ll “WWI"“I“: MIMI: I: :‘l: l I: CC‘MP ERA“! “viii": 1313.": 5L 5?:: W {)5 IUE‘K A}: {2135; 501'?€.EB Mimi 5:5: brat: CI. 555;; 4 w t: .5 "a (z a: 5“. z as at ‘THESIIsI II 0-169 - ”'1‘ ‘ -A __. - - .l .1. A, L- '._ .2-fl-._._L._ ‘4 -.; .. _':'“A—A_.AL- --.. This is to certify that the thesis entitled A Comparative Bacteriological Study of Bulk Milk Versus Bottled H111: presented by lanflce Go 13118 has been accepted towards fulfillment of the requirements for M degree in M91037 Majtk professor Date MM ‘-‘.—._._ .h-- .H‘- A GOIPARLTIVE BACTERIOIOGICAL STUDY OF BULK MILK VS. 30m MILK llauriee 0.;ng A THESIS 8nh1ttedtothe8chool “Graduate Shudieeornehigan State College or Agnculture and Applied Science in partial fulfillment of the requirements for the degree of mm 0]? SCIENCE Depart-ant of Bacteriology 1950 The writer viehee to apron his appreciation to m'. I. L. mm for bill guidance and helpful critici-I given this work. 1339289 TABIEOF GONEN'I‘S Wimeoocoooeoeeeeeeeeeoeeoeoeeeeeeeeeeee Deecriptian or Intern]. and Phases of Study“... Exporilflntal,Procednr.eeeeeeeeeeeeeeeeeeeeeeeeee 35'1"'at,Dat‘eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee Di'cufllianoeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee 00901n'10n800000eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee Litiritur. Citheeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee 2h .0..... IOIQOooelioouw.OOOOOOQOOOCIOG- COOOQODOOICOOOIQ... ODOlioootooneee- 0.....0.......I‘..O.v ....-.........5..IO'. .I...OOO .000.- 3.0.9.0000 .QIGOOCI‘IQOOCOQOOI‘I.CDC-ICU... INTRODUCTION “me greateranountofthenarketnilkaoldintheflnited States today is required either by law or regulation to be sold or dispensed only in sealed containers filled at the dairy (7). The primary reason for the enac‘hnent of such laws or regulations is to assure the consumer Iholenilk, not skinorpartlyakilnilk. hentyyearsagoitias cos-mpractice tosellhulknilktore— tail merchants for resale to the public. The methods used for dispensing thenilktothe individualuereusuauyoneoft'o: (l) dippingthenilk from the bulk container and pouring into the patron's container, or (2) pouringthefilkintoanumanddispensingthroughafcucet atthe bottcn. In either case if the milk were not stirred frequently, the creantoruld rise to the top and some patrons would not receive whole milk. The milk bottle was introduced to expedite retail distribution and to show the cream line (1;). {this assured the patron that he us getting nilkvithagood creamcontent andnot skin-ilk. Inotherierds, the nilkbettle helpedtoprevent fraudinthe sale ofailk. I‘tnsnot until later that regulatory measures for the sale of milk were incorporated into some villages and city codes. is early as 1896 an Ohio circuit court upheld an ordinance pro- hibiting the sale of bulk milk (6). Bauer, it was not until 1915 that the State of lichigan recorded on its statute hooks an act Ihich had refer- ence to milk bottles (1). .2- Two years previous Act 222, Session me, 1913, of the State of mchigan, was adopted which provided for the prevention of the sale of unclean and insanitary milk and cream. This is the first reference to action taken in connection with nil]: in the interest of the public health by the State of llichigsn. Itwas not untill919tha:t aState Councilofiisalthwas author- ised (2). thus itwould appearthatintheState offlichigantheuse of the individual nilk bottle was not initiated booms bulk milk was considered a public health hazard, but rather, as a convenience for the processor. To further substantiate this, Bulner (1;) states that when the nilk bottle was first introduced "many public health officials . . . ...grsvely frowned upon this departure fro. the dip nilk"." In 1931 it was the opinion of sanitarians and public health authorities that nil]: dipped frcn large containers for sale to individuals "is a potential nenaee to health‘ (5)." today it is generalJy accepted that nilk drawn through faucets or valves or dispensed fra nilk pupa, is a potential menace to the public health, the reason being that the seal of pasteurization has been broken. his is done when the milk is either exposed to the dipping utensil or is passed through a valve or milk pulp. If the equipment