are varied, ('cppndinc: on tho specific crravity of tho
liqni<l or liquids to be tested. The ereatcr tlie ■-•pi'cific frravity of the liquid or liquids
to be testinl, the heavier must th" hydrometer be loaded iu iirojuirtien to the size of
the hojldw bulb. The readinir on the f-'r.idualed stem Is t;tl>eu just at the surface of the
liquid. Thi> llKliter the l;<|uid, tlic t'ariher will tlie hydrometer aink into the liquid,
and the hi>avii>r the liquid, the hijrlier nj) will the hydrnmeter float.

Since liquids exjiand and become lishter when wanned, and contract and bcoine
heavier when cooled, an hydrometer will oidy civo a correct rcadiii,;; when used iu a
liquid at the temperature for which it is constructed to be used.

TiiK Lactometeh.

Such an hydrometer is used to determine the specific gravity of milk and is usually
called a lactometer. The lactometer conmionly used in determining' the specific

gravity of milk in known ai the " Qucvenno " lactometer. It ii uiuolty conttrueted
•• ■ combined thomiometcr nnii hydromctfr (Fig. 1) and i* coiwtructed to give a
correct reading when ukmI in milk at a tcmperoturo of 60* F.

r\t. \.

The icalo on the lactometer i» usually grudiiatod from 14 iit the tup to 42 nt the
bottom and each of the twcnty-eiisht (i8) divinionsi botwcoii 14 luul 42 i» called a
lactometer degree.

If the lactometer i» constructed an a combined thermometer and
hydrometer, the thermometer scale should bo above the lactometer KcaU-
in the stem, ho thot the temjierature may be read when the laotomet.'r i»
at rest in the milk.

In taking the lactometer reading of milk, a glass ct tin cylinder
about one and one-half inches in diameter and twelve inches hitjli
(Fig. 2) is noci'sxnry. To bo strictly accuriite, the milk should be brouKJit
to a temperature of 60° F. and after carefully pouring the milk from
one vessel to another a few times to distribute the fat unifi>riiil,v
throughout the milk, the cylinder is filled with milk to within iibKiit
one and one-half inclu* of the top. The clean, dry lactometer is mnv
gradually lowered into the milk in the cylinder until it comes to n-st
and lloats steadily in the milk. The point on the scale which showi* nt
the surface of the milk is at once road and recorded and is known ii»
the lactometer reading (L.R.).

PIK. 1

The Influence of Diffkhent Tempkhatures.

As pointed out above, ns the temperature of the milk is raised above 60' F.
(the tempcnitiire ut whic-li the luctomotor is constructed to be used) the milk expands
and becomes less dense; consequently, the lactometer must sink farther into the
milk to displace its own wciRht. Since the luctometcr scale is graduated from 14
nt the top to 42 at the bottom, the lower the lactometer sinks the lower is the readini.'.
On the other hand, as the temjM r \ture of the milk is reduced below 60° F. the milk
contracts and becomes more dciisi- and the lactometer does not sink as low in the milk
to displace its own wcJKht. This means a hiffhcr reading on the graduated stem than
if the temperature was nt 00° F.

In order to avoid the necessity of brinping the temperature of each sample of
milk to exactly 60' F., the lactometer rcadiuR is us\ially taken at whatever temperature
tho milk may lie. provided it is between 50° F. and 70° F. It will bo found that if a
sample of milk has a lactometer readiiiff of say 30 at a temperature of tW° F.. the
sample will have a lactometer rending of approximately 29 at a temperature of 70°
F. and of approximately 31 at a temi>erature of .50= F. From this, a rule for making
corrections to the lactometer reading on account of the temperature bciuR above or
1k>1o-v 60° F. is deducted, viz., for cnch dejrree that the temi)oraturc exceeds 60° F.
add A or -1 to the lactometer ro.idini? and for each degree that the temperature
is less than 00° F. substract t\, or -1 from the lactometer rending.

For example, if a nuuiple of milk .=hows a lactuiiicter rc-adiiif; of 20-5 nt 67° F., the
correct rending would be 29-5+ .7 = .'iO-2; and if a sample shows a hictometcr reading
of say 31-6 at 52° F.. the correct reading would be 31-5 — S = 30-7. This rule gives
fairly accurate results when the temperature at which the reading is taken ranges

b«tw««n SO* F. •11(1 70° F., but U itrictly accurate re«iilt« are deiired. it U beat to
brinff the temperuturo of tho milk to exactly t)0' F. before taking the lactometer
reading lo that no rurn>ction nn nccuunt of ti'mprrature will bo ncceaaary.

In taking a lactunictur reuding of n lampio of milk, the milk «houlil not bo
allowtxi to *tand after jioiiri „•, lirfiiro taking the rradiiig. If tho fample is allowed
to itand, tho cream rites to tho tnrfaoo and tho bulb of tho lactuniotor will be in par-
tially akimmed milk which in hcuvier than the whole milk ; conMNjucntly, tho lactomtter
will not link n^i do<i'y into tho milk, giving a laotonieter reudiii;( whiih will bo too
high.

Tho lactonictcr roiidinK ot' milk xhoiild nut In- takon until milk i« at leant two or
three hoiir« old, a* tlii> liu'tonictiT rcudiiur of (rv*\\ i 'I'; will UHiially Ihi uIhiuI ono
degree lower than it will 1m' from two to tlirco lumrs hi er.

Uht\imnu Tin; Si-Kciric (Jhavitv jhom tiik Lactomktkh l{K\niNii.

In onlir to obtain tin- j-iticitic gravity of milk from tho hu-tomctor reading,
1,0(K) in adili-il to tlir Ijn'tonictcr ri'a<liiiK and the result i* dividt-d by l,tK)«). That it:

L. It. -f ino()

KXW — Sp<>eitlo Orauty.

For exiiniiili', if tlio correct laetoiueter reiulini; in 30, the .•ipecitic gravity will k'
30 -f 1(M)0 lOMO

lOOU

1000

«10.TO

In order to obtain the lactometer reailintr of milk from the spccilic jtravity, the
specific gravity is multiplied by 100<i and lOlMl is suhtriictcd from the result of the
multiplication. That is, the Inctunieter rcndiiiK - (Sp. Or. ■ l(MMI) — KHMI.

For exni.iple, if the .spei'itic (travity of a ^aluplc of milk is t -Oi!*. tlie lactometer
roadiuR is (1 .(Il'U x lOiHO — KMM)- IOl'9 — lOfltt - 2!t.

The specific (travitj of normal milk us\iii'ly rauRes hetvccii 1-020 and l-OUl,
and will average about 1-0316.

DKTKinilXATION OF TIIK VKll TKNT OF SOLIDS NOT FAT AXD TOTAL

.SOLIDS I\ MILK.

The solids of milk consist of fat. casein, alliuincn, sntrar and nsli. These total
solids (T.S.) are fre.iuently diviilcd and referred to as "Fat" anil "Solids other than
Fat,"' or " Solids not fat " (S.N.F."). The percciita«e of fat is iletcrinin(>d by means of
the Babcock tcst-

C.*LCta.\TISO IIIK PkIK KXTACK OF Sol. IDS NOT FaT IX JIll.K.

The perccnt.aBc of solids not fat in milk is calculated from the percentafje of fat
and the lactometer reading. Several dilTcrent forinnlas have been worked out for
calculatinjt the pereentafre of solids not fat. The one most commonly used is to add
the percentage of fat to the lactometer readinir at fiO° F. and divide the result by 4.

c S N F - li' l!^ "*■ ^•'^- "* ^'^'

For esample, if ■ Mmple of a Ik tmU 3-5 per cimt of fat with u luutoiuetor readinf
of SIS at to*

% 8.N.F.

3S>31K .ia

-8.76.

