186 W. T. GORDON 
bar across the end of the trace. The bay or the groove now becomes an island of 
parenchyma. 
In Diplolabis the xylem bar is well developed and the size of the island is constant, 
as is also the case in Zygopteris. In EHtapteris the tongues of xylem do not meet, but 
break away and unite after becoming detached; this is rather a specialised type of 
pinna-trace departure. Dineuron and Metaclepsydropsis, however, exhibit quite a 
different type. The small groove is bridged across much more quickly than in 
Mplolabis and Zygopteris, but subsequent growth causes the island to become larger 
and larger until it reaches a maximum at the level of the departure of the pinna-trace- 
bar. In Dineuron, it is true, the island does not reach the proportionate dimensions 
that it does in Metaclepsydropsis, but the increase is distinct, and I believe that these 
two genera must be grouped together. 
Such a grouping would necessitate the division into two groups of the first sub- 
division of the Zygopterideze with quadriseriate pinne, as given by Kipston and GwynnE- 
VaucHaNn. In the first group would be included al] forms with well-marked arms and 
the bay between them always constant in size :— 
Diplolabis rémeri (Solms). 
Zygopteris promaria (Cotta). 
The second group would include forms where distinct arms are not developed, and 
where consequently the bay is reduced to a mere groove. The island of parenchyma 
formed by the bridging of this groove becomes gradually larger until a maximum is 
reached just before the departure of the pinna-trace-bar :— 
Dineuron ellipticum, Kidston, and D. pteroides, Renault. 
Metaclepsydropsis duplex (Williamson). 
This subdivision, however, is based entirely on the mature form of the petiole-trace. 
In certain members of each division (Diplolabis and M. duplex) it has been shown that 
in early stages of petiolar development the traces are distinctly similar in appearance 
to that of Dineuron. 
In a recent paper on Diplolabis rémeri* I drew attention to an hypothetical type 
of petiole-trace (Protoclepsydropsis) from which some other Zygopterid traces might be 
derived, and, in a tabular form, indicