Abstract: The Giemsa C-banding procedure was successfully modified to stain the nucleolus by prestaining the root tips with 1% aceto-carmine for two to three days before squashed. After band differentiation, the nucleolus was positively stained, and the hands and heterochromocenters remained their speciality and good resolution. Two pairs of chromosomes (chromosome Ⅵ and Ⅷ) were banded in the genome of maize line DMo-17. Four heterochromocenters appeared in the interphase cells correspondingly.With the band closely neighbouring the secondary constriction of chromosome Ⅵ (NOR-chromosome) as marker, it was found that both NORs in all the root tip cells were active and participated in nucleolus formation, and that uucleolus was exclucively associated with the banded NOR-chromosome (s) in mitotic as well as in interphaae cells.In cells with single nucleolus, the nucleolus was always associated with the pair of bands or heterochromocenters on the NOR-chromosomes but not with those on chromosome Ⅷ. It was clear that the single nucleolus was actually a fused body which was made up by the two NOR contributors.In cells with double nucleoli, it was observed that each nucleolus was always related with a single band or heterochromocenter on one of the NOR-chromosomes, but not with those on chromosome Ⅷ. It was apparent that one NOR-chromosome was responsible for one nucleolus formation in this type of cells. The one for one relationship provided us a direct explanation to Gates'dogma that the maxium number of nucleoli in a species was contant and equale to the number of secondary constrictions in the genome.It was also observed, in the cells with double nucleoli, that the two "nucleolus-band" structures were constructed in a polarized manner, with a band or heterochromocenter joining to the small end of a pearlike nucleolus. The two polarized structures were in a parallel distribution and oriented in the same direction inside the nucleus. This could be interpreted alterlatively by: 1) the chromosome arrangement within the interphase and prophasc still remained the telophase configuration in the mother cell; 2) the parallel distribution and the one-direction orientation could, with some unknown mechanism, have prevented the two polarized stuctures from fusing into a complex one.