436 



The Journal of Heredity 



althoujih kafir seeds are specked with 

 brown and the seeds of feterita are 

 often marked with brown beneath the 

 glumes. It is usually thouj^ht that the 

 brown color on feterita seeds arises 

 from a discoloration of the seed 1)>' its 

 contact with the j^Iinne or hull. There 

 may, howe\er, be some color in the 

 seed itself. 



Segregation in the F2 generation 

 showed many widely diverse types 

 from dwarfs about two feet tall to plants 

 over six feet tall; the population varied 

 also in leafiness, shai)e, and compact- 

 ness of the panicle, size and color of 

 the seeds, juiciness and sweetness of 

 the stems, and many other characters. 

 The several types selected for further 

 trial were intermediate in most charac- 

 ters to the two parents resembling in 

 many wa>s Dwarf hegari. (See Fig. 1.) 

 In Table II is given in parallel columns 

 a comparison of the F7 of one of these 

 hybrids with its two parents. Most of 

 the t>pes selected for increase and plat 

 trials have bred true from the third 

 generation. (Frontispiece.) 



The general practice which is now 

 being followed in testing these hybrids 

 at Chillicothe, Tex., is as follows: 



First generation. — The cross-polli- 

 nated seeds are planted in rows and all 



the heads of the Fi plants are bagged to 

 prevent cross-pollination. 



Second generation. — All the seeds pro- 

 duced by the Fi plants are sown in rows. 

 As these Fo plants mature selections 

 representing the most promising t>pes 

 are matle and the heads bagged for 

 continued testing in c(jmparison with 

 the parents and otiier commercial 

 varieties. 



Third generation. — Head to row seed- 

 ings of the F2 selections are made, each 

 row being duplicated at least three 

 times with rows of the parent varieties 

 included as checks. From the rows of 

 hybrids, selections are made of the 

 best plants and the heads bagged. If 

 some rows appear to be particularly 

 promising a sufficient number of heads 

 are bagged so that a plat test of such 

 strains can be made the following year. 

 The other selections made are con- 

 tinued in row tests. In the third 

 generation the row is judged as a unit 

 if it shows decided uniformity in its 

 general characters. 



Fourth generation. — The most prom- 

 ising strains from the Fs rows are seeded 

 in duplicate plats which are protected 

 from cross-pollination with other sor- 

 ghinu varieties by surrounding the plat 

 with rows of some tall growing corn 



Taui.f. li: ( O.Mi'AkisoN of a Seventh CiESEKATioN Selected Hvhrii) with the Parent 



Types 



• Exsertion of the head in feterita is often due to the bending of the peduncle away from 

 the leaf sheath as in milo rather than extension beyond the leaf sheath. 