Another formula comnior' . j !« to multiply the per c«ut of fat by 0'!, dlvida
the lactometer rpndinir hy 4 nna ddd the two renulti: —

%8.N.F. = Fatx0.2*^'"'

Fnitiir thi« formula v.th n lumplo tc>i<tiiij{ 3S pur pent fut and having u lactometer
renditiff of 31 -S, the cuioulutioii it at follow* :^ —

n 8.N.F. = 8.6x08 +

3t 5

= 0-7 + 7-875
= 8-578.

Thia lattqr formula i» iliRhtly more compliciitod to w*c, Imt in more art urat<« than
the former.

Dktkiimination i)K tiik Pkh Cy\T «v Totai. Si)UI)s.

The prr coiit of total Bolidn may be ohtniiied by iidiliii(r tin- piT cent of ».ilid»
not fat to the per cent of fat. For oxamiilt-', if a naniplo of tnilk is found to contnin
4 per cent f;it and 9 p«r -cnt solid-* not fat, tlio per i-i-rit of total ixi'.idi is 4 + 9 = 13.

THE DKTKCTIO.V OF .\I)r[,TKHATIf)\ OF MTLK BY MKANS OF SKIM-
MINO, WATKHINO. Olt ROTH. AND CALCULATION OF THE KXTKN'T
OF ADIM.TKRATIOX.

The detection of adnltt'ration by nifan-i of .-kiniminir. Wiiti-rinc, or lioi'n, u. ■ tlio
i-;ili-iil.itiou of till- i-xti'Mt of tlio luliiltcr.-ilion i» ilc|n-iiili-iit on the ilfi-ct of »uch
ndiiltcration on tlic per i-i-nt of fat, on tho l:i.-toiiii-l'r roadiiitr or spi-oific (^ravily, :i!id
on tiie per coiit of mtlUh not fat of tli<' milk.

The s|M^cifip jfruvity of water i» 1-0, of fat aliout "•■0, of wliole milk ll^^alty from
1-(1l'9 to 1-<':11, with jin av(ra|.'(' of i:l)o\it l-i>:!l.'>. of .-kinmird milk from I-ii.-.l' to aliout
l-():iT.

Tin: I-Ikkcit ok Skimmim; oi! PviiTUi.i.v Skimmim; Mn.K.

It will 1)1' readily iiiuler-'ood tliat reniovlncr part of the en-am from milk will oaiise
the milk to test lower in fat content. Since fat has a speeilic (gravity of 0-9, whi.li is
(-onsiderably Ii-ss than the speeitie ftravily of milk, removinir i>art of the fat from milk
by mean-* of skinnninir will cause the partially skinniied milk to have a hi;,'her speeitie
L'ravity or hi^rlier laelooiefer n-adimr than the whole milk had before beinj; partially
skimmed. Konsrhly speakiiiff, t'.ir each one per (-ent of fat removed by ^^kimininir, the
lactometer reading of the partially skimmed milk will be increased by about one det^ree.

For example, if a sample of milk test.^ 4 per cent fat .nd has a lactometer
readiiitr of 'M, rc-nioviiifr 1 per ci-nt of fat, that is, skimmintr it down to ,'5 \WT cent of
fat, will i'lcnjiise t)<o Ini'tometer n-ailiTitr to alioiit .";?, ,\piilyiiiL^ tin- formula

<■; S.N'.F

'.' F. fL.It. a! 00"
4

to hoth tho piiiti (ample nml tho partinlly iikimnuHl ■uniplo, wo iind tbo jwr cwit of
■olid* not fat in in eni-li imi«i»». Thnt it. if Ihu luctomrttT reudinx imffiKM 1 di>iri««<
with wii'h 1 |K'r cent of fnt nmnvi'd hy ■kimtninir, lh« |>«>r wnt of nolid* not fnt would
not be iifff><>t)>d. In aptuiil pnirtifp, it ii unuiilly fount! tliitt imrtiiilly nkinuninir milk
•litrhtly iniTcnitHi tim pi<r it«nt of iintidit nut fnt.

Hkimminir milk, tluTtforc nclu«'<<n tin- jht wnt of fnt, incrrniM the Inrtomotcr
rMdinir, iiml unv.» the [mt c.nf of lolid lot fnt norni.il or tlinhtly hiifh.

TiiK KrrwT or Anm.Mo WATrn t« Mu.k.

If n ■iimplo .if milk ••,>ntiiin< 4 jht ictit of fnt with n liiitMm..ti.r r.-in'itin of 32
fth;it i« II (Jr. ..f l.(i:i'.>) iin.l !» |ht .cut nolidn not fat. ono tfiill..ii of ni.-h milk will
weiifh 10. ;i-.' pniimlK. If one Knllori of thi-< milk is mix.'<l with otir ^Mon of Wiit«'r,
wo will hav«i two irnllons of wiitori><l milk which will oontiiti 2 \)er •'■iit of fnt (om-hulf
of 4 per cent) nnd 4-5 per crnt iioliiU n..t fnt (onc-hnlf of 1» p«'r rnil ). Siiico n gnllon
of water weiuhj 10 pound* the two (fnllon* of milk nnd wnti<r will wii^h 20.;t2 |K)undii.
«nd ono M^ill'>'i of «iich milk uml wafir will wi'iirh 10. IB pmitidK, or hnvf n Sp. (Jr.
of 1 0|(i „r ii In.tonu'f.r n-mlinK of 1(1. From thin we m.. thnt nd<liiijr wnfor to milk
rrdiifM the jKT oent of fnt, tin- hu'tomctcr rriulinir and tho jut ffiit of sol" U not fnt,
all three hvino mluciii in equal pruiwrltonn.

TlIK KirKCT OF HoTII SkIMVINO AM) W.»TKHIJIO Mir.K.

If a Hnmpio h both nkimnu'il nnd wntcrcd, tlio wntcrinir ndiiccn tho ix-r cent of
fnt, the Ini'lomotrr rondlti« an<l the per cint of Moli.li not fnt, nnd all aro rodiicod in
tho snnii- proportion. Tlio skimm'nK would rodii.-i' the fnt nlill furthor, inorcnse the
Inctomcrnr icnilinif nnd oiflior not alTcil or only xliuhtly iniTiM^*!.' tho iior cont of
j-olids not fnt. Tho roinlt will ho n low piT nut of fnt, a Inotomi-t'T nndiiiii normul
or low (aioro frcpiontly low) nnd n low por cont of solidt not fnt, ivilli Ihc fat rcdurtd
in yreatcr pinportinn than vtlhrr Ihr larlomfUr rua<li)i!i or pir r.iil of unlUlii not fat.

To roon|)ilulato: —

Skimminy; i» itidicHtod by (a) low por cent of fat.

(h) IiIkIi lac'f;ini('tor re idinir,

((■) normal or slifilitly l.ijth por i-cnt of mdids not fnt.

Wutcriii^' i-4 iiidiriiHv! hy (a) low por oont of fat,

(h) h)w Lwtometor rcatlinjf,

Waloriiii: iiiol .»klniininK' \* iiidientod hy («) low por cent of fat,

(/») Inctometor roadinjr may he nor-
mal, hut is more usually low,
(c) Inw jii'r cont of solids not fnt. the
fat Ixiiiji reduced in i/realer
prnporliiin than either the lac-
tnmi'lcr read'nii] or pi r cent of
aoJidx not fnt.

