8 BOTANICAL GAZETTE [JANUARY 
In one of the uppermost members of the aggregation (fig. 10) may 
be seen a darker dot, representing the transverse section of the 
outgoing foliar trace.. The trace is obviously extremely small in 
size. Fig. 14 illustrates a portion of fig. ro under a higher degree 
of magnification. The composite character of the foliar ray is now 
more easily seen, for rays, fibers, and vessels are all evident as 
component parts of the aggregate foliarray. Fig. 11 is a tangential 
view of the foliar ray of H. tuberosus, on the same scale of 
magnification as fig. 10, representing a similar aspect of the ray 
in H. hirsutus. The contrasts between the two figures are both 
interesting and important from the standpoint of the question of 
the origin of the herbaceous type. The leaf ray in fig. rr is much 
more clearly developed than in fig. 10. The fibers are much reduced 
in quantity and very few vessels are to be seen. Another important 
contrast is presented both by the size of the foliar ray and of the 
foliar trace to which it is related. The leaf strand appears as a 
dark, somewhat triangular spot in the upper part of the ray, as 
shown in fig. 11. In fig. 9 is shown a portion of the ray of 
H. tuberosus, under somewhat higher magnification than fig. 11. 
By comparing figs. 9 and 14, which represent details of the foliar 
rays of H. tuberosus and H. hirsutus under the same degree of 
magnification, it is clear that the ray in the former and more 
herbaceous species is much larger, and also contains more parenchy- 
matous storage tissue than in the latter and more woody species. 
The greatest contrast in every detail of organization, as well as 
in relative size, however, is presented by H. annuus, the most 
herbaceous of the three compared species. In fig. 12 is shown the 
tangential view of the leaf ray of this sunflower, under the same 
magnification as in figs. ro and 11. The huge size of the ray is 
clearly seen. Not less striking is its homogeneous organization, 
resulting from the virtual disappearance of the fibers and vessels, 
which betray the aggregated character of the rays in H. hirsutus 
and H. tuberosus. The foliar ray in H. annuus is of the type which 
we have called compound, since with a homogeneous structure 
it still betrays evidence (from the comparative anatomical stand- 
point) of having been organized by the fusion of a number of 
ordinary wood rays in the vicinity of the foliar trace, with the 
