366 MALIS, ROSE, KRUGER AND BAKER 



per cm- is made clear by the consideration (see Fig. 6) that the number of 

 ions produced per unit length of path varies for different sectors of the path 

 and that it is the number of ions produced per unit length of path and not 

 the number of particles which traverse the cortex which is responsible for 

 the laminar destructions. 



One does not avoid this essential disadvantage by computing the dose in 

 rads, if the computation is made on the basis of number of particles, their 

 energies, and their range in tissue. The figure which emerges from such cal- 

 culation represents an average dose in rads, which would be an adequate 

 measure only if the energy loss along the beam path were uniform. Since 

 this is not the case, it is desirable to estimate the actual doses delivered at 

 chosen points along the beam path. We have routinely computed such doses 

 at two levels. The first is the dose in rads delivered to the surface of the 

 cortex {surface dose) ; the second is the dose delivered at the ionization peak 

 {peak dose). For calculation of the surface dose, the rate of energy loss must 

 be known. For brain, this rate was considered equal to the appropriate values 

 tabulated for water (Rich and Madey, 1954), since the rates of energy loss 

 for brain and water cannot be significantly different. Once the surface dose 

 is determined, the peak dose (or any other dose at a desired depth) can be 

 estimated from the ionization curve (Fig. 6), if a reasonable assumption is 

 made that the shapes of the ionization ciuves in brain and aluminum do not 

 diflfer materially. 



The peak dose is, we believe, the most appropriate measure of the radia- 

 tion effect in our material, despite the fact that its estimate is technically 

 less reliable than that of other measures. The peak dose is the highest dose 

 delivered over a short distance in tissue. It can be argued from the ionization 

 curve in Fig. 6 that values close to the peak dose can be expected to be 

 delivered to a strip of cortex about 50 p. wide. 



The peak dose, the surface dose, and the a\erage dose have constant rela- 

 tionships to each other if the range of particles in tissue remains the same. 

 If the range is shortened in tissue by introduction of an absorber into the 

 beam path, the ratios change, since the value of the peak dose remains con- 

 stant, while the average dose and the surface dose become higher. In our 

 irradiation arrangements, the peak dose was usually almost 5 times larger 

 than the surface dose and about 3 times larger than the average dose. 



As is known, some time elapses before a radiation lesion manifests itself 

 histologically. The latent periods tend to be long with marginal doses and 

 become greatly reduced as the dose increases (Malis et al., 1960). We will 

 consider the dose necessary to produce a laminar lesion in rabbits which 

 survived from 3 weeks to 18 months after irradiation. Table I assembles such 

 data for 177 lesions. Each lesion was studied in serial sections cut at 30 fx, 

 and every section was mounted. 



