GENETIC AND CYTOLOGICAL EFFECTS 259 



lated mutants in a single treated gamete. In addition, three of the other 

 mutants occurred in the progeny of plants segregating for defective pol- 

 len. The independent inheritance of the mutants, in these and several 

 similar cases, has been shown by Sprague (1942). The degree of coinci- 

 dence is far beyond that expected by chance, if all the tested gametes 

 received an equal dose. But the treated gametes must receive 

 quite unequal doses, for the sperm nuclei are eccentrically located 

 in the spheroidal pollen grains, and the loss by internal filtration must 

 vary widelj^ with the casual orientation of the individual grains. If this 

 is the explanation of the coincidences observed, the frequency of mutation 

 in the most effectively exposed pollen grains must be very high. In later 

 experiments with more effective ultraviolet treatments, mutation rates of 

 about 20 per cent have been reported (Stadler, 1941b). These are sub- 

 ject to the same limitation by casual orientation of the pollen grains 

 treated, and the results suggest that the effect of ultraviolet radiation on 

 mutation frequency in maize, in individuals that are effectively treated, 

 may be well beyond that produced by X rays. 



A technical advantage of the maize material is the availability of many 

 endosperm characters of known inheritance. In matings with appro- 

 priate marker genes, the loss of the effect of dominant alleles present in 

 the male parent may be detected at once by phenotypic changes in the 

 endosperms of the seeds produced. Linked endosperm characters, deter- 

 mined by genes located on one arm of one of the chromosomes, permit the 

 detection of deficiencies in this region. 



Ultraviolet treatment applied to the pollen greatly increased the fre- 

 quency of endosperm deficiencies, as detected by loss of the linked factors 

 C and Wx. The frequency of loss of other dominant genes for endosperm 

 characters {A, Pr, Su) was similarly increased. There are no linked 

 genes for endosperm characters suitable for determining whether A, Pr, 

 and Su losses represent deficiencies. The fact that ultraviolet-induced 

 loss of C and Wx is usually coincidental indicates that the endosperm 

 effect is due usually, if not always, to deficiency rather than to gene muta- 

 tion. Since no genetic analysis can be made, the identification of an 

 individual case as a recessive mutation rather than a deficiency could not 

 be positively established. The cases are referred to as "endosperm 

 deficiencies," with the reservation that there may be included among 

 them an unknown proportion of losses of dominant characters as a result 

 of gene mutation. 



The endosperm deficiencies resulting from ultraviolet treatment of pol- 

 len included a large proportion of fractionals, in contrast to those resulting 

 from X-ray treatment of pollen, which are largely deficiencies affecting 

 the entire endosperm. 



The occurrence of deficiency in the Fi plants was not determined by 

 means of marker genes or cytological examination in this experiment. 



