1911] SINNOTT—FOLIAR BUNDLE 267 
The protoxylem is always attached to the primary centripetal 
wood, which forms the great bulk of every bundle, and it is clearly 
in seriation with it, the smallest and ringed elements being on 
the very outside and becoming successively closely ringed, spiral, 
and pitted as we approach the inner or adaxial edge of the strand 
(fig. 4). Between the protoxylem and the centrifugal wood there 
is in nearly every case a row of parenchyma, one or two cells wide, 
just as in the cordaitean leaves which possess centrifugal wood. 
The petiolar bundles may vary in shape from narrowly triangular, 
with the protoxylem carried up on the apex, to almost semi- 
circular, with the first formed tracheids in the middle of the flat 
upper side. 2 
The resemblance of a cycad leaf bundle to the foliar trace of 
Lyginodendron has been much emphasized by Scorr and others. 
There are apparently two very vital differences between the two, 
however, for the former has no centrifugal primary wood, as has 
the latter, and in the cycad leaf the protoxylem is distinctly seriated 
with the centripetal xylem, and not with the centrifugal, as it is in 
Lyginodendron. The one is exarch and the other mesarch with a 
tendency toward endarchy. Scattered bundles were observed in 
several species, where the protoxylem connected the inner with the 
outer wood, being attached to both. Such cases are doubtless 
reduced and due to the partial abortion of the separating zone of 
parenchyma. In all of them the protoxylem remained in seriation 
with the centripetal wood. 
In the leaf blade the centrifugal secondary wood is still further 
reduced, especially when the bundles become small at the end of a 
Pinna. In several species, such as Dioon edule, it is altogether 
absent, and in no case is it represented by more than three or four 
cells. These often lose their radial arrangement and tend to 
cluster round the outer face of the protoxylem cluster. This gives 
the mesarch appearance, which is probably similar to what WIE- 
LAND saw in the blade of Cycadeoidea. In the small cotyledonary 
traces of Bowenia, WorsDELL observed a true mesarch structure, 
which may perhaps be a reversion to the very primitive botryop- 
teridean condition. 
