1098 
Thompson . — The Anatomy and 
The vascular system of the leaf consists of two bundles, the continua- 
tions of the double trace whose course we have already examined. The 
two strands continue distinct and unbranched to the apex of the leaf. They 
are visible at low magnification in Fig. 38, and one of them is shown more 
highly magnified in Fig. 39. Each bundle is without a sheath of any kind 
and is flattened in a dorsiventral direction. The tracheides, which are very 
few in number, are arranged in radial rows. There is no trace of centripetal 
wood, all the larger elements being developed centrifugally to the proto- 
xylem. The phloem is very poorly developed. At the edges of the bundle, 
as Strasburger 1 has noted, there is an occasional transfusion tracheide, one 
of which can be distinguished in Fig. 39 by its bordered pits and thickening 
bars. The bundle figured is better developed than is usually the case, 
for as a rule the tracheides are very few in number, the bast practically 
absent, and the transfusion tracheides rare. Such a bundle with its lack of 
centripetal wood and possession of transfusion tracheides is exactly like 
a reduced Coniferous bundle, and quite unlike that of the Cycads with its 
large development of centripetal wood. 
In their basal portion the long needle leaves of the seedling have the 
same vascular organization as the typical adult leaves. But in their tips 
this organization becomes modified. The secondary wood of the two 
bundles diminishes (Fig. 40) until it quite disappears (Fig. 41). Meanwhile, 
the transfusion tissue increases laterally, and also develops centripetally. 
In the latter position it increases in amount towards the tip until it forms 
a solid mass between the two fast vanishing groups of centrifugal wood 
(Fig. 40). As it increases in amount, the form of its cells also changes ; 
they become longer and more like true tracheides. Finally, in the extreme 
tip of the leaf the centrifugally developed tracheides have completely dis- 
appeared, and only a group of the centripetal ones remain (Fig. 41). Some 
of the latter are like typical transfusion tracheides, and some of them are 
more like ordinary tracheides, and might for this reason be considered 
centripetal wood. 
The conditions just described might be used to support WorsdelFs 
contention 2 that the transfusion tissue has been derived phylogentically from 
the centripetal wood of the ancient Gymnosperms. The transitions between 
centripetal wood and transfusion tissue which he observed in several forms, 
notably in Cycas and Araucaria , are duplicated here. The evidence is thus 
the same as he presents. It should be emphasized that the leaves in which 
these conditions are found are the abnormal seedling ones, and that the elon- 
gated tip to which the centripetal wood is confined is not ordinarily present. 
The structure of the vascular bundles of the leaf strongly indicate 
Coniferous affinity. The development of centrifugal and typical absence of 
1 fiber den Bau, &c. 
3 Origin of Transfusion Tissue. Trans. Linn. Soc., 1895. 
