III. Inheritance in Shirley Poppy 
415 
The followiug table gives a summary of the results hitherto reached: we have 
included the Shirley as if the sister plants were true half-siblings : 
Half-Sibling Resevihlance. 
Thoroughbred Horse, through dam (3 series) "36 
Shorthoi-n Cattle, through sire and dam (6 series) ... '32 
Basset Hound, through dam (1 series) '22 
Shirley Poppy, Co-ovarial Plants. Colour (5 scries) ... '38 
„ ,, „ All characters (32 series) '24 
In our opinion the horse and the shorthorn are the most reliable of these series, and 
in both cases the values are sensibly larger than half the whole sibling values. 
The Shirley when we deal with all the characters is fairly close to 'io and the 
Basset Hound value. But the colour resemblance in the Shirley, which we 
consider least affected by homotyposis and selection, is in closer agreement to what 
we find for horses and cattle. VVe feel therefore that differential local environment, 
maternal and offspring selection, varying personal equation in estimating plant 
characters, and the complete absence at present of any quantitative allowance for 
positional differentiation and homotyposis, prevent us from saying more than that 
the results so far obtained for fraternal correlation in Shirleys are not inconsistent 
with those found for animals. 
Much enlightenment would be reached could we deal with definite pollen 
parents in very large numbers. Further what is badly needed for our general 
guidance at the present time is a measure of the relative intensity of whole and 
half-sibling resemblance in some quantitatively determinable character. Possibly 
the record of the cephalic indices of whole and half brothers in public primary 
schools would be a comparatively easy method of reaching a definite result. 
(8) Oeneral Conclusions. 
(a) We consider that in colour distribution and tint as well as in other 
characters there is continuity in the variation of the Shirley Poppy. 
(6) The Shirley Poppy under ordinary conditions of environment and season 
is cross-fertilised and probably there is a very considerable variety of pollen parent 
to sister plants grown from the seed of one capsule. 
(c) The very great influence of environment upon nearly all the characters 
dealt with, the variation of character within the same plants, and the unknown 
allowance to be made for homotyposis, even when we confine our attention to 
apical flowers, render the problem of plant heredity of very great complexity. 
(d) As suitable measurable characters in a plant are extremely few, one 
is compelled to select qualitative characters ; and their description, especially when 
up to the present no scale of patterns has been available, is difficult, and liable to 
considerable modification by the influence of personal equation. 
The experience of the present series of observations by no means accords with 
that of those observers who find it easy to formulate simple alternative categories 
