Xo. 51S] 



VARIATION 



increased gametic variability results. Johannsen 10 has 

 shown that there is no such increase in fluctuation when 

 close-pollinated plants are crossed. I have crossed 

 several distinct varieties of maize where the modal num- 

 ber of rows of each parent was twelve, and in every 

 instance the F x progeny had the same mode and about the 

 same variability. 



Finally, a possibility of gametic coupling should be 

 considered. Our common races of flint maize all have a 

 low number of rows, usually eight but sometimes twelve ; 

 dent races have various modes running from twelve to 

 twenty-four rows. When crosses between the two sub- 

 species are made, the tendency is to separate in the same 

 manner. 



Attention is not called to these obscuring factors with 

 the idea that they are universally applicable in the study 

 of supposed continuous variation. But there are similar 

 conditions always present that make analysis of these 

 variations difficult, and the facts given here should serve 

 to prevent premature decision that they do not show 

 scgri'gation in their inheritance. 



Table VI shows the results from several crosses be- 

 tween maize races with different modal values for number 

 of rows. Several interesting points are noticeable. The 

 modal number is always divisible by four. This is also 

 the case with some twenty-five other races that I have 

 examined but which are not shown in the table. I suspect 

 that through the presence of pure units zygotes having 

 a multiple of four rows are formed, while heterozygous 

 units cause the dropping of two rows. The eight-rowed 

 races are pure for that character, the twelve-rowed races 

 vary but little, but the races having a higher number of 



