Correct identification is one of the major problems facing those working with milk flavors. This committee sought first to develop a practical classification for off-flavors. Next, methods for producing some of the more common off-flavors were studied. Finally, a bibliography covering the literature from 1950 to 1976 on off-flavors was prepared. This report is intended to aid in training both research and quality control personnel in the hope that more uniform terminology will result.

Ground beef from two locker plants using beef trimmings, and two retail stores, using 4.5-kg chub packs of coarse-ground centrally packaged beef as a source of meat, was evaluated microbiologically on the first Monday and Wednesday of every other month for 14 months. Only slight variations in coliform, fecal coliform, coagulase-positive staphylococci, and aerobic plate counts by plant or retail store were noted. Both retail stores using chub packs showed average total aerobic plate counts of 3.5 × 106/g while one locker plant averaged 1.7 × 106/g and the other 7.7 × 106/g. Source of meat for ground beef and holding time of the meat contributed most to differences in microbiological quality. Overall, ground meat packaged on Monday had higher (P<05) average total aerobic plate counts (5.7 × 106/g) than that packaged on Wednesday (2.5 × 106/g) when more recent beef shipments were available. Coliform, fecal coliform, and coagulase-positive staphylococcus counts followed the same patterns as noted for total aerobic plate counts. Total aerobic plate counts in beef increased slightly in the locker plants where game was processed in the fall, in addition to beef. However, game meat did not cause Salmonella contamination of ground beef in plants where both game and beef were processed. Three Salmonella positive samples out of the 112 total ground beef samples and the 112 swabs from used grinders were isolated and serotyped.

Addition of 100 ppm of malathion or trichlorfon had little effect on lactic acid production in reconstituted non-fat dry milk by Streptococus diacetilactis or Streptococus thermophilus. Variation in cell size of Streptococus cremoris 3, Streptococcus lactis W 8 and several strains of S. thermophilus occurred following repeated transfers in litmus milk to which 100 ppm of malathion or 500 ppm trichlorfon had been added. The Lactobacillus bulgaricus A culture, after 15 serial transfers in litmus milk containing 500 ppm of trichlorfon, exhibited increased frequency of chain formation and cell structure 30 to 50 μ long with few visible cross walls. A similar effect was observed with L. bulgaricus A and L. bulgaricus GA after repeated transfers in litmus milk which contained 100 ppm malathion. One of five strains of L. casei studied showed similar changes in morphology after repeated transfers in litmus milk which contained 100 ppm malathion. Adding 500 ppm trichlorfon produced a lesser effect. Changes in morphology of this strain of L. casei were not permanent since cells resumed their usual appearance following three to five transfers in plain litmus milk.

The effect of 0.05 and 0.10% potassium sorbate on acid and acetaldehyde production and growth of three commercial yogurt cultures was studied. Either amount of potassium sorbate retarded growth of both Lactobacillus bulgaricus and Streptococcus thermophilus strains in the commercial cultures and also decreased the rate of acid production of cultures incubated at 45 C. Acetaldehyde in these cultures decreased during storage and decreased more in samples containing potassium sorbate. In all trials 0.1% potassium sorbate had a greater effect than 0.05% on the above characteristics of the yogurt cultures.

Fifty randomly selected consumer house-holds evaluated the flavor of homogenized, standardized whole milk pasteurized for 17sec at 72.2, 75.6, 78.9, 82.2, or 85.6C. As a group, the consumers preferred the flavor of milk pasteurized at 78.9C over that processed at the lower or higher temperatures. The tendency to note differences in flavor of milk processed at various temperatures appeared to increase with increasing age of the respondent. Those over 19 years of age exhibited the strongest preference for milk pasteurized at 78.9C versus the other temperatures used. Children under 13 years of age were least discriminating in evaluating cooked flavor of milk. The intensity of the cooked flavor in milk stored at 1.7-4.4C decreased within a few days after processing.

Summary Methods of preventing mold growth on the surface of rindless Swiss-type cheese, cured in an unsealed film wrap, were investigated. Painting the cheese with a 15 to 20% aqueous solution of potassium sorbate, 1 oz/20-lb. block, resulted in very slight mold growth and small cleaning losses during curing periods up to 9.5 mo. Holding for longer periods allowed slightly more extensive mold growth and cleaning losses. The use of 10% solutions of this compound was less effective. A 1 or 2% solution of sorbic acid in propylene glycol failed to inhibit mold growth sufficiently to be acceptable. The use of oils or greases was unsatisfactory. Mold growth occurred on all blocks treated with Confecto grease, a candy-slab releasing agent. Mineral oil was inconsistent in its ability to protect the cheese surface against mold growth; at times its use resulted in extensive cleaning losses. The use of Saran film dusted with 2.0 mg. of sorbic acid per 100 sq. in. consistently failed to inhibit mold growth. Cryovac film was preferred slightly over Saran for wrapping the cheese because the former, after shrinking, maintained a closer contact with the cheese surface.

Rindless Swiss-type cheese was made in 20-lb. blocks by modifying the procedure previously used in making the cheese as 5-lb. wheels. Manufacture of the cheese in 20-lb. blocks did influence eye formation. Less Propioni-bacterium shermanii culture was used and the eyes formed were larger and more characteristic of Swiss cheese as compared to those found in the 5-lb. wheel.

The production of a coagulum in milk with compounds which hydrolyze slowly to produce acids has been investigated. Of the compounds tested (including anhydrides, esters, lactones, and lactides) D-glucono-delta-lactone and meso-lactide produced a uniform coagulum suitable for the manufacture of Cottage Cheese. Meso-lactide and d-glucono-delta-lactone at weights equal to 8.8 and 12.3% of the SNF, respectively, gave a pH of about 4.6 after complete hydrolysis. It was necessary to use about 0.02% calcium chloride with the lactide to obtain normal expression of the whey during cooking. The coagulation time varied with the temperature of the milk. Lactide required 2 hr. at 25° C. and 47min. at 37.5° C. Glucono-lactone required about 15 hr. at 20° C. and 3.5 hr. at 40° C. At higher coagulation temperatures the curd was firm enough to cut at a higher pH than is normally used. The addition of rennet made it possible to cut the curd at pH's higher than normal and thus reduced the setting time. A coagulation temperature above 37° C. caused excessive matting of the curd. The Cottage Cheese produced by this method had a bland flavor and was very similar in appearance to that produced by starter culture.