EXPERIMENTS IN BREEDING SWEET CORN. 275 



ing- the average for the 5, produced 20.68 per cent, of their total 

 shelled corn on ears of Ai cjuality, while the progeny of the good 

 ears of Fig. 226 produced 20.40 per cent, of their total shelled 

 corn on such ears. In other words, the progeny of the ears of 

 Fig. 226 included on the average no more ears of Unc quality 

 than did the progeny of the ears of Fig. 22/. 



How are such results as these to be interpreted? It is, of 

 course, open to one to maintain if he chooses that the cases illus- 

 trated are merely isolated instances — exceptions which prove 

 the rule as it were. As a matter of fact these cases are not 

 exceptions. They agree essentially with the great bulk of data 

 regarding inheritance accumulated by the experimental work of 

 recent years in supporting the generalization that the external, 

 visible characteristics of a plant or animal furnish an exceed- 

 ingly unreliable criterion of its probable behavior in breeding. 

 The fact that an ear of corn is of especially good type and 

 appearance is no guarantee that its progeny will also be better 

 than the average. These corn ears illustrate exactly the same 

 principles that have been brought out in the studies on breeding 

 for egg production at this Station.* The force of such facts as 

 are here set forth and the general principles upon which they 

 depend, is making itself felt in practical corn breeding work in 

 the western states. The writers gather from their reading of 

 such papers as the Breeders Gazette (and they are informed by 

 competent authority that their conclusion on this point is quite 

 correct) that among the most careful and thoughtful of the corn 

 belt farmers and breeders of seed corn there is developing a 

 marked reaction from the belief in the great value of the "show" 

 type of ears for seed raising purposes. It is being found by corn 

 breeders everywhere that such ears do not always or even in the 

 majority of cases produce the highest yields or the largest pro- 

 portion of perfect ears. 



Facts of the character brought out by these protographs of 

 corn ears are capable of satisfactory interpretation on the basis 

 of Johannsen's ** concept of genotypes. According to this view 

 the ears shown in Fig. 22/ represent very poor indivi(Uial spcci- 



* Cf. Me. Agr. Expt. Stat. Bulletin 166 and Bureau of .A.nimal Ind. 

 Bulletin no, Part I. 



** Cf. this author's recent hook "Elemente der e.xakten Erblichkeits- 

 lehrc." Jena (Fischer), 1909. 



