434 BOTANICAL GAZETTE [JUNE 
the old leaf nature may be found in its entire ontogeny. The tube of 
gamopetalous flowers is another structure which in some cases has 
attained to approximate independence, and the style and stigma yet 
another. All of these parts may become centers of independent 
(though usually limited) metamorphosis, altering in size, shape, num- 
ber, building new structures (as the corona) all independently of any-: 
thing they did before they acquired their present more or less 
completely independent membership. It is so with the ovule, orig!- 
nating in a macrosporangium, but now an independent member. The 
same is true of the embryo-sac, which, originating unquestionably as a 
germinated macrospore, has, after a long history, become a new member 
with a high, though not acomplete independence. Some of its features. 
are to be explained as a persistence of its ancient nature, but it is use- 
less to attempt to interpret all its actions upon that basis, for it does 
some new things upon its own account—as a new member, the 
embryo-sac. To return for a moment to the flower as a whole, it is 
important to notice that the study of its morphology is in one sense 
complex and difficult, partly because it is a composite structure with 
various degrees of independence in its component parts, partly because 
its development in different families has been so independent that it 
has given rise to homoblastic rather than homogenetic homologies. 
Hence it is impossible to make distinct categories of members apply- 
ing to all flowers, but each group must be considered by itself, a feature 
indeed which applies not to the flower alone but to the vegetative parts 
as well. Hence we must in theory recognize as potential members all 
organs, one may even say all recognizable parts, although in practice 
it is needful to take account for the most part only of those most con- 
spicuous and distinct. 
The realistic system, with its infinite gradations and limitless pos- 
sibilities, is much harder to grasp and to apply, and is less pleasing to 
teach than ‘the idealistic system, with its few distinct categories and 
their involved limitations. But it is truer to nature, more stimulating 
to research, and more replete with promise of great results.— W. F. 
GANONG, Smith College, Northampton, Mass. 


