THE ORIGIN OF ANGIOSPERMS. 37 
been used in a great variety of senses *. We would, however, restrict its 
application to the Angiosperms alone, since it was from these plants that the 
idea expressed by the word originally arose. In our opinion extra-seminal 
pollination, in which the carpel or carpels play the chief part in the pollen 
collection, is the essential feature of a hermaphrodite, or female flower. We 
regard a flower as typical when it possesses both micro- and megasporangia, 
as well as a perianth which in many cases has an attractive function. 
А flower, on our view, is a special form of a type of strobilus, which is 
common both to the Angiosperms and to certain Mesozoic plants, and which 
may be termed an anthostrobilus. The anthostrobilus of hypothetical Mesozoic 
ancestors of the Angiosperms, and of their supposed near relatives the Ben- 
nettitese, differed from the flower of the Angiosperm in certain important 
respects, especially in the presence of direct pollination, in which the mega- 
sporophyll played no part. Tt may perhaps be useful to distinguish it as a 
Pro-anthostrobilus, and the Flower proper, a term here restricted, as an 
Eu-anthostrobilus. 
Pro-anthostrobilus of Mesozoic Ancestors and Bennettitez. 
Anthostrobilus 
Eu-anthostrobilus (Flower) of Angiospermez. 
The necessity for these new terms arises from the fact that the word “ flower ” 
has been applied in many different senses, for instance even to the strobili of the 
Coniferales. Also because, as we hope to show here, the Angiosperms are 
descended from Mesozoic ancestors nearly related to a group of fossil plants, 
whose fructification is now well known, and indeed has been, though as we 
think inaccurately, termed a flower. 
We shall discuss at some length at a later stage the evidence for the 
derivation of the Eu-anthostrobilus from the Pro-anthostrobilus—types of 
cones which we believe represent different stages in the evolution of the 
fructification of one and the same line of descent. 
The strobilus or cone is of course a very ancient type of fructification, 
common to many distinct, and only very remotely related lines of descent. 
Other forms of strobili were borne by plants which flourished at a very much 
earlier period in geological time than the anthostrobilate races discussed here. 
The anthostrobilus is distinct from any of these, and it is, in all probability, the 
newest modification or creation of the strobilate form of fructification, in point 
of geological time. Tt differs from all other strobili in that it is typically 
amphisporangiate, by the megasporophylls being invariably aggregated on the 
axis of the strobilus above the microsporophylls (i. e. nearer the apex of the 
cone), and by the presence of a distinct perianth, below the fertile sporophylls, 
whose function is apparently wholly, or partly, of a protective nature. In 
* Coulter & Chamberlain (1904) p. 9. 