were properly sanitized no public health hazard would result, unfortunately such equipment is generally improperly sanitized. In addition, if the nilk is nixed and poured from the can into a pitcher, and then into glasses, it is thrice enosed to possible contamination. -3- ntheabove oases theniseithereneornorebreaksinthesealofpas- teurisation, exposing the milk to the possibility of contamination with bar-m bacteria (3). The first definite reference requiring that nilk be served to the restaurant trade in the original container is found in Regulation 78h, adopted July 25, 1939 by the lichigan Department of Agriculture. In 1910 Regulation 78h was replaced by Regulation 500 which new states (paragraph 17) IWill: shall be served to the customer in the ori- ginal individual container ......" not simply the original container. This regulation was rescinded July 12, 191:9 by the adoption of aRegu- lation No. 525 Governing Food Establishments“: ." These regulations were pronulgatsd under the authority given to the Omissioner of Agriculture, The State Health Omissioner, and the inchigan liquor Control con-isomer, by the following enactments of the nchigan legislature: Act No. 3m» Public Acts of 1917, as amended, Act No. 11:6, Publiclets ofl919, asanended, andlctllo. 8, PublicActs of 1933, Extra Session, as amended. It is well established that the original individual container guards against both fraud and public health hazard. With the introduc- tion of homogenized milk, it would be possible to use bulk milk without danger of fraud as it pertains to ski-ing or uneven distribution of milk fats due to mechanical separation. If homogenised milk were dis— pensed fl‘on a properly sanitized can, stored in a refrigerated unit, it -lt- wouldseenthattheproductfrau suchacontainerwouldbe equaltothat obtained free an individual container or bottle as measured by either quality or sanitary significance. The researches presented in this thesis were planned to obtain sane of the answers using equipment available on the narket at the present tine. DESCRIPTION or serum. dim mass or arm)! Toundsrtalnethisstudy'twodispenserswereused. Dispenseru' andDispenser"B'I bothusedrefrigerated'units for storageofthecan whflninnse. Dispenser 'A' (Plate I) uses a can fitted with a recessed ca- pression Joint (l) with a lock nut (2) on the side at the bottom. A a stainless steel tube (3) is used in the dispensing apparatus with a single service rubber compression gasket (1;) and a single service rubber stopperG). These parts are assembled as shown in Plate I, (_6_). The assembled parts (6) are inserud into the can and secured in place with the wrench (7). Having been secured in the can a dust cap (8) is snapped over the spend and of the dispensing tube. To put the dispenser into operation the can is set in the refrig- gerated unit (1) shown in Plate II. The dust cap is taken off and the tube is retracted with tool (9) shown in Plate I. has tube is placed -s- in the dispensing mechanism (2) Plate II and the natal tabs on the rubber stopper are secured with traction claws to the control handle (3). Dispenser '3' (Plate III) uses a can fitted with a small nipple (1) in the bottom. The single service rubber tube (2) which is plugged atborth ends andwrappedinaplastic envelope, is cut inhalf. One half of this tube is attached to the nipple and secured in brackets around the bottom of the can (3). The dispenser is put into operation by setting the can into the refrigerated unit (1) Plate IV. The rubber tube is taken fro: the brackets and threaded through the dispensing nechanisn (2) which con- sists of a weighted pinch-cock arrangement. The protruding end of the tube is then cut off flush with the bottom of the mechanism. The first phase of the experimental work involved the use of the dispensers in a student domitory. Dispenser "A" was tested first and Dispenser '8' at a later date. Both machines were operated by the one people and under identical conditions. The cans were washed, sanitized and filled at Dairy “P. llilk was refrigerated at the domitory in the coldboxes untilreadyforuse. The canswers thenplacedasnsededinto the refrigerated units and the milk dispensed. Hmogenized milk was used. The second phase of this work consisted of checking the every day run of bulk silk in cmarison to the bottled milk at four dairies, namely, "A”, '3“, "HI and “11". This was done in order to determine whether the PlateI Photograph of Dispenser Parts for Dispenser "A“ ‘I‘. I .v “I g e.