88 
Ethel N. Thomas. 
the cotyledon traces, the protoxylems of which assume the endarch 
position ( Calycauthus , Linum, etc., diarch ; Acer, Ricinus, etc., 
tetrarch). 
( d.)—Collateral Evidence. 
The very good case for reduction in Gymnosperms adds 
probability to the case for Dicotyledons, and further, Miss Sargant’s 
work on Monocotyledons supports this view. She saw reason to 
derive all the modifications met with in the Liliaceae (including 
diarch forms), and probably in all Monocotyledons, from the tetrarch 
type as seen in Anemarrehena. Here each of the two massive 
traces found in the single cotyledon is believed to represent the 
vascular system of an originally independent cotyledon, and 
each is shown to be a double structure. There is no doubt that 
each represents a four-fold cotyledonary trace massed together as 
in the petiole of the cotyledon of Althaea rosea and many other 
Dicotyledons. A tetrarch root is formed precisely as in the 
Dicotyledonous forms. 
It may perhaps be mentioned here that the Gnetacese— 
suspected by some of Angiospermic tendencies—show an absence of 
midrib in cotyledons and foliage leaves, except in a few species of 
Gnctam. 
The embryos in Bennettitean seeds are so beautifully preserved, 
that the paired bundles (usually six) of the cotyledons, are obvious 
and unmistakeable. Recent accounts of the Bennettiteaae lead one 
to suppose, that among them, or in an allied group, will be found 
at least a candidate for the position of ancestor of the Angiosperms. 
V.— Broader Phylogenetic Considerations. 
If then the double bundle, which is so common in the 
cotyledons of Angiosperms, be the homologue of the two central 
bundles found in some Dicotyledons and in most Cycads, we have 
in the cotyledons of Angiosperms the reduced equivalent of the 
parallel series of paired bundles composing the foliar trace, not only 
of the cotyledons, but also of the foliage leaves of the older 
Gymnosperms. 
Matte’s work, previously referred to, was chiefly concerned 
with the elucidation of the vascular system of Cycadean leaves. 
He found that however complicated the vascular system of the petiole, 
its bundles were all derived from two initial strands. Worsdell 1 had 
shown this to be the case for the sporophylls and scale leaves. 
In all cases there is an absence of any central bundle throughout 
1 W. C. Worsdell. Vascular Structure of the Sporophylls of 
the Cycadaceie. Annals of Botany, XII., 1898. 
