SEEDLING STRUCTURE IN THE LEGUMINOS.;E. 95 
of the central cylinder into an elliptical form, and the consequent obliteration 
of the pith. 
It might perhaps be expected that if this were the true explanation proto- 
xylem would be found all along the major axis of the ellipse, marking the 
line of fusion of the two lateral bandle systems ; whereas it is found that the 
protoxylems are located at the foci of the ellipse *. When we consider proto- 
xylem in the light of its formation and physiological function, however, it 
seems clear that, if only two groups arise, they would tend to be developed 
at the foci of an elliptieal desmogen strand ; these being the points which in 
an early stage of development would serve as the best distributing centres 
for water and foodstuffs. | 
The reasons for this compression of xylems to the centre and obliteration 
of pith are not far to seek. The typical “stem position ” of vascular tissue 
surrounding a pith is commonly ascribed to the mechanical need of resistance 
to bending strains; the “ root-position,” on the other hand, freely allows 
bending to take place. Since the difference between root-position and stem- 
position is primarily a mechanical adaptation, it might be expected that plants 
of the habit of the Vicies might show some peculiarities of structure. The 
majority of the Vicieæ are climbing plants without any assured vertical 
position ; consequently, the part of the plant just above the ground must be 
flexible. Since the germination is hypogeal, this mechanical need falls upon 
the first few epicotyledonary internodes, and it is these internodes which 
exhibit the so-called root-structure f. 
The ease of Vicia Faba supports this view, for here we find that the stem 
is erect and does not depend upon external supports, while there is never any 
trace of the “medullary xylem” in the lower internodes. The function of 
the base of the epicotyl differs from that in the majority of Viciew, and 
necessitates rigidity rather than flexibility $. 
* The protoxylems of the central mass are distinct from those of the polar leaf-trace 
bundles, and are often separate ; though frequently they may be continuous and form a 
single protoxylem group at each focus of the ellipse. 
+ The only other recorded example of medullary xylem in the vegetative stem of a Phane- 
rogam is that of Cephalotarus Koraiana, described by Rothert (Ber. d. Deutsch. Bot. Ges, 
xvii. p. 275, 1899). But the physiological function of the centripetal wood here appears 
to be that of water-storage, and the case is quite unlike that in the Viciew. 
t Since the above account was written the work of Chauveaud (1911) has been published. 
This author also arrives at the conclusion that the pith is obliterated by the compression of 
the lateral stelar bundles towards one another (p. 348). Не attributes this “ rapprochement," 
however, to the presence of the cortical fibrovaseular bundles, an interpretation which fails 
in the case of Vicia Faba, and also in the higher internodes of those species whose lowest 
internodes possess medullary xylem. 
