1050 
Benson.—New Observations on 
There is no evidence that the vascular bundle supplying an aphlebia ever 
branched. In no single case has an aphlebia been seen in connexion with 
a diarch trace (cp. PI. LXXXII, Figs. 11-13). Thus we may regard it as fairly 
established that diarch petioles without any aphlebia occur, succeeded by 
petioles with a monarch vascular bundle and a monarch aphlebia trace. 
This contrast between the petioles of B. antiqua is interesting, and probably 
will prove to indicate some distinction in function to which we have as yet 
no clue. 
The aphlebia is simpler in structure than the bodies described under 
this name in the Zygopterideae, for the latter receive a bundle which ‘ divides 
into two, three, or four strands V whereas in Botryopteris antiqua I have met 
with no example of the branching of an aphlebia bundle. 
In one respect the diarch petiole resembles the monarch petiole with 
its aphlebia, e. g. they each receive a vascular supply, including two pro- 
toxylem groups. Perhaps on this ground we may regard the diarch petiole 
as equivalent to the monarch with its appendage. 
As is well known, Dr. Paul Bertrand regards Botryopteris as having 
been derived from Anachoropterid ancestors by the shifting of the two 
fundamental poles of the leaf-trace to an adaxial position. Such a theory 
involves the assumption that the leaf-trace in Botryopteris was primarily 
diarch. The new observations appear to introduce some difficulty in the 
acceptance of this view because, if the diarch type were the older, we 
should be under the necessity of regarding it as already undergoing dis¬ 
integration or reduction in B. antiqua —the oldest known species of the 
genus. 
The argument in Bertrand and Cornaille’s recent account of the 
characteristics of the Botryopteridean trace is based primarily on the 
‘ forensis 3 type. They suggest the * ramosa 3 type to be a reduction form, 
since they compare it 1 2 with the reduction form met with in the upper part 
of the frond of the ‘ forensis 3 type. They also refer to the B. antiqua trace 
as having ‘ le facies tridente ’. If B. antiqua , ‘ d’Esnost,’ has a tridentate leaf- 
trace it cannot be identified with B. antiqua of Pettycur. Except in 
branching diarch petioles, I have never seen a tridentate trace from Pettycur. 
If the authors of the above note homologize each of the minute cusps 
limiting a group of protoxylem with the tooth of B. ramosa , then the 
B. antiqua trace is respectively either quadri- or bidentate. The identifi¬ 
cation of a single protoxylem group in B. antiqua with a correspondingly 
single protoxylem group in the latter type of trace seems absolutely 
necessary, even though in the former it may be slightly mesarch or sunk 
between two cusps, and in the latter it is endarch and projects as a single 
point. It will be simpler, therefore, to refer solely to the number of pro- 
1 Scott: Studies in Fossil Botany, p. 314. 
2 Bertrand et Cornaille : loc. cit., p. 1022. 
