172 Sargant.—The Reconstruction of a Race of 
Monocotyledon. The two cotyledons of Cycads and of the Bennettiteae 
would be derived from ancestors in that line of descent. 
But the case is very different if we are to postulate a monocotylar 
Primitive Angiosperm. We must then either give up the connexion with 
the Cycads and the Bennettiteae altogether, or suppose that the mono- 
cotylar form of seedling was evolved in an ancestor of the Primitive Angio- 
sperms more recent than that which they have in common with the 
Bennettiteae. From the monocotylar seedling of Primitive Angiosperms 
derived in this way, a dicotylar race must be again evolved from which 
modern Dicotyledons are descended. Further, some descendants of this 
comparatively modern dicotylar race must reproduce the tetrarch seedling 
structure of the Cycads. The least improbable form which this hypothesis 
could take would be to suppose the seedling of Primitive Angiosperms 
to resemble that of Anemarrhena . That is, it would be completely mono- 
cotylar externally, but its vascular skeleton would remain tetrarch. But if 
we are to credit the Primitive Angiosperms with this partial assumption of 
monocotylar symmetry, we burden ourselves with the necessity of explain- 
ing why the main body of its descendants should have reverted to the 
dicotylar form. All these difficulties are escaped if we suppose the Primi- 
tive Angiosperms to have possessed two cotyledons, united in living 
Monocotyledons to form a single member. 
The antiquity of the dicotylar type is suggested by all the evidence 
which we possess at present. But so far the suggestion that the pair 
of cotyledons characteristic of Primitive Angiosperms united to form 
a single member in one branch of its descendants depends only on the 
evidence for the comparative antiquity of the Anemarrhena type of vascular 
symmetry. This is, indeed, the chief argument in its favour. But corro- 
borative evidence is found in the vascular structure of some Ranal seedlings 
with cotyledons united almost to the top. Eranthis hiemalis has been 
described elsewhere (Sargant, 71 and 72). Its seedling skeleton is very 
similar to that of Anemarrhena. The stele of the primary root is indeed 
diarch in Eranthis. But at the level where the root meets the tuberous 
hypocotyl there are marked indications of a tetrarch structure. Podo- 
phyllum peltatum has a similar seedling (Holm, 45). Its anatomy has not 
hitherto been described in detail. The main features of its vascular 
skeleton are given in Figs. 19-21 (p. 170). 
Just above the plumular bud the united petioles of the cotyledons are 
almost solid. The outline of the section is elliptical, the bore of the tube 
very narrow, and the two traces occupy the foci of the ellipse. Each trace 
is clearly double: the protoxylem is already becoming external (Fig. 19). 
The hypocotyl is not swollen into a tuber, though its tissues are packed 
with starch. As the cotyledonary traces approach the centre, they are 
