124 Sargant . — The Reconstruction of a Race of 
That such forms were once present is inferred from the sudden appearance 
of certain characters in this group. Angiosperms possess structures which 
are unique : for instance, the flower, the carpel, the endosperm. These mem- 
bers are found in Monocotyledons and Dicotyledons alike, but are absent 
from all other groups. The history of their origin has not yet been reco- 
vered : it is not even clear what members represent them in the lower plants. 
The flower, the carpel, and the endosperm then have been evolved by 
Angiosperms and their ancestors since the epoch when they were repre- 
sented by a race of Pteridosperms. The question is, whether these members 
appeared in one line of descent or several. Monocotyledons and Dico- 
tyledons must have had a common ancestor at some time, but was that 
ancestor in essentials an Angiosperm ? Did it — to take the simplest 
test — possess a true carpel ? 
The problem so stated is reduced to its simplest terms. We suppose 
that Angiosperms were derived from two races only, one of them giving rise 
to Monocotyledons, the other to Dicotyledons. We agree to consider for 
the moment one character only among those peculiar to the two classes, 
neglecting not only the other two unique characters, but also the numerous 
minor features which they have in common. 
If the common ancestor X did not possess true carpels, then carpels 
must have been evolved independently by Monocotyledons and Dico- 
tyledons. Naked seeds were transmitted by X to its branches V and Z. In 
the course of generations the seeds of both races have become enclosed in 
a carpel. There is no evidence to show how this occurred. The sporophyll 
might gradually close round one or more ovules, or some sort of cupule 
(Oliver, 62 ) might become a carpel. But though the sporophyll or the 
cupule is inherited from X , each step in its transformation must have been 
taken independently by Y and Z, in response no doubt to the demands of 
the environment. 
In this way each race might well acquire a carpellary organ, but 
the chances are certainly much against the evolution of precisely similar 
members in both. The Angiospermous carpel combines several functions. 
It protects the ovule, and also collects pollen-grains and directs the course 
of their tubes. But in other groups these functions are not always per- 
formed by the same structure. In the Bennettiteae the ovules are sheltered 
by peltate scales, which serve as a very efficient protection to them and also 
to the seeds. The pollen, however, is still collected by the integument of 
the ovule, and germinates on the micropyle. 
If in response to their surroundings both races developed organs which 
would at once protect the ovule and secure its pollination, we should still 
expect to find some constant difference in the carpels of the two races, due 
to their distinct race-history. But no such difference exists. There are 
indeed single carpels and carpels united into a pistil ; superior and inferior 
