We previously described inhaled compounds with properties that deviate markedly from those predicted by the Meyer-Overton hypothesis. Specifically, their anesthetic potencies do not correlate with their solubilities in olive oil [1-3]. These "deviant" compounds include nonimmobilizers (previously called nonanesthetics) and transitional compounds. Nonimmobilizers do not produce anesthesia when given alone, nor do they lower the anesthetic requirement (minimum alveolar anesthetic concentration [MAC]) for conventional inhaled anesthetics. Transitional compounds also do not produce anesthesia when given alone. However, they add to the effect of anesthetics given concurrently, and these data can be used to estimate the partial pressure of the transitional compound that would produce anesthesia. These concentrations are >3.5-fold greater than those predicted by the Meyer-Overton hypothesis . That is, transitional compounds are far less potent as immobilizers than predicted from their lipophilicity.
Nonimmobilizers and transitional compounds differ from other inhaled compounds in ways beyond their absent or lesser capacity to produce immobility. They have a lower polarity, as indicated by saline/gas partition coefficients generally less than those found with conventional inhaled anesthetics . When administered alone, nonimmobilizers and transitional compounds produce twitching and convulsions, usually at concentrations that would be expected to produce anesthesia by the Meyer-Overton hypothesis . That is, the product of the oil/gas partition coefficient of the anesthetic multiplied by the alveolar partial pressure of anesthetic required to produce convulsions usually seems to equal a value between 0.5 and 6 atm, a range of values that encompasses those found for conventional and most experimental anesthetics . Confirmation that such a relationship existed would imply that a critical number of molecules of nonimmobilizer or transitional compounds must reside at a certain nonpolar site to produce convulsions. To confirm or refute these initial impressions, we examined the relationship among lipophilicity, hydrophilicity, and convulsant activity for various gaseous and volatile compounds, including nonhalogenated [5-7] and halogenated [1,2] alkanes, nonhalogenated cycloalkanes , halogenated ethers [9-11] [including cis-trans isomers ], halogenated aromatic compounds , and noble gases .
Much of the data presented in the present report are drawn from published results for particular compound series (e.g., n-haloalkanes). Some data are from reports submitted for publication or that are in press. Several values are of unpublished data. In some cases, additional data have been gathered since publication; thus, the values may slightly differ from published values. Experiments were sometimes limited because of the availability or high costs of compounds. In some cases, only one or two rats were studied. In some cases, the convulsant 50% effective dose (ED (50)) was not obtained by bracketing the anesthetic concentrations immediately surrounding that which caused convulsions: only one concentration (that producing convulsions at an initial step or as part of a continuously ascending concentration) was given. We assumed that the alveolar concentration equaled the inspired concentration of poorly soluble compounds. For more soluble compounds, we converted inspired values to alveolar values by multiplying the former by the arterial to inspired partial pressure ratios (separately determined).
After approval from the University of California, San Francisco committee on animal research, we studied the convulsant properties of various inhaled anesthetics in adult specific pathogen-free Sprague-Dawley male rats (300-500 g). Each animal was caged individually and had continuous access to standard rat chow and tap water before study.
Most compounds were purchased from PCR Incorporated (Gainesville, FL) or Flura Corporation (Newport, TN) and had purities >97% and usually >99%. Ross C. Terrell synthesized and provided four convulsant ethers [CF3 CH2 OCF2 Cl, CF3 CFCIOCF2 Cl, CF3 CCI2 OCF2 Cl, CFCl2 CF2 OCF2 Cl], while one of us (TH) synthesized and provided one convulsant ether [CF2 ClCF2 OCF2 CI]. TH also provided enriched forms of the cis (91% cis to 9% trans) and trans (8% cis to 92% trans) isomers of 1,2-dichloroperfluorocyclobutane. The 3M Corporation (St. Paul, MN) donated the mixture of the structural isomers, CF3(CF2)3 OCH3 and (CF3)2 CFCF2 OCH3 (approximately equal quantities of each in the mixture). One compound (CF3 CHFCHFCF2 CF (2) CF3) was obtained from Van Waters and Rogers (Los Angeles, CA).
Each rat was placed in an individual plastic cylinder with a holed rubber stopper through which passed the tail and a rectal temperature probe. Tape secured the tail to a plastic extension of the cylinder. Thus, the rat could move within the cylinder, but the secured tail prevented it from exiting through the "head" end of the cylinder.
