MANNER OF PRODUCTION OF MUTATIONS 487 



which recei\'e radiation. The number of these units within a given 

 amount of tissue will necessarily vary in direct proportion to the dose; 

 hence the frequency of gross structural changes must vary in this manner. 



Few experiments have involved doses which were low enough in totality 

 or in intensity (i.e., doses which delivered a small enough amount of radi- 

 ation within the time necessary for the process of union of ends to com- 

 plete itself) to allow this linear relation of the frequency of gross structural 

 exchanges to dosage to become observable. This condition was, however, 

 fulfilled in certain experiments of Catcheside, Lea, and Thoday (1946) 

 and of Sax (1950) on Tradescantia microspores. In the first-mentioned 

 work the time of exposure was held constant, and the frequency of cyto- 

 logically observed gross exchanges between broken ends would therefore, 

 under ordinary circumstances, have been proportional to the square of 

 the dose, but the time of exposure was so long, 251 minutes, i.e., the 

 intensity was so low, that at doses of 25 and 50 r the frequency followed 

 the first power of the dose almost exactly. Thus the great majority of 

 the exchanges at these doses must have had both their breaks arising as 

 effects of the same projected particle. Even at 150 r, as the curve 

 showed, about one-third the exchanges must still have been derived from 

 nonindependent breaks. Less direct evidence of the operation of the 

 linear rule for interchanges at low doses is moreover provided by the 

 fact that, with very low intensities of radiation or with very small but 

 often repeated doses, some of the subtler forms of long-term damage, 

 such as the lowering of the expectation of life, appear to be simply pro- 

 portionate, in amount, to the total accumulated dosage. In this con- 

 nection it may be noted that the greater the specific ionization of the 

 radiation — or, more accurately, the greater the number of ions which are 

 produced within a critical volume as effects of the same projected parti- 

 cle — the sooner this linear rule will come into operation as the dosage is 

 lowered, since the frequency of independent tracks falling within a critical 

 volume in the required time will be lower. 



In the work on breakage and structural changes, as in that on gene 

 mutations, tests were made of the relative effectiveness of different 

 wave lengths. Definitive evidence was obtained by Muller (1939a, b, c, 

 1940), wdth the collaboration of Ray-Chaudhuri, showing that 7 rays 

 from radium and ordinary X rays, when applied to Drosophila sperma- 

 tozoa, produce translocations with sensibly the same frequency.^ More 



^ Eberhardt (1939) obtained data indicating that this was true of the special class of 

 translocations which give the cubitus-interruptus position effect. However, these 

 results are not cited because the questionable results on dosage which were obtained 

 with this material at the same time raise serious doubts (see p. 481) concerning the 

 reliability of the methods of culturing and of scoring used for this very variable 

 character. The same doubts apply to the peculiar results later reported by Eber- 

 hardt and Zimmer (1940) on the effects of differently timed doses on the frequency of 

 translocations of this special kind. 



