Liquid Industrial Wustes cided t h a t most of the waste liquors could be improved by compositing, neutralizing, settling, aerating, and diffusing into receiving waters. The initial cost of the plant for these operations, including a 10,000,000-gallon compositing basin and a 60,000,000gallon settling lagoon, was reported t o be $325,000.
Carrent Developments Further extension of methods for treating liquid residuals from this industry is represented by a plant under construction, which projects a n expenditure of $1,600,000, for 1% aste disposal installations, including $460,000 for a deep disposal well. This installation also represents the grom-ing practice of management not to approve installations for productive operations until adequate provision for waste disposal has been incorporated in the design. The same company is pushing into new frontiers of m-aste disposal methods, as represented by the development of a catalytic method for destroying low concentrations of organic wastes in water vapors. It is reported t h a t this method also gives promise for dilute water solutions.
There are many individual or gioups of similar waste liquors which must be specially treated a t the process sources, because they cannot be treated satisfactorily in processes for general mixtures or because they are harmful to the processes for these mixtures, such as toxic substances in biochemical processes. Furthermore, manufacturing plants which are restricted as to available land for treatment of large volumes of process wastes place relatively more emphasis on elimination of existing and prevention of new pollution a t the process sources by recovery, process modifications, and treatment. That part of the problem in this industry remaining largely unsolved is comprised of a large number of waste liquors which do not present much of a problem individually but which are of considerable concern collectively. Individual treatment is not practicable and treatment of composites is made very difficult by exceedingly complex propelties of the mixtures. Treatment as a whole or as selected mixtures is further complicated by using process sewers for cooling n atel in many cases. RECEIVED for review September 6 , 1931.
ACCEPTED
Jsnuary 15, 1952.
PACKING OUSES F. W. MOHLMAN, The S w a i t a r g District of Chicago, Chicago, Ill. W. V. HILL, GreeZeg und H # 6 n S @ n YClaicugo, Ill. Frequent investigations of treatment of packinghouse wastes have been made, but the processes in full-scale use are largely limited to use of activated sludge or trickling filters, provided there is sufficient dilution with human sewage to support biological processes. Prior to treatment, salvage processes such as screening, evaporation, or flotation reduce the concentration of the wastes to be treated. Surveys to determine unit losses are of value to show the degree of recovery and the remaining load that
must be handled by treatment processes. The largest treatment plant for packinghouse wastes is at Chicago, the Southwest Treatment Works of the Sanitary District of Chicago. This activated sludge plant is successful because of ample dilution and careful control of air and sludge flows. A recent process, anaerobic digestion, is being tested at Austin (Hormel C o . ) and in New Zealand. This promises to give satisfactory results at lower costs but so far has been tested onlv on an experimental basis.
P
dustry in research on processes and improvements, but until recently has ignored waste disposal. However, for the past two years papers on waste disposal (3, 7 ) have been included on the program of t h e annual conference. Also, the program of the Department of Packinghouse Practice of the institute included a paper (5) in 1950, and those a t the meeting evinced considerable interest in the subject. Thus, there n o v appears t o be an awakening interest in waste disposal, or a t least it has been publicly acknowledged t o exist.
