HARRIS: RELATIONSHIP OF OVULES TO SEEDS 245 
The ovules are about fifteen one-hundredths more numerous in 
the sample collected from the few large trees. This is certainly 
not a difference which would have been detected by other than 
biometric methods, and some might consider it neglible, but it is 
slightly over eleven times its probable error and so unquestionably 
. TABLE I 
SEEDS 
Ovules I 2 3 | 4 5 é 7 
* 3 a a — _ eee — 3 
: nb 39 te vit wee Saas aa 49 
3 mW Iog 203 — — —_— — 349 
4 61 292 701 933 — — 1,987 
5 32 202 472 887 830 | _ 2,423 
6 19 40 106 223 260 | aon 29 1,042 
7 5 i 17 | 3 140 
8 pe hc See — 2 | 2 7 
oe 162 687 1,489 2,062 Harr aey a 6,000 
TABLE II 
SEEDS 
Ovules z | 2 | 3 4 5 6 7 
. rt eee — paid Guan eat ng = 4 
4 14 8 — — | — fo me 103 
3 29 | 147 254 gene re Eh ace ae 430 
as 34 160 435 678 | | — — | 1,307 
> 17 105 207 485 568 tee —_ 1,382 
6 4 19 46 122 210 | 277 — 677 
7 — 3 2 £40 o:| 17 | 26 36 95 
8 _ _ — — to I — 2 
tor | 523 | 944 1,296 | 796 | 304 36 4,000 
TABLE III 
, PuysicaL CONSTANTS FOR RANDOM SAMPLES 
4 
Constants 6,000 Pods 4,000 Pods | 10,000 Pods 
Mean WOE oe ia tes 4.7498 = .0080 | 5952 +.0112 | 4.6880 + .0062 
Standard deviatiol of ovules. 9274 *.0057 | 1.0532 +.0079 | .9826.0046 
— } 
4 regs of variation of | wee 
Seas He cee! | Ae Rangers Uae .7786 =.0102 | 3.8047 +.0131 | 3.7891 +.0081 
Standard deviation of seeds...| 1.1781 +.0072 | 1.2377.0003 1.2024 .0057 
Coefficient of variation of seeds 31.1792 } 32.5339 | 31.7337 
Cotrelation ovules and seeds .|  .5763 =.0059 | -6782 =.0057 6133 =.0042 
eee tee ent 
