HYBRIDIZING AS A CAUSE OF VARIATION. 



57 



the differentiation of a single original species into subspecies, races, 

 and varieties, or whether it is in part due to hybridizing between 

 original species, may be questioned. It is also a remarkable illustra- 

 tion of the intricate relationships existing in nature even among plants 

 apparently exclusively self -fertilized. 



In figures 2 to 7, inclusive, with accompanying text and notes, the 

 fact of tlie creation of variation in yield of wheat plants by hybridiz- 

 ing is also illustrated. 



In 1893, from a floret of Blue Stem wheat pollinated from a Fife 

 plant there resulted a seed which in 1894 developed into a plant, No. 

 1814 in figure 19. In 1895 a centgener of plants was grown from the 

 1894 mother plant. Of these, 30 per cent had smooth chaff, resem- 

 bling the Fife parent, and 70 per cent had hairy, velvet}^ chaff, resem- 

 bling the Blue Stem parent. In the succeeding years smooth-chaffed 

 plants were chosen for mother plants from one stock selected for the 

 development of a smooth-chaffed variety, and plants with velvet}^ 



1893 



1894 



1895 



1896 



1897 



1898 



1899 





-Smooth- 



-100- 



-90- 



—80- 

 —7 0- 

 -60- 



^814^ 



E 



"5 0- 



-10- 



-OX)- 



-^20- 



— 10- 



0- 



10- 



-Hairy- 



30- 



-^ — 40- 



^50- 



^60- 



-It 



-87 



—90- 



-10,0- 



FlG. 19.— Graphic expression of the results of an experiment in developing from a single hybrid 

 plant. No. 1814 (produced by crossing a plant of Fife with one of Blue Stem) two varieties, one 

 having smooth and the other hairy chaff. 



chaff were chosen as mother plants fr(mi another stock selected foi- 

 the dev<dopment of a haiiy-chaffed variety. 



In figure 19 it is shown graphically that in tlu^ thii-d generation each 

 was nearly ti-ne to type, and that it remained i)ra(*t ically true to type. 

 Other hyl)rids under exi)eriment are not all reduced to type so rap- 

 idly, as plants vary greatly in their tend(»ncy to continue departing 

 from type; but if liybrid plants can thus easily be reduced to type in 

 regard to larg(;ness of yield, content of nitrogen, etc., and afterwards 

 or simultaneously be in like manner <nisily made uniform in appear- 

 ance, the improvement of wheat by (•.i-6ss bi-eeding will not appear 

 v<M'y (linicult. 



BREEDIN(i CORN. 



Corn breeding illustrates .some of the principles of practice appli- 

 cable to species wh' nature are open-pollinated. Since the male 

 and female flower Darate portions of the plant, the female 



