Evolution of Monocotyledons. 291 
with the idea of the primitiveness of the arboreal habit (cf. 5, 10, 
18). He contends that while the Angiospermous condition may be 
regarded as an argument for intimate relationship between the two 
groups, this condition may have arisen more than once, in the same 
way as did the seed-habit. 
According to Wettstein (75, p. 865), the Angiosperms are 
monophyletic, or possibly diphyletic (Diagram III). The chief 
point of his system in the present connection is that he derives all 
the Monocotyledons through the Helobieae from the Polycarpicas 
(a group corresponding broadly to Engler’s Ranales). The Poly- 
carpicae in their turn he considers as having descended from 
Gymnosperms by way of the Monochlamydeae, as the diagram shows. 
Lotsy (42, pp. 863, 864) also believes that the Dicotyledons are 
the older type (Diagram IV), but he derives the Monocotyledons 
from them at two different points—the first in the neighbourhood 
of the Piperales, through which the Spadiciflorae are considered to 
have evolved; and the second in a hypothetical group, the Prora- 
nales, from which lines of evolution are traced to the Liliiflorae and 
Helobieas. A similar diphyletic view of the origin of Monocotyledons 
was suggested by Hill in 1906, in a paper on Piperalean seedling- 
structure (33, pp. 173, 174). According to Lotsy, the whole of the 
Angiosperm phylum is derived from the Bennettiteae by way of a 
hypothetical Proangiosperm ancestor. 1 
Hallier (26, 27, 28) has advanced a system which he varies in 
different degrees from time to time—as Lotsy remarks, it is free 
from the reproach of being too rigid. In all his variations, however, 
he treats the Dicotyledons as the older form and derives them from 
Cycadophyte ancestors. The Monocotyledons he refers to ancestors 
amongst the Polycarpicae, which group he gives thirty-two reasons 
for considering as primitive. 
With regard to the primitiveness of the Polycarpicaa and their 
relation to the Monocotyledons, the following investigations are of 
interest. Fries (22) in studying the leaf-arrangement on lateral 
axes, found that among the Anonaceae and Nymphaaceae, the 
Monocotyledon type, in which the first leaf is adaxial to the main 
axis, is more common than the Dicotyledon type, in which the two first 
leaves are transversely placed (see 22, fig. 1, p. 294). Mez and Gohlke 
(48), from serological investigations, decide that the Magnoliaceae 
include the most primitive Angiosperms, and that the Alismataceae, 
Butomaceae and Juncaginaceae are primitive among Monocotyledons. 
1 Cf. Arber and Parkin (2) and Hallier (27). 
