Primitive Angiosperms. 183 
of these localities should reach the others and spread all over the globe 
as Monocotyledons have done. But conditions may have been more 
favourable in an earlier geological epoch. During the glacial periods the 
winter in Europe and Northern Asia was much longer and colder than 
at present : the summer shorter and warmer. The Alpine flora of this 
continent is supposed to represent the survivors of a flora which spread all 
over it during these epochs. On such a stage as this the great drama 
of the evolution of Monocotyledons might well be played with success. 
The geophilous habit does not, so far as I can see, explain the ternate 
floral symmetry of Monocotyledons. Whorls of three and six parts are not 
uncommon in the flowers of the Ranales. It is possible that they are 
inherited from an ancestor which this alliance has in common with Mono- 
cotyledons. I agree with Messrs. Arber and Parkin in believing that the 
stamens and carpels of the Primitive Angiosperm were indefinite in number 
and spiral in arrangement. The perianth may, however, have been in 
whorls of three or six. 
The hypotheses which have been considered with regard to the 
number of cotyledons in the Primitive Angiosperm and their morphological 
nature have led to two phylogenetic schemes only which attempt 
to explain the evolution of Monocotyledons and Dicotyledons from a 
common ancestor. The first is that of Professor Henslow which derives 
Monocotyledons from a dicotylar race by suppression of one cotyledon and 
adaptation to an aquatic habit. The second is the fusion hypothesis. The 
two cotyledons of Primitive Angiosperms have united to form the single 
member in Monocotyledons. This union, as well as the peculiarities of 
stem anatomy, and many of the minor characters in Monocotyledons, 
are due to adaptation to a geophilous habit. 
Both schemes, it will be observed, give the Primitive Angiosperm two 
cotyledons and the stem anatomy of a Dicotyledon. Most of the minor 
characters which distinguish Monocotyledons are also treated in both 
schemes as departures from the primitive type. It is probable, therefore, 
that the Primitive Angiosperm resembled Dicotyledons much more nearly 
than Monocotyledons in their general features, as well as in stem anatomy 
and the possession of two cotyledons. 
Quarry Hill, Reigate. 
Jan. 20, 1908. 
References to Literature. 
1. Anderssohn, S. : Ueber die Entwickelung der primaren Gefassbiindelstrange der Monokotylen. 
Bihang till K. Svensk. Vet. Akad. Handl., xiii, 1888. 
2 . Anonymous : Discussion on the Vascular Structure of Seedlings. New Phytologist, v, p. 182, 
1906. 
