GENETIC AND CYTOLOGICAL EFFECTS 275 



In experiments with the discharge-tube radiation (largely wave length 

 254 m^), it was possible to show the relation of internal filtration to the 

 dosage curve directly (Table 7-4). A given dose may be applied with 



Table 7-4. Frequency of Endosperm Deficiencies from a GrvEN Dose 



When Applied to Both Sides of Pollen as Well as to One Side 



Only (Wave Length 254 M/j.) 



(Stadler and Uber, 1942.) 



Frequency, 

 Dose per cent 



One unit (one side) 18.9 



Two units (one side) 24 . 9 



Two units (one from each side) 35.4 



equal effect from either above or below the layer of pollen. When the 

 dose is doubled by applying a second unit of dose from the same direction, 

 the added frecjuency of induced effects is considerably less than that from 

 the first, and thus the yield from 2 units is considerably less than double 

 the yield from 1. But if, instead, the second unit of dose is applied 

 from the opposite side, the added yield of induced deficiencies is as great 

 as that from the first unit, and thus the yield from 2 units of dose is 

 double the yield from 1. 



When the measure of radiation effect is not a specific result (e.g., a 

 given deficiency, death of the irradiated individual) but rather an indefi- 

 nite group of results, any number of which may be observed in the single 

 treated individual (e.g., mutations at miscellaneous loci), the result 

 expected from internal filtration is not a flattening of the dosage curve. 

 Instead it is a tendency toward coincidence of independent effects in the 

 single treated gamete, such as was noted in the maize experiments men- 

 tioned in an earlier section of this review (see also Meyer et al., 1950). A 

 flattening of the dosage curve for mutation frequency would be expected 

 as a result of variations in internal filtration only if the accumulation of 

 mutants and other radiation effects is a factor in eliminating individuals 

 from the population tested. 



In treated populations in which there are large variations in exposure 

 to radiation injury, gross distortion of the dosage curve ma\^ occur. For 

 example, if the treated population were a mass of pollen grains more than 

 one layer deep, the lower layers would be almost wholly shielded from the 

 radiation. With a sufficiently heavy dose, the pollen grains of the top 

 layer might be largely eliminated from the population tested, and the fre- 

 quency of genetic effects in the surviving population would be materially 

 lower than that found with lighter doses. The correlation between 

 genetic effects and killing would be a spurious correlation, but any correla- 

 tion of genetic effects with elimination among the individuals treated 

 would tend to flatten or reverse the dosage curve. 



The dosage data for maize pollen treatments are therefore of interest 