‘ . «as: D 1 Compression Joint 6 Assembled parts 2 Lock nut 7 French 3 Stainless steel tube 8 Dust cap h Rubber compression gasket 9 Retraction toll 5 Rubber stopper Plate II Photograph of Refrigeration Unit for Dispenser "A” l Refrigeration Unit 2 Dispensing mechanism 3 Dispensing handle Plate III Photograph of Dispenser Parts for Dispenser "B" l Nipple 2 Double length dispenser tube 3 Dispensing tube in place PlateIV Photograph of Refrigeration Unit for Dispenser "3" l Refrigeration Unit 2 Dispensing nechanisn -6- ordinary methods of washing and sanitizing milk cans would yield a pro- duct of coaparable quality to bottled milk. mm PROCEDURE Samples were taken periodically and during each period two or more samples were taken from the dispenser and two or more bottles of homogenized milk were collected at the dairy. The bottled and bulk milks were ofthe sane day's processing atthe sane deiryandwere taken in this manner in order to remove as many variables as possible. The couples were immediately refrigerated after collection. The sane method of collection was used in phase two. The bacterial counts were made according to the procedure set forth in Standard Methods for the hanination of Dairy Products, Ninth Editions Following the accepted procedure for averaging the data obtained, the logarithic average counts were determined for each test period. REVIEW OF DATA PhaseI Dispenser "A" During the first ll periods of comparison, the dispenser cans were washed in the ordinary manner. The stainless steel tubes were -7- washed by hand, then.placed in a.pail of sanitizer (quaternary ammonium compound) with the single service parts. The parts were assembled while wet and secured in the can. The cans were then filled and delivered to the dormitory. The data for these 11 periods are shown in Table I and Graph I. During these 11 periods, the bacterial counts of the dispenser samples were higher than the bottle samples in 9 (81.8%) of the test periods. The average count for the bulk milk in the 11 periods was 33,000 bacteria per ml. and for the bottled milk the count was 15,000 bacteria per ml. The bulk milk had an average count 120 percent higher than that for bottled milk. The increase in count of the bulk milk was assumed likely to be due to improper washing and sanitizing of the dispenser cans. Based on this assumption the dairy was asked to change their handling pro- cedures. After the cans had been washed and steamed in the can washer, each one was rinsed'uith'the same type of sanitizer used in the assembly of the parts. The reason for doing this was to insure proper sanitizing of the cans. This procedure is termed "special treatment." The data for the 15 test periods following the change to "Special treatment" are presented in Table II and Graph II. The bacterial counts of the dispenser samples were higher than those of the bottled samples in 9 (60%) of the test periods, thus a --8— drop of 21.8 percentage points over the counts obtained without "special treatment." The average count for bulk milk in these 15 periods was h?,000 bacteria per ml. compared.with hl,OOO per ml. for the bottled.milk. The average count of the bulk milk was only lh.6 percent higher than the average count for the bottled milk. This demonstrated that with a properly sanitized can the bacterial countS'were comparable to the counts for bottled milk. A8 a consequence of the above results, it was considered advis- able to¢:ontinue the "special treatment" during the test periods involv- ing Dispenser "B". Dispenser "B". The data for the 31 test periods are shown in Table III and Graph III. The dispenser samples gave higher bacterial counts than did the bottled samples in 6 (19.3%) of the test periods. 