138 Sargant . — The Reconstruction of a Race of 
foliage of the leafy plant was Angiospermous : it bore flowers and seeds : 
its ovules were enclosed in carpels. The male gametophyte was represented 
by three nuclei within the pollen-tube. The female gametophyte before 
fertilization consisted of six cells and two free nuclei, orientated in a very 
characteristic way within the embryo-sac. The endosperm was formed 
after fertilization and arose from a fusion of three nuclei — two of them 
already free in the embryo-sac, and the third a generative nucleus from the 
fertilizing pollen-tube. All the nuclei of the endosperm must in consequence 
have possessed a number of chromosomes in excess of the sporophytic 
number. 
These are for the most part very definite characters, but one of them 
opens out a question of great importance. What sort of flower was borne 
by the Primitive Angiosperm ? 
The classification of living Angiosperms is largely based on the external 
morphology of their flowers. The mass of observations on this subject ac- 
cumulated by taxonomists represents the labour of many generations. As the 
natural system is now avowedly based on the principles of evolution, its divi- 
sions claim to represent degrees of kinship. The form of flower which the 
elder systematists would have called typical of an order, for example, is now 
understood to represent — at least in its main features — the flower of the 
ancestral stock from which all the genera included in that order are de- 
scended. The question of descent underlies all the problems of systematic 
botany. 
Thus the evidence of taxonomists on the question of the primitive 
form of flower is by far the most weighty to which we can appeal. If they 
were substantially agreed on the form of flower which their comparative 
researches showed to be the most primitive, their collective authority would 
be overwhelming. But since they are divided, their conflicting opinions 
may be tested by those results of systematic botany which are the common 
property of all botanists, as well as by the knowledge derived from other 
sources. 
Messrs. Arber and Parkin have undertaken this task (4). They agree 
with the elder systematists who considered the primitive flower as an 
elaborate structure, its numerous tepals, stamens, and carpels arranged 
spirally on the axis in acropetal succession. Such flowers occur in living 
species — for example in the Magnolia and other members of the Ranales. 
According to this view, all other floral types are derived by reduction from 
this. 
The alternative view is held by most modern taxonomists, but not 
by all (Arber and Parkin, pp. 38, 39). It is agreed that a large number — 
perhaps the majority — of simple unisexual flowers are derived by reduction 
from more complex bisexual forms. But these complex floral types are 
supposed to be themselves built up from unisexual ancestors, and some 