ACKIKGHOUSE operations have been supervised for many years by scientific chemical control, and salvage of recoverable by-products has been associated wit.h this industry from its beginning. Innumerable by-products have been developed, many into quite profitable materials, and this salvage record is still progressing, stimulated by the discovery of cortisone, varieties of insulin, and various animal protein factors. It seems surprising, therefore, tmorealize that there is still a packinghouse waste problem, which has apparent,ly eluded the efforte of packinghouse chemists, but it is probable that management has the final decision concerning the money required for treatment facilities and perhaps t,he allocations for such facilities are not made until conditions become so bad that, something must be done. However, investigations of wast,e treatment are nunierous in the packinghouse industry, by employees of individual plants, municipal or state officials, or by the U.S.Public Health Service or similar agencies. Centers of t,hese investigations are the Sanitary District of Chicago ( 1 2 ) , the cities of Ft,. K o r t h ( 9 ) , Mason City, Ia. ( 6 ) , Sioux Falls, S. D. (a),South St.Paul and Fargo, N. D., Madison, Wis., Austin, XIinn. ( 4 ) , the Department of Health of Pennsylvania ( I d ) , and the City of huckland, N e x Zealand (13). Auckland has made a thorough, up-to-date survey of the entire subject of disposal of packinghouse ivastes, and some of the data from this survey will be quoted later. An aerial view of the Chicago stockyards is given in Figure 1. The investigations have not included such agencies as the American Meat Institute, which has long represented the in-
498
The Waste Disposal Prsblem Degree of efficiency of salvage determines the magnitude of the waste treatment problem and therefore losses per unit of raxy material .(live weight) shovi the magnitude of the problem and the degree of success t h e packer has att’ained in preventing losses t o the sewer of recoverable materials. Unit, quantities have been reported frequently, even when losses were expressed per “hog unit,” since replaced by the unit “1000 pounds live weight.” When placed on this basis losses average 14 pounds of biochemical oxygen demand, 12 pounds of suspended solids, and 2 pounds of nitrogen per 1000 pounds of kill, and grease varies widely, possibly as low as 0.1 pound and up t o 6 pounds per 1000 pounds of kill (8). Mortenson (10) says t,he losses of floating fat can be kept as low as t h e minimum value-0.1 pound-with well-designed and operated grease-skimming basins. Note t,hat this does not mean total grease, but only the floating grease. Experience indicates t h a t most houses have much greater over-a,ll grease
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 44, No. 3
LZquOd IndustrOal Wastes losses than this figure, and it appears t o be more a n objective rather than a standard that can be and is being maintained. T h e flow of water comprising the industrial wastes from slaughter of 1000 pounds of animals varies widely-from 1000 t o 1500 gallons per 1000 pounds at packinghouses where water must be conserved and is used frugally, and up t o a t least 4000 gallons where water is used lavishly or even wastefullv and none is recovered. T h e flow usually varies with t h e kill, and temperature also affects the unit volume of flow t o some extent, as shown by Hill (5). I n Chicago, excessive losses of flow have been reduced by a second survey, where the first survey showed high lossee. Usually more efficient grease basins , more evaporative capacity, or more effective screens have been responsible for the decreased losses. Treatment of t h e residual waste has usually followed the procedure required for handling municipal wastes, because of the organic nature of the wastes and experience in the use of screens, tanks, or filters for preliminary treatment of sewage. The presence of excessive amounts of grease usually is responsible for a n y differences in t h e layout of the treatment plant, and one characteristic of packinghouse wastes-namely, high nitrogen content-is responsible for flotation of sludge in activated sludge final settling tanks, requiring t h e use of special arrangements for removing sludge more rapidly than under normal circumstances.
Standard Processes of Treatment
Figure 1. The Heart of the Meat Packing Industry in the United States
There are a number of full-scale treatment plants for packinghouse wastes, usually in combination with more or less domestic sewage. This combination is desirable, as operating difficulties are avoided by having a sufficient excess of sewage. T h e most widely known and largest of these combination plants is t h e Southwest Sewage Treatment Works of the Sanitary District of Chicago, where some 850 million gallons of total pumpage per d a y includes approximately 50 million gallons of packinghouse wastes, An aerial view of this plant is given in Figure 2. The combined flow is treated by activated sludge and t h e plant reached capacity just during t h e past year. Results during this period are shown in Table I. These results show a n average pumpage of 828 million gallons per day, with a biochemical oxygen demand of only 147 p.p.m. and suspended solids of 193 p.p.m. T h e packinghouse waste was much more concentrated, but the rather weak sewage diluted it sufficiently t o permit successful activated sludge operation and stabilized the organic content so that operation would not be
-
Table I.
