152 L. J. MULL1NS 



being introduced into the membrane. How difficult it is, therefore, to conclude 

 that because a drug modifies junctional transmission and yet the response to 

 topical ACh is unaltered, the drug is acting on the release of ACh. The great 

 dispersion in membrane interspace size that appears necessary to account for 

 the action of drugs and ions should also result in a greater instability than that 

 found in the axon; it would not be unreasonable to find small leaks developing 

 spontaneously. Whether these are pre-junctional and involve quantal ACh 

 release, as suggested by Fatt and Katz, or whether they are post-junctional 

 and limited by membrane 'viscosity' effects seems not at all clear. As suggested 

 previously, curare may be supposed to stabilize the post-junctional membrane 

 by reducing the number of ACh receptors; it may also stabilize against ther- 

 mally induced ion leaks. Charge is transferred across the post-junctional mem- 

 brane considerably more rapidly than is the case for the axon; this may reflect 

 a thinner membrane but also suggests that ions such as sodium with two shells 

 of hydration may be of importance in effecting this rapid charge transfer. Thus, 

 it is suggested, the release of ACh results in a small number of molecules filling 

 'receptor' sized interspaces, the distortion surrounding such an interspace 

 results in a large number of (Na + ) 2 size and larger interspaces, and the ACh is 

 removed by hydrolysis or diffusion. 



Variations in the physiological response to ACh are understandable from the 

 model if the mean interspace size and the membrane viscosity are considered 

 as variables from tissue to tissue. It is then possible for ACh to inhibit function 

 by hyperpolarization of the membrane, and for it to have its excitatory action 

 integrated over periods of many minutes. Certain generalizations with respect 

 to inhibition and synergism are worth commenting upon. If an excitatory sub- 

 stance requires sites that are small as compared with the mean of a distribution, 

 the inhibition of such excitatory action directly (e.g. by occupying the sites 

 that are to be acted on by the excitant with another molecule) is impossible 

 unless substantially all of the sites in the membrane are occupied with inhibitor. 

 Indirect inhibition, (e.g. by shifting the distribution of interspace sizes so as to 

 diminish the number with small size) by condensing the size of a few, large 

 sized interspaces, can be brought about without occupying very many inter- 

 spaces. Specific inhibition, (e.g. the inhibition of one junctional transmission 

 system without influencing others) can be effected if the dispersion in inter- 

 space sizes is different for two systems, because a molecule that is very large, 

 compared with the dispersion on which it acts as an inhibitor, may be equal to 

 the mean size for the second dispersion, where it will be relatively inactive. 

 Molecules of a size such that they may be taken up by half of all the interspaces 

 in the membrane may be expected to decrease the mean interspace size. Such 

 a change may be expected to give variations in the number of interspaces that 

 can act as receptors for chemical transmitters or for K + . Large decreases in 