0n the other hand the counts of the bottled samples were higher than those for the dis- penser samples in 13 (h1.9%) of the test periods and in 1.2 (38.7%) of these periods the counts were comparable. The average count of the 31 periods for bottled milk was 32,000 bacteria per ml. and that for bulk milk was also 32,000 bacteria per ml. The results obtained with these two diSpensers when the cans were properly washed and sanitized indicate that the quality of the milk delivered to the consumer was identical to that of bottled milk, from a public health standpoint. Phase II The second phase of the eXperimental work was to determine whether the can washing and sanitizing procedure now generally prac- ticed will yield cans properly sanitized so that milk put in these cans will be comparable in sanitary quality to bottled milk. The four dairies selected were not chosen for any definite rea- son. They represent the leading dairies in the area under study. Daigz "A" The bulk milk handled by this dairy is standard pasteurized. The samples were taken within 15 minutes after the cans or bottles were filled to avoid cream separation. The data obtained are shown in Table IV and Graph IV. The results of the 13 test periods show bulk milk to be higher in bacterial count than the bottled milk in h (30.7%) of the test periods. The bottled milk counts were higher than those for the bulk milk in 8 (61.5%) of the test periods and of comparable counts in 1 (7.79%) of the test periods. The average count for the 13 periods was 28,000 bacteria.per ml. for the bulk milk and 36,000 bacteria per ml. for the bottled milk. These data show that the average count for bottled milk is 28.5 percent higher than the average for the bulk milk. -10- Dairy "H" The bulk milk at this dairy was homogenized. SampleS'were col- lected within one hour after filling of the cans or bottles. During these 13 test periods 2 (15.3%) show bulk milk to be higher in count than bottled milk, 8 (61.5%) of the test periods show bottled milk to be higher in count than.bulk milk,'While 3 (23.1%) periodS'were of comparable counts. In the 13 test periods 5 (38.1%) show bulk milk to be higher in count than bottled milk. The bottled milk counts were higher than bulk milk also in 5 (38.14%) of the test periods, while 3 (23.1%) per- iods were of comparable counts. The average counts were 23,000 bacteria per ml. for the bulk milk and 2h,000 bacteria per ml. for the bottled milk. These averages are comparable. Dairy "M" This dairy used standard pasteurized milk for bulk shipment, therefore, prompt sampling was again practiced. Table VII and Graph VII show the data collected. These results of the hh test periods shOW'the bulk milk to be higher in bacterial count than the bottled milk in 16 (36.3%) of the test periods, while bottled milk counts were higher than those for bulk milk in 22 (50%) of the test periods, with 6 (13.6%) periods -11.. showing comparable counts. The average count for the bottled milk in these hh.periods was 21,000 bacteria per ml. and for the bulk milk the average is 19,000 bacteria per ml. The data collected from these four dairies show that the bacter— ial counts of the bulk milk were lower or at least comparable to the bottled milk. Therefore, if these four dairies had handled bulk milk in dispenser cans at this time the sanitary quality would be comparable or better than the bottled product. DISCUSSION The study of the two bulk milk dispensers show exactly what one would expect. That is, if a good quality of milk is put into a properly “washed and sanitized container it will retain its original quality when properly stored. These studies demonstrate also that in many cases the bulk milk was of better sanitary quality than the quality of the corresponding bottled milk. This may be due to better washing and sanitizing of the cans than of the bottles. If washing and sanitizing of mans and bottles are comparable the bulk milk would carry less organisms than would the .12 .