Pilot studies in 31 rats to detect gases in the brain during anesthesia demonstrated the presence of gas microbubbles in electron photomicrographs of brains taken from anesthetized and unanesthetized control animals. Gas microbubbles were numerous in specimens preserved with near-isotonic xatives, but nearly absent in brains xed with a substantially hypertonic xative solution. Hypertonic xation appears to deplete extracellular and intracellular uid to an extent that tissue entrapped bubbles collapse and are ushed away. Using near-isotonic xative solutions, it was shown that gas microbubbles may play important roles in Toxicology (CCl4-induced hepatotoxicity), Physiology (oxygen transport), and Pharmacology (inhalation anesthesia). Excessively hypertonic tissue xative solutions are unsuitable for the histological study of tissue gases.Initially, the purpose of this study was merely to determine whether microbubbles of gas could be found in the brains of rats anesthetized with gas and volatile anesthetics. The premise was that anesthetic gas bubbles might interfere with normal brain function by simple physical mechanisms. A pilot study, conducted on rats anesthetized with methoxy urane, halothane, ethyl ether, or nitrous oxide, con rmed the presence of gas microbubbles in brains xed in 2% glutaraldehyde. Surprised to nd gas bubbles in the brain of an unanesthetized control rat, the author reviewed earlier literature to determine which xatives were optimum for rat brains. In 1969, Sumi found that mitochondria were swollen and vacuolated and the neuroprocesses were swollen in 24-h-old ether-anesthetized rats perfused with 50 to 100 mL of 1.2% glutaraldehyde buffered to 280 mOsm/kg (isotonic). These ndings were attributed to poor tissue xation, since they disappeared in rats perfused with 50 to 100 mL of 4% glutaraldehyde buffered to 710 mOsm/kg (»2.5 £