If a P'linplp of milk npponrs to ho ndultcratod, a control sample should, if possible,
bo secure ; that in, a sample frniii tlio SMOir liiTd, iiiilked in tiu' presonco of the
inspector or jpcrson niakiuL' the ti st. The control •simple scrvis as a basis of com-
parison nnd if tiio ofi-inal sample shows a mark.-.l inferiority to the control sample,
it will be fairly I'onehisive proof of ndulterntiou. Tliere niny be, umior ordinary

conditions, in the inilk of individual cows, variations from day to day, of at least one
per cent of fat aud one-half per cent of solida not fat. The variations will not bo so
KTcat in the milk of herds, and the lartfor the herd, the less will the variation be.
In eompnrinff the ori(;inal sample with the control sample, reasonable allowance,
taking into consideration the nnniber of cows, must be made for this natural variation
from day to day.

At times there will be a wide variation in the fat content in the night's and
morning's milk of herds. Particularly is tliis the case wlicn the milking is not done
at ihe same hour night and morning. When the periods betwot^n milkiugs are not
equal, the richer milk will be obtained after tlie shorter jM'riod, and tlie variation in
fat content between the two milkings will frequently <'xceed one per cent. The effect
of irregular milking on the lactometer reading is, however, not so marked.

Deter-mixation of the Extent of Adulteuation nv Skim.ming.

In samples which are simply skimmed, the j)ounds of fat removed fr;;;?i each
hundred pounds of the milk is obtained by subtracting the per cent of fat in the
skimmed sample from the per cent of fat in the pure sample. For example, if the
skimmed sample tests 2-7 per cent fat and the control sample tests 4-0 per cent fat:
4-0 — 2-7 = 1-3 pounds of fat removed from each hundred pounds of the skimmed
milk. Therefore, the pounds of fat removed from each 100 pounds of milk = jier cent
fat in control sample — per cent of fat in the adulterated sample.

Dktekminatio.n of the Extent of Adulteration nv Waterinc.

In calculating the per cent of foreign or extraneous water present in a watered
simple, either the fat, the lactometer reading, or the solids not fat might be used
as a basis of calculation since all are reduced in the same proportion, luit since the
solids not fat are less subject to variation from day to day tlian either the per cent of
fat or the lactometer reading, the calculation is based on the solids not fat. To
illustrate, if we have a watered sample showing 3-0 per cent fat, a lactometer reading
of L'4 and (!-7.') per cent solids not fat, and a control sample showing 4-0 per cent
fat, a lactometer reading of 32 and !) per cent solids not fat, to determine the per
cent of foreign water, the calculation is as follows: —

There are 9 lb. solids not i.it in 100 lb. pure milk.
" is 1 " " " 100

" aret>-75" " " lot) x O-T.")

= T5 lb. piin' milk.

That is, f|.7.'5 pounds solids not fat are present in 75 pounds pure milk, and since
we have 6-7.5 pounds solids not fat in 100 pounds of adulterated milk, it is evident that
7.') pounds of puri^ milk liave been increased to JOO pounds by the addition of water.
That is in each lOO jjounds of watered milk there are TOO - 7.5 — J."> pounds of foreign
water or 25 per cent of foreign or extraneous water. Since the calcidation is similar
in all cases, we may deduce the following formula: —
The per cent of extraneous water —-

per cent solids not fat in adulterated sample x 100

100

per cent sol'ds not fat in the p\ire sample.

That is, to find the per cent of extraneous water, multiply the per cent of solids not
fat in the adulterated sample by 100, divide the product of this tuiiltiplication by the
per cent of solids not fat in the pure sample and subtract the result of this division
from 100. The difference will be the per cent of extraneous water present.

Dktkiimi.vation of Extknt ok Aim i.tkiution ok Milk Both Skimmko and Watkred.

In case a samplf i* l)oth skiiniiUHl and watcrcil, tlio above forin\ila for (li'tcrniininff
the IXT «>nt of for«'i>?n water present, liolds Rood.

To illustrate, if we have an adnlU'rated ^an)Ille showinjr l"r fat, n liictoniofer
reitdinK of 22, mid fl% solids not fat, and a eontrol sample showiuK 4% fat. a lacto
meter reading of 32 and d^i, solids not fat. it is plainly evident that the adiilt.Tnted
soniple is both skimmed and watered since while the fat is reduced by one-half (from
4% to -2%) the lactometer is only n-dueed about one-thir<l (from 32 to 22) and the
solids not fat are only rcdnerd o'.- -^: '-.[ f •'rom 9'"^ to (1^ ; ).

Applying the alwvc formi
C X 100

«■'■ hav

II r cent of extraneous water =

1(H)-

•=100-Gb3

The deti'rmination of the per cum «.; 'it distracted l)y fekimining ilepi iids on the
facts, that skimming does not to any extent affei't tli«' per cent of solids not fat. and
that watering reduies the fat and the solids not fat in the same proportion. Tlie
problem is. thcrelore. one of jiroportion.

With (!'; of solids not fat we have S';,' fat.
** 1 r^ t. .4 t( ■»/

2j«_!l-:.T-; fat.

That is, the water which will reduce the per cent of solids not fat from 9',; to Ct^-'r
will reduce the fat from ..'■,', to 2';,', or the :!:iV; of water is responr.ible for redu<'iMg
the fat from 3^; to 2% and the sample was skimmed from i% to 3',', that is, 1% of
fat was skimmed off.

From this we may de<lnce the following fornnda: -Tn a skimmed and wiitered
sample, the per cent of fat abstracted = <;;. fat in the pure sample -

T?> fat in adult, sample x Cr solids not fat in pure sample.
% solids not fat in adult«Tat<^l sample.

That is, if a sample is both skimmed and watered, to find the per cent of fat
abstracted, nmltipiy the per cent of fat in the adulterated sample by the jier cent of
solid.s not fat in the pure sample. Divide the product of this midtiplii-atioti by the
i>cr cent of solids not fat in the adulterated sample and stdttraet the result of this
<iivision from the per cent of fat in the pure tample. The difference will lu the jier
cent of fat alistractcd.

Tf a sample appears to be both skimmed and watere<l, and one cannot tell by
inspection whether the fat and solids not fat are reduced in the same proportion or
not, this formula may be applied. If the milk is from a herd, and the c;ilculntion
shows over •.'i<~;, fat abstracted, it is fairly good proof of skimming. If the fat
abstracted show^ under .^i"";. the difference may be <lue to the natural variation in
the per cent of fat from day to day, and one would not be safe in saying dcfitiitely
th.it the milk had been skimmed ns well as watered.

THE DETERMINATION OF THE PER CENT OF ACIDITY IN ?fILK.

The detJ^rmination of the per cent of acid in milk is ba?ed on the facts that aciil^
and alkalis neutralize each other in definite proportions and that certain chemicnis
known as indicators may he used to denote by a change of color in the liipii.l lieing
neutralized, the exact point at which all the acid and all the alkali are neutralized.
This ife known as the neutral point. The method of determining the per cent of acid
in milk consists of neutralizing the acid in a definite volume of the milk by means of
nn alkaline solution (usually caustic soda) of hnnirn sfren<ith and using a solution ol

rhenolphthalein which ig colorless in aoid.s and pii>k in alkalis, as an indicator. By
ineabunng the <iunntity of alkaline solution of known strength used to neutralize the
ncid in a given volume of milk, and b.v knowing the proportions in which caustic soda
and Inctic acid neutralize each other (which is 40 grams of caustic soda to 90 grams
of lactic acid) the per cent of acid in the milk may be calculated. In order to avoid
tlie necessity of ealculntlng the per cent of acid in each sample tested, the alkaline
solution Ks made of such a strength that each cubic centimetre of solution used will
neutralize A^' or -Kl of lactic acid in a 10 c.c. *nmple of milk. This strength of

solution is known as a one-tenth normal ^"^ solution, and consists of four and four-
ninths (4 %) grams of chemically pure caustic soda in l.nOf) c.c. of solution.

The standard alkaline and the indicator solutions may be obtained from the dairy
schools and agricultural colleges throughout the country.

