660 Worsdell. — A Study of the Vascular System in 
to the fact that each has abstricted off a bundle on to the ventral side in an 
imperfect manner, so that the two have remained in contact ; or, stating the 
matter conversely, the sole remaining bundles of the ventral part of the 
cylinder have imperfectly assimilated themselves with the dorsal portion of 
the latter, each bundle having fused with an end-bundle of the dorsal arc 
without revolving on its axis ; hence the fusion is with the entire inner face 
of the dorsal bundle, instead of, as is usually the case, with one side of 
it only. So that each concentric bundle here represents a kind of little 
remnant of the primitively complete cylinder. It is precisely the same 
phenomenon as we shall presently find reappearing in the case of the con- 
centric bundles of Paeonia. 
H. niger , L. 
Leaf'. A perfectly horseshoe-shaped arc here obtains, the end-bundles 
of which come to lie quite on the ventral side of the petiole and are hence 
inverted ; it is interesting to note that these tend to become obliterated, 
being very small ; and in one case the bundle was quite rudimentary in 
structure, possessing a few cells of phloem only and no xylem. These facts 
tend to show that the arc has been derived phylogenetically from a cylinder 
of bundles. 
H. lividus , Ait. 
Leaf'. The bundles of the petiole are arranged in a wide and very 
open arc. In passing upwards from the basal region inverted bundles 
appear upon the ventral side of the arc ; they are mostly smaller than those 
of the dorsal arc, and lie close to the latter ; a bundle of the arc may have two 
bundles lying obliquely on the ventral side, one at each side of the 
protoxylem ; another bundle is seen with a bundle immediately on its 
ventral side and perfectly inverse in its orientation. In passing both into 
the basal region, and also upwards into the lamina, some of these bundles 
pass in and unite with those of the arc, others appear to die out in situ in 
an oblique position (i. e. at the half-way stage from the purely ventral and 
inverted state to union with the bundle of the arc) ; immediately below 
the lamina the two end-bundles of the arc have each an inverted ventral 
bundle closely connected with its protoxylem, while the vestiges of another 
occur further out on the ventral side. 
H. foetidus , L. 
Leaf'. This is anatomically one of the most advanced species, as a very 
open dorsal arc only prevails throughout the petiole ; but traces of the 
primaeval cylinder occur here and there, as in about the middle region 
of the petiole a bundle about half the size of most of those of the arc lies at 
one side of the inner half of the xylem of one of the arc-bundles and is 
orientated sideways, and is obviously in a position representing a half-way 
stage on its transit to the ventral side of the arc. Again, in the lamina> 
