12 MESSRS. NEWELL ARBER AND J. PARKIN ON 
Angiosperms, are those of the Neocomian (Lower Cretaceous) of Portugal 
and the United States. In these rocks, what appear to be Dicotyledonous 
and Monocotyledonous leaf-impressions occur together. There have, of 
course, been many attempts to show that Monocotyledonous leaves are to be 
found in Mesozoic sediments of pre-Cretaceous age, or even in the Paleozoic. 
None of these, however, appear to us to afford trustworthy evidence, and 
in many cases such fossils have been already claimed as members of other 
groups, such as the Cycadophyta and Cordaitales. 
It seems evident that the earlier Angiospermous fossils afford practically no 
help in attempting to trace the ancestry of the race. Such plant remains consist 
almost entirely of detached leaf-impressions, which furnish little or по trust- 
worthy evidence, beyond the fact that they are of undoubted Angiospermous 
origin. In the Tertiary rocks, seeds and fruits, also detached, occur on 
certain horizons, but impressions of flowers are almost unknown, or at least 
extremely rare. On the other hand, petrified woods, showing the typical 
structure of Dicotyledons and Monocotyledons, especially Palms, аге found 
in the Upper Cretaceous and Tertiary formations. These fossils are usually 
of considerable size, but on the whole hardly advance our ideas in respect to 
the phylogeny of the group. 
On the other hand, the Bennettitez, the near relatives of the hypothetical 
Нетіапріоѕрегтеғе, afford some evidence in this connection. Аз was first 
pointed out by Solms-Laubach some years ago, the embryo of Bennettites has 
two cotyledons. We imagine that the Hemiangiospermez also possessed two 
cotyledons, and that the Dicotyledonous type was thus more primitive than the 
Monocotyledonous. 
Turning to the living Monocotyledons, we regard this race as one which 
has become largely specialised, in part to a geophilous, and in part to a 
hydrophilous * habit. The best explanation of the monocotyledonous embryo 
is, in our opinion, that put forward by Miss Sargant f. We consider that it 
is more than probable that the single cotyledon of Monocots, and also of 
some Dicots, is due to the fusion of the two cotyledons originally present, in 
response to the geophilous habit. 
During the course of evolution there would seem to have been considerable 
“play upon,” or modification of, every unit of the flower. And this appears 
to us to be true also of the embryo. Late, or far from primitive adaptations 
are to be found among embryos, just as among flowers. The cotyledonary 
tubes of some Ranunculacez and other families, and the division of labour 
exhibited by the cotyledons of certain geophilous Peperomias, recently 
described by Mr. A. W. Hill f, appear to us to be cases in point. 
In regarding the Angiosperms as a whole as monophyletic, we are in 
* Gardner (1883), Henslow (1893) p. 527. t Sargant (1903, 1904, 1905). 
1 Hill, A. М. (1906). 
