A quantitative Study of Variation, Natural and Induced, ete. 215 
this subject of food supply as it may influence variability. Jennings!) 
states that in Paramecium the characters do not all respond in the 
same way, some showing an increase and some showing a decrease 
in variability as the nutrition is changed. Reinohl2) found in studying 
the effects of nutrition on the variability of the number of stamens 
of Stellaria media that the increase of the food supply gave a larger 
index of variability. In other words, the plants which had the excess 
food supply were more variable than those poorly fed. Davenport?) 
after conducting extensive experiments with corn (Zea Mays) concludes 
that “variability is not markedly increased by fertility.” Love‘) 
in studies with peas, buckwheat, asters, and corn under three con- 
ditions of food supply: namely, sand untreated soil, and heavily 
manured soil, finds that the increase in food supply has increased 
the variability in the great majority of cases under observation. 
We would expect that as variability increases the correlation 
between characters would decrease. We find that this relation exists 
in the Silene plants studied. This is shown by referring to tables 22 
to 27 inclusive. The following figures represent a summary of the 
correlation constants for these tables. 
250—11U. . .Z Height subject Width relative r is 590 
250—11a FF 3; er x r is 384 
250—I1IU...Z “ ki Branches ,, T is 447 
250—11a ‘ ry > u T is 311 
BO DINU i 3) 3 Z, a oe Seed Pods relative r is 539 
250—I1a en = ER BS r is 225 
We see here as stated above that as the variability increased 
the coefficient of correlation decreased, or in other words, that the 
garden plot (good soil) gave less variability and greater correlation 
of characters than did the plot on the Mitchell farm (poor soil). 
It is of course recognized that in this Silene experiment reported, 
the difference between the Plant Breeding Garden and the Mitchell 
Farm plots does not rest alone in a difference in food supply. It 
includes also the points of temperature, moisture, physical condition, 
etc., etc. We know however that under one set of conditions the 
1) Jennings, H.S. Heredity, Variation, and Evolution in Protozoa. II. Proc. 
Amer. Phil. Soc. Vol. 47:479. 1908. 
2) Reinohl, F. Die Variation im Androceum der Stellaria media. Bot. Zeit. 
1903 page 159. 
3) Davenport, E. and Rietz, H.L. Type and Variation in Corn, Bul. Ill. Exp. 
Sta. No. 119. 
4) Love, H. H. Studies in Variation. Doctorate thesis, Cornell University. 1909. 
