III. Inheritance in Shirley Poppy 
401 
We liave then the following points which must always be borne in mind and which 
render in our experience the problem of inheritance in plants extremely complex. 
(a) The wide-reaching effect of a changed environment, whether this be a 
change of climate, soil, or season. 
(6) The effect of selection in modifying the intensity of inheritance ; the 
variability of the actual mother plants is for a number of characters far less than 
the variability in the general population. 
(c) The difficulty as to the multiplicity of like organs in the individual and 
the choice as to which of those organs is to represent the individual. 
We propose to consider these points in the following section.s. 
(3) On the Inflaence of Environment on the SJiirleT/ Characters. 
In the first place let us consider the type and variability in number of 
stigmatic bands — a character which we can compare with the previous crops. Of 
course this comparison can only be partial, because in the earlier crops this charac- 
ter was recorded for every flower on the plant, while in 1903-4) only the first or 
apical flowers were dealt with. Further, for the earlier crops the mean and 
variability are those found for the homotyposis tables, while in the 1903-4 crops 
they are taken from the fraternal correlation tables. 
TABLE I. 
Change of Type and Variabiliti/ tuith Environment. 
Crop 
Mean No. of 
Stigmata 
S. D. 
C. of V. 
Parkstono, 1903 
14-03 
1 -562 
11-13 
Enfield I, 1900 . 
13-78 
1-518 
11-02 
Kidderminster II 
1903 ... 
13 -.58 
1-880 
13-84 
Crockham, 1900 . 
13-39 
2-055 
15-35 
Kidderminster I, 
i900 ... 
13-18 
1-702 
12-91 
Bookham, 190(t . 
13-14 
2-330 
17-73 
Crewe, 1904 
13-06 
1-687 
12-92 
Enfield II, 1900 . 
12-72 
1-717 
13-50 
Hami^den, 1899 . 
12-61 
1-885 
14-95 
Chelsea, 1899 . 
12-37 
1 -680 
13-58 
Arncliffe, 1903 . 
12-12 
2-112 
17-43 
Chelsea II, 1900. 
12-04 
1 -694 
14-07 
Highgate, 1900 . 
12-03 
1 -890 
15-71 
Oxford, 1900 
11-39 
2-173 
19-08 
Streatham, 1903 . 
11-39 
1-507 
13-20 
Considering tiiat the probable error in these cases is a small quantity in the 
first or second place of decimals, we reach a very wide range of difference in 
type and variability occurring in different localities and seasons. We see at 
once that plants growing under diversities so great as are evidenced above are 
not directly comparable. 
Without laying too much stress on this table we may remark that the mean 
variation of the first seven entries is 1-82, and of the last eight 1'83. In other 
Biometrika ly 51 
