528 Sinnott and Bailey . — The Significance of the ‘ Foliar Ray ’ 
equal, has a direct relation to the degree of herbaceousness of the axis ’,are 
invalidated by many important facts in the comparative anatomy of the 
Dicotyledons. 
Not only have these investigators erred concerning the distribution of 
so-called foliar storage rays, but they also appear to be mistaken in their 
interpretation of the origin of these structures. According to their view, 
the foliar storage rays originate in woody herbs as shallow, radially elongated 
masses of parenchyma. With increasing herbaceousness these structures 
become extended downwards and bifurcate, forming sheets of parenchyma 
which flank the leaf-trace segments in the internodal portions of the stem. 
Careful study of a wide range of plants, however, indicates that the evolu- 
tionary history of the ‘ foliar ray’ has been quite different from this. We 
have already stated that vertically elongated types of ‘ foliar storage rays ’ 
occur in many arborescent Dicotyledons. That the flanking portions of these 
masses of parenchyma actually are ordinary multiseriate rays which have 
fused at the node or have been united by the formation of intervening 
parenchyma is indicated by a considerable body of facts. In many trees 
and shrubs, the long flanking portions of the so-called foliar rays are present, 
but the radially elongated confronting masses of parenchyma are absent. 
In other words, the pairs of high multiseriate rays which flank the leaf-trace 
segments do not fuse at the node (Figs. 21 and 22), but merge into the 
parenchyma of the foliar gaps, forming structures which, in tangential 
longitudinal sections of the stem, resemble inverted tuning-forks. Various 
stages in the fusion of such flanking rays, or of their apparent union by the 
metamorphosis of intervening woody tissue, may occur in different parts of 
a given individual or in different representatives of a particular genus or 
family. Furthermore, in multifasciculate herbs, as shown by many of 
Jeffrey and Torrey’s own figures, the long ‘foliar storage rays 5 are not 
homogeneous bifurcated masses of parenchyma, but consist of two struc- 
turally distinct portions, the flanking parenchyma and the confronting 
parenchyma (Fig. 19). The latter is composed of vertical parenchyma, or of 
a heterogeneous mass of vertical parenchyma and of fascicular ray paren- 
chyma. It is clearly differentiated from the two sheets of interfascicular ray 
parenchyma which flank it on either side and which extend downwards into 
the internode (Fig. 20). Such facts as these indicate very clearly that the 
flanking portions of the so-called foliar storage rays are not prolongations of 
originally shallow masses of radially disposed nodal parenchyma, but are 
a pair of high multiseriate rays whose upper extremities have fused with the 
independently formed confronting parenchyma. 
Among vines and herbs, as among trees and shrubs, there are forms 
which have narrow rays and essentially continuous woody cylinders (Figs. 3, 
5, 11, and 14), as well as forms with high multiseriate rays and steles which 
are dissected into a series of discrete woody segments (Figs. 4, 6, and 8). 
