Abstract:Anthers with mid-unlnucleate microspores were cultured on W5 medium supplemented with 0.5 mg/l kinetin, 2 mg/l 2,4-D and 9% or 3% sucrose. At a series of interval (0, 1, 1.5， 2， 14 days) after cultured, the anthers were labelled with 3H-thymidine (4 MCi/mi) for 24 h, fixed, and then performed autoradiography according to conventional method. Results show that after cultured for 24 h, 3H-thymidine was incorporated into some late-uninucleate microspores (see Plate I, 3), and after for 2.5 days, vegetative nuclei in pollen grains were la- belled (see Plate I, 4). Usually, vegetative nuclei were labelled frequently and generative ones were labelled rarely. Sometimes generative cell which could synthesis DNA might develop suspensor-like structure individually (see Plate I, 13). During early stage of development of a multicellular pollen grain, the DNA synthesis in the cells were synchronized. With pollen development, the synchronism of DNA synthesis was destroyed. When anthers cultured on medium with 3% sucrose, DNA in microspores could be synthesized normally, and the number of labelled microspores was more than that of anthers cultured on medium with 9% sucrose. However, on medium with 3% sucrose, the nuclei in microspores stopped dividing after one or two divisions and the cell wall of them could not be formed and multicellular pollen was not observed. It seems that the absence of multicellular pollen on medium with 3% sucrose was primarily due to the block of cell division and cell wall formation, not due to the interruption of DNA synthesis.