The similarity of the pattern of methyl ketones obtained from the steam distillates at atmospheric pressure of cheeses made under controlled aseptic conditions, despite the wide differences in bacterial flora, led to the suspicion that the methyl ketones were being formed as artifacts. This was confirmed by steam distilling cheeses from 1 day to 13 months old when the qualitative patterns and quantitative amounts differed little with the age of the cheese. Evidence was produced to show that the greater part of these methyl ketones must be formed during the heat treatment of milk fat. The maximum quantities of methyl ketones obtainable from cheese and from milk fat, determined by exhaustive steam distillation at atmospheric pressure, averaged from 14 p.p.m. for 2-undecanone to 46 p.p.m. for 2-pentadecanone. Some artifact formation of methyl ketones also occurred, although to a greatly reduced extent, when dairy products containing milk fat were steam distilled under reduced pressure at 40°. As methyl ketones in low concentrations could be extracted from mature cheese at room temperature by solvents or by flushing cheese suspension with nitrogen, milk fat appears to contain precursors which break down to methyl ketones slowly during cheese ripening, this breakdown being accelerated at higher temperatures. Two possible modes of formation of the methyl ketones with an odd number of carbon atoms, found in limiting quantity in the steam distillates of Cheddar cheese, were considered:- (a) From precursors, probably β-keto acids, bound in milk fat. (b) The β-oxidation of free fatty acids, formed by the lipolysis of milk triglycerides, and subsequent decarboxylation of the β-keto acids formed. The use of radioactive milk fat from a lactating cow which had been injected intravenously with carboxy-14C acetate allowed a direct comparison to be made between the labelling patterns of fatty acids and the corresponding methyl ketones from the same milk source. The similarity of the labelling patterns suggests that the C6 to C16 β-keto acids and the corresponding fatty acids have a common precursor (or that one is the precursor of the other) and are together incorporated into the triglycerides. Only butyric acid of the C4 to C16 fatty acids in all 3 milkings had a higher specific activity than the corresponding methyl ketone. This suggests that the acetone found in steam distillates of milk fat is formed from a compound (probably D-β-hydroxybutyrate), derived almost entirely from a precursor other than acetate. The finding that the saturated C18 acid in all 3 samples of radioactive milk fat had an extremely low activity was in agreement with the fact that no C17 methyl ketone was detected in any of the numerous steam distillates from milk fat or cheese. This supports the generally accepted view that, in the biosynthesis of milk fat, the fatty acids up to c16 acid are synthesised from an acetate pool, whereas C18 acids and above are obtained from the blood triglycerides. The possibility that methyl ketones were being formed in Cheddar cheese from the β-oxidation of free fatty acids, as well as from a slow breakdown of bound β-keto acids in milk fat, was shown to be improbable. Triglycerides of acids (undecanoic, nonanoic, and heptanoic), which occur normally only in traces in milk fat, were synthesised and incorporated in Cheddar cheeses. On steam distillation of these cheeses when mature, no methyl ketones corresponding to the acids in the added triglycerides were obtained, although the normal range of methyl ketones with an odd number of carbon atoms was found in the distillates. A detailed study of the metabolism of fatty acids and synthetic triglycerides by spores and mycelium of Penicillium roqueforti was undertaken, this fungus being chosen as a general representative of lipolytic organisms that might be of importance in producing Cheddar flavour. The effect of the growth medium, pH of solution, concentration of acid and inorganic ions on both oxygen uptake and methyl ketone formation was determined. The rate of methyl ketone formation suggested the synthesis of adaptive β-keto acid decarboxylases after a lag period of 1 to 2 hours. An hypothesis based upon the possible toxicity of the C6 to C12 β-keto acids can explain a number of the experimental results: (1) Only one methyl ketone was formed (2) The most toxic acids were those which gave the least amount of methyl ketone (3) Concentrations of the C8 to C12 fatty acids that markedly inhibited the respiration of mycelium were nevertheless oxidised to considerable quantities of the corresponding methyl ketone. The relationship between the toxic action of fatty acids and chain length was found to be dependent upon pH. The chain length of the most toxic acid increased with pH, being C10 acid at pH 2.5 and C12 at pH's 5.2 and 6.0. At pH 6.8 none of the acids from C4 to C18 inhibited oxygen uptake but C14 acid was the most toxic acid at pH 8.0. There appeared to be no sharp dividing line between the metabolic activity, with respect to fatty acids, of spores and mycelium. Mycelium oxidised fatty acids rapidly giving varying amounts of methyl ketone but considerably more CO2 than spores. In general spores formed higher amounts of methyl ketones than mycelium but showed also a slight but definite ability to form CO2 from octanoic acid. Evidence for a β-oxidation mechanism in the fungal metabolism of fatty acids was obtained by the use of 1-C14 and 2-C14 octanoic acids. A relatively slow movement of intermediates through the T.C.A. cycle was also indicated. High concentrations of triglycerides were oxidised slowly by spores to methyl ketones when equivalent concentrations of the free acid (up to 66 μmoles/ml) inhibited methyl ketone formation. It seems probable that the very slow rate of formation of methyl ketones is due to the inhibitory effect of the ketones themselves on the lipases. Methyl carbinols were not detected in significant amounts as products of the metabolism of methyl ketones, evidence being obtained on the contrary that the carbinols were possibly precursors of the ketones during the oxidation of fatty acids by spores. Cell free extracts obtained from mycelium were able to oxidise low concentrations of octanoic acid (0.5 μ-moles/ml or less) after a lag of up to 3 hours. The supplementation of the extracts with several coenzymes, known to be associated with fatty acid oxidation, or with T.C.A. cycle intermediates were unable, however, to decrease the lag before oxidation started. No methyl ketones were detected after the oxidation of fatty acids by cell-free extracts but were formed when cell debris from the Hughes Press was used. This suggests that the β-keto acid decarboxylases were tightly bound to the cell walls.