136 Adkinson . — Some Features of the Anatomy of the Vitaccae. 
a ring of bundles separated by multiseriate rays and without intrafascicular 
linear rays. But below the node where the leaf-traces begin to pass out at 
five regions conspicuous for reduction in number and size of vessels (a cross- 
section of the seedling of V. cordifolia , Michx., is shown in Fig. 5), a departure 
from the adult condition is found. Within the leaf-trace segment (Fig. 7) 
only linear rays occur. The leaf-trace, with the reproductive axis and the 
root, is known to be one of the most conservative regions of the plant, where 
primitive structures persist long after they have disappeared in both stem 
and root. The fibro-vascular bundles of the petiole of V. Labrnsca , L., cut 
a short distance from the branch where it is most woody, showed no vestige 
of uniseriate rays. The presence of the linear rays in the leaf-trace of the 
seedling in Vitis species, where all trace of them has vanished in the adult, 
indicates that linear rays are a primitive feature of the Vitaceae. 
This interpretation of the facts is made stronger by the occurrence, 
even in the adult stem, of vestiges of linear rays. In V. Coignitiae , Pulliat, 
only multiseriate rays are present in the mature wood (illustrated in Fig. 8), 
but in the leaf-trace segments of the first year’s growth (shown in detail in 
Fig. 9) the primitive linear rays persist. The primitive rays appear also in 
the leaf-trace of V. Doaniana, Munson (Fig. 10), the bundles of which show 
more abundant wood parenchyma than does V. Labrnsca :, L. ; the leaf-trace 
segment of the first annual ring exhibits radial rows of parenchyma which 
appear in cross-section like linear rays. Such rows of parenchyma are to be 
interpreted as a transformation of ray parenchyma. In V. arizonica and 
V. cinerea , Noronha, the multiseriate rays within the departing leaf-trace 
are formed by the union of very short linear rays. In V. cinerea , Noronha, 
moreover, tannin-filled cells in the xylem of the central cylinder indicate the 
persistence of wood parenchyma as scattered cells. In the leaf-trace, these 
tannin-filled cells, arranged in uniseriate radial rows, are a last vestige 
of linear rays which penetrate the wood and are present in the multiseriate 
compound ray. In the tendril of V. Labrnsca , L., linear rays are found in 
the trace of the tiny bract which subtends the longer branch of the tendril. 
In these forms where vestiges of linear rays still persist, there can be no 
doubt that primitively the structure of the wood resembled V. calif or nica in 
the presence normally of linear rays within the fibro-vascular bundles of the 
central cylinder. 
This conclusion is confirmed by evidence from the condition found in 
wounded wood of V. Labrnsca , L. One general principle of morphology is 
that primitive structures which have disappeared in the normal wood are apt 
to reappear after a wound. The more severe the wound, the more complete 
is the reversion to primitive conditions. Through the wounded wood of 
V. Labrnsca , L., the multiseriate rays were found to persist unchanged except 
by the strong development of tannin. But at a little distance from the 
injury, linear rays make their appearance between the multiseriate rays, 
