254 Co)itribf(tio)i to a Statistical Study of the Grucifer^m 
intimately correlated with tlie male reproductive organs than is the corolla which 
serves as an attraction for insects as well as a protective covering of the bud. 
(3) The two whorls of the calyx are not highly correlated, i.e. they vary inde- 
pendently of one another. 
(■i) The two whorls of the androecinm also are not highly correlated. Morpho- 
logically this means that when there are two constituents in one organ, each having 
the same function, they may vary independently of one another, so that although 
an increase or decrease in tVie number in either may be correlated with an increase 
or decrease in the number in any other constituent of the flower, the same does 
not hold true with regard to the two constituents. 
(5) The outer whorl of the calyx is most highly correlated with the corolla, 
next with the outer whorl of the androecium and lastly with the inner whorl of 
the androecium. The reason why the outer whorl of the calyx is more highly 
correlated with the outer whorl than with the inner whorl of the androecium is 
not at first sight very evident, but may be explained on the basis of its protective 
power. The members of the outer whorl of the androecium lie in a plane parallel 
to that of the outer whorl of the calyx, and are much more widely separated in 
this plane than are the members of the inner whorl of the androecium. Con- 
sequently any increase in the number of stamens in the outer whorl would involve 
a much greater inci'ease in volume within the flower than a corresponding increase 
in the number of stamens in the inner whorl. Thus we are not surprised to find 
that such an increase in the outer whorl of the androecium is more intimately 
associated with an inci-ease in the outer whorl of the calyx than a corresponding 
increase in the inner whorl of the androecium would be. 
(fi) There is very low variability in the inner whorl of the calyx and it is 
almost equally correlated to the two whorls of the andrcecium. The morphological 
explanation of these facts follows as a corollary to that given above. 
(7) The calyx is much more highly correlated with the andrcecium as a whole 
and with its two whorls than is the corolla. 
As we have already said the calyx is the predominantly protective organ and 
consequently this higher correlation has a physical basis. The corolla being partly 
attractive does not enter so closely into space economics. 
(8) The outer whorl of tlie androecium is more highly correlated with the 
other components of the flower than is the inner whorl of the androecium. This 
again follows on the basis of space considerations. Any increase in the number 
of members in the inner whorl uf the androecium does not involve so radical a 
change in the volume of the flower as does a corresponding increase in the outer 
whorl of the androecium. 