The test for determining the per cent of ncid in milk is termed the acidimeter
and the apparatus (Fig. 3) employed in mnking such a test consists of:

Fie. 3.

A 10 ^^.c. burette, preferably one with a blue line down the back to ensure
iiccuravy of reading, and a {rla.«s stop-cock. The burette should be
graduated to 01 c.c.

A damp for holding the burette.

A 10 c.c. i)ipette known to be correct. Since many of the c.c. pipettes
sold arc not accurate, the pipette should be tested for a.'curacy by
comparison with t!ie 10 c.c. burette.

A delf cup or glass beaker.

A glass siirring rod.

A dropper bottle for the indicator solution.

A bottle lor the standard alkaline solution.

Protecting the Alkaline Solutiov from the Am.

The alkaline solution must be protected from the nir sinoe if exposed to air it
weakens, due to the neutralization of the alkali by the carbonic a.id of tho air. The
most common form of acidimeter is arranged to syphon solution through glass and
rubber tubing from th bottle into the burette. The air admitted to the bottle to
replace the solution drawn out, is first passed through the alkaline solution in a small
wash bottle to neutralize the carbonic acid. This method of protecting the solution
from the carbonic a'cid has not been entirely satisfactory. Prof. W. O. Walker of
Queens University, Kingston, Ont., has recommended covering the surface of the
solution in the bottle with kerosene which prevents air from coming in contact with
the solution.

Making the Test.

In determining the per cent of acid in a sample of milk, on. should first see
that all glassware is perfectly clean. The sample is poured to ensure uniformity
and by means of the 10 c.c. pipette, 10 c.c. of milk is measured and delivered into
the cup or beaker. The pipette should be rinsed with a few cubic centimetres of
distilled water or clean rain water, and the rinse water added to the cup. Three to
five drops of the indicator solution are now delivered from the drop^jcr bottle into
the cup and the burette filled with solution to the mark on the graduation scale,
care being taken that no air remains in the tip of the burette. Solution is now
carefully dropped from the burette into the milk in the cup, and the milk and
solution constantly mixed by stirring with the glass rod. The solution is added until
a faint pink colour is obtained uniformly throughout the mixture, which indicates that
the neutral point has been reached. If one ceases adding the solution at the correct
point, this faint pink colour will disappear in a few seconds due to the action of the
carbonic acid on the air. Sufficient solution should not be added to make the pink
colour permanent, as the neutral point will have been passed and the mixture h, the
cup will be decidedly alkaline. The number of cubic centimetres of solution drawn
from the burette is now noted, and each 1 c.c. used represents -1 per cent of acid in the
milk; therefore, to determine the per cent of acid in the milk, multiply the number of
cubic centimetres of solution used by -1. For example, if 2-1 c.c. of solution have been
used, the per cent of acid i 'Ik equals 21 x .1 = .21.

Prepai • OF THE Alkaline Solution.

As previously stated, the alkaline solution is a caustic soda solution of strength

known as-, which means that in each 1000 c.c. of the solution there are 4% grams

of chemically pure caustic soda. Owing to the difficulty of getting caustic soda
absolutely pure and free from moisture, the solution is not prepared by weighing
out a definite quantity of caustic soda and dissolving the same in a definite volume

of water. A- acid (usually hydrochloric) solution is prepared by a trained chemist

and the alkaline solution m^ A such a strength that the^ acid solution and the

alkaline solution neutralize each other in equal volumes; that is, 10 c.c. of the -
acid solution will exactly neutralize 10 c.c. of the alkaline solution. The alkaline
solution will then be- .

The cheese or butter maker who wisi.es to prepare the standard alkaline solution
for himself should secure a quantity of the standard -^ acid solution from one of the

dairy schools or agricultural colleRos or from a trained chemist. Obtain the best
quality of caustic soda and distillwl watrr or clean rnin water. If a delicate scale and
gram weights are available, 5 grams of caustic soda may be weighed out for each
1,0(10 c.c uf solution it is desired to make. For every 5 grams of caustic soda used,

1,000 c.c. of water is measured out and the soda dissolved in the water. Since u " ■

caustic soda solution consists of 4% grams of caustic soda in 1,000 e.c. of the solu-
tion, the solution obtained by dissolving 6 grams in 1,000 c.c. of water will be too
strong, provided the caustic soda is reasonably pure. In standardizing the alkaline
solution, it is best to make it too strong at first and then rediu-o the strength by
adding water, rather thnn to have it too weak at first and have tn incrense the streii^'th
by n<lding more caustic soda. Two 10 c.c burettes should bo used in standardizing the
solution, one to measure the acid solution nnd one to meiisiire the nlkiiline solution.
If two burettes are not available, the acid solution may be measured with the pip<-tte
and the iilknline solution measured with the burette. Have all glassware thoroughly
cleiin and after the caustic soda is thoroughly dis.solved and nii.Ncd in the water, the
burette to measure the acid, or the piix^te, if only one burette is available, should 1m<
thoroughly rinsed with the iicid solution. Ten cubic eentimotro^ of tlie ii<'i<l solution
are delivered from the acid burette or pipette into the cup or beaker and three to five
drops of indicator solution added. The burette for measuring the alkaline solution is
riii>ed two or three times nnd filled to the O mark with the alkaline solution already
prepared. The alkaline solution is slowly dropped into the acid and constantly stirred
until tlie niuhal point is reached. The number of cubic centimetres of solution used
siiould lie noted and the neutralization pt^rformed a second time to verify the accuniey
of the result. ]f the alkaline solution is too strong less than 10 c.c. will be required
to neutralize the 10 c.c. of standard ncid solution. For example, if 9-2 c.c. of alkaline
solution neutralize 10 c.c. of acid solution, the alkaline solution is too strong and -8
c.c. of water must beadded to each 9-2 c.c. of solution. By measuring the volume of solu-
tion prepared, tlie necessary v.dume of water to add may l>e ealeulated. For example, if

"iOOO
there are 5,000 c.c. of the solution, the water to be added would be ; .,S = 435

y ■ j5

(almost). This volume of water is now added to the solution, which is thoroughly

inixi-d and tested for accuracy. If tiie work has been carefully done, the solution

should be of th:? correct strength, that is 10 c.c. of the alkaline solution should exactly

neutralize 10 c.c. of the acid .solution. If the alkaline solution proves to Ik- still too

struiir. more water must be added, if too weak more caustic soda must be added. In

standardizing the alkaline solution, two or more tests should be made each time to

ensure accuracy. As soon as the correct strength is obtained, the alkaline solution

should 1)(! tightly corked in a glass bottle or buttles. Earthen jugs arc not suitable

as containers for the solution. If the solution is not being prepared in a bottle, but

in some other vessel, the bottle or bottles to receive it should be thoroughly clean

and rinsed with a little of the solution before being tilled.

If a delicate scale and gram weights are not to be had, the caustic soda may be

dissolved in a little water, making a strong solution. This strong solution may be

gradually added to the water, testing after each addition of the strong solution.' If

too much of the strong solution is added and the solution being prepared is made too

strong, it is reduced by adding water. The solution can soon be brought to the

correct strength in this manner.

STRENGTTI OF IXDICATOU SOLUTION.

The strength of the phenolphthalein solution used as an indicator may bo varied
considerably without affecting the results of the tests and different strengths of solu-
tions have been recommended by different authorities. Prof. Walker has recommended
the use of a 1 per cent solution in connection with the Walker method of determining

the per cent of casein in n.ilk unci nuch a solution is aUo satisfactory for use in c.m-
nectjon w.th ho ac.d.„,c.tor. A 1 per cent solution is prepared by dissolving Vra«.
of p!, .ioli.thulein m 100 c.c. of 95 per cent alcohol. ssoiving i grau.

