248 CHANDLER—AERIAL Roots OF 
We have here, then, an interesting illustration ofa pith cambium 
ultimately becoming continuous with a wood cambium, and 
developing with it phlcem and xylem tissues in normal sequence, 
although in order to do so the relative position of these to the 
pith cambium in its primary position in the root has to be 
inverted. 
What causes the death of the root tip I cannot say. The cells 
in the centre of the root appear as if injured, and the presence of 
fungus mycelium has been detected in them. But the material 
available so far for investigation does not enable me to express 
an opinion as to whether this mycelium is to be regarded as a 
stimulating cause in the production of the phenomenon. 
Where the lobing at the end of the root has a fasciated 
character, the method of development is in essential the same as 
that I have described. The mother root simply opens out 
unsymmetrically, splitting at the same time along one side, and 
an obliquely-mouthed depression appears at the base of the 
fasciated lobes. : 
- In the cases where the circular cushion swelling is formed 
surrounding the mouth or an apical depression (Figs. 3, 8), the 
construction of the cushion is on the same lines as that of the 
transition stage shown in Fig. 18. The trumpet-shaped depression 
is really the symmetrically opened out root end, lined by an inner 
cambium producing cells which become cuticularised on the free 
surface and collapse, and through which mycelium ramifies. In 
this formation the mother root does not break up into lobes, but 
there is a formation of branch rootlets in a circlet, as has been 
already described. I am not yet able to speak with certainty 
about the origin and development of these rootlets, owing to the 
failure of material, but there are indications in the structure of 
the roots I have examined suggesting points of interest for 
investigation when material serves. 
There is nothing to notice particularly in the origin of the 
aerial roots upon the stem. They come off from the pericycle 
and pass out in normal fashion (see Fig. 17). But the branch 
rootlets from the swollen end of the arrested aerial roots offer a 
problem for solution. In F igs. 14 and 16, at ~ there may be seen 
in the cortex outside the epidermis small groups of cells which 
one might take to be the initials of those lateral rootlets. If the 
suggestion be confirmed we must regard this cortical mass as a 
callus meristem formed in correlation with the death of the root 