upset by rapid fluctuations in sludge index or dissolved oxygen content of t h e aerating mixture. Removals of 92.5 and 90% of the biochemical oxygen demand and suspended solids were accomplished and troubles usually associated with more concentrated Kastes were not encountered. However, this does not mean t h a t operating difficulties were not noticed. This d a n t has t o be operated with the utmost care t o avoid overloads, and the strong midweek industrial load and t h e
Chicago stockyards
amount of air applied and solids recirculated have t o be adjusted almost hourly t o avoid disaster when t h e strong Monday and Tuesday packinghouse load appears. The effect of more concentrated packinghouse waste treated in admixture with sewage is illustrated b y results a t Sioux Falls, S. D., where t h e load of packinghouse waste exceeds t h a t of sewage ( 2 ) and the strong mixture has t o be treated by a variety of pretreatment processes in order t o obtain results satisfactory for discharge t o a stream with limited flow. The analysis of packinghouse waste a t Sioux Falls for the year 1948 showed 1053 p.p.m. of biochemical oxygen demand, 1065 p.p.m. of suspended solids, 90 p.p.m. of nitrogen, and 297 p.p.m. of grease. This waste is skimmed for grease removal and adequate evaporators are installed for nitrogen recovery, but the treatment processes at t h e municipal plant include a holding basin t o accumulate excess day flow, preliminary settling, roughing filter, final
Activated Sludge Treatment at Sanitary District of Chicago, West-Southwest Sewage Treatment Works April 1950-March 1951 First full year
Month, 1950-51 April May June July August September October November December January February March Av. a
Flow, Million Gal./Day 909 770 884 826 866 846 791 779 752 825 854 836
828
Activated Sludge Treatment, % ’ 95.3 92.9 94.2 88.2 $4.2 95.0 91.1 98.4 99.7 99.2 98.7 100
...
5-day Biochemical Oxygen Demand, P.P.M. Final Reduction, Raw eff. % 127 11 91.5 90.2 14 144 92.5 11 146 86.0 131 18 89.0 14 129 91.8 11 135 17 88.8 153 163 8 95.0 95.0 164 8 95.0 8 164 94.7 153 8 95.7 6 139 147 11 92.5
Pop. Equiv., Thousandsa 5782 5543 6487 5422 5619 5707 6079 6329 6151 6788 6554 5786 8086
Suspended Solids, P.P.M. Reduction, Raw Final % 181 20 88.3 23 87.7 189 19 90.7 204 23 87.5 184 18 181 80.0 17 179 90.5 27 85.7 188 17 209 92.0 91.7 209 18 17 91.4 198 202 15 92.5 193 15 92.4 193 19 90.0
Air, Cu. Ft./Gal. 0.56 0.72 0.65 0.74 0.68 0.71 0.82 0.75 0.72 0.72 0.61 0.63 0.69
Based on bioohemioal oxygen demand.
March 1952
INDUSTRIAL AND ENGINEERING CHEMISTRY
499
-I;i@uid
Industrial Waste-
filters, and activated sludge, in series, in order t o cope with the mixture of sewage and waste. This plant encountered the difficulty of rising sludge in the final activated sludge settling tanks in its earlier years of operation and could not discharge an effluent sufficiently low in suspended solids. This difficulty was an example of the phenomenon discovered by O'Shaughnessy (11)namely, liberation of free gaseous nitrogen by reduction of excessive nitrates in the final settling tank. These bubbles of ni-
Figure 2.
Southwest Treatment Works of the Sanitary District of Chicago
trogen lifted particles of sludge, to discharge with the final effluent, so that it was found necessary to keep the content of solids very low in the final tank and t o remove the sludge in a period of less than 15 minutee, rather than 30 to 60 minutes common with the conventional type of sludge removal. Installation of the Tow-Bro sludge remover by the Chain Belt Co. of Milwaukee was found to be one of the features of successful rapid removal of sludge, and this installation was largely responsible for the greatly improved results reported by Bradney (I) in 1947, as compared with 1942, as shown in Table 11.
Table 11. Final EfTluent and Activated Sludge at Sioux Falls
Month January February RIaroh April May June July August September October November December Av.
Before and after rapid removal of sludge Suspended Biochemical Oxygen Demand, P.P.31. Solids, P.P.M. __ 1942 1947 1942 1947 20 0 132.0 16 0 164 0 I3 0 97.0 18.0 100.0 67.5 $1 6 11 1 88.0 12 0 11.1 37.7 31.5 13.0 37.0 29.6 13 0 9 2 52.0 13 5 23.0 8.2 37.5 28.5 11 7 37.0 12.0 11 0 24.0 37 0 11 0 37.0 13 0 73.9 11 9 54.0 11.8 13 8 12.0 75.0 89.0 13.8 243.0 16.0 140.0 63.2 13 3 81 3 12 3
-
Various chemical retardants were used to prevent accumulation of too much nitrite, which %-hen reduced to gaseous nitrogen caused the flotation of sludge, but all of the chemicals, including copper sulfate and chlorine, either failed to reduce the nitrite or allowed so little nitrite to be formed that the quality of the effluent was impaired. This discovery by O'Shaughnessy of the detrimental results of too much nitrite or nitrate from high nitrogen wastes has other
500
practical applications and explains why sludge floats in final tanks, even after the liquid is thoroughly treated.
