1. A heavy ingestion of frothy emulsified fat by rats and bullfrogs does not increase susceptibility to bubble formation when the animals are decompressed 2 to 72 hours later. This indicates that gaseous films (bubble nuclei) initially present do not pass across the intestinal wall with the digested fat, and also that high fat content per se in the lymph and blood does not increase susceptibility to bubble formation. 2. Liquid caprylic acid injected into veins of bullfrogs crystallizes when the frogs are cooled. The crystallization causes bubbles to form without muscular activity on subsequent decompression. Cooling normal bullfrogs to 1-2 degrees C. fails, however, to crystallize any substances occurring naturally in the animals that might act in a similar manner. 3. When bullfrogs are cooled (e.g. to -5 degrees to -10 degrees C.) until ice forms in the blood vessels, and are then warmed and decompressed, bubbles form in the absence of exercise. Crystallization of water in the body thus forms nuclei or even small bubbles that persist. If only one foot is frozen, bubbles originate in the frozen foot. In some cases visible bubbles were observed in thawed feet at sea level (i.e. without decompression). When frog's blood is partly frozen in test tubes or in tied off sections of veins, bubbles will appear on decompression in the absence of mechanical agitation. The practical relation of this phenomenon to flight at high altitude should not be overlooked. 4. Fracturing a leg bone (tibia or femur) in a frog induces bubble formation on subsequent decompression. Bubble nuclei, which persist for (1/2) to 1 hour, are probably formed as a result of the intense mechanical disturbance when the bone snaps. Fracturing of bone is considerably more effective than crushing muscles for producing bubbles in frogs.