IOO 
Ford. — The Anatomy of 
lying on each side of it. On the under or convex side four 
or five bundles only are present. In Engler and Prantl’s 
‘ Pflanzenfamilien 1 * it is stated that the number of bundles 
present in the petiole of Ceratopteris varies considerably with 
its age and development. In a mature plant two concentric 
rings of bundles are present, an inner and an outer, each ring 
containing from one or two up to forty steles. In young and 
immature petioles a single ring (the outermost) may alone be 
present, and the numbers may be as low as four or five. 
I have examined a large number of petioles of all ages, and 
I have always found the arrangement of the bundles to be in 
accordance with the description given by Engler and Prantl. 
Even in small petioles, measuring only 2-3 mm. in diameter, 
I have found a fairly regular ring of bundles towards the 
periphery, whilst only one or two, sometimes none at all, of 
the inner ring are present. The single large median bundle 
described by Thomae is certainly not a constant feature. It 
is true that at some levels bundles larger than the rest are 
found, but these may occur in any position in the outer ring, 
and are only found when a bundle is about to divide. 
Sections taken a few millimetres above and below the level 
of such a bundle show either two strands instead of one or 
a single strand of ordinary size. I cannot help thinking, 
therefore, that Thomae has taken a petiole in which there 
has been, by chance, a single median bundle about to divide 
in this way, and lying towards the upper side, and has 
described this arrangement as being constant and typical of 
all Ceratopteris leaf-stalks. 
Secondly, in regard to the structure of individual bundles : 
Thomae states that they are all collateral, but in the majority 
of the bundles that I have examined the arrangement is 
markedly bicollateral, sieve-tubes being well developed on 
both sides of the xylem. In the bundles forming the inner 
ring, and in any immature vascular strand, it is often difficult 
to state definitely whether the arrangement is concentric or 
bicollateral, for it is not easy to distinguish between the 
1 Engler and Prantl, 1900, p. 341 . 