Features of Standard Processes One of the most fundamental processes in preparing packinghouse wastes for treatment is the screening of paunch manure, which unless removed from the waste would cause difficulties in succeeding treatment plant operations. An ordinary revolving drum screen of t h e Xort,h, Dorr, or Green Bay type has been successful for this .ir-ork and still is installed by packinghouses. Occasional washing .ivith hot water, kerosene, or steam removes adhering grease. The paunch manure comprises a small fraction of the biochemical oxygen demand, as much of the semiliquid content of the paunches passes through the screen. However, removal of hay, straw, or fleshings is helpful t o further treatment. Equalizing tanks are in rather common use for receiving excess flow during the working day and discharging the stored waste a t night, These tanks must be emptied daily and not allowed t o accumulate sludge. Various types of grease basins, mechanically skimmed, are in use. One manufacturer removes grease and scum with a rotating ribbon-type skimmer. Other manufacturers utilize a tipping circular trough, cut out t o permit the entrance of the scum. Both. devices are operated in conjunction vith chain-propelled flights which push the scum t o the point of withdrawal. Washable trickling filters have been developed as roughing filters for packinghouse wastes, in order to prevent clogging and to remove grease. This type of filter \vas first developed experiment.ally b y Levine (6) in 1927, and was later extended to the full-scale plants a t Mason City, South St,. Paul, West Fargo, and Sioux Falls. The washable feature has been a successful development in preparation of packinghouse wastes for further treatment. Rates of biochemical oxygen demand application are quite high, even up t o 17,000 pounds per acre foot per day. Consequently, biochemical oxygen demand removals are low. High-rate trickling filters in two stages usually follow the roughing filters, with biochemical oxygen demand loading rates from 6000 to 9000 pounds per acre foot for the first &age and 1000 to 2000 pounds per acre foot' for t.he second stage. The final effluent still may have a biochemical oxygen demand from 50 t o 100 p.p.m. If dilution is insufficient t o stabilize the higher biochemical oxygen demand of the trickling filter effluent, it may be necessary to add a final stage of activated sludge treatment t o reduce this further. If activated sludge treatment is necessary, care must. be taken t,o avoid flotation of sludge in the final settling tank, as has been discussed for the arrangements a t the Sioux Falls plant. Also, moderate recirculation has been found helpful with both primary and secondary filters, with rates not much greater than 300% of the flovc. of settled effluent. Activated sludge treatment is not well adapted t'o packinghouse wastes, unless the wastes are considerably diluted with domestic sewage, as a t Chicago and Madison. The final stage (activat'ed sludge) a t Sioux Falls has given excellent results, as s h o m in Table 111for the year 1948. In general, high-rate filtrat,ion is satisfactory, and free from the difficulties noted with act'ivated sludge operation. Sludge disposal usually follows oi-thodox principles of digestion and sand-bed drying, although mechanical filters are used for undigest,ed sludge a t several plants, including the chemical treatment plant of Oscar Mayer a t Madison, TVis. The disposal of the large volumes of filtered sludge has become rather burdensome, and digestion is under consideration.
Unnsnal or Experimental Processes Packinghouse waste treatment development has had more than it,s share of novel or temporary processes, among which are superchlorination, chemical precipitation, flotation, vibrating screens, and anaerobic digestion. Some of these have passed into the
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 44, No. 3
Liquid Industrial Wastesdiscard while others are very active a t the present time and may develop into practical treatment processes. Superchlorination, or the Halvorson process, was widely publicized for a time some ten years ago. Production of a sludge which could not be sold and a nonstable effluent caused the process t o be abandoned (IS).
Table 111.