Each rat, enclosed in its individual plastic cylinder, was placed into a closed system that allowed for administration and sampling of the inhaled anesthetic, administration of oxygen (typically maintained at 0.5-1.0 atm), removal of carbon dioxide (typically <0.01 atm) with soda lime, circulation of gases through the closed system with a fan, and monitoring of rectal temperatures. One compound (CF3 CHFCHFCF2 CF2 CF3) broke down in the presence of soda lime, and we omitted the absorbent when testing this compound. For this compound, removal of carbon dioxide was accomplished through the use of a high inflow rate. One closed system was a 3.4-L clear plastic cylinder (permitting full visualization of each rat) with metal ends that contained O-rings and clamps to provide a seal . This system held two rats and was used when a limited amount of the experimental compound was available and/or when it was required to administer hyperbaric pressures of up to 10 atm. A second closed system enabled testing at very high pressures (e.g., for measurement of the convulsive potential of helium or neon or perfluorinated n-alkanes.) This system used a stainless steel chamber with a viewing window at one end . The chamber allowed application of pressures as high as 300 atm. A third closed system  allowed concurrent testing of four rats and was used when there was less concern about the availability and/or expense of the anesthetic and/or when hyperbaric pressures were not required.
Because the excitation associated with administration of experimental compounds often increased body temperature, we applied bags of ice to the individual plastic cylinders containing the rats before placement in the closed system and to the outside of the closed chambers to maintain rectal temperatures <39[degree sign]C. The chambers were flushed with oxygen for 10 min (chamber concentration >98% O2) before closing the system and introducing test compounds. Ports into each system permitted the introduction of gaseous or liquid anesthetics and sampling of chamber gases.
For most experiments, the test compound was initially added to the chamber to produce a partial pressure that was approximately 25%-50% of the predicted MAC value (which was calculated by dividing 2 atm by the oil/gas partition coefficient of the compound). The animals were observed for 20 min after the compound was introduced, and a gas sample from the chamber was taken for analysis, usually by gas chromatography. For a few gases [e.g., helium ], we determined the test compound concentration by measuring the concentration of oxygen by using a Pauling meter and subtracting that concentration from 100%. After the 20-min observation period, additional test compound was added (typically an amount that increased the previous partial pressure by 30%-50%), the rats were observed for an additional 20 min, and a gas sample was again taken from the chamber for analysis. This process was repeated until the rat exhibited a clonic convulsion, defined by strong contractions of muscles over most of the body, with the muscles alternately contracting and relaxing, the eyes tightly closed, and the mouth widely opened with teeth bared. We did not quantitate more subtle characteristics of excitatory phenomena (e.g., tremors or twitching). The convulsant effect of the experimental compound was estimated for each rat by averaging the inspired partial pressures at which the convulsion occurred and the immediately previous partial pressure measured after a 20-min equilibration period during which there was no convulsion. Mean +/- SD values for these convulsive ED50 values are given in Table 1, Table 2, Table 3, Table 4, and Table 5.
The testing of convulsant ED50 values of certain cycloalkanes and aromatic compounds required the intraperitoneal administration of these compounds . Dose-response relationships were established by noting the effect of a given intraperitoneal injection. Arterial blood was taken for analysis of the test compound using gas chromatography. The values obtained were converted to partial pressures based on known blood/gas partition coefficients and concentrations of test compound in the gas phase equilibrated with the extracted blood.
Saline/gas and olive oil/gas partition coefficients for the compounds were taken from published articles or were determined using standard techniques .
Correlation coefficients were determined for the relationships between the ED50 for convulsions and the oil/gas and saline/gas partition coefficients. Values are reported as means +/- SD.
Data for convulsant ED50 values and for solubilities were available for 3 n-alkanes and 18 n-haloalkanes (Table 1), 3 cycloalkanes and 3 halocycloalkanes, including the cis and trans forms of c(CClFCClFCF2 CF2) (Table 2), 13 halogenated ethers (Table 3), 3 halogenated aromatic compounds (Table 4), and 2 noble gases, He and Ne (Table 5). The most potent compounds (CClF2[CClF] (2) CClF2, CClF2 CClFCCIFCClF2, CF3 CHFCHFCF2 CF2 CF3, flurothyl [CF3 CH2 OCH2 CF3], cyclooctane, and perfluorotoluene) each caused convulsions at <0.005 atm, whereas the least potent compounds (He and Ne) required approximately 90 atm, a >20,000-fold range of partial pressures. CF (3) CHFCHFCF2 CF2 CF3 was a nonimmobilizer that, at 3 times the convulsant ED50, did not decrease or increase the partial pressure of desflurane required to produce MAC.
