I I 2 
Ethel Sargant. 
masses at the same time fuse in pairs, and each double group places 
itself outside one of the four xylem masses (fig. 12). The stele 
narrows very quickly here, for the bundles leave the periphery and 
approach the centre. A diarch root-stele is formed: two opposite 
groups of protoxylem disappearing (fig. 12). 
The resemblance between figs. 2-5 and 8-11 needs no comment. 
The structure of Eranthis is rather more symmetrical than that of 
Anemarrhena, owing to the one-sided position of the stem bud in the 
latter (fig. 3.) The cotyledonary traces of Eranthis also reveal their 
double character more clearly than those of Anemarrhena. Com¬ 
paring fig. 8 with figs. 9-11, no one would hesitate to consider the 
two bundles of fig. 8 as representing in fact four, united in pairs 
throughout the length of the cotyledons, but betraying their real 
character at the base of the cotyledonary tube (fig. 9.). They 
behave as four throughout the transition (fig. 11), following Van 
Tieghem’s first type except that the resulting root is diarch not 
tetrarch. (Traite de Botanique, Vol I, p. 782). 
In Anemarrhena the corresponding traces never separate so 
completely into four, at least in the upper part of their course, but 
they are all represented in the root-stele which becomes and remains 
tetrarch. 
In the formation of a diarch root from a system of traces which 
suggests a tetrarch symmetry, the seedling of Eranthis follows that 
of the allied genus Nigella. The vascular structure of the seed¬ 
ling Nigella damascena has been described at length by MM. Gerard, 1 
Dangeard, 2 and Sterckx (loc. cit. pp. 5-24 and PI. i.-viii. in the 
separate copy.) 
It is at first difficult to understand why the cotyledonary traces 
of Anemarrhena, which are more reduced in the upper part of their 
course than those of Eranthis, should retain their independence 
more completely in the root. The explanation perhaps lies in the 
power possessed by Dicotyledons of increasing their root-girth by 
secondary thickening. A Monocotyledon has no such resource. 
If its primary root is to become a stout and long-lived member, the 
root-stele must be designed on a generous scale. 
The seedling of Fritillaria imperialis offers a good example of 
adaptation to this necessity. The primary roots of F. alpina and 
other species of the genus are slender and comparatively short¬ 
lived. They possess a diarch stele which is formed from two coty¬ 
ledonary traces in the manner characteristic of the Tulipeae. 
1 R. Gerard, Arm. des Sci. Nat. 6'*\mn s£r. tom. xi. p. 279, i88r. 
2 P. Daugeard, be Botaniste. iiSre ser. p. 75, 1889. 