3.95
Activated Sludge Resulis, Sioux Falls, 1948
3.78
251
983
569
446
79
14
Chemical precipitation has its advocates, with several plants using ferric salts, others alum or lime, and one plant using zinc chloride and lime for a number of years. The latter was later reduced to lime alone and reports have recommended changing to Figure 3. Anaerobic Digestion of Packinghouse Wastes, sedimentation, filters, and sludge digestion. Another process Hormel Co., Austin, Minn. described by Sanders (14) comprises the use of ferric sulfate and Pilot plant studies sulfuric acid a t p H 4.8, followed by sedimentation and concentration of the sludge for 12 t o 24 hours. The sludge was heated t o digestion a t 90" t o 95" F. for a displacement period of 1 or 2 days. 200' F. and further dewatered t o 15 to 25% solids by vacuum Development of an effective culture was a lengthy process, refiltration. When dried, the sludge contained 35 t o 45% protein quiring from 4 t o 6 months and continuous feeding of raw waste, and was considered salable, but inability t o remove fat from the to prevent overloading the digester contents. The experiments sludge injured the value of the protein as feed, and the process finally indicated that a 24-hour turnover might be expected; has not developed. following this observation a larger pilot plant was built and Vibrating screens are enjoying considerable development in operated. packinghouse waste treatment. Screening media vary from 20 This plant included a holding tank of a 440 cubic foot capacity, to 80 mesh and capacities from 10 t o 50 gallons per square foot per minute. Bradney (1) describes results a t Sioux Falls with a a digester with a capacity of 880 cubic feet, a settling tank of the same size as the holding tank, and a heat exchanger for main70- by 32-mesh screen which removed 73% of the suspended solids taining the desired temperature-about 90" F. from paunch manure a t 100 to 150 gallons per minute. The The rate of feed during the starting period, added to 3000 screen proved practical and was later used for screening activated gallons of sludge containing 2.89y0 solids, was 100 cubic feet per sludge prior to centrifuging. day; an additional 1500 gallons of seeding sludge was added 22 Vibrating screens, either mechanical or electrical, appear t o days after the digester was started, The daily addition of waste have a bright future in packinghouse use. They can be kept clean by occasional washing with hot water or kerosene and are was increased whenever gas production rose, and after 131 days the amount of feed could be increased t o 880 cubic feet per day, rugged. Flotation, using the Bulkley-Dunton Colloidair or the Birdor 1 day's detention time. Gibbs flotation unit, is being investigated widely by various Liquor and sludge was recirculated from the final settling tank packinghouses, with no appreciable publicity concerning results, t o the holding tank a t ratios of 0.63 and 0.49 of the feed, during the efficiencies, operating difficulties, or costs. Apparently it is second and third tests, from December 1, 1950, t o February 1, necessary t o use coagulating chemicals in various amounts up t o 1951. No ratio was determined for the first test, from October 20 grains per gallon, in order t o get a flotable precipitate, and this 17 t o December 1, 1950. Analytical results for the three tests requirement militates against widespread use of these types are as follows: of units. They have the advantage, however, of compactness and of producing a relatively dry Third Test First Teat Second Test sludge, and may be of value in certain restricted ReduoReducReducRaw Eff. tion, % Raw Ed. tion, % Raw Eff. tion, % installations. I n some cases the reported removal of biochemical oxygen demand is rather remarkably Total sohds, p.pm 5622 4402 21 6 5630 4329 23 1 5480 4208 23 2 high. T h e Vacuator furnished by the Dorr Co. is used in treating wastes from other industries and should be applicable for packinghouse wastes. I
Anaerobic Digestion
Volatile solids, p.p.m. 1363 Suspendedsolids, 848 p. .m. Totaynitroeen m o m . 145 Ammonia nlitroge;, p.p.m. 30 Organionitrogen,. 115
390
71 5
1665
354
78.7
2077
384
81 5
223 126
73 6 13 4
807 170
139 137
83 19 8
816 170
170 136
82 8 19 I
105 -22
81.3
...
25 146
119
...
25 145
114 21
84.7
18
87.7
...