Each test compound had a measurable solubility in saline and olive oil. Excluding nine compounds (see Discussion), a correlation was found (r2 = 0.99) between the solubility of a compound in olive oil and its convulsant potency (Figure 1). For 36 of the compounds (those following the correlation described in the preceding sentence), the product of the convulsive ED50 times the oil/gas partition coefficient was 2.80 +/- 1.24 atm. This value differed significantly (P < 0.0001) from those of four compounds below the line of correlation (0.134 +/- 0.062 atm) and five above the line (10.0 +/- 1.7 atm). Convulsant potency did not correlate (r2 = 0.003) with the solubility of a compound in saline (Figure 2).
The Meyer-Overton hypothesis proposes that the potency of inhaled anesthetics directly correlates with their lipid solubility such that the product of anesthetic requirement (MAC) x the oil/gas partition coefficient is nearly constant with a value of approximately 2 atm . The implication of this hypothesis is that anesthesia results from an action at some nonpolar phase (e.g., a lipid site or the hydrophobic interior of a protein). The discovery of certain lipid-soluble inhaled anesthetics (the nonimmobilizers) devoid of at least one anesthetic property-the capacity to produce immobility-or of compounds far less potent as immobilizers than predicted from their lipophilicity (transitional compounds) undermines this hypothesis [1-3]. Although the anesthetic potencies of nonimmobilizers and transitional compounds do not follow the hypothesis, it seems that the convulsant potency for most of these compounds do follow the hypothesis: the ED50 values for convulsions correlate with the oil/gas partition coefficient (Figure 1). This correlation also may be seen from the relative constancy of the product of the ED50 and the oil/gas partition coefficient in Table 1, Table 2, Table 3, Table 4, and Table 5 (Figure 3). For 36 of 45 compounds, the value for the product lies between 0.5 and 6 atm (i.e., within a factor of 3 from the 2-atm value for the Meyer-Overton constant); for 29 of 45 compounds, the value for the product lies between 1.0 and 4 atm.
What are we to make of the exceptions? One suggestion comes from data for the effect of some nonimmobilizers on GABAA receptors. Two nonimmobilizers, c([CFCl]2 [CF2]2) and CF3(CClF)2 CF3, do not alter the effect of GABA on the GABAA receptor [14-16]. The products of the ED (50) and the oil/gas partition coefficient of these compounds are 2.4 and 2.3, respectively. However, flurothyl (CF3 CH2 OCH2 CF3) is a noncompetitive antagonist (or negative allosteric modulator) of GABA. It acts to decrease the maximal effect of GABA; thus, its effect is not surmountable by high concentrations of GABA (Matthew Krasowski, personal communication; [17,18]). The product of the ED50 and the oil/gas partition coefficient of flurothyl is 0.073, 1/30th the values for c([CFCl]2 [CF2]2 and CF3(CClF)2 CF3. This suggests that compounds that are more potent convulsants than would be predicted from their lipophilicity (i.e., CF3 CHFCHFCF2 CF2 CF3,CF3 CH2 OCH2 CF3 [flurothyl], perfluorotoluene, and c[(CFCF3)2 (CF2)2] may primarily produce convulsions by their capacity to block the action of GABA. Conversely, compounds that are less potent convulsants than would be predicted from their lipophilicity (i.e., c[CH2]6, c[CH2]7, c[CH2] (8), CH3[CH2]6 CH3, and CH3[CH2]7 CH3) may be less potent because, like conventional inhaled anesthetics [19-23] and alkanols [24,25] they may be able to enhance the effect of GABA. This would counteract (but incompletely) their inherent capacity to produce convulsions by an action at a nonpolar site.
