436 
Hill and de Fraine. — On the 
the presence of four strands at the apex. This feature has, however, been 
commented upon by Miss Thomas, 1 who draws attention to the similarity, 
in this respect, to some of the Araucarias. 
The actual transition takes place along fairly simple lines and with 
some degree of rapidity. The bundles from each seed-leaf travel gradually 
towards the centre of the axis, and the metaxylem, if it had previously 
joined to form a continuous tissue, divides first into two parts (Diagram i, 
Fig. 9) ; then each portion again divides into two, so that at each end 
of the central cylinder there occurs an uninterrupted arc of phloem bounded 
on its inner side by two more or less well-marked pairs of groups of xylem 
elements. The individual masses of each pair then commence a rearrange- 
ment, which is generally a distinct rotation towards one another and out- 
wards, which tends to bring the protoxylems into an exarch position 
(Diagram 1, Fig. 10). While these latter movements are taking place 
the arc of phloem divides into three parts, the division occurring opposite 
the place where the protoxylems are becoming exarch (Diagram 1, Figs. 11 
and 12). Of these three portions of bast, the central one remains in situ , 
while the two lateral parts travel towards the intercotyledonary plane and 
effect a junction with the corresponding tissue of the plumular strands 
(Diagram 1, Figs. 13 and 14). The completion of these movements 
results in the formation of a tetrarch root-structure. 
In no case have we observed a rotation of the protoxylem within the 
cotyledons themselves as indicated by Tansley and Miss Thomas 2 : at the 
top of the hypocotyl each seed-leaf-trace is a large tangentially elongated 
structure from which most of the centripetal elements have disappeared, so 
that it is practically endarch. As before mentioned the number of cotyle- 
dons is usually two, but among the plants at our disposal there was one 
which had three seed-leaves. Hitherto our experience has been that in 
similar cases in other plants, the transition is the same as in an example 
with the normal number of cotyledons, and the root instead of being diarch 
is triarch. Applying this experience to Gmkgo we should expect a trico- 
tyledonous plant to have a hexarch root since a dicotyledonous one has 
a tetrarch root-structure. 
But this was not found to be the case in this tricotyledonous specimen 
of Ginkgo biloba. The facts are as follows : as in the normal seedlings, 
a single tangentially elongated bundle enters the axis from each cotyledon 
and, omitting unessential features, the xylem of each divides into two 
principal parts which rotate towards each other and outwards so as to form 
the three protoxylem poles of the triarch root-structure. The phloem 
of each strand divides opposite each protoxylem, and each half passes 
1 Thomas : a Theory of the Double Leaf-Trace founded on Seedling-Structure (New Phytologist, 
vi, 1907). 
2 Tansley and Thomas. Discussion on Seedling-Structure. Brit. Assoc. York, Sect. K. 1906. 
