Primitive A ngio sperms. 1 6 7 
as in Dicotyledons, but directly from the roots to which they are respectively 
attached. 
The primary root then naturally withers away at the same time with 
the cotyledon. They rarely survive the first season of growth, and their 
structure— external and internal — is therefore affected by those conditions 
only which immediately succeed germination. With the later life of the 
plant they have nothing to do. This peculiarity accounts for the much 
greater variety of internal structure found in the young seedling of Mono- 
cotyledons as compared with that of Dicotyledons. 
In consequence of this variety the results of such an examination 
cannot be so clearly defined. But, on the other hand, the vascular characters 
of the seedling are of more taxonomic value among Monocotyledons than 
among Dicotyledons. A genus or a group of genera is very often separated 
from its fellows by the characteristic vascular skeleton of its seedling. 
Two distinct types of vascular skeleton are often connected with each 
other by a series of intermediate forms. When such forms constitute 
a single series — such, for instance, as that connecting the diarch with the 
tetrarch type among Dicotyledons — there is usually no way of determining 
which of the two extreme cases is the more primitive. But when, as some- 
times happens, several series start from a single type, when, that is, this 
single type is connected with two or more very distinct forms by two or 
more distinct series of links ; then there is very good ground for supposing 
the type common to all the series to be primitive. For it is generally 
acknowledged that the descendants of a primitive stock may be modified 
in various ways, until several races are evolved which differ very considerably 
from each other both externally and internally. But it is far more difficult 
to imagine causes sufficiently potent to lead even two distinct forms to 
complete identity in internal structure. And this difficulty is enormously 
increased when we start with more than two. 
By comparison of seedling symmetry in the genera belonging to 
a single family, it is thus possible to pick out a type of seedling structure 
as relatively primitive within that family. This has been attempted in the 
Liliaceae (Sargant, 72). The seedling of Anemarrhena, a monotypic genus 
belonging to the Asphodeleae, possesses a very definite and characteristic 
vascular skeleton (Sargant, 70). The vascular skeletons of seedlings belonging 
to other genera of the Asphodeleae can be derived from the Anemarrhena 
type ; many of them are, indeed, merely variants on it. The vascular 
symmetry of seedlings belonging to another tribe, the Allieae, is linked to 
the Anemarrhena type through Arthropo diuni . Thus comparative evidence 
suggests that Anemarrhena is primitive as compared with other genera 
found in these two tribes. But this is not all. The Anemarrhena structure 
reappears in Galtonia and Albnca which belong to the Scilleae. The very 
numerous types of vascular symmetry found within this tribe, and also the 
