Notes on the Anatomy of Stangeria paradoxa. 25 
four or five lignified elements, but sometimes tracheids are entirely 
absent. As one passes up the petiole it is found that the amount 
of centrifugal xylem does not decrease, as has often been stated, 
but at a level just below the insertion of the first pinnae there is, 
on the average, two or three times as much centrifugal wood as 
there is at a centimetre or two above the base. This large amount 
of centrifugal xylem persists up the rachis and into the midribs of 
the pinnae. In the terminal leaflet, whose tip is represented in 
Fig. 6, two sections were cut, one above and one below the point at 
which the midrib suddenly and characteristically becomes very 
faint. In “ b ” the amount of centrifugal xylem, formed in connection 
with a well-marked cambium was very little less than that of the 
centripetal xylem (see Fig. 9), while even in “a” the midrib showed 
Fig. 7.—Longitudinal section of tracheids from wood of stem, 
about five lignified centripetal elements and one well-marked 
centrifugal tracheid. Only in the lateral veins in “a” were all signs 
of centrifugal tracheids missing. 
It is obvious from this that the centrifugal xylem system is not 
a thing (in Stangeria at any rate) that gradually dies out from the 
base to the tip of the leaf. Its amount increases over a large part 
of this length and it persists in the lamina almost as long as the 
centripetal portion. 
Then we have its invariable formation in rows in association 
with a cambial layer, and in the light of Le Goc’s proof that 
secondary growth occurs in Cycad leaves, we are almost certainly 
justified in calling the Stangeria centrifugal wood secondary. At 
the base we have a centrifugal xylem, not arranged in rows, which 
