ON THE STRUCTURE AND AFFINITIES OF METACLEPSYDROPSIS DUPLEX. 185 
| _ enumerate the reasons for believing that the stem was a rhizome :—(1) The xylem of the 
stem is small compared with that of the petiole ; (2) the internodes (if we may call the 
distance between two petioles by this name) are long; (3) what evidence there is re- 
garding the distribution of adventitious roots seems to indicate that they are irregular 
in their occurrence; (4) the petiole-trace is at first small and then grows larger, as 
though it were supported in the lower portion either by overlying vegetable matter or 
soil, and did not attain its maximum development until it got above the substratum ; 
(5) there is no sclerenchyma in the cortex of the stem or of the lowest part of the 
petiole ; (6) the pinne are in four orthostichies on the petiole, and the latter must there- 
fore have been held erect. 
The study of the petiolar development has demonstrated certain points of interest. 
The term ‘‘arm” has sometimes been used to characterise the portion of the petiole- 
trace between the protoxylem groups and the “ horizontal bar,” and while such arms do 
exist in Diplolabis and most other Zygopterid petioles, they do not occur in Metaclepsy- 
dropsis duplex. The increasing size of the island of parenchyma at the emergence of 
a pinna-trace-bar from the petiole no doubt gives the petiole-trace at certain levels the 
appearance of having such arms (PI. II. fig. 18), but this stage is not constant. As has 
been already pointed out, the ends of the trace in M. duplex have become dilated 
instead of being produced into arms. ‘This has permitted the insertion on the petiole 
of more robust pinne. In Dinewron we find a petiole-trace quite similar to that of the 
early stages in MW. duplex, but in no case have petioles of the former genus been dis- 
covered which had dilated ends. 
Etapteris, on the other hand, appears to be a case where the dilatations have increased 
to such an extent that they have become quite arm-like. ‘These arm-like processes are 
quite different from the arms in Diplolabis ; they are mere swellings similar to the 
dilatations on the outer ends of the arms in Diplolabis and on the ends of the trace in 
M. duplex. Dr P. Berrranp has pointed this out in his memoir on the Zygopterid 
‘petiole-trace, and has also demonstrated the manner in which the pinna-traces depart from 
the petiole-trace. As Dr Brrrranp has further pointed out, there is some slight trace 
of arms in Htapteris comparable with those in Diplolabis. This will be made clear by 
a glance at text-fig. 4. 
In order to decide whether arms are present on the petiole-trace or not, it is 
hecessary to examine a section immediately above the level of a pinna-trace-bar 
departure. In the case of Diplolubis very distinct arms may be seen, with protoxylem 
groups at the ends, and the same applies to Ktapteris; but Metaclepsydropsis and 
Dinewron do not exhibit such arms, the ends of the trace have grooves in them, and in 
_ these grooves lie the protoxylem elements. he grooves of Metaclepsydropsis and 
_ Dineuron are therefore equivalent to the wide bay in Diplolabis, Zygopteris, and 
Htapteris. In all cases except Htupteris the development of new pinna-traces is 
| — essentially similar; small tongues of xylem elements are developed round each proto- 
| xylem group, and these grow towards one another until they meet and form a xylem 
~_ 
itt 
