Bluhm1 and Danforth,2 using somewhat different experimental methods, have reported a rise in the sex ratio of mice from alcohol treated fathers. However, these ratios are based on small numbers (182 ♂ : 149♀ Bluhm; 210 :164♀ Danforth) and, although the probable errors of the deviations from the control ratios (+10.60 per cent ±2.03 and +5.36 per cent ±1.99) may be taken to mean that random sampling alone is not responsible for the results, various other possible influences besides alcohol, such as season, mother's age and parity, have not been eliminated by the methods of these investigators. Moreover, Gyllenswärd's3 more completely controlled experiments with alcoholized male mice show as great a change in the opposite direction (—10.4 per cent ±3.4), and MacDowell and Lord4 have called attention to the fact that when all question of the modification of the sex ratio by prenatal mortality was removed by using only complete litters (so judged by the number of the corresponding corpora lutea) the primary sex ratio given by litters from heavily alcoholized fathers and normal mothers (50.3 per cent males, based on 308 mice) showed no significant deviation from the primary sex ratio from normal parents (49.9 per cent males, based on 523 mice).\ud \ud The present report gives the sexes of 2133 mice from normal mothers by fathers given completely anesthetizing doses of alcohol fumes five days a week, beginning at the age of four weeks and continued over a year, to the end of the experiment; and 2322 mice from the same normal mothers by normal fathers, brothers of those treated. The treatment was given by inhalation, in one pint milk bottles; for each treatment 3 cc. 95 per cent alcohol was poured on a piece of atisorbent paper which was placed in the bottle with the mouse, a regular milk bottle cap inserted and the bottle inverted