254 Harris: RELATIONSHIP OF OVULES TO SEEDS 
Here 1o-/E7oz = 3.15. Possibly this value is statistically sig- 
nificant, but considering its extremely small magnitude I think one 
should be cautious in attaching any biological significance to it. 
TABLE VI 
SEEDS ’ 
Ovules I 2 3 | 4 5 6 7 
2 _ 2 a — —w caida oma 2 
3 I 24 65 — _ — 90 
4 2 47 234 384 — sete 667 
5 — 27 427 350 361 — — 865 
6 —_ 12 44 112 174 143 _— 485 
7 as — 4 II 29 27 19 - 
8 _ —~ “ I — — I 
3 II2 474 857 565 170 19 2,200 
TABLE VII 
SEEDS 
Ovules | ° I 2 3 < fe 6 7 | 8 | 9 ; 
3 I 2 23 23 — — |} — j-|- 49 
4 | 38 27 | 106 | 194 162 is Gres eG re Pens Bale 
ee a 41 182 | 320 504 381 —j|—|—| 1,464 
6 | 26 17 86 212 205 420 265 —~{— | 2,gat 
ae Bae 4 25 4I 95 IIo F220 | 74) — 474 
Sere er I Ane a 9 12 24 |18| 6 |— 81 
9 ee — SRT HR oacas I _— I : a ae 4 
_| 87 | 92 | 424 | 798 | 1,066 | 923 | 410 loa! 7 || 3,900_ 
D. Comparison of Constants from Three Series of Cercis 
A detailed comparison of the characters of red bud from various 
regions of the United States or from different habitats falls outside 
the scope of this paper. I will, however, lay the results from the 
total materials of the three side by side for a casual comparison 
in Table VIII. 
It is clear without further arithmetic that many of these con- 
stants differ significantly from series to series, that is to say, the 
difference between them is several times as large as can be attrib- 
uted to the errors of sampling from a homogeneous population. 
This fact does not, however, necessarily indicate that the three 
series are “genetically,” “racially” or ‘“‘genotypically’’ distinct. 
Each general collection is composed of a (relatively) small number 
of trees. These individuals are, as will be shown later, differenti- 
