Adkinson . — Some Features of the Anatomy of the Vitaceae . 137 
extend for some distance as linear rays, and finally become bi- or tri- 
seriate. 
From the persistence of linear rays in the seedling and the leaf-trace, 
and the reversion to the uniseriate condition after wounds, V. Labrusca , L. 
and the other species studied may be considered as forms which have lost 
the linear rays by a process of reduction of ray tissue within the so-called 
bundle. If this be the correct interpretation of the presence of the linear ray, 
V. calif ornica, Benth., retains the earlier condition and is the most primitive 
member of the genus. 
In a less thorough examination of two other genera of the Vitoideae 
some interesting conditions were found. The wood of Ampelopsis Veitchii , 
Hort. (shown in transverse section in Fig. 11), resembles closely the structure 
of Vitis in the large vessels, the broad rays, and the absence of linear rays. 
The wood is reduced to a few scattered libriform fibres, and to segments 
terminating the annual rings. In this species, as in V. cor difolia, Michx., 
the leaf-trace segment reveals the primitive condition by the persistence of 
linear rays (Fig. 12). The genus Cissus is indigenous in tropical forests, and, 
according to Schenck , 1 differs from the other Vitaceae in the extreme 
reduction of the lignified tissues of the stem. Peculiar forms are described 
in which the stem is transformed by the excessive development of parenchyma 
and ray tissues into soft, stout ‘water reservoirs’. In Cissus nummulifolia , 
the most woody of the species examined, the rays are only four or five cells 
broad, the wood is segregated as in Ampelopsis , and the vessels are large. 
In the leaf-trace segment of this species, vestiges of linear rays were found 
in the first annual ring. But in Cissus cordata y Roxb., and Cissus discolor , 
Blume, the reduction had progressed further. In the latter, the rays are 
wider than the bundles, which are reduced to a single row of large vessels 
surrounded by libriform fibres. Some of the fibres, particularly near the 
pith, are transformed into parenchyma. Here no trace of the linear rays 
was found. The stem might almost be taken for a woody herb. 
From the study of these four genera, it seems probable that the ances- 
tors of the Vitaceae were originally woody perennials without the climbing 
habit. The xylem cylinder of the stem was interrupted by both multiseriate 
and linear rays. For a time the two types of ray existed side by side, the 
broad ray serving for storage in connexion with the leaf-trace. The uni- 
seriate rays lay within the bundles of the wood. In this condition, Leea 
still remains a tropical shrub, and is on anatomical grounds the most 
primitive genus in the Vitaceae. 
From forms like Leea the other genera are probably derived. With 
the taking on of the twining habit came a need for mechanical strength 
combined with flexibility. The formation of parenchyma by the fibre ele- 
1 Schenck, Heinrich : Beitrage zur Biologie und Anatomie der Lianen, Jena, 1892-3, Part II, 
P- I 37 - 
