216 Contribution to a Statistical Study of the Cruciferm 
ground in his garden, and the results of these observations form the basis of the 
present contribution to the study of the variation in the Cruciferge. 
Botanical problems, which have been hitherto attacked from the biometric 
standpoint, have been comparatively easily handled, because the material has 
been more or less homogeneous in character. For example, variations in the 
number of sepals of Anemone nemorosa* or in the number of ray-florets of 
Ghrysanthemum leucanthenmm-f , and the consequent distribution of these are 
capable of direct treatment by Pearson's well-known method of fitting frequency 
curves. 
The only work comparable to the one in hand occurs in Biometrika, Vol. II. 
p. 145 (Variation and Correlation in the Lesser Celandine), but in this case the 
numbers of member's in the calyx, corolla and androecium have been examined 
as a basis for a study of homotypic correlation and in this flower each of these 
organs consists of a single constituent with numerous members. 
The problems studied in this paper, however, are more complex inasmuch as 
they deal not with one organ of the flower but with all the organs, their con- 
stituents and members both separately and collectively. 
It is also, I believe, the first biometric work of its kind on a cruciferous flower 
and embodies a study of chorisis, that is, " the splitting up or division of one or 
more components of a flower into two or more equal or unequal parts '"' — a factor 
which is supposed to have been of the utmost importance in the evolution of the 
natural order — Cruciferse. A complete discussion of this phenomenon is reserved 
until the flower is studied in detail. 
It would be well here to emphasise the fact that the flowers examined for this 
study were not taken from different plants but, on the contrary, were obtained 
from several inflorescences growing on stems which had arisen from buds on the 
roots of a single parent plant. This mode of reproduction is rather unusual, but, 
in the present instance, is of particular interest inasmuch as it gives greater 
homogeneity to the material. 
The parts of the flower which have been considered are (a) the perianth, which 
consists of (1) the calyx and (2) the corolla, (h) the androecium and (c) the 
gynaecium. 
The functional differentiation of these organs is of great importance in the 
interpretation of results so that it might be well to recall the particular rdles 
which these play in plant economics. 
The gjma^cium and the androecium are respectively the female and male organs 
of reproduction and consist of carpels and stamens, while the perianth forms a 
protective covering for these delicate structures. The calyx or outer organ of 
* Yule, Biometrika, Vol. i. p. 307. 
t Biometrika, Vol. ii. p. 309 et seq. 
