Sorting and blending of unit characters. 219 



Whatever may be the nature of the lack of a reciprocal working 

 of the reproductive mechanism in hyhrida nutella it cannot be attri- 

 buted to irregularities in the meiotic division preceding the formation 

 of the pollen such as have been shown to take place by Juel (639 — 641, 

 1900) in Syringa rothomagensis (a hybrid of S. vulgaris and S. persica 

 according to Focke 255, 1881), where typical mitotic figures are rare 

 in the first division of the gouotokonts. In the atypical cases a process 

 intermediate between mitosis and constriction occurs, a number of chro- 

 mosomes appear to go to the daughter nuclei undivided, some are left 

 in the cytoplasm. He suggests that there is a lack of mixing of the 

 nuclear substance. Irregularities have also been shown by Rosenberg 

 (1903, 1904, 1909) in a h^'brid between Drosera rotundifoUa and D. 

 longifolia where there is a difference in the number of chromosomes in 

 the two parents; by Cannon (133, 1903) in a Gossypimn hybrid (G. 

 harhadense X herbaceum) where some pollen mother cells divide normally, 

 others abnormally; and by Metcalf (109, 1901) in Gladiolus hybrids 

 where two spindles were observed, presumably due to a repulsion bet- 

 ween the paternal and maternal chromosomes so that they remain in 

 separate groups. These and similar abnormalities in meiotic division may 

 explain the sterility of pollen in certain hybrids. Strasburger (609, 1904) 

 makes the interesting suggestion that the difficulties resulting in the 

 formation of imperfect pollen occur in synapsis of the heterotypic di- 

 vision. But, as stated above, this cannot account for the sterility of 

 hyhrida mdella, unless during synapsis (or other critical moments in 

 the formation of the gones) there is such an association of physiological 

 "properties" or "tendencies" in the pollen and embryo sac of such a 

 nature as to prevent the reciprocal working of the sperm and Q^g of 

 hyhrida nutella when brought together, but permits reciprocal working 

 when either sperm or egg mates with a germ cell from related forms. 



The blend, hyhrida nutella, is thus a striking instance of an 

 intermediate hybrid which is self sterile, but in which the sterility is 

 not due to an impairment of the fertile condition of either the pollen 

 or e^^ cells. 



The three hybrids present an interesting cori-elatiou between their 

 hybrid constitution and their relative state of sterility or fertility. 

 Hyhrida nutella inherits the tendencies or dispositions for all the vege- 

 tative characters of both parents. Or if we put it in form of a meta- 

 phor, it inherits all the talents of its parents and attempts to make 

 use of all, but onlv reaches a mean between them. The vegetative 



