STRUCTURE OF NERVE CELL MEMBRANES 131 



low mobilities for the ions concerned. If two ions of the same size and opposite 

 charge are alternately loaded into a channel, it would appear that the ions in 

 the channel would escape the influence of the electric field imposed by the 

 membrane potential and move outwards by diffusion forces alone. 



The Action of Narcotic Agents 



In a previous study of narcosis (Mullins, 1954) it was suggested that the 

 experimental data could be explained equally well by representing the mem- 

 brane as a homogeneous lipid film or as an aggregate of pores formed by macro- 

 molecules since, in either case, a constant thermodynamic activity for narcosis 

 implied that the narcotic molecules fitted into a structure at their site of action 

 as they did into their own pure liquids (Brink, 1951). Some of the reasons for 

 believing that the membrane has a pore structure have already been presented, 

 and others will be given later in the discussion. The size of such membrane 

 pores or interspaces can be estimated in the following way. Data for the narcosis 

 of organisms by homologous series of paraffins, alcohols or ethers quite fre- 

 quently show that the threshold for narcosis expressed as thermodynamic 

 activity is constant from member to member up to some point, and then rises 

 such that log activity vs. number of carbon atoms are linearly related. The 

 point where a discontinuity in activity vs. size of narcotizing molecule takes 

 place may be interpreted as indicating some limitation imposed by the mem- 

 brane on the uptake of molecules of larger and larger size. If all membrane 

 interspaces were the same size, the discontinuity ought to be abrupt, while it 

 is usual to find increasing activities required for molecules with 5 to 10 carbon 

 atoms. Since we are dealing with molecules with an alkyl chain of increasing 

 length, it is possible to adopt the view that the size of the molecule in a par- 

 ticular orientation (with the chain perpendicular to the surface of the mem- 

 brane) is not changing, but only the concentration of molecules in a particular 

 orientation. Somewhat more certainty with respect to size attends the use of 

 rigid molecules as narcotics. Thus benzene has a threshold for narcosis for the 

 grain weevil of 0.02, while methyl benzene has a threshold of 0.08 (Ferguson, 

 1951). We can specify a size for both molecules and consider that the difference 

 in concentration is because the methyl group on toluene imposes on this mole- 

 cule a limitation on penetration not found in benzene. The rigidity of both 

 molecules makes it likely that the nature of the limitation is an orientational 

 one. Thus, toluene can pentrate in an orientation where its methyl group does 

 not interact with the walls of the interspace but as the number of such orienta- 

 tions is only a fraction of total number of collisions with the surface, the con- 

 centration of toluene must be increased for narcosis. We may expect, therefore, 

 that for paraffins there are relatively few interspaces larger than benzene. 

 When molecules that are practically spherical are used as narcotics, it is some- 

 what easier to analyze the situation; they have the same profile in all orienta- 



