486 JOHN W. GOWEN 



unit. By plotting these m values a measure of the X-ray effect is 

 obtained and the equation for survival of the biological material 

 against dosage may be calculated. The method is not very precise 

 but is the best available in many instances. 



Another method of interpreting X-ray investigations is the analysis 

 of variance (49), where the experiment is analyzed as a whole. Statis- 

 tical design of the experiment frequently adds greatly to the precision 

 of these interpretations. There is a necessary word of caution, how- 

 ever. X-ray experiments very generally do not meet the require- 

 ments of variance analysis, in that the variance of the individual 

 treatment is generally greater for the control and successively less 

 until it reaches its lowest value at the highest radiation treatment. 

 The coefficient of variation rather than the standard deviation tends 

 to be a constant. Such complications make the variance analysis 

 inapplicable. Special methods for transformation of data will be 

 necessary. 



H. INTERPRETATION OF RADIATION EFFECTS 



In what has gone before, the emphasis has been on the physical 

 effects and viewpoint in radiation experiments rather than on the bio- 

 logical variabilit}^ of the material. Different schools of thought have 

 interpreted X-ray experiments from either one of these viewpoints. 

 Some have looked on the progressive nature of irradiation effects as a 

 selection of the innate variability of the biological material to the 

 radiation, such that the weakest are killed off early and the strongest 

 survive for the longest times. As the reader has already noticed, the 

 physical viewpoint is that the radiation is distributed more or less 

 at random over the biological material and that it is this random 

 variation that produces the progressive X-ray effects. Evidence is 

 accruing for the importance of both these elements in the interpreta- 

 tion of biological data. 



The biological variation of the material has, until recently, been 

 largely postulated rather than proved to exist. The postulate seems 

 good because of the known biological variability of the treated mate- 

 rials as exposed to many other agents. The presumption is that 

 similar variability exists for X rays despite the fact that they pene- 

 trate more readily than do the other agents. Recently there has 

 been rather definite evidence indicating that certainly a small amount 

 of biological variability to X-ray radiation does exist. Intraspecies 



