Benson.—New Observations on 
1048 
grooves, which appear like shallow pits bordered by a pair of minute cusps 
in transverse sections of the petiole, persist after the branching of the main 
rachis (PI. LXXXIII, Fig. 24). Each groove with its projecting margins 
probably represents a so-called tooth of the later species, B. hirsuta and B. 
ramosa , although in these species the adaxial protoxylems are clearly endarch. 
In this view I differ from MM. Bertrand and Cornaille, who refer 
to the Botryopteris antiqna type of petiole as tridentate. 1 This interpre¬ 
tation may possibly have prevented the earlier diagnosis of B. antiqua , 
‘ d’Esnost \ 
As is shown in the succeeding paragraph, it is only during incipient 
phases of branching that the Pettycur species could be described as tri¬ 
dentate, and it was due to the prevalence of petioles in this condition in my 
earlier material that I was unable to clearly distinguish the earlier from the 
later species until comparatively recently. 
V. The Branching of the Diarch Petiolar Bundle. 
When a diarch petiolar bundle is preparing to branch, a third pair of 
cusps appear, exactly like the original two. This, which probably owes its 
origin to the bifurcation of one of the two existing protoxylem groups, 
travels for a considerable distance in the petiole before the separation of the 
two metaxylem masses takes place. Hence, as already stated, in a long 
series of sections (Slides 390. 11-21, vide Fig. 24) of a branching petiole, we 
may have a fair semblance of the triarch type of petiole characteristic of 
later species. Finally, the branch, which is generally smaller than the parent 
rachis, passes off laterally. No equal dichotomy of a diarch petiole has been 
met with—the only record of a case being apparently due to an oversight. 2 
VI. The Monarch Petiolar Bundle and Aphlebiae. 
I have no instance of two successive nodes giving off diarch traces— 
monarch traces apparently precede and follow the diarch trace. The 
monarch traces, however, apparently never come off, pass through the cortex 
and enter the petiole, without being accompanied by aphlebiae. 
In four of the specimens figured (Figs. 8-9, 10, 19-20, and 21), 
monarch petiolar traces are obviously associated with minute monarch traces 
which come off from the stele almost simultaneously with their respective 
petiole traces. 3 In three of these cases (Figs. 8-9, 19-20, 21), the 
monarch petiolar trace with its aphlebia trace follows a diarch trace, whereas 
in Fig. 10 there is evidence that a large trace succeeded such a pair. This 
figure also indicates that the aphlebia trace diverged from the petiole trace 
1 Bertrand et Cornaille: loc. cit., p. 1022. ‘La trace realise alors le facies tridente des 
B* hirsuta , B. antiqua .’ 
9 Kidston : loc. cit., Figs, n and 12, which appear to represent a node of a stem. 
8 Cp. Kidston: loc. cit., Fig. 4, where * pet' ’ appears to represent the aphlebia corresponding 
to the petiole trace pet". 