Thus, the most economical view of how nonimmobilizers and transitional compounds produce convulsions postulates two mechanisms. The first is an inherent, unexplained mechanism that involves a nonpolar phase. Unstated is the parallel fact that these compounds usually have a low affinity for a polar phase (see the saline/gas partition coefficients in Table 1, Table 2, Table 3, Table 4, and Table 5), unlike anesthetic compounds, which generally have an affinity for both polar and nonpolar phases. The second is a variable capacity to affect GABAA receptors. The hypothesis postulates that most nonimmobilizers and transitional compounds do not affect GABAA receptors. However, some (e.g., flurothyl) oppose the effect of GABA, thus decreasing the natural protection against convulsions afforded by the inhibitory stream of impulses conveyed by GABAA receptors. This leaves the first mechanism unopposed in its tendency to cause convulsions, and a convulsion ensues at a far lower concentration than would be predicted from the compound's lipophilicity. Some compounds are postulated to have a converse effect on GABAA receptors (e.g., c[CH2]6). Like conventional inhaled anesthetics, they enhance the effect of GABA. However, this effect is insufficient to provide the inhibitory input that prevents convulsions. Nonetheless, it is sufficient to partially oppose the convulsions initiated by the first mechanism, and the result is that a concentration higher than would be predicted by the compound's lipophilicity is required to produce convulsions. One nice result of this dual hypothesis is that it is immediately amenable to testing.
At least one clinically used anesthetic, enflurane (CHF2 OCF2 CHFCl), predisposes subjects to convulsions [26,27]. However, convulsions are rarely seen with this anesthetic, and they usually must be provoked by either repetitive auditory stimulation  and/or by application of deeper levels of anesthesia and hypocapnia [29,30]. Enflurane enhances the effect of GABA on GABAA receptors [22,23,31], which may explain the rarity of convulsions with enflurane. That is, enflurane may represent an example of a compound in which the enhancement of the effect of GABA opposes the capacity of the effect on a nonpolar site to evoke convulsive activity.
Some compounds tested in the present study were known from previous studies to be convulsants. Compounds other than those tested in the present study also have convulsant properties (Table 6 and Table 7). Although the definition of a convulsion in these earlier experiments may differ from that applied in the present experiments, such differences are not likely to be crucial. However, earlier studies differ in experimental design from that used in the present experiments (e.g., definition of the convulsant ED50 by administration of compounds over a range of concentrations, accurate measurement of inspired-arterial partial pressures differences, and monitoring and control of body temperature). For both the present (Table 1, Table 2, Table 3, Table 4, and Table 5) and earlier compounds (Table 6 and Table 7), molecular structure would not seem to predict convulsant activity. Many convulsants tend to have extremely low saline/gas partition coefficients, but flurothyl and perfluorotoluene are exceptions to this trend.
Our studies have several potential limitations. Despite usual purities of >97% (and often >99%), it is possible that a small contaminant in certain compounds produced the convulsant effects. Similarly, we cannot exclude the possibility that convulsant effects of some compounds may have resulted from a metabolite. We assumed that the behavioral manifestations of a clonic convulsion arose from increased repetitive firing of central nervous system neurons, but we did not measure cerebral electrical activity. We do not believe that the convulsions resulted from hypoxia, because the animals usually breathed 100% oxygen and did not appear cyanotic before convulsive activity.
In Table 1, Table 3, and Table 4, we include values for eight compounds that did not produce convulsions in all test animals. We assumed that the values for these animals only slightly underestimated the true ED50. This assumption follows from the observation that the dose-response curves (for dose versus the cumulative incidence of convulsions) for other convulsants are steep, as indicated by the small coefficient of variation (standard deviation divided by the mean value) for those compounds (see Table 1, Table 2, Table 3, Table 4, and Table 5). Thus, the concentration producing a convulsion in only one animal would only be slightly lower than the true ED50. Omitting the few compounds in which this assumption applied would not change our present results or conclusions.
Some of the data for the present report were gathered in studies preliminary to determination of anesthetic potencies. In those studies, the focus was not on the definition of the ED50 for convulsions. Thus, the resulting tabulation of data for the present study includes compounds for which only one or two rats were studied and for which data were obtained during the application of an ascending concentration of test compound rather than steady-state, bracketing determinations. The resulting ED50 values probably overestimate the true ED50 values. However, because of the steepness of the dose-response curves, the resulting ED50 probably does not overestimate the value by more than 20%-30%, an amount that would not undermine our conclusions.