5-$$$oohemioal
The mostrecentprocess to assail the packinghouse Oxygen demand 1461 76 94.8 1948 86 95.8 1959 108 94.5 waste problem is anaerobic digestion of the liquid The remarkable reduction in biochemical oxygen demand is wastes, with production of effluents of low biochemical oxygen noteworthy. A great increase in ammonia nitrogen is noted, but demand-reductions up to 95% with one day's digestion. There this is not included in the 5-day biochemical oxygen demand. seems t o be a flurry of interest in this procedure because of reSludge rose to the surface in the final tank and escaped with the markable results reported by the George A. Hormel Co. ( 3 )a t Auseffluent. A mechanical stirrer was therefore installed in the tin, Minn. Figure 3 gives a view of this process a t the Hormel Co. final tank; operated at.5.4 r.p.m., i t appeared t o overcome this The first tests made a t Austin were laboratory investigations, difficulty. The sludge had a tarry odor with a slight odor of using oil drums for digestion vessels. These studies indicated a hydrogen sulfide. removal of 90 t o 95% of the biochemical oxygen demand by
March 1952
INDUSTRIAL AND ENGINEERING CHEMISTRY
50 1
Liquid Indusirial Wastes Visitors have reported the effluent to be free from odor, an observation which needs confirmation over a continuous period of operation. Further checking is necessary with reference t o the actual net digestion time, the success of stirring equipment in the final t a n k d e s i g n e d for quiescent settling-and the lack of odors. However, the results are excellent from the standpoint of biochemical oxygen demand reduction and the production of gas. Estimates attempted t o compare costs u i t h similar costs for activated sludge treatment, much t o the advantage of the anaerobic process. The plant is continuing iri operation, to investigate some of t h e factors mentioned. Hovever, the 1-day detention time is quite attractive and it is hoped that the short time of digestion will accomplish the high biochemical oxygen demand rcductions and produce an odorless effluent. More information is needed on the efficiency of clarification of the final effluent. The plant so impressed the sanitary engineer from Auckland, Kew Zealand, who visited it in 1949, that he has checked the laboratory work and is prepared to operate a pilot plant to confirm the larger scale work. These tests are now under n a y and the results will be observed with interest This experiment has been described in detail because it is something new in the disposal of packinghouse wastes, although it has been applied experimentally t o other wastes and used in two or three instances for full-scale treatment of yeast Jyastes. However, continuous operation of these plants has developed some troubles, particularly in the maintenance of an effective uniform bacterial flora, and in plant control to avoid overloading the digestive capacity. G e n e r a l Survey of Problem
,
The treatment of packinghouse waste is still a largely unsolved problem, particularly in the smaller packinghouses or slaughterhouses, Studies made in Pennsylvania point t o chemical precipitation as the most practical procedure for small plants, using large amounts of chlorinated lime plus filter alum. Whether this procedure has been followed by small packinghouses in Pennsylvania is not known, but precipitation for such wastes has had long study and not much practical application. For large plants, as stated heretofore, chemical treatment has had many unsuccessful trials on a large scale, and this type of treatment is not favored by packers nor approved universally by authorities. Activated sludge has had several trials on a small scale (Decker, Kuhner) but is not favored unless it is combined with other biological processes, notably high-rate filters, for practical attack on the packinghouse waste problem. However, the treatment of
.
all Chicago packinghouse wastes a t the southwest activated sludge plant of the Sanitary District of Chicago is the world’s largest example of successful treatment and the highest percentage purification of any packinghouse wastes, but the success of this plant is due t o the ample dilution of the packinghouse wastes by a large excess of domestic sen.age. More concentrated mixtures have encountered difficulties in application of the activated sludge process. Combinations of high- and low-rate filters, plus activated sludge in the flow cycle, seem t o offer the best and most practical method for disposal of these wastes. There are still larger unsolved prohleiiis in packinghouse waste treatment on a large scale, particularly \?Tastes in the Sioux City, Omaha, St. Joseph, Kansas City, and St. Louis areas, where this industry has shown much grov-th in recent years, taking away much of the supremacy once held by the industry in Chicago. Also the smaller packinghouse and slaughterhouse waste problem is scattered well over tlie T’nited States, thus offering a challenge t o the sanitary engineer conceiiiing the abatement of this widespread pollution. It is in this field that chemical treatment appears most reasonable, thus maintaining the need for further study of this problem by the chemist. Possibly there is still opportunity for the packinghouse chemist t o add his contribution t o the development of a cheap, simple, and economical process for the treatment of u astes discharged by his industry.
Literature C i t e d Bradney, L., Selson, TT‘., and Bragstad, R. E., Sewage and Ind. Wastes, 22, 807 (1950). Bragstad, R. E., and Bradney, L., Sewage W o r k s J . , 9, 959 (1927). Fullen. W. J.. Am. Meat Inst.. Proc. Cons. Research (1950): Public Works, 82, 44 (July 1951). Hansen, P., and Hill, I
