58 Proceedings 
subject is too technical for treatment here. Podophyllum 
however may be mentioned as possibly a transitional case. 
It is a geophilous Dicotyledon in which the anatomy of the 
erect stem resembles that of a Monocotyledon, while the 
elongated horizontal rhizomes are nearly Dicotyledonous. 
Proceeding to the other characters which distinguish 
Monocotyledons, two are clearly connected with a geo- 
philous habit. 
The first of these is the fibrous crop of roots which 
commonly replaces the primary root of Monocotyledons 
at an early age. It is characteristic of geophytes to produce 
a fresh crop of roots from the tuber each year in con- 
nexion with the year’s crop of leaves. 
The second character is the linear shape of the leaves 
and their parallel venation. They are usually inserted on 
the underground stem with very broad bases. Such leaves 
are well adapted to grow rapidly when the genial weather 
recurs, and they readily pierce the soil in which they are 
buried. 
It is only fair to add that this character and the preceding 
one are found chiefly among that great majority of Mono- 
cotyledons which are still geophilous. They may be con- 
sidered as direct adaptations to that habit rather than the 
repetition of ancestral features. Still, considering how 
many well marked geophytes are found among Dicotyle- 
dons, it is rather remarkable that the two characters just 
mentioned should be general among Monocotyledons and 
rare in Dicotyledons. 
The connexion between the tripartite floral symmetry 
of Monocotyledons and the geophilous habit is not obvious. 
I believe however that such connexion does exist, for 
tripartite flowers are occasionally found among Dicotyle- 
dons, and species exhibiting it are almost always geophytes. 
The account of the evidence on this very important 
subject which I have given to-night is brief, and necessar- 
ily incomplete. I hope it is sufficient to show that the bal- 
