Sorting and blending of unit characters. 235 



may account for a small measure of what appear to be fluctuating 

 variations in other characters. 



4 th. The zyg-otaxis theory. This interesting theory was pro- 

 posed as a provisional hj^pothesis by Swingle in 1898, and further 

 elaborated by him in 1913, for the different types of hybrids appearing 

 in the first generation of interspecific crosses. It rests on the assump- 

 tion of the chance arrangement of the parental chromosomes in different 

 positions in different zygotes, at the time of their fusion, or association, 

 this being the period when a rearrangement would be more likely to 

 occur. According to this hypothesis the chromosomes situated near the 

 cytoplasm would be better nourished and would because of their position 

 and better fed condition exercise a greater influence on the formative 

 processes of the cell than those chromosomes more distant. Thus in 

 some zygotes chromosomes representing certain factors or qualities would 

 lie in a more favored position. In other zygotes chromosomes represent- 

 ing other qualities or factors would be the favored ones. The arrange- 

 ment of the chromosomes which is assumed in any given zygote is 

 maintained during the ontogeny, and thus different hybrid types arise 

 according to the different chromatin plans determined in the zj'gote. 

 The association and configuration of the sperm and egg chromosome 

 complex in the zygote Swingle (1913) terms, zygotaxis. 



If we now apply tliis hypothesis in explanation of the Fi hybrids 

 obtained in the reciprocal crosses between Oe. pijcnocarpa and Oe. nutans, 

 it will emphasize their peculiar composition and behavior in contrast 

 with the usual first generation hybrids. In the case of the blend, 

 hyhrida nuiella, which is a true intermediate hybrid so far as the vege- 

 tative characters are concerned, all the homologous contrasting chromo- 

 somes representing these characters would be paired and all have an 

 equally favored relation to the cytoplasm. Their arrangement would 

 correspond to that of "interspecific hybrids" resulting from "violent 

 cross breeding' (Swingle, 391, 392, 1913). The hyhrida nutella is self 

 sterile in accordance with the expectation in such crosses, but the self 

 sterility is not due to "incomplete synapsis'" since the pollen as well as 

 the egg cell is efficient in other combinations, and the entire repro- 

 ductive mechanism is unimpaired so far as one can determine. 



In hyhrida pycnella the chromosomes representing the broadness 

 and toothedness of the rosette leaves of nutans would occupy the favored 

 position in the zygote, while those representing the crinkledness and 

 convexity of the nutans rosette leaves would occupy an unfavorable 



