288 C. T. GAFFEY 



is known (Zirkle, 1954) that the linear energy transfer (i.e., the stopping 

 power or rate of energy loss) along a particle's track varies as the square of 

 its charge. The linear energy transfer on an alpha particle is 4 times that of 

 a deuteron of the same velocity. Since biologic eflfects in general vary with 

 the linear energy transfer, it would be expected that the RBE of alpha 

 particles with respect to deuterons would approach 4 as a limit. 



Membrane Model 



In the following section a membrane model for nerve is outlined with the 

 hope that such a model may suggest how the function of nerve is affected by 

 radiation energy. 



Direct evidence of neural membrane structure, in terms of lipid and 

 protein components, must await a detailed study of lipids and lipid-protein 

 systems. Whatever may be the ultimate interpretation of the molecular or- 

 ganization of the axon membrane, it is probably safe to say from electron 

 microscope studies that the unit membrane includes two protein monolayers 

 allied with a double layer of lipid molecules (Schmitt, 1959) . 



If it is assumed that the protein molecules of the neural membrane are 

 helical in nature and form an oriented structural layer, certain insights into 

 membrane properties are revealed. When 3 protein molecules of macro- 

 molecular diameter are closely packed, a 4th element is created — an inter- 

 stice or fault which for convenience will be referred to as a "channel." When 

 3 protein macromolecules 28.2 A in radius are most efficiently packed, an 

 intermolecular channel about 4 A in radius is obtained (Fig. 10). 



It is known that models of membranes based on the concept of a continu- 

 ous lipid layer are untenable because experiments reveal that biologic mem- 

 branes are crossed by molecules of water and numerous compounds insoluble 

 in fat. This is a property of a membrane with channels rather than a solution 

 process in a lipid film. Comparison of rates of water entrance into cells 

 under the influence of osmotic pressure gradients and simple difTusion 

 gradients gives a rough indication of what may be the "equivalent channel 

 size" (KoeflFed- Johnson and Ussing, 1953; Prescott and Zeuthen, 1953). 

 Values of channels range from 5 A in red blood cells to 16 A in squid axons 

 (Nevis, 1957; Solomon et al., 1957) . When frog nerve is placed in a medium 

 labeled with deuterium, tritium or O^^, the half time for equilibration is 

 only 1 minute (Tobias and Nelson, 1959). Hence, the existence of a channel 

 pathway through the ultrastructure of cell membranes to water and small 

 ions seems likely. 



It has been suggested by Mullins (1956) that the number of water mole- 

 cules associated with each ion in traversing the neural membraine is limited 

 to the same minimum, say one. In physiologic solutions sodium ions are 



