488 Harris: FACTORS INFLUENCING THE WEIGHT OF BEAN SEED 
TABLE II 
Series Regression Equation End Points of Line 
ya W =.3522+.0023 P I=.354, “3t'.422 
ib W =.3201+.0012 P I =.330, 27 =.360 
VY =.4167+.0012 P I=.418, 26=.447 
A V =.4298+.0027 P I=.433, 33 =-543 
GGH, V =.4427-+.0022 P I=.445, 17=.479 
GG Vy =.4409+.0010 P I=.442, 15=.456 
V =.3385-+.0079 P %=.3947, 8.402 
GGD2 V =.3609+.0128 P 1=.374, 7.451 
GGDD V =.3670+.0111 P 1=.378, 9 =.456 
H J =.2127-+.0007 P 1=.214, 46 =.246 
HH VY =.2184-+.0008 P I=.219, 67=.271 
HHH V =.1825+.0017 P =.184, 34=.241 
HD V =.1772+.0058 P I=,t83; 14=.258 
HDD V =.2102+.0027 P. I =.213, 16.253 
D V =.1723+.0008 P t=,173, 9=.18 
DD V =.1757+.0065 P I=,182, 14=.2 
DDD V =.1976+.0014 P =,190, 13 =.210 
DH V=.1961-+.0009 P =.107, ° §0 =.239 
DHH V =.1797-+.0020 P =.182, 27 =.232 
USH V =.3182+.0017 P I =.320, 39=.385 
USHH V =.3018-+.0019 P I =.305, 15 =.331 
VY =.2243+.0108 P Y=,235, . 7.300 
USDD W =.2806+.0049 P I =.286, 16=.358 
W =.2001+.0000 P I=,200, 50=.201 
FSHH =.1645-+.0003 P =.165, 34=.-175 
=.1501-+.0032 P =,153, 10=.182 
Loeee oo V =.1632+.0012 P =-164, 14 =.179 
W =Weight in grams, P =Pods per plant. End points of lines give calculated 
mean seed weight in grams for number of pods indicated. 
weight will vary widely because of differences in the variability of 
number of pods per plant. 
The irregularity of the means of seed weight for plants with 
different numbers of pods is attributable to several factors. Note 
the following: 
Taken as a whole the correlations are of a very low order. In 
such cases the empirical means-must be expected to deviate con- 
siderably from the theoretical means as predicted from the re- 
gression formula unless the number of observations be very large - 
indeed. 
Number of pods per plant is an exceedingly variable character. 
In consequence, the means for seed weight must, in experiments 
made under favorable conditions, be based upon a relatively small 
number of seeds, even in series in which the total number of seeds 
weighed is very large. It is impossible, if all series are to be 
treated alike, to increase the number of seeds upon which means 
