42 MESSRS, NEWELL ARBER AND J. PARKIN ON 
the presence of antipodals, or the time of endosperm formation, from that of 
other Dicotyledons, as Treub supposed. It has been further suggested, within 
the last year, on embryological evidence, that Casuarina is closely related to 
Carpinus, and may be placed within the Betulaceze as а group of equivalent 
rank with the Corylez *. 
Consequently, on such arguments as have been brought forward, we fail 
to find conclusive evidence that the flower of Casuarina is essentially of a 
primitive nature. Nor does the position of the group now appear to be isolated. 
Fagales.—The perianth of the female flower, when present, is superior and 
somewhat gamophyllous. These are not likely to be the characters of a primitive 
perianth. Besides, the syncarpous inferior ovary does away with any idea of 
primitiveness as regards the whole flower. The inflorescence is also especially 
complicated. Similar considerations apply equally to the Juglandales. 
We are inclined to adopt Hallier^s } view, that this group can be connected 
with the Hamamelidacez, and thus with ancestors possessing hermaphrodite 
flowers and biseriate perianths. On this supposition, the perianth of the 
Fagales may probably be regarded as a survival of the calyx, the corolla 
having disappeared completely. 
Monocotyledons. 
Among the Monocotyledons we find certain genera which may very possibly 
be ancient types, without, or with a only very insignificant perianth. Hence 
the question arises whether these plants were originally without such an 
envelope. 
Pandanales.—To this group, regarded as among the most primitive by 
Engler f, and also by Coulter and Chamberlain $, we think the same arguments 
apply as in the case of the Piperales and Amentiferous families. The 
inflorescence is of a very dense and sharply defined nature. In the Pan- 
danaceæ the individual flowers are difficult to make out, bracts and 
bracteoles being absent. It seems much more probable that in this case the 
perianth of the individual flowers, as well as the bracts and bracteoles of 
the inflorescence, have totally disappeared, and the internodes of the floral 
axis become greatly reduce d, with the result that the individual flowers, 
especially the male, have become so merged together that they can hardly 
be distinguished from one another. Consequently we are inclined to think 
that the Pandanacez branched off at some early period from the main line 
of the Monocotyledonous descent, and are thus c: ipable of being derived from 
an ancestor with hermaphrodite flowers and a well-developed perianth. 
Aracew.—lIn the interpretation of the Araceous flower we are in general 
agreement with Engler. He regards those members of the family with few 
* Benson, Sanday & Berridge (1906) p. 43. + Hallier (1903). 
t Engler (1897) p. 360. $ Coulter & Chamberlain (1904) p. 298. 
