526 Sinnott and Bailey . — The Significance of the ‘ Foliar Ray ’ 
herbs, the rays are all narrow and are more or less uniformly distributed. 
The secondary xylem forms a continuous, unbroken cylinder — except for 
the foliar gaps at the nodes — regardless of whether the primary elements 
tend to be aggregated into more or less distinct strands or not (Figs. 1, 3, 5, 
11, and 14). In another group of trees, shrubs, vines, and herbs, having 
sharply defined primary bundles, the gaps between the strands of primary 
xylem are confronted in the secondary wood by wide sheets of ray paren- 
chyma. Thus, the stele, both in the nodal and in the long internodal 
portions of the stem, is dissected into a ring of discrete woody segments 
(Figs. 2, 4, 6, 8, and 9). In the first annual ring these wide rays tend to have 
a very considerable height, i. e. longitudinal extension (Figs. 19-22). There- 
fore the leaf-trace segments, which are commonly set off some distance below 
the points of entrance of the leaf-traces, are flanked on either side by wide 
sheets of ray parenchyma. At the node these pairs of flanking rays unite 
with the parenchyma of the foliar gaps (Figs. 21 and 22). (2) In most Dico- 
tyledons there is a tendency for the vessels of the secondary xylem to curve 
around, and thus to avoid, the entering leaf-traces (Fig. 18). In extreme 
cases the segment of xylem, confronting the leaf-trace in the stem, may be 
devoid of vessels for a considerable distance below the node. This phe- 
nomenon appears to be closely associated with the conduction of water to the 
upper levels of the stem, and the vessel-less tissue cannot be regarded mor- 
phologically as ray tissue. (3) In many Dicotyledons the rays, in the 
segments of wood confronting the leaf-traces, tend to be more numerous or 
wider, particularly in the vicinity of the node (Figs. 7 and 1 5). (4) In certain 
trees, shrubs, vines, and herbs, the fibres in the segments of wood confronting 
the incoming leaf-traces tend to be replaced by vertical parenchyma 
(Fig. 19). 
The last three phenomena may occur in plants which have narrow rays 
and an unbroken vascular cylinder, as well as in those which have wide, high 
rays and a ring of discrete woody segments. In stems of the latter type, 
the substitution of vertical parenchyma for fibres, in the segment of secondary 
xylem confronting the incoming leaf-trace, bridges the interval between the 
large flanking rays (Fig. 19). Thus, in tangential, longitudinal sections, the 
leaf-trace enters the stele through a jacketing mass of heterogeneous paren- 
chyma : ray parenchyma, vertical parenchyma, and parenchyma of the foliar 
gap (Figs. 13 and 19). 
Jeffrey and Torrey interpret the second and third phenomena, outlined 
above, as stages in the evolution of large storage rays, and actually refer to 
segments of the cylinder in which they occur as foliar rays. The ‘ com- 
pounding ’ of a homogeneous mass of storage parenchyma is supposed to 
occur through the intervention of the fourth phenomenon. Their principal 
conclusions concerning the evolutionary history of the foliar storage ray are 
the following : 