TESTlNti CIIEAM, SKIMXfll.K, BUTTERMILK AND WHBY.
n.ilk'^'l'!ntr,"','lir"' i' f '" ■""! '" ^^■'^•""""' the per cent of acid in crean,, skin.-

DETER.M1NAT10X OF THE PER CENT OF CASEIN IN illLK BY ME VNS
OF THE WALKER CASEIN TEST. -UtANS

fl,n ■'"• ,'l'-''7""''"« t''«. I«"f ''^t of casein in milk by moans of tb.. Wnikor method

he ac.d.meter as descnbed in this bulletin is usc.l. In ad.lition to the aeid "m tor"

It IS necessary to have: ""luuKror,

1. A ICS c.c. pipette.

2. A 2 c.c. graduate.

3. A bottle of neutral formaldehyde solution (40 per cent).

The neutral formaldehyde solution is prepared by adding a few drops of phenolph-
thalein indicator to the bottle of commercial formaldehyde and then adding ,bo n
caustic soda solution until a faint pern.anent pink colour is obtained.

Making tiik Tkst.

Af,.!"*!?'''''''''?',"''' ^'''•' ^'•*'.^«'^^"<' " r..presentative portion of the milk to be tested
After thoroughly pounng th,s portion, t.ke a san.,.lo witli the ]0..'5 c.e. pipette nd
deliver the sample mto a white ,lelf eup or beaker. Add about 1 c.e. o m ieato,
1 per cent solution) and add the alkaline .solution as in testin, for aeid ty U a
de.Mded permanent jiink cojonr is obtaine.l. Now add ^' ec of the ne.tr;! f ,

hy.le solutio,. whieh destroys the pink eolour in the snn.lc The burette ;„f'''"fl^;^^^
to the mark with the alkaline solution, aft.r whieh thl^tdut^ £ S h " So he
cup or beaker unt.1 a pern.anent pink eolour, of the same shade as in the fi^ t
operation, is obtained. The number of eubie centimetres of solution used Tn 1 e
se..,.„. operation is noted and will represent the per cent of casein in the , ilk Fo
e-xampe ,f ■>. c.c. of solution are use.l. the percent of casein in the milk i 1^5

bottle t''! t".""f ■'"''•"'','"'"""" ^'""^''^ '"^ '^"'^"t preferably in a plass-stoppered
bottle. In tnno the a.nt pink colour of the formaldehyde solution will dislpe'r
owmj; to the action of the .-arbonic aci.l of tlie air, when a few drons of tl.n u T
solution should again be added to the formaldehyde tl res^;; t'^olou "

I sing slightly more than 2 c.c. of formaldehyde solution in making the test wil'
not in any w.iy affect the results. " ' ^^'"

In case a 16-3 c.e. pipette is not obtainable, the 10 c.c. pinette m,v be „».,] ,„ i
the te,t performed as outlined. The number of cubic ocntim'trefof " utionT .ir"
in the second neutrahzat.on must be multiplied by the factor 1.63 to give the pe n
of casein when the 10 c.c. pipette is used. '^

To avoid the necessity of this mnltiplioMtion for rnnh 8nmpl« tcslpd when usinff
the 10 o.c. pipette, the following tnble has heen pn.p„mi by the oriRinntor of the
test •*""

;ilkiili iiM-().

1-00
I'On

1'15

f20
f25
1-30

•it wilt cawin

1T,3

ITl

1-7!)

1-87

f9.1

2-04

2-:2

(| »lku]i iiHt'd.

1-36
f40
1'4B
1-BO
1-S6
I'SO

I't-r ci'iit ciiw'iii

2-20

2-28

2MG

2-44

2Ti3

2-61

DETERMINATION OF THE PER CENT OF CASEIN IN MILK BY MEANS OF
THE HART CASEIN TEST.

The apparatus employed in determining the per <ent of casein in milk by means
ot the Hart casein tester (Fig. 4) consists of:

Flc. 4.

1. Centrifuge.

2. Metronome.

3. One pipette, 20 c.c. capacity.

4. One pipette, 5 c.c. capacity.

5. One graduate, 2 c.c. capacity.

The chemicals employed in making a test are chloroform of the best quality and
a 0-25 per cent solution of acetic acid.

The. centrifuge is constructed to receive either six or twelve bottles, the bottoms
of which describe a 15-inch circle when revolving. The bottles are revolved 2,000
times per minute and the machine is so geared that the crank must revolve 55 to 66
times per minute to impart the necessary speed to the bottles.

The metronome is used to enable the operator to turn the crank at a uniform
speed of 55 to 56 turns per minute.

The 20 c.c. pipette is used to measure the dilute acetic acid solution, the 5 cc
pipette IS used U, measure the milk sample end the 2 c.c. graduate is used to measure
the chloroform m making the test.

The bottle is constructed with the graduated stem on the lower end and the bulb
on the upper end. The graduation of the scale is similar to that of the 10 per cent
Bal .uek milk test bottle.

Preparation or Acetic Acid Solution.
The dilute acetic acid »olution i» prepared by adding 90 c.c. of distilled water
to 10 c.c. of chemically pure glacial acetic acid. To 25 c.c. of this 10 per cent
w ution are added 976 c.c. of distilled water, making 1,000 c.c. of a .26 per cent
solution.

Making the Test.
In making the test, the temperature of the milk, the chemicals, and room in whicn
the work is being performed should be between 65° F. and 75° F.— preferably of
70 r ,

By means of the small graduate 2 c.c of chloroform are mea8ure<l and tran?<"erred
to the test bottle, and by means of the 20 c.c. pipette, 20 c.c. of the dilute acetic acid
are also added to the test bottle. The sample of milk to be tested is carefully poured
to ensure uniformity, and by m«ins of the 5 c.c. pipette, a 5 c.c. sample is taken and
added to the chloroform and acid in the test bottle. The mouth of the bottle is
tightly covered with the thumb and the bottle inverted and carefully shaken for
16 to 20 seconds, which is timed by means of a watch.

Tests should be made in duplicate and should be whirled as soon as possible after
being shaken. If more than one sample is being tested, the chloroform and acid are
added to nil bottles and then the sample of milk added to each bottle, after which
all are shaken.

The bottlfs arc placed in the machine in such a manner that the machine is
properly balanced. The metronome is now set to beat 55 to 50 beats per minute uiid
the crank is turned 55 to 56 revolutions per minute for seven and one-half to eight
minutes.

The bottles are now taken from the mnchino and placed in a perpendicular po.sition
iu a rack made for the purpose.

In the lower part of the graduated tube, there is now a liquid which is comiwsed
of the chloroform and fat. Above this is the casein, which should appear as solid
white column free from ragged edges, and above the casein is the acetic acid mi.xed
with the remaining jnirtion of the milk.

After the bottles have been allowe«l to stand ten minutes, the per cent of casein
is read from the graduaUnl scale.

The test is verj- sensitive to changes in conditions under which it is operated
Increasing the temperature of the milk, the chemicals, or the room in wliicli the work
is performed, will give lower readings, while decrensing tiie temperature of the milk,
the chemicals, or the room, will give higher readings. Decreasing the speed of the
centrifuge will give higlier readings, while increasing the speed will give lower
readings.

THE TESTING OF BUTTER.

Determinatiox of the Pkr Cent of Water ix Butter.
As the principles which dotermlne the porcent.ige of wnter incorporated in butter
liave become better understood by buttermnkers, a gradual incroiise in the water
content has been apparent. Since it is unlawful to manufacture or to sell butter in
Canada containing more than si.\teen per cent of water, n simi)le, rapid and reasonably
accurate means of determining the percentage of water in butter has become a neces-
sity for the guidance of the buttermaker as well as for inspection purposes.

