Two inbreds of mutant origin differing in the number of bracts were crossed to obtain the F 1 hybrid. One mutant line had 24.5 ± 1.01 bracts while the other 78.6 ± 1.69 bracts. The F 1 hybrid had an average value between parents which practically did not differ from the mean value in the F 2 population. The variability of the trait under study in the F 2 population was continuous varying from 20 to 84 bracts. This indicated the probable participation of several genes with an additive effect in the control of the number of bracts. Assuming that the differences between the parental lines are due to two pairs of genes the F 2 plant population grown in 2016 was divided into 5 classes. In that population the observed classes ratio turned out to be close to the theoretically expected ratio of 1 : 4 : 6 : 4 : 1. Over the next two years F 2 populations were tested in a similar way. In all the cases the calculated chi-square value did not exceed the critical value for 4 degrees of freedom and 5 % significance level. This gave reason to talk about the two-loci control of a such quantitative trait as the number of bracts. The participation of two non-allelic genes in the control of this trait is also proved by matching the actual number of plants in the parental classes to the theoretically expected number of plants. Thus the number of bracts depends on the number of dominant alleles of two different genes in the genotype.