448 MR. 8. G. SHATTOCK OX THE SCARS 
conceivable supposition as to their nature has been made), were ca- 
ducous appears certain from the fact that the scars are in all cases 
completely formed, and no remnants of the separated organ per- 
sist in connection with them. And it may be assumed that the 
disarticulation proceeded from below upwards ; the great size of 
the scars points to this. For it is a fact of some importance, 
and which I have not found noticed, that the size of the scars by 
no means necessarily represents the actual area of the separated 
surface of the disarticulated organ, since the scars, as in other 
exogenously growing trees, must have grown commensurately 
with the part on which they were seated. This may be readily 
observed in the case of ordinary leaf-scars. In ZEsculus Hippo- 
castanum, for example, the leaf-scars come in the older branches 
to exceed beyond all bounds their original size ; the seven corky 
points, indicating the position of the subjacent ruptured fibro- 
vascular bundles, retain their relative, but not their original, 
distances from one another and from the margin of the scar. 
The pits and furrows on the scars of UTodendron, as was shown 
by Presl and Goppert, are the ends of fibro-vascular bundles; 
and these authors regarded the scars as those of branches. This 
is also the view taken by Renault (‘ Cours de Botanique fossile, 
1880), who, however, places as an alternative that the scars may 
be those of the fructification, the strobiles or lepidostrobi. 
Carruthers’s view that the scars result from the separation of 
aerial roots* (or, as now regarded, rhizophores) is based on the 
fact that in one instance described by him the scars have a down- 
ward inclination, as proved by the opposite direction of the 
remains of the leaves. 
Among living forms of plants the scars of aerial roots are by no 
means easy to find. 
I have examined Philodendron, Vanilla, and different Orchi- 
dacee, but in none have I ever seen a perfectly formed scar ; the 
roots that die are not shed, but, with the greatest tenacity, 
remain connected with the rest of the plant. The central fibro- 
vascular system of the dead root projects from the centre of the 
scar after all the surrounding parenchyma has been removed by 
decay, and the surface subjacent to the latter has been smoothly 
healed by cork-formation. 
In Philodendron the cortical parenchyma of the roots gene- 
rally heals by layers of tabular cork-cells produced by a sub- 
jacent phellogen, which is formed by subdivision of the cells 
* Monthly Microscopical Journal, March 1, 1870, p. 150. 