Method of Determinixo the Per Cext of Water ix Bitter.
Some of the moisture tests which have been put on the market have heen
complicated and lacked durability. The method .nost commonly used in Canada
at the prest-nt time, consists of the evaporation of the water from a definite weight of
butter and the determination of the percenatge of water from the loss in weight.

•o-l J°k"^°'u **"' "«^«'«''i*J' °f computing tho por ernt of «at.T from the loss in weight
T.tZ. .r ^" '^"""r"''""' t" «i-^. when n d.-finite weight of button ' u"od a, a
.ample, the per cent of water directly from the iH-nn, nju, bnlnnce l.l."! „fter

market'whtrH ff''"' "*"','" '""" "'" ""••'>^- '*"^'— ' "-"k- """« - on the

i. trf-jit.. Jiwllfc «r ^;.

FK. S.

:rh. apparatus (Ki^. :,) used in making a moisture determination, consists of:
(a) The scales;

(<f) An alcohol lamp with which to heat the sample.
Sampling Butteb fob Testing

t I'SZ ""'"" """" ■""■' " "' '"•» '»■'' "'«"• -" «« which °ft."e"

Preparing the Sample yoR Testing.

If the .«a,n,>!« become, o.ly, > i. difficult to get duplicate tests to agj" "^^

Af*»ciNO the Test.

,H l"i^"''i"v " T- '^^- '"^- '" ''1''*^'' '^^ '"""P'^ >« ^^«^ should be ne^..*lv ^I„n
r J ^ . * -^^ •' ""'""■" "^"''^'^ '"'"■ "«' alcoiiol lamp t„ drv it fhorouyhK oj^^
placed on the pan of the ^.le. After the c^ is cool. theLle isMLre^ j'idiu^t

•K the woiKht on the tare beum. Ton Rrann of butter are weighed into the cup from
U.e 8«,„ple .., the jar 1 h« cup in now heated .lowly over the lamp to eva,«rHto the

T 1m I. •* '^P"''^-. ^^^ *'^''*'"» •''""''I P'"*"^** •'"^''y "o »» not to char the fat.
and while being heated the butter should be agitated by shaking the cup with a
rotary motion. Care must be taken that no fat is splashed out of the cup The
evaporation will be accompanied with more or Ws noise from the sample. As soon
as th,. Moiso reasos and just as the butt-T pomr.i.-n.v. t.. tak.. „„ „ more uimIht .-olour.
tnc h.u ing should bo discontinued, as all the water will have been evaporated. Should
the heating !«. .■ontinue.l beyond this point, the re-idue will Ih> ..hurnd aiul the result
ot tho tcs will be too high As soon as the heating is concluded, the cup is placed
on the scale pan and allowed to cool. Tho scale does not balance now since the sample
H i^'hter. ou'.n^' to the ovaporntio.i of the water. The weights on the b.uni an- now
tS henm'" balances exactly and the percentage of water is read from

When the hot eup is j.la-od on the scale pan, an upward current of nir is set
up. duo to the cup healing the surrounding air. This upward current of air tends to
raise he scale pan Consequently, if the reading is taken while tho cup is still hot.
It will be too high, since the weights must be moved farther over on tho beam to
counteract the effect of the air current.

The high pressure oven described under the " Determination of the Per Cent of
Water in Cheese furnishes a most satisfactory method of evaporating the water
irom the sample of butter.

A method of heating the sample by using a paraflSne bath has also been advo-
cated and widely used. The cup containing the butte. to be heated is placed in a
second cup which is surrounded by paraffine. A small copper kettle contains the
paraffine which ,s boated to a temperature of 175° C. (847° F.) which temperature
18 inaintained during a period of five to seven minutes. This method avoids danger
ot over-heating the sample, but lengthens the time required to make the' test. E.x..eri-
ence has shown that heating direetly ever the lamp, when ..arofullv done is quite
accurate as compared with chemical analysis.

As with cream testing scales, the scales used for the testing of butter should
be kept in a dry place.

DETERMINATION OF THE PER CENT OF SALT IN BUTTER.

H.t«3°""''^^'* °^ ''*'*"" ;'"* '> ^' '^^"' "^ *"'' '" ''""«' »* somewhat similar to
determining the per cent of acidity in milk and cream

.»!. A '"f^'""^ •"/ •^«'^'™'"'"» tl"" silt content of butter is based on the fact that
salt and silver nitrate neutralize each other in definite proportions. The salt con-
!nTnf t'" f definite quantity of butter is washed out and dissolved in a definite
quantity of water. A definite volume of the salt solution thus obtained is measured
out by means of a pipette and the salt in it neutralized by titrating with a silver
nitrate solu ion of known strength, using a potassium chromate solution as an indi-
cator, to determine when the neutral point is reached.

Pipettes of diffnrent volumes have been recommended for measuring the salt
solution and correspondingly different strengths of silver nitrate solutions. Of the

Ci^c^W^ if'/T'n^-'^"-'""'!;!^'""'^ *"'^'' '^^ suggested, that outlined in
S S ♦ ^' F»'^«">ty °f Wisconsin by J. L. Sammis. seems to be the most

adapted to our conditions. In this method, a silver nitrate solution, prepared by
h nnir^' A%T "l""*""^ /^-^^ ^'■"'"' "^ chemically pure silver nitrate in two
sohitinn 'r A u !^ '^l^^'-f-s -^ 'li^till«l water, and a potassium chromate

chrorTt'e h,T ♦ fi ^'^^"ving seven and one-qnarter (7-2.5) grams of potassium
chromate in twenty-five (25) cubic centimetres of distilled water are u=ed

Phkpamtiom or tiir Mii.vk> Nitiim S..llti»x.
procured from any .ood dru7hJ«, xl ■ . "1 potaiiium chromate may be

W cubic SiSri^l, taw'" ""^ i° '""'"1 "'"' '"••"'•"' '» «•• l"'«J'^

.oule and U I wShi on to Z n "'''«'""""« ''i" ^ squired to balance the

place oneTthehtvier ic^.hU u/ed in ^'T""f, '" *'''«'' " -•"""•« »^ ^""-' «"'
per cent n-,t<.h of^h^LZ \T1ZZ '''' u' ^^ "T "^ '""''"""' "" ••'" '^o
pn- and ad<l .hat or othrheuvr,£ „c^\rt'h '""" ^^ T'"''?"'" P'""" "" ♦»•«
p.. - "ntiltlm.c«leiHal„.o'tbl .^ C™^^^^^^^ uudern.niU the

on the end of the beam R^m Jl Un^ i omi'kte the ., lance by means of tho poise

until the Bcale tlunc™" exaX Tho u7 I 7*^ f "^ u'^'^' "'*"*'" '" ^'^ '•■"-'
nitrate are transferred to a Xa^browL,h;.7;/T «"""' "*' '"^•"'

Five hundred (500) cubic c^^ti'^/dtn^^^^^

now nddeil to the bottle and „., „. »k *"*' "f« ^•"tc'' or clour rum water nro

bottle shaken, the sot t ion 'V. „: ,o use "'' "" """""'^'''^ ''"^°'^-'"' ""'' *»'«

Protect.nc. Silvrk N.tratk Somt.on- from Svsucit.

the ..nlight. In addition to this th '^ T'\T 't '° P'''''''''^' "'" «"l"'i"" »>"'"
-own paper and kcp °n u d rk h ,u d "'tI, ' "''" "'' '" "■'*' "" """"""••' "'
to protect the silver nitrut' crystals Zli.h^n'''"'''' ^'T'T ^'""''^ *"-' »='''^'«
solutiof. it is advisable for 7}l t ^ """'' '° '^^ ^'^'''■' "^ ^i^f'" "" the

that five an,l one" nth ^l^l^Z^ k""" -"."i""!*^ f ^'" "^ ''^""' ""'"'"^ -
nnd tifty (250) cubic centimetre; o 2t " "''lif "''• /^'I'^l' «•'" "'"k«> two l„„„lrod

fr«,uentiynni there w^ttT li^^r JJ^,: ■';::"::;i:" ♦'- '^^ -"'^ ■'-^

Preparation ok Potassmm Ciihomatk Soli t.o.v.

