I 
Anatomy of the Leaf in the Osmundaceae. 491 
. % In considering the relative primitiveness in the structure of the trace in 
the different regions of the leaf a great deal of caution must be exercised. 
It seems clear that the highest complexity in structure is to be expected, and 
indeed is actually to be found at the base of the free petiole, for this region 
must negotiate the water-supply for the whole leaf and must also bear the 
strain of the total weight unassisted by the stem. The leaf-trace in the 
cortex of the stem is relieved of the duties of mechanical support and is free 
to retain more primitive characters. In the upper region of the rachis, at 
any rate above the first lateral branches, and in these branches themselves 
the duties both of support and of conduction progressively diminish, and here 
again a simpler construction would at any rate be feasible. 
So far as the structure of the trace itself is concerned this certainly 
holds good in the Osmundaceae. Starting from a point just above its base 
the petiole may be said to indicate broadly its phylogeny in two directions, 
both upwards and downwards. 
With regard to the manner of branching, however, other factors come 
into play and the matter is on a different footing. Speaking generally the 
method of branching becomes more primitive on passing higher up the main 
rachis and in its lateral branches. Still it must be remembered that in some 
species the lowest branches of the rachis and of the primary branches are 
themselves reduced in size, being markedly smaller than some of those 
higher up. This reduction affects the method of branching so that it may 
present features of a more or less primitive type. 
The apical pinnae of the rachis and its branches are about the same 
size and have about the same work to perform as the earlier leaves of the 
young sporeling. In consequence, as we proceed down from the apex of 
the leaf we may expect to meet very much the same series of changes in 
the vascular system as we should if we were examining the young leaves of 
the sporeling passing from leaf to leaf up the stem. So far as we have been 
able to follow it this is indeed the case in Osmunda , with one distinction, 
that the thickening of the xylem at a branching and the immersion of the 
protoxylem are omitted in the sporeling. It may be that the phylogeny of 
the earlier leaves is distinct from that of the later leaves and that, owing to 
the small size of their traces, these features could not appear. 1 
In the largest branchings of Osmunda rcgalis the median axis of 
symmetry of the branch-trace is almost at right angles to that of the parent 
trace. It is rather less inclined in the other species of Osmunda. In all 
the smaller branchings it becomes still less inclined and may become quite 
parallel to the axis of symmetry of the parent trace. In Todea the median 
axis of the branch-trace is actually at right angles to the parent trace in 
large branchings and is still inclined to it at a wide angle even in the smaller 
branchings, though it may become almost parallel in Todea hymcno - 
phylloides. 
