IV" iio.nkkhk.nck iNïKii.wrutwi.i; m-: (.km-tiuie. 



THE RESULT OF SELECTING FLUCTUATING VARIATIONS 

 DATA FROM THE ILLINOIS CORN BREEDING EXPERIMENTS ' 



By 1)' Frank M. SURFACE 



lliolosisl, Kr/nliicky AgriciilUual Ex|ieiiiiifMl St^aioii. 



'l'hero ure i'ew liiologiral piiiljlciiis wliifli arc altractiiifi: more aUeiilioii al 

 liie présent timc Ihan Ihat regarding llie elTect of selecting small, lUiclualintf 

 variations. Until recently the efïecliveness of such sélection lias Lcen acco])lcti 

 almost witiiout question. However, tlie récent worl; of Johannsen, Jcnnings 

 and others lias caused many biologists to entertain serious doubts as to llio 

 real evolutionary significance of so-called lluctuating variations. If onc 

 searches through Ihe literature for clear eut cases in which plants or animais 

 liave been modified by Ihe graduai accumulation of small variations bo is 

 surprised at the small number wbicli are supported by adéquate data. Of 

 thèse few cases there is one which has been referred to frequenlly as a classic 

 example, of what can be done by simple sélection. I refer to the work of 

 breeding corn (maize) for chemical constitution carried out by the Illinois 

 Agricultural Experiment Station. 



Over two years ago while engaged in some corn breeding work at the 

 Maine Experiment Station, the writer had occasion to work out the pedigree 

 tables which are given in the présent paper. (cf. pp. 'i25-!2t.'9). ^^'hen dis- 

 played in this way, the results of thèse extensive experiments appeared sugges- 

 tive of the results which actually come from an attempt to seleci fluctualing 

 variations. The writer is well aware that no definite conclusions of far reacliing 

 importance can be drawn from this data alone. It was for this reason and willi 

 the hope of accumulating more delinite data Ihat publication has been delayed 

 so long. In view of the intcrcst manifested in this subject al the présent 

 lime il has seemed worth whilc to publish thèse tables together wilh a brief 

 discus'^ion. 



In 190s Ihc Illinois Experiment Station pulilishetl^ the delailed évidence of 

 very careful and long continued experiments in selecting corn wilh référence 

 lo the chemical constitution of the grain. Four definite experiments were 

 carried oui simultaneously, viz : (1) Selecting to increase the protein content; 

 (2) Selecting to decrease the protein content; (li) To increase Ihe oil content 

 and (i) to decrease the oil content. For the détails of thèse experiments the 

 readcr inusl be referred to the bulletin mentioned and others by the same 

 Station. For Ihe présent discussion il will be sufficieni to mention only a few 

 of the more important points regarding the melhods used in thèse experiments. 



After some preliminary work there were selected in 1896 one hundred and 

 sixly-lhrec ears from a standard variety of white dent corn. Chemical analyses 

 were niadi' from saniples of Ihese ears, sbowing the protein and oil conlenl of 

 each. From thèse one iuuidred and sixiv-lhree cars the twenlv-four showiiii;- 



I. Co llllli(\lll<>Jl r.lilr ;i l;l m'c !,■ scillKV ili' la CuillV'lvilcc. 



-1. Ton tieiierations uf Coi-n Crofdint;, by l.uuis 11. Smitu. ///. Aijr. Expci: Slaliun ISiilL, >u 1-28. 



