The solubilities of carbon dioxide in soybean oil, olive oil and linseed oil, and those of nitrogen, hydrogen and oxygen in soybean oil were measured at pressures ranging from 0.2 to about 1 atm. of gases and at temperatures between 30 and 70°C. In most cases, the solubility of gases in oils follows Henry's Law, that is, the solubility of gas increases linearly as the pressure of gas increases. Generally, the solubility of gas in oils decreases as the temperature increases, except for the case of oxygen. The solubility of oxygen at 70°C is as large as about twice compared with that of 30 or 50°C. This might be due to the reason that the oil is oxidized slightly by oxygen at higher temperatures.Another expression of the solubility of gas, the Bunsen's absorption coefficient may be calculated as the Henry's Law rules in every case. These coefficients are shown in Table-2. The temperature dependence of logarithms of the Bunsen's absorption coefficients enables the calculation of the differential heats of solution of gases. These values are listed in Table-3.Although the values of the differential heat of solution of carbon dioxide in fatty oils do not vary so greatly with the kind of fatty oils, it seems that the value depends either on the molar volume or on the iodine value of the fatty oils. The theoretical considerations based on the quasi-lattice model of the solution of gas in oil suggests that the effect of variation of the molar volume of fatty oils on the differential heat of solution of gas is negligibly small, and the effect of the iodine value is predominant.In spite of the low solubility of hydrogen in fatty oil, the differential heat of solution is larger than that of other gases.The viscosities of soybean oil under vaccum or at saturation with gas at a definte pressures were also measured. When carbon dioxide is dissolved in oil, the viscosity, or the flow out time of oil, decreases lineally as the amount of dissolved gas increases. But, when nitrogen is dessolved, the flow time of oil is somewhat longer than that under vaccum, slightly exceeding experimental errors. This might be partly due to the larger differential heat of solution of nitrogen than that of carbon dioxide.