1912] HARRIS—STAPHYLEA 401 
3. THE HYPOTHESIS OF THE MUTUAL DEPENDENCE OF FRUIT 
LENGTH AND FERTILITY UPON OTHER CHARACTERS 
OF THE PLANT 
Two characters both positively or negatively correlated with a 
third are correlated with each other. This is doubtless the source 
of much of the correlation found between organs. Possibly 7, and 
rj; indicate no physiological interdependence between / and o or 
/ and s, but are due merely to /, 0, and s being influenced by the 
position of the fruit on the inflorescence or by the number of pods 
per inflorescence. The correlation between the length of the pod 
and the distance of the flower-bearing node at which it is produced 
from the base of the inflorescence? is shown for the 1906 series? in 
table XIV, 1906, and for number of pods developing per inflores- 
cence and length of pod in table XV, 1906. These give: 
Position and length, 7,,= —0.0580+0.0148 
umber and length, r,;= —o.1828+0.0144 
Statistically these values are certainly trustworthy, for r)/£,= 
3.91, and 7,,/E,=12.70. They indicate that both a more distal 
position on the inflorescence and a number of fruits above the 
normal is prejudicial to the maximum development of the fruit.‘ 
If it be found that the number of seeds developing is also nega- 
tively correlated with the position and with the number of the 
TABLE XIV, 1906 
LENGTH OF FRUIT 
Position 3 4 5 6 7 8 9 10 It Totals 
i EO I 65 192 | 322 | 259| 125 62 {| 20 I 1047 
Pic. 4 50 167 | 197 | 143 31 9 689 
mea e ee r 25 69 75 54 31 14 2 271 
Be ae 2 9 13 6 4 5 t 4° 
Stein, 3 .* 3 
Totals 6 142 437 | 610 | 462| 248 | 112; 32 I 2050 
: ie serial position of the node on the inflorescence counting from the base. 
a for position are not available for 1907. Because of a severe frost many of 
the inflorescences produced only a single fruit, and it seems idle to study the number 
per inflor 
4 pean with larger numbers of fruits will have more distally placed 
fruits. I have not worked out a — between these two characters, - “tf it 
is considerable the two constants just discussed are not independent, and probably 
one or both should be, aserally panteinry more nearly o. 
\ 