.eveJIiLfrS't:?'?^^ Sm"':;;r'"^^' ^'? "^°^^- '^ ^-"--^ ^^ <i-'i-«

centimetres of later If H fl'' , 'r^*"^^'"'" ''^^'^T'" '" t^^'«"fy-fivc (25) cubic
potassium chromate may be wcXl o'^^^^^^ T- T '" '? '"'l \"' ^"^ '^^•^'"^ "' ^^e
(35) cubic centimetres of wat^Si'^eL rT'""" ""'"':"'';' "'^'"■'' '" '^''''-^'■^
two measures of the pipette wm ^^^ L'^a ;;^'voir':f ^^7' "''"" ''■' '^

Tht> Bppliancpi riH|uirt><l for mukiiiK n »nlt tiiit nro;-

(a) A t«alo for wrinhinir « ton (10) uriim natiiplf of butter. The moi»tur«

•ntle ii uicd for thi« pur|)o»«»;
(6) B oylindriciil iiu'imuritifr kIhhk nUmt <.iir itml ..m-hnlf (1() iiiilu-x in

diHinet.'r mid twflv.' (12> imhos hiirh, ifniduiitwl to hold two hundn>d

and fifty (250) cubic ct-ritlinvtri-i;
(c) A 170 c.c. pipette;
id) a RmaU ghiis breaker;
(e) a ten (10) cubie centiniptre burette witli ii\a*-> utop-eoek irmduiited to

one-tenth (Mii) of one (1) rubic ocntinictrc-' and chinip for liolding

the burette;
(/) a dropi)er bottle for the potaiwium rhromnto indioator;
(.'/) a one-pint wide nioutli gh** Uittle,

Making tiik Teht.

A sample of butter is sociir.!.! an<l pn-parfd for testinu in the xanio manner ni
f..r a n.o.-tur.. tent. The »fult. is balan.vd with « Mnall pi.r.. of parchnuMit pain-r on
the pan and tun (10) grams of the prepared sample weighed out on the pa|M<r. The
pajwr and butter are transferred to the pint bottle and two hundred and fifty (250)
lubic centimetres of water (preferably soft) at a teni|)erature of 110' F. to 120^ F.
measured in the graduate and added to the bottle containinK the butter. The bottle
IS thoroughly .shaken to melt the butter and wash out the salt. After allowin« the
bottle to stand a few minutes, it is attain shaken to ensure an even distribution of the
salt throi.ifliout the water. The botth- is tlien allowed to stand until the fat comes to
the .surface. The 17-0 c.c. pipette is inserted into the solution of salt, blowiuK through
the i)ipette until the end of the pipette is below the surface of the water to prevent
the liquid fat risini? into the pipette. The pi|H.-tte is filled to the ffraduation mark
with the salt solution and this quantity is transferred to the bt>aker. To the salt
solution in the Iteaker is added one drop of the potassium chromate soluti.-.i from the
dropper bottle. The burette is fillcl to the top of the scale with tlie silver nitrate
solution, care bcinR taken that no air bubble, remain in the tip of tlic burette The
silver nitrate solution is now slowly dropped from the burette into the beaker, which is
shaken constantly to mix the silver nitrate solution with the suit solution \s soon
as a permanent faint reddish-brown ••olour is obtained, the addition of the silver
nitrate solution sliouhl case as the salt is all neutralized. Th.. number of cubic
centimetres of silver nitrate solution reciuiri'd to iieutrali/e the salt is read from the
burette. Each cubic centimetre of solution usod rcpn-sonts one p.T .•cut of salt in the
butter. Thus if two and nine-tenths (i>.9) cubic centimetres of solution arc u.scd, the
butter contains two and nine-tenths (2.9) per cent of salt.

Nearly all well wat.Ts contain more or less salt and for this reason are not
suitable for use in making the silver nitrate solution. Condensed steam from a
boiler also contains impurities which render it unfit for use for this purpose. Clean
rain water will ^ive more satisfactory results than either well water or condensed
steam from a boiler.

If well water is used to remove the salt from the butt.r in making the test it
should be tested for salt in the same manner as the salt solution from the butter is
tested. Whatever perceiitatfo of salt is present in the well water should lie dclucted
from the result of the test of the butter. For example, if the well water shows 0.5 in-r
cent of salt and the butter shows .T ,--, per rent of salt, the correct test of the butter is
3-5 — 0.5 per cent = 3.0 per cent since 0-5 per cent of salt has been added with the
water. Condensed steam from a boiler shoul.l not be used at all in making the test.
\\cll water to be mixed with the sample of butter siiould not even be heate<l by turninf?
live steam into it. The well water may be conveniently heated by surroundinjr a small
uail or jar of the well water with the hot water.

TIIKTKSTINOOFCJIEKSE.
Dktkbmimatiom or the Pm Crwr or Fat in Cimiwr

Sani'lino Ciiecmk r<m Tkmiko.

cWw taking ,h.. pl^B. from Jhl ll„, 7'''"' r " ""• /'"'" •'"^•'' «"<» "^ the
portion of each plug uL £,^11 .'"r '"/' "'"ffi*"" '" 't'i|"» B"d « «m.ll

.pread o. and ^i,!^. .J^i^a^^^/niiirr ';,:L:^ ^^'i 'ii^.r '""-'^^''

Makixij tiik Tkst.

•nd all lump, of chi-ese d Un .Dour Tl " """^ ''"'" "" thoroughly ,nixed

uf acid add.!!l and .^o toH Tnthne. r;r' 'V'"" "'"'*"V" ^"* *"• ""'« ^-S c.o

the per cent of f.a. If gJrn i./n .1 I',' T' "" """, "*"''' "^ '^^ *^'"'' «''' »>«
reading must bo multiplicl by'", i'," • r ,;:;' fT '" ;"'./'^-'"""' b-.ttlc. the
ha. boon u«cd in a 10 ,M.r .^-nt mUk bo fl , i- ' "'"' 'f " ^-^ '^''"" «'""•''«

that i*. the i^r cent of fnt ; XnT, ..I v , I . """'^'7 ""''.' ^ ""■I'ipli'-"! kv 4;
gramn for which th,. bo Uc i. ".I n 't "^ '">' tij.Iy.n^ tho reading by the nun.bcr of
gran., u.ed in nmking the i;,;"""*""-""' "'"' <''^"l"'*r the rc.ult by ,1... n,,,..,,.,^ of

Dkte..m,nat.on o. t.,k IV:,, c,nt of Wateu ,n Ckme

If 8tonm "n.icr prc4 1h r " i , ' "'T •'™.' /'°"' '^e loM i„ wci«h,.

f-r cn„t of n,oi!tur. i^ IL"!" '•!''u:."l^"'^"' «atKsf„cto_,y n.ethod of dCcrnu.ung

the ocr cn„t of n-oi^turc ^n ^.^^ e h by Irlrlh 'f '7 ""'^'"' "' d^'ernu.ung

ntisi, IS Dy means of tho liigh pressure oven (Fig. 0)

3- ^i;"s^s:— -j-^--,--.^ ™ £..^

wall, and a„ outlet to .Jniin of! .),« ron.|pn.wl w.itBr. The rhiirober ii fittixl with a
perloMlcU rack on wh..h to plm-t' th« •nmplM and whi.-h permit. fre«. circulntion of the
warm air oround the lamplri. The ovin ii fitt.d with a • !o.o-attitiir door .,nd a
thermometer extendiiiR into the phamber. The tempt-nitur,- m.iintaiiiP.| in the oven
may be reirulated by the itcnni pre«.iro appli»l in the hollow walU. In ttddillon to
the o»en, a fine balance i* nocewnry or n good moi.tur.- »<alo and u »vt „f umm

*/o«!.*" """5 ''"' '■'""'"' »""'•»""> ■"*«'« with n 10-Bram inmplo. a rondintf

of 38^8 per c-nt of moi.tur., miiy be taken, but with a Toriion toule only 30 por cent
may be road without the ixtra ttram weighta.

