THE ORIGIN OF ANGIOSPERMS. 41 
The Megasporophylls and Megasporangia. 
The dietum of Goethe that the carpel is a fertile leaf, more or less modified, 
has stood unshaken, and there appear to be such ample evidences of its truth 
that it need not be considered further here. Among the Angiosperms, the 
condition of аросагру seems to us to be primitive. The spiral arrangement 
of the monocarpellary ovaries on a long receptacle, a state of affairs which 
still survives in the Magnoliaceze, and in certain other members of the Ranales, 
may be regarded as a primitive feature of the flower. From this we derive, 
by suppression of the internodes, the whorled arrangement so characteristic 
of the great majority of Angiosperms, and often common to all parts of the 
strobilus. We regard the verticillate grouping as due partly to a tendency 
to cohesion and adhesion, which has always been marked among the Angio- 
sperms, and partly to a proneness to а dissimilarity in the size and shape 
of the different organs of the strobilus. The fact that the protective function 
is, in this type of cone, relegated to sterile members at the base may also have 
had some bearing on the question ; more efficient protection being perhaps 
afforded where the axis is reduced in length, and the various organs arranged 
in whorls. The result has thus been a tendency to a horizontal rather than 
a vertical distribution of the organs. 
There are numerous instances shown by many families, e. g., Ranunculacee, 
Crassulaceze, and Rosaceæ, of how syncarpy has arisen from apocarpy. In 
the great majority of the Angiosperms there has been a distinct bias in this 
direction, with various modifications, the significance of which is to be sought 
for in the fruits. Bi- and multicarpellary ovaries have been the result. 
We regard the carpel as а megasporophyll, present in the ancestor of the 
Angiosperms as an open leaf, bearing several ovules on its margins, and 
not unlike the megasporophyll of Cycas. With the shifting of the pollen 
collection from the seed itself to the carpel, it became possible for the 
latter, both to afford more efficient protection to the developing seeds, 
by eompletely closing over them, and also, at the same time, to fulfil its new 
duties as а pollen receiver, by adopting some mechanism for the purpose at 
the apex. The necessity for some protection for the ovule is well seen in 
Bennettites, where, however, it is effected in a totally different manner. The 
style, probably non-existent at first, may be looked upon as a later adaptation, 
connected with the perfection of the method of insuring cross-fertilisation. 
The stigmatic surface was, in the early stages, simply a localised portion of 
the carpel, adapted possibly by some sticky secretion for the collection of 
pollen. 
It need hardly be mentioned that we are in agreement with Bessey * 
and other recent writers, in deriving all syncarpous ovaries from apocarpous 
* Bessey (1897). 