In summary, the molecular structures of nonimmobilizer and transitional inhaled compounds do not seem to correlate with their ability to cause convulsions. These anesthetics may produce convulsions by two interrelated mechanisms. The first of these correlates with the lipophilicity of the compounds, implying an action in a nonpolar phase. This first effect is altered by a second mechanism, the capacity of each compound to modify the action of GABA on GABAA receptors: to block the effect of GABA, to do nothing, or to enhance the effect.
1. Eger EI II, Liu J, Koblin DD, et al. Molecular properties of the "ideal" inhaled anesthetic: studies of fluorinated methanes, ethanes, propanes and butanes. Anesth Analg 1994;79:245-51.
2. Liu J, Laster MJ, Koblin DD, et al. A cut-off in potency exists in the perfluoroalkanes. Anesth Analg 1994;79:238-44.
3. Koblin DD, Chortkoff BS, Laster MJ, et al. Polyhalogenated and perfluorinated compounds that disobey the Meyer-Overton hypothesis. Anesth Analg 1994;79:1043-8.
4. Taheri S, Halsey MJ, Liu J, et al. What solvent best represents the site of action of inhaled anesthetics in humans, rats and dogs? Anesth Analg 1991;72:627-34.
5. Liu J, Laster MJ, Taheri S, et al. Is there a cutoff in potency for the normal alkanes? Anesth Analg 1993;77:12-8.
6. Liu J, Laster MJ, Taheri S, et al. Effect of n-alkane kinetics in rats on potency estimations and the Meyer-Overton hypothesis. Anesth Analg 1994;79:1049-55.
7. Taheri S, Laster MJ, Liu J, et al. Anesthesia by n-alkanes not consistent with the Meyer-Overton hypothesis: determinations of the solubilities of alkanes in saline and various lipids. Anesth Analg 1993;77:7-11.
8. Fang Z, Sonner J, Laster MJ, et al. Anesthetic and convulsant properties of aromatic compounds and cycloalkanes: implications for mechanisms of narcosis. Anesth Analg 1996;83:1097-104.
9. Koblin DD, Laster MJ, Ionescu P, et al. Polyhalogenated methyl ethyl ethers: solubilities and anesthetic properties. Anesth Analg. In press.
10. Fang Z, Laster MJ, Gong D, et al. Convulsant activity of non-anesthetic gas combinations. Anesth Analg 1997;84:634-40.
11. Fang Z, Laster MJ, Ionescu P, et al. Effects of inhaled nonimmobilizer, proconvulsant compounds on desflurane minimum alveolar anesthetic concentration in rats. Anesth Analg 1997;85:1149-53.
12. Halsey MJ, Koblin DD, Eger EI II, et al. Convulsant and anesthetic properties of cis-trans isomers of 1,2-dichlorohexafluoro-cyclobutane, and of 1,2-dichloroethylene. Br J Anaesth. In press.
13. Koblin DD, Fang Z, Eger EI II, et al. Minimum alveolar concentrations (MACs) of noble gases, nitrogen, and sulfur hexafluoride in rats: helium and neon as nonimmobilizers (nonanesthetics). Anesth Analg 1998;87:419-24.
14. Mihic SJ, McQuilkin SJ, Eger EI II, et al. Potentiation of gammaaminobutyric acid type A receptor-mediated chloride currents by novel halogenated compounds correlates with their abilities to induce general anesthesia. Mol Pharmacol 1994;46:851-7.
15. Harris RA, Mihic SJ, Dildy-Mayfield JE, Machu TK. Actions of anesthetics on ligand-gated ion channels: role of receptor subunit composition. FASEB J 1995;9:1454-62.
16. Kendig JJ, Kodde A, Gibbs LM, et al. Correlates of anesthetic properties in isolated spinal cord: cyclobutanes. Eur J Pharmacol 1994;264:427-36.
17. Richter J, Landau EM, Cohen S. Anaesthetic and convulsant ethers act on different sites at the crab neuromuscular junction in vitro. Nature 1977;266:70-1.
18. Wakamori M, Ikemoto Y, Akaike N. Effects of two volatile anesthetics and a volatile convulsant on the excitatory and inhibitory amino acid responses in dissociated CNS neurons of the rat. J Neurophysiol 1991;66:2014-202.
