IQQ THE BOTANICAL MAGAZINE. LVui. xxxv No. m. 



unite together at the approached poles of the two different spindles, 

 while the opposite poles remain free. This results in the formation 

 of the three nuclei in a pollen mother cell instead of four (PI. Ill, 

 Fig. 24). As the consequence an abnormal large diploid pollen grain 

 and two normal pollen grains are produced in such a mother cell 



(Text-fig. 1. e). 



From the various modes of development of pollen grains, which 

 I have described here, we see that all the pollen grains formed in a 

 pollen mother cell may differ from each other, not onh' quantitatively 

 but also qualitativel}'. 



At about the anaphase of the second division or earlier than 

 that, there appear many vacuoles in the cytoplasm in some pollen 

 mother cells, and degeneration of the nucleus and cytoplasm takes 

 place, the result of which is the abortion or sterility of the pollen 



grains. 



3.' Cytomixis.' As is shown in PI. Ill, Fig. 2, the nuclear sttb&tances 

 of a mother cell in prophase, may pass into the cytoplasm of 

 the neighbouring pollen mother cell or cells, through the pits of the 

 cell membrane. This phenomena were already observed by Koernicke 

 (in Crocus, 190.1), Gregory (in Sweet pea, 1909), Gates (in Oeno- 

 thera, 1911) DiGBY (in Primula, 1912) and others. Some of them 

 ascribed this phenomenon to an artificial production, but others, e.g., 

 Gates, as a natural process in the meiosis. I have seen the same 

 phenomenon also in the preparations of ,the pollen mother cell in 

 synapsis of Papaver orientale and in diakinesis of P. somniferum, so 

 that it seems not seldom to find this phenomenon in the meiotic 

 phase in fixed materials ; but on account of the fact, that it does 

 not alwa3's occur even in the same plant materials and in the same 

 stage of cell division, I am inclined to the view that this phenomenon 

 is to be taken for an artifact which is caused by the fixing liquid, 

 when the protoplasm is in a state particular to the stage of mitosis. 



General Considerations 



1. Hybrid nature of the plant and the irregular behavior of 

 chromosomes. In the prophase of the meiosis of the hybrid plants 

 artificially raised between the parents, which have different numbers 

 of chromosomes, the bivalent and univalent chromosomes are formed. 

 The former pass the metaphase regularly, while the latter may or 

 ma}^ not follow the same route. In either case, it is chiefly the 

 behavior of these univalent chromosomes, which cause the irregular 