SIakino the Tut.

determ.nut.on. The anrnple i. thoroughly ipreud out by moan, of « knife nnd t.h.to

-hch h«. been thorouBliIy dried by heating and then bulan-.-.l on the *.-..le . t '
cooling. The cheew la .preiid iii. thinly and an evenly .„ poMibie over the di.h Tlie
work of •ampl.ng. pulverizing and weighing the ehee«. into the .ii.h »hould W- done
at quickly a» powible. to avoid lo.. of water by evaporation. The di*h containing the
Bonft t?»r '''«"-^^"' »»"• °^«'"- t!"" 'Jo'-r eh..ed. and 45 to .',0 pound, .team pre.,ure
applied to the oven. fhi. should give a teni|KTnturo of 22:,^ K. to S.'.O" F whi.-l, ,vlll
be«uffi.,entl.vi.igh. The sample in lieaU.! until all tl... waf.-r ha. UviiVv™"'.!
which point IH dcterminwl by .uoee.sive weighings followed by further heating in'
the oven of course aUowiiig the .ample to cool before weighing ..a.li time When
the .ample cease, to oso weight, the water h all evaporated and the iK>r cent of
moi.ture ,. dctermine.1 from the 1o,h in weight. The gnimn of moinure evaporated
multiplied by 10 will give the per ..ent of moisture. For example, if the ]0 gr^,n
of chee.e lose 3.6 gram, of water from evaporation. 100 «r,.n.8 would lo.e 1m„
divided by 10 and multiplie.1 by a..'.. whi<li equals .1.1 grams, or 3.'. per cent
1 u, •"?'" V"''*" !"'«^^»s"'-c '" not available, nn oven construct.Hl of tin or co,.,.er
double-jacketed on favo .ides fitted with « perforated rack and close-fittin;, Lr
will answer. The oven is filled with water between the walls, which water is ken[
boding by mean, of a gas jet or alcohol lamp. Since n lower temperature (i.oilin'o
. employed with thi. oven than with the high-pres.ure oven, longer time is required
to dry to constant weight. '

In Conci-psion.

To one who ha. carefully read the foregoing pages it will, no doul.t, seem th.t
the manipulation of the tests described is -luite simple. While such is the case
extreme care and accuracy must bo exercise,! in all details of the tests in order to
secure accurate nsults. Many of the details in connection with the making of
these test, seem unneeessarj- to the careless operator, but the neglect of n few minor
detail, in making the tests means the difference between accuracy and inaccuracy
It 1. not uncommon for instructor, of milk testing to find students without previous
cxp<-rience in testing doiii^ more accurate work during the tir^t few davs of their
training than they do some week, later. " Freedom breeds contempt." and as some
operators become more familiar with the tests, the work is less carefullv and' less
accurately performed. To slightly change an old axiom: "Eternal care i. the price
of oecuracy." '

inx Of vrmioiAizon of thi saut aib oold iiokaoi
oomranons's sebzes.

— SUf*^"^

V JamaMTf 1006, «•

i0otiijss.?8». njoftsfiirisss: i%ri?i0i4.

Buuivrara.

v*.

1000

10(16

1006

10OT

IT
•M
•00

I 111 )

t 111 1 '

»Si

1011

101S

•00

1010

1018

1010

101S

nit

41
40

1014

Datwalastlmi •£ F»« •md WmUm la

Pw e rt yttT*

1010

1011.

1011
1010
1014
1014

1014
1016

1016

•14

1018
1016

10OT

1011

Iilat «t 0«M« BHtfak iMporten o< Tmm V»*4«ota.

XJat at a M gwto w of Bom* Ouuidiaa V^torta. . ^ ^ .

0«M« •< &• raaton tluit Oomtvol «k« W»*m 0«ai«mi •<

Batter.
0w*«t Ommb Batter.
A»M3atea for taa

Battar.
B atUr io MaE oa tka Farm.
Tka Vaa of laa oa tha Farm.
Tka Oool^tB of mik for OkaaaoaaUafr
moaort oa loato Trial 0UMMate of Oold 0toH«o Ayylaa.
OoalowBter OluMaa, BoaMllotaa oa Ite llaaaf aataro.
Trial 0UMaoate of.Faaohaa. 1010.^ ^ ,
Tha Dawylac ladaatry, aa Hlatorlaal aad

Aaaoaat.
Botaa for Faatorj Okaaanudkara.
Tka Cara of Oroaat for Batto ri a Ma ».
Oow Taatlas (aaparaadaa Bo. 10).
0aall Oold 0toracoa, Bta.
Oold 0toaM(aa for OraaaiaHaa.

Tka lalaad of OrioMM Okaaaa. -..._.. ^

T.I.* «c Okaaaa Faaterlaa, Oraaatariaa, 0ktauHlac>8tatteaai

aU> Ooadaaaed KUk aad CHtr Flaafa, ate., la Oaaada.
Okaaaa Faaterr aad OraaatOTT^PUata wltk 0»aeUaat!«aa.
Tka JhSrr ladaatry^Aot, 1014, aad Ba«rfatliM.
Tka Oold 0torasa Aet, 1007, aa Aataadad la 1000, and

Baaalatloaa.
Tka Oold 0toraaa of Food Frodaete wltk aoMO aotaa oa la-

aalatloa aad Warakoaao Maaaseaiaat.

<]IB01TLABS.

Tko mik Taat A«t.

Tko Oatlook for Oaaadlaa Tomatoaa la Oraat Brltala.

Oraaiary Oold Sterac* Boaaaaa.

Botaa oa Ooar Tastlas.

Bavlaad Uat of A*pla Daalara la Xortkara Oatarlo, Maal-

teka, 0aakataka«aa aad Alkarta.
Tka Braadla« of Dairj Batter. _ . . ._ ^.

iliOFroeooUas aad 8terace of Fralt for tko

Ooautorelal

of 1016.
Oaaaaa of Tarlatloa

0oparater Oraam.
Okarrr Praeoollac PoadlUltlaa.
Oow Toattac Botaa.

la tka Feraaataca of Fat la Haad

0FBOIAI. FUBUOATIOira.

Kap akowlaa tko Loeatloa of Okeeae Faetorloa aad Oroaai-

^ j|0g fa Oaaada.
Bapert of tka Tklrd Domlaloa Ooaf araaaa of Dairy Baporta,

Daaemkar 6 aad 7, 1011.

Aay of 4k»— paUloatleaa will k« leat froa of ekargo oa appllaatloa te tko
Sairr aad Oold Oteraxa Ooaudaaloaar, Ottewa, Oat.

^^7i aSloteat aamSor of kaUatlaa 0O, 80 aad 88 aad alrealar 14 arlll ka saat
«• tka maaaaer of aay ekaaao f aetorr or eroaaorjr te aapply oaa te eaek patroa.

Boa. 3^88 aad 40 ralatiac te tka fmlt (rowlac ladaatry kara kaoa traaa-
f arrad te tk» Fralt Oeaualaateaar'a Braaek.