Primitive A ngio sperms. 169 
and other genera of the Scitamineae recall the tetrarch symmetry of 
Anemarrhena (72, pp. 38-40, and pp. 50-52). 
The Anemarrhena type, then, may very well be primitive among 
Monocotyledons in general as well as among the Liliaceae proper. But 
if so, it will probably reproduce the main features of the seedling in the 
primitive Monocotyledons. The exact structure of this type is therefore of 
importance in the present inquiry. 
The cotyledon in Anemarrhena , Albtica , and Galtonia is erect and 
green. The apex is enclosed within the seed, and commonly carries the 
shell of the seed above ground with it (Sargant, 70 and 72). The primary 
root is long and stout ; externally it appears as the direct prolongation 
of the cotyledon, from which it is divided only by the slight swelling which 
marks the position of the future bulb or tuber. Two massive bundles 
traverse the cotyledon from apex to base. Above the plumular bud they 
occupy the foci of the elliptical transverse section (Fig. 16). They are 
symmetrically placed in the sheath and below the plumular bud — the 
traces of which remain embryonic long after germination — the two cotyle- 
donary traces meet in the centre of the circular section to form the stele of 
the hypocotyl (Fig. 17). As they approach, each trace opens out into 
a double bundle, resembling in essentials that of Delphinium (Fig. 2). 
Before the double traces meet, the xylem of each branches in three direc- 
tions (Fig. 17), and for a very short distance there are six protoxylem 
rays and four phloem groups in the stele. Below this the stele becomes 
root-like and tetrarch by union of two pairs of adjacent lateral protoxylems 
(Fig. 18). 
The resemblance to the vascular skeleton of Althea needs no comment. 
It is startling when the diagrams are compared (Figs. 13-15 and 16-18). 
But Althea is a mere variant, very slightly reduced, on the tetrarch structure 
which is certainly primitive among Gymnosperms (Figs. 10-12), and prob- 
ably primitive among Dicotyledons. We have seen that the seedlings of 
primitive Dicotyledons must have been either diarch or tetrarch in their 
vascular symmetry. Comparison with the seedlings of Gymnosperms 
inclines the scale in favour of the tetrarch type. This comparison, how- 
ever, cannot be held to settle the question, because living Cycads are very 
remote from living Dicotyledons (ante, p. 165). The presumption is un- 
doubtedly in favour of the tetrarch symmetry of seedlings among the 
primitive Dicotyledons. 
This presumption is increased almost to certainty by comparison with 
Monocotyledons. We have seen that the tetrarch type of symmetry 
reappears among them. It reappears in seedlings believed on independent 
grounds to represent a type of vascular symmetry primitive among Mono- 
cotyledons. In other words, the evidence suggests that the seedling of the 
primitive Monocotyledon was tetrarch. 
N 
