304 EQUISETALES 
essential to a proper understanding of the embryogeny to fix the attention 
primarily upon the origin of the apex of the axis, which is defined at a 
very early stage in Lgudsetum. The epibasal half of the embryo is 
described as dividing into octants by walls at right angles to one another: 
one of the octants then takes the lead over the others, and it is this one 
which gives rise to the axis, with its tetrahedral apical cell like that in the 
mature plant: the product of this octant soon constitutes the greater part 
of the epibasal region (Fig. 214 1v.). A little consideration of the facts 
thus stated will show, first, that the octant-walls are the natural preliminary 
steps to the definition of a tetrahedral initial cell centrally in the epibasal 
hemisphere: the octant-walls might even be held to be themselves the 
first segmentations in the definition of that cell; secondly, that, con- 
sistently with the initiation of a conical initial cell, the origin of the axis is 
in the closest possible relation to the point of intersection of the octant- 
walls, just as it is found to be in the Lycopodiales. There is, however, 
this difference, that the apex asserts itself very early in Zgudse¢um, which 
is in accord with the early dominance of the axis over the appendages. 
These arise as three (or sometimes only two) leaf-teeth, borne upon a 
coalescent sheath, which is described as originating partly from the 
remaining three octants of the epibasal half, but partly also from the 
lower portion of that which gives rise to the apex of the axis itself. It 
seems quite unnecessary in such a case as this to attempt to allocate the 
several parts to definite octants: clearly if the leaf-sheath be partly derived 
from the stem-octant, this is not rightly so named. Probably the allocation 
of parts to definite octants would not have been attempted in Lgudsetum 
had it not been found to apply with apparent success elsewhere, and 
especially in the Ferns. In the present case it seems more natural to 
regard the whole epibasal hemisphere as formative of the shoot: from this 
the stem-tip originate# at the central point by the simplest course of 
segmentation, which happens to involve octant-walls, while the peripheral 
region of the epibasal hemisphere gives rise to the first leafsheath with its 
three teeth. The shoot thus established continues its apical growth 
directly upwards, forming successive three-leaved sheaths, followed soon by 
the appearance of accessory branches. The hypobasal half of the embryo 
meanwhile becomes slightly distended, as the “foot,” which remains in 
contact with the prothallus after the young plant emerges. The root 
originates laterally in the hypobasal hemisphere in £. arvense and palustre 
(Fig. 214), but in 2. Atemale it appears to be formed laterally at some 
distance from the base, and even from the epibasal hemisphere. 
This embryogeny accords readily with a strobiloid theory. The apex 
of the axis arises early at the usual point in close proximity to: the 
intersection of the first octants, and it is dominant from the first. The 
leaves, which are minor appendages in the mature shoot, arise relatively 
late, and are not prominent features in the embryogeny. The branching 
is clearly accessory, as it is also relatively late in the time of its appeararice. 
