458 JOHN W. G OWEN 



with a standard ionization chamber. The whole box is heavily 

 shielded against X rays entering at any place other than the thimble 

 chamber. 



The threshold erythemal dose is equivalent to about 500 r. since 

 this dose delivered at one sitting to a local skin area is capable of 

 producing perceptible reddening in 80% of patients within four weeks 

 of treatment. Similarly, the dose necessary to kill 50% of the Dro- 

 sophila eggs under the standard condition is 180 r. 



2. The Depth-Dose Problem 



The close relation between the biological effects and the X-ray 

 dose in r. values in quantitative experimentation justifies the meas- 

 urement of X-radiation doses in r. units. For actual dose measure- 

 ments in r. units to correspond to the absorbed energy in the experi- 

 mental tissue, certain conditions must be satisfied. If the layer of 

 material to be treated is thin, as, for instance, a film of virus on Cello- 

 phane, the experimental arrangement must be such that the film is 

 treated with a sufficient layer of air between it and any other more 

 dense material to prevent the return of scattered radiation from the 

 dense substances to affect the virus. 



This "back-scatter" effect becomes particularly important in 

 treating deep tissues. Problems are presented not only by the scat- 

 tering characteristics of the tissues that surround the object of inves- 

 tigation but also by the absorbing power of intervening layers of tis- 

 sue. For example, if a bone is located close to deep tissue, scattered 

 radiation from the bone would change the dose from that expected 

 if the bone were not present. Computations of such an effect are 

 very difficult because the X rays are not monochromatic and the 

 several materials present of different atomic number contribute in 

 various amounts to the back-scattered radiation. With an appro- 

 priately constructed "phantom" (see 5, p. 179), dosage measurements 

 to take such effects into account are not too difficult to make. At the 

 same time it can correct for the layer of material over the treated ob- 

 ject, which is likewise important in order to find the dose incident on 

 the object under investigation. The total "absorption" in this 

 layer depends on several physical phenomena, including photoelectric 

 effects and two types of scattering. These different phenomena af- 

 fect the absorbing power of a substance with respect to its atomic 

 number in different ways. The photoelectric absorption, for example, 