- bottled milk. This would be due to the fact that there would be less surface area in contact with the bulk milk per volume than with the bottled milk. The four dairies used in this study were selected due largely to the fact that they are the leaders in this area. Even though they show excellent results it cannot be assumed that all dairies will do the same. There are some dairies in the country that do not do as good a job of washing and sanitizing as the four studied. The question of the use of bulk milk dispensers resolve down to the fact that the container is not the important element in producing a quality product but rather that the washing and sanitizing of the con- tainer into which the milk is placed is the paramount issue. The acceptance or rejection of a bulk milk dispenser in any area should be on the basis of how well the dairies of the area handle the 'washing and sanitizing of their cans. It is also necessary to consider the personal hygiene of the milk handlers in each dairy. In the final evaluation the machine which is to be accepted should be selected on the basis of single service parts. These so called single service parts must be absolutely single service. They" must be so designed that after one usage it is impossible to use them -13.. again. Dispenser "A"'s single service partS'were not of this kind, for it was possible for them to be re-used. It'was noted that if the milk handlers are rushed, it is very easy for them to leave the once used single service parts in the dispenser can and re-use them. Also in the final evaluation of the machine it should be cone sidered whether the dispenser cans must be used in the refrigerated cabinet or if it is possible for these dispenser cans to be used without the cabinet. There is yet one final consideration, that is, will the volume of milk be too large for the establishment which uses the dispenser. All milk delivered should be served within 72 hours after pasteuriza- tion. All cans should be labeled with the date of filling. All milk used in bulk dispensers must be homogenized. CONCLUSIONS The results of the comparisons of the dispensers and bottled milk indicate that if the dispenser cans are not properly cleaned and sanitized the product will be of higher bacterial count than that in the bottles. ~11:- If properly cleaned and sanitized a can with the least possible number of avenues of ccntaminatialwfllyieldaproduct equaltothat of bottled milk, as measured by either quality or sanitizing significance. The problem of producing a good quality product resolves itself to the fact that proper washing and sanitizing of the containers, se- gardless of whether they are bottles or cans, is an absolute necessity. Therefore, if single service dispensing equipment is used, the cans must be properly washed and sanitized and the product must be preperly refrigerated. Under these conditions it will be possible to dispense bulk milk of the same quality as bottled milk. Table l. The Comparative Bacterial Counts of Hucgenized Bottled Ink and llilk Dispensed frm Bulk llilk Dispenser ”A" Washed and Sanitised Without aspecial Treatment'I Ira Dairy up Bulk Bottled Test Period Log Average Log Average 1 18,000 10,000 2 16,000 6h,000 3 28 ,000 7,000 h 50,000 7,000 5 2h,ooo 6,000 6 69,000 22,000 7 311,000 5,000 8 63,000 “.000 9 58,000 75,000 10 27,000 13,000 11 25,000 8,000 ”01186 33,000 15,000 0H eoaaoa shoe N. xHHE xddm xHHE reappom D I =5: hhflwfi xaws oouwcomosom anmSPwohB Hwfioomm paonvfla =4: Hmmfimamfim H sodas 10H ION uom tom r8 10» tom (spuesmqm) °o°o .Ied eueqoeq GBEJIQAV Bo]: Table 2. The Cmparative Bacterial Counts of Honogenized Bottled ~16- lflk and Hill: Dispensed from Bulk lilk Dispenser "A" Washed and Sanitised'With Special Treataent' from Dairy "11" Test Period p Fugfissooowmmrum 51 Average Log Average 31,000 73,000 10,000 5h,000 27,000 103,000 57 .000 55,000 . £0,000 30,000 3h,000 58,000 88,000 151,000 H.000 10,000 Battled Dog Average 116.000 6h,000 25,000 01,000 hh,000 69,000 3h,000 10,000 £56,000 35,000 19,000 28 ,000 8h,000 202,000 27,000 111,000 eoaaoa shoe ma 4H ma NH AH 0H m m s b m a m m .n o 1 ON L _ ,. a . _ - u 8 T Ow I I 00H 1 cud saw: MHsm nu sass eoapeom _- - 04H u cod :2: hon—“ma l OQH Mafia oouficowosom hostess Heaooom fig 1 8m :4: noncodofin HH eases (spuesnoql) °o°o Jed eraeqoeq eBeJeav 301 .17.. Table 3. The Oulparative Bacterial Counts of Hoaogenised Bottled Hill: and lilk Dispensed fro- Bulk Milk Dispenser '3' Washed and Sanitiaed With Special Trea‘hlent' From Dairy '1!" Bulk Bottled Test Period Log Average Log Average 1 10,000 10,000 2 19,000 27,000 3 36,000 h8,ooo h 3h,ooo 116,000 5 311,000 36,000 5 37,000 36,000 7 31,000 28,000 8 21,000 29,000 9 77,000. 