Sept., 1921.] YASUI :— ON THE BEHAVIOR OF CHROMOSOMES 161 



division of the pollen mother cell. The similar process was first 

 observed by Rosenberg in Drosera hybrid (1904) and afterwards by 

 Gates (Oenothera, 1907, 1909), Geerts (Oenothera, 1909, 1911), 

 Rosenberg (Hieraciuw, 1917) Kihara (Triticum, 1920, 1921) and 

 others in artificially raised hybrids. Lately Tacholm, and also Black- 

 burn and Harrison studied several forms of roses independently, 

 and the latter two authors highK- emphasized the presence of 

 univalent chromosomes in the prophase of the first division of meiosis 

 as an indication of hybrid nature of the plant. My Papaver hybrid 

 is an additional example of the presence of univalent chromosomes as 

 well as the irregular behavior of chromosomes in the meiotic prophase 

 in experimentally raised h\-brids. But even the fact that hybrid plants 

 show the irregular chromosome behavior is universal, yet the converse 

 can not be always true, so that it does not necessarily follow that 

 plant showing irregular chromosomes behavior is of hybrid origin. 



2. End to end-union and parallel-union of chromosomes. There 

 were two different types of the gemini formation described, the end to 

 end-union and the parallel-union. Authors who have studied artificially 

 raised hybrids, generally describe end to end-type. My material also 

 shows the same. Though I am not yet prepared to give a definite 

 account of comparison abo at the same stage of cell division in my 

 hybrid Papaver and the parent plants, it may be suggested here that 

 the end to end-union is the phenomenon which occurs when the 

 afiinity or similarity of geminal chromosomes is in a lower degree. 



3. Behavior of the univalent chromosomes. They undergo the 

 longitudinal split either in the first or the second division. The case 

 of Triticum figured by Kihara is the one, which is typical to the 

 former instance. The case of the Papaver hybrid is ver^' much like 

 his result ; the univalent chromosomes on the spindle of the first 

 division of meiosis split into two, and in majority' of cases each 

 halves of the univalent chromosomes follow the halves of the bivalent 

 chromosomes for the opposite poles, but in the case of Papaver, there 

 is more irregularit}^ in the behavior of univalent chromosomes and 

 sometimes it resembles the diagram C of the meiosis in the 

 Pisosella of Rosenberg, in which some univalent chromosomes pass 

 undivided to one of the opposite poles. In the second division of the 

 meiosis, however, ordinary univalent chromosomes part themselves, 

 not undergoing the longitudinal division for the second time, as is 

 the case with Rosenberg's Hieracium, in unequal numbers for the 

 opposite poles of the spindle. Consequently the distribution of 



