128 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



the equatorial plate stages of the heterotypic division Geerts 

 claims to recognize seven double and seven single chromosomes. 

 The double chromosomes divide and go to their respective poles, 

 while four of the single chromosomes may go to one pole and 

 three to the other. A'ot infrequently one or more of the single 

 chromosomes fail to reach the poles and are left on the spindle. 

 In the second division the three or four single chromosomes may 

 divide but more commonly they disintegrate. Here again a 

 chromosome may be left on the spindle and form a dwarf nucleus. 

 Still, Geerts reports that pollen grains with ten or eleven chromo- 

 somes are not uncommon. Geerts regards his results as in perfect 

 accord with that of Rosenberg on Drosera obovata and as opposed 

 to those of Gates ('09) on the same hybrid. In both heterotypic 

 and homoeotypic divisions of male cells of Oenothera gigas, Davis 

 ('11) found chromosomes that fail to reach the poles and form 

 supernumerary nuclei. 



It is plain that the cases of chromosomes which fail to reach 

 the poles are not confined to hybrids with parents having the 

 unequal number of chromosomes. 



In the cases so far considered there is a question as to whether 

 the bodies appearing in the spindle in early telophase are really 

 chromosomes. It is well known, however, that other deeply 

 staining bodies may be found on the spindle in both the reduction 

 and somatic divisions. Zimmerman ('93, '94) reported the ap- 

 pearance of bodies on the spindle and in the cytoplasm which in 

 appearance and staining reaction agreed in all essentials with 

 the nucleoli of the resting nuclei. Zimmerman believed that 

 these bodies came from the nucleolus as a result of its disinte- 

 gration during karyokinesis and called them extranuclear nucle- 

 olcs. He held further that these bodies go into the daughter 

 nuclei and fuse to form a new nucleole — a view subsequent in- 

 vestigation has not confirmed. Allen ('05) in Lilium canadense 

 described and figured extranuclear nuclcoles in the prophases of 

 the pollen mother-cells. These bodies he claimed disappeared 

 by the time spindles were formed. They reappeared and later 

 were numerous in the equatorial plate stage. Gregory ('05) 

 studied similar bodies in the fertile variety of the sweet pea, 

 Emily Henderson. He found that the nucleolus fragments and 

 spherical bodies, three to four in number, appear in the cytoplasm 