19. Lin LH, Chen LL, Zirrolli JA, Harris RA. General anesthetics potentiate gamma-aminobutyric acid actions on gamma-aminobutyric acid A receptors expressed by Xenopus oocytes: lack of involvement of intracellular calcium. J Pharmacol Exp Ther 1992;263:569-78.
20. Longoni B, Demontis GC, Olsen RW. Enhancement of gamma-aminobutyric acid A receptor function and binding by the volatile anesthetic halothane. J Pharmacol Exp Ther 1993;266:153-9.
21. Kira T, Harata N, Sakata T, Akaike N. Kinetics of sevoflurane action on GABA- and glycine-induced currents in acutely dissociated rat hippocampal neurons. Neuroscience 1998;85:383-94.
22. Jones MV, Brooks PA, Harrison NL. Enhancement of gamma-aminobutyric acid-activated Cl- currents in cultured rat hippocampal neurones by three volatile anaesthetics. J Physiol (Lond) 1992;449:279-93.
23. Nakahiro M, Yeh JZ, Brunner E, Narahashi T. General anesthetics modulate GABA receptor channel complex in rat dorsal root ganglion neurons. FASEB J 1989;3:1850-4.
24. Wafford KA, Burnett DM, Leidenheimer NJ, et al. Ethanol sensitivity of the GABA-A receptor expressed in Xenopus oocytes requires 8 amino acids contained in the gamma 2L subunit. Neuron 1991;7:27-33.
25. Harris RA, Mihic SJ, Brozowski S, et al. Ethanol, flunitrazepam, and pentobarbital modulation of GABAA receptors expressed in mammalian cells and xenopus oocytes. Alcohol Clin Exp Res 1997;21:444-51.
26. Virtue RW, Lund LO, McKinley P Jr, et al. Difluoromethyl 1,1,2-trifluoro-2-chloroethyl ether as an anaesthetic agent: results with dogs, and a preliminary note on observations with man. Can Anaesth Soc J 1966;13:233-41.
27. Botty C, Brown B, Stanley V, Stephen CR. Clinical experiences with compound 347, a halogenated anesthetic agent. Anesth Analg 1968;47:499-505.
28. Stockard J, Bickford R. The neurophysiology of anaesthesia. In: Gordon E, ed. A basis and practice of neuroanaesthesia. Amsterdam: North-Holland, 1975:3-46.
29. Neigh JL, Garman JK, Harp JR. The electroencephalographic pattern during anesthesia with Ethrane: effects of depth of anesthesia, PaCO2 and nitrous oxide. Anesthesiology 1971;35:482-7.
30. Joas TA, Stevens WC, Eger EI II. Electroencephalographic seizure activity in dogs during anaesthesia. Br J Anaesth 1971;43:739-45.
31. Mihic SJ, Ye Q, Wick MJ, et al. Sites of alcohol and volatile anaesthetic action on GABAA
and glycine receptors. Nature 1997;389:385-9.
32. Robbins BH. Preliminary studies of the anesthetic activity of fluorinated hydrocarbons. J Pharmacol Exp Ther 1946;86:197-204.
33. Krantz JG, Rudo FG. The fluorinated anesthetics. In: Handbook of experimental pharmacology. Berlin: Springer-Verlag, 1966:501-64.
34. Rudo FG, Krantz JC Jr. Anaesthetic molecules. Br J Anaesth 1974;46:181-9.
35. McNamara BP, Krop S. Observations on the pharmacology of the isomers of hexachlorocyclohexane. J Pharmacol Exp Ther 1948;92:140-6.
36. Koblin DD, Eger EI II, Johnson BH, et al. Are convulsant gases also anesthetics? Anesth Analg 1981;60:464-70.
37. Terrell RC, Speers L, Szur AJ, et al. General anesthetics. I. Halogenated methyl ethyl ethers as anesthetic agents. J Med Chem 1971;14:517-9.
38. Speers L, Szur AJ, Terrell RC, et al. General anesthetics. II. Halogenated methyl isopropyl ethers. J Med Chem 1971;14:593-5.
39. Terrell RC, Speers L, Szur AJ, et al. General anesthetics. III. Fluorinated methyl ethyl ethers as anesthetic agents. J Med Chem 1972;15:604-6.
40. Speers L, Szur AJ, Terrell RC. General anesthetics. IV. Methyl pentahaloethyl and methyl heptahaloisopropyl ethers as anesthetic agents. J Med Chem 1972;15:606-8.