452 DR. ERIC DRABBLE ON THE 
forms a root-gap which persists for some time, being still open when the cylinder of the 
lateral root has become complete. 7 
Through the cortex of the extremities of some roots of Areca Catechu break many 
large lateral roots, giving to this region a swollen and irregular appearance 
(Pl. 51. fig. 114). 
In some cases the growing-point of the parent root aborts, and one or more of the large 
lateral roots grow downwards and continue growth in the direction of the parent. Two 
lateral roots are seen doing this in Areca Catechu (Pl. 51. fig. 116). 
A better example of the same thing is illustrated for Actinorhytis Calapparia in 
Pl. 51. fig. 117, where the downwardly growing portion o is a large lateral root proceeding 
from the main root b, the other large lateral roots at c ceasing to elongate. 
Iriariia presents one of the few known examples of spiny roots, and is referred to by 
Goebel in his * Organographie. PI. 51. fig. 112 illustrates a large aerial root arising from 
the slender stem. The lateral rootlets are all reduced to short spines, which present a 
curious interna] structure. The outer layers of the cortex are parenchymatous, as are 
also the inner, but between these two zones occurs a lignified sclerenchymatous ring 
(PL 51. fig. 113). The parenchymatous elements just inside this ring show clear evidence 
of secondary division, having quite a cambiform appearance (fig. 115). 
The central cylinder of the spiny rootlet is bilobed, but otherwise it is normal in 
structure. 
The well-known contractile nature of the roots of Palms is splendidly shown in certain 
roots of Corypha umbraculifera; a small portion of one of these is figured in 
Pl. 51. fig. 120. 
THEORETICAL CONCLUSIONS. 
One result of the examination of the roots of Palms has been to show that the central 
cylinder in these organs is not the simple structure which Van Tieghem considered the 
root-cylinder to be, and upon which he based his conception of monostely. 
The base of the root is occupied by a number of separate cylincers, each essentially 
“ stele "-like in structure, as evidenced by the presence of regularly-arranged strands of 
xylem and phloem embedded in the fibrous ground-tissue of the cylinder, although no 
definitely marked endodermis is present. 
Any objection to the term “stele” as applied to these cylinders, based upon the 
apparent absence of an endodermis, loses much of its force in view of Strasburger’s 
demonstration that the histologically differentiated endodermis, even when present, does 
not always represent the innermost cortical layer, but may occur in any zone, while the 
first ** extra-stelar” layer may differ in no respect from the ordinary ground-parenchyma. 
| The base of the root would therefore be termed, in the language of the stelar theory, 
** polystelie," and the normal root-cylinder, formed as it is by the fusion distally of the 
| proximally free e steles," would be ** gamostelic "—ruly a curious result. 
Cormack's “ polystely," likewise produced by fusion distally of the free * steles," would 
be termed * gamostely "—a number of “ gamosteles ” traversing the ground-parenchyma. 
D however, it be still urged that a histologically differentiated endodermis is necessary 
