III. CURVES OF RAY-FLORETS OF THE COM- 



POSITAE AND OF RAYS OF UMBELS 



IN THE UMBELLIFERAE. 



§ 6. THE OBLITERATION OF THE EFFECT OF SELEC- 

 TION BY NUTRITION. 



We must now inquire whether the conclusions arrived 

 at by our study of the fruit-lengths of Oenothera La- 

 mar ckiana apply to other species and other characters as 

 well. I propose to confine myself to a single case — the 

 operation of nutrition and selection in opposite direc- 

 tions, a case exactl}^ parallel to that studied in the short- 

 fruited race of Oenothera Lamarckiana. The effect of 

 selection is pretty accurately known ; so that the separate 

 effects of selection and nutrition can be directly inferred 

 from the result of such an experiment. It is a question 

 of the effect of selection in a minus direction ; of how a 

 character will behave when we try at the same time to 

 impair it by selection, and improve it by nutrition. 



To make the significance of the point at issue still 

 clearer I will give a short summary of the results de- 

 tailed in this and the two following sections. They 

 show that under the above conditions of experiment the 

 effect of nutrition exceeded that of selection in Anethiim 

 graved ens (§6), that in other cases both had about the 

 same effect, as in Chrysanthemum segetiim. Coreopsis 

 tinctoria, and Bidens grandi flora (§7), and finally that 



