136 Bower . — On the limits of the use 
cylindrical upper portion the structure is symmetrical round a 
central point, and even corresponds in detail to that of the axis 
below the inflorescence. Accordingly the upper portion of the 
leaf would, on anatomical grounds, fall under the definition of 
an axis. Thus one and the same member, which on other 
grounds is regarded as a leaf, shows in its lower portion those 
anatomical characters which are ascribed to the leaf, in its 
upper portion those ascribed to the axis 1 . The tubular leaves 
of species of Allium present similar difficulties, the sheathing 
lower portion conforming to the foliage type, while the tubular 
upper portion has the vascular bundles corresponding in 
position and arrangement to the type of the stem. Again, in 
the ensiform leaves of Iris, Tofieldia , etc., and the phyllodes 
of certain Acacias , it would be impossible to tell from the 
transverse section alone, and judging by the arrangement and 
orientation of the bundles, whether the member were a leaf or 
an axis. 
The exact converse of the case of J uncus or Allium is shown 
in the phylloclades of Ruscus androgynus 2 . If transverse 
sections be cut at the base of the phylloclade, the arrange- 
ment and orientation of the vascular bundles is according to 
Van Tieghem’s type for an axis, being symmetrical with 
reference to a central line ; but if sections be cut successively 
further from the base, it will be seen that the arrangement 
and orientation of the bundles gradually passes over into Van 
Tieghem’s type for the leaf. As Professor Dickson has pointed 
out, the phylloclade undergoes a twist at the base, so that the 
morphologically lower surface is directed upwards, and this 
actually upper surface bears no stomata, though they are to 
be found in large numbers on the morphologically upper but 
downward directed surface ; all the vascular bundles have 
their xylem directed upwards, i.e. towards the morpholo- 
gically lower surface. In passing from the base where the 
1 This example has been cited by Goebel as showing that the distinction of 
members on anatomical grounds is untenable. Vergl. Entw., p. 128. 
2 The structure of these is described by Professor Dickson (Foliage Leaves in 
Ruscus androgynus') in Trans. Bot. Soc. Edin., vol. xvi. 