28,000 10 1:13.000 118,000 11 22,000 26,000 12 77,000 16,000 13 61,000 16,000 In 9,000 13,000 15 17,000 18,000 15 13,000 18,000 17 28,000 11,000 18 3h,ooo 36,000 Test Period 19 20 21 22 23 2h 25 26 27 28 29 30 31 Average -18- Table 3 (continued). Bulk Log Average 16,000 25,000 27,000 39,000 16,000 36,000 15,000 15,000 53,000 50,000 83,000 116,000 118,000 32,000 Bottled L08 m 91,000 25,000 26,000 76,000 17,000 h7,000 h2,ooo 111,000 53,000 10,000 92,000 111,000 53,000 325000 888 some Hm on mm mm E 8 mm .3 mm mm S 8 9” ed as ma g Mag U and: voflvpom I =2: hHfiwQ has poaaeomoaom escapees do.“ 00% 5.3., cm: 90983.3 HHH eases OH O O O —:f m N O m 53 58 (spuesnoqm) 0000 red ewaqoeq GSBJQAV 801 O CO 8 .19... Table )1. The Comparative Bacterial Counts of Pasteurized Bottled Hill: and Bulk Milk Tran Dairy “A” Bulk Bottled Test Period Log Average Log Average 1 10,000 15,000 2 55,000 59,000 3 77,000 115,000 h 112,000 19h,000 5 1,500 175,000 6 119,000 127,000 7 17,000 13,000 8 57,000 58,000 9 28,000 17,000 10 8,500 11,000 11 18,000 21,000 12 20,000 217,000 13 88,000 88,000 Average 28,000 36,000 MH NH HH eoaaoa pods Mag 033m U a: voHpPom I g posagopmom e«=_haaoo sH geese lo ION Io: (spuesnoqm) '00:) .red uueqoeq eBesenv 301 -20- Table 5. The Couparative Bacterial Counts of Hanogenised 301?th Milk and Bulk [ilk Fran Dairy '3' Bulk Bottled Test Period Log Average Log Average 1 20,000 20,000 2 2h,000 23,000 3 127,000 15,000 h 18,000 21,000 5 800 3.000 6 600 12,000 7 23000 [£23000 a 172,000 50,000 9 17,000 28,000 10 26,000 h0,000 11 1h,ooo 21,000 12 16,000 ' 3h,000 13 12,000 12,000 Average lh,000 20,000 OH eoaaoa toe m p b is :53 xHHE MHdm xHH: oodpvom :5! U I fies poaBomoaom =m= Earn a cameo ON on 0m 8 OwH % (spuesnoqm) 00's 19d spleqoeq,efiesenv 301 -21- Table 6. The Comparative Bacterial Counts of Honogeniaed BottledmkandBulklinfrmDairy'L' Bulk Bottled Test Period Log Average Log Average 1 13,000 37,000 2 72,000 75,000 3 113,000 115,000 8 118,000 82,000 5 5,000 2,000 6 5,000 27,000 7 9,000 9,000 8 6,000 h7,000 9 25,000 27,000 10 171,000 11,000 11 91,000 21,000 12 ‘ 8,000 13,000 13 10,000 11,000 Average 23,000 211,000 O\ m coflom ghee as has 0 one: 338m I F 0 a I a: voufiGowofiom =9: haw-an He sodas ION I o: 10w IOOH TONH T 04H I8H 18H (spiresnoqg) 0000 .xed epseqoeq 981219177 301 -22- Table 7. The Caparative Bacterial Counts of Pasteurised BottledliJkandBulkmlkmnairy'l' Test Period \OGJ'QOVlF-‘UJMH BIZ-'5 C'. Log Average 22,000 2h,ooo 18,000 10h,000 163,000 35,000 36,000 26 ,000 50,000 27 ,000 19,000 7,000 28 ,000 17,000 23,000 16,000 33,000 17,000 52 ,000 87,000 Bottled Inshore:- 7,000 9,000 105,000 3,000 7,000 6,000 28,000 33,000 33,000 113,000 9,000 105,000 20,000 55,000 101,000 16,000 18,000 66,000 66,000 -23- Table 7 (continued) Bulk Bottled Test Period Log Average Log Average 21 80,000 13,000 22 22,000 131,000 23 8,000 8,000 2h 20,000 20,000 25 9,000 6,000 26 1h,000 8,000 27 11,000 10,000 28 15,000 22,000 29 17,000 20,000 30 10,000 5,000 31 26,000 18,000 32 10,000 15,000 33 7,000 12,000 3h 10,000 28,000 35 10,000 16,000 36 7 ,000 33,000 37 9,000 65,000 38 7,000 35,000 39 5,000 10,000 ho 9,000 12,000 bl 19,000 28,000 02 17,000 2h,000 is 35,000 20,000 M: 20,000 28,000 Average 19,000 21,000 poaaom toe H ON on om OOH ONH 0.: ea: 685 D 81“ one: 2:38 I 00H wa2 bonansopmmm =3: hafiwm HE sodas spuesnoqg) Btéeqoeq OBBIGAV 30¢ °o°o Jed 1. 2. 3. h. 5. 6. 4hr LIMATURE CITED Act 1511, of Public Acts of 1915 of the State of Hichigan, Compiled Lm, llichigan, 1918, Vol. 2, Chapter 288 (288-351 - 288353). Act 1176, of Public Acts of 1919 of the State of Michigan, Calpiled 1m, llichigan, 191,8, Vol. 2, Chapter 3325 (325.1 - 325.13). An abstract of, Septic Sore Throat Spread by Bull: lilk, Baltimore Health News, July, 1938, Jour. llilk Tech. 2: 86, 1939. Bulner, L. 0., The Probln of Reccntsninaticn of Pasteurized lilk and its Products. Jour. llilk Tech. 1: 10-13, 1938. Is loose llilk a Health Hazard? Report of the Hill: Connissicn, Health Depart-eat, New York City, 1931. Tobey, Janos A. , Legal Aspects of 11.11: Sanitation, [ilk Industry Foundation, Washington, D. 6., 2nd Edition, 19117. 38M USE ONLY MICHI AN STAT G E 31293 IVERSITY LIBRARIES 3 T 015 6730 fl 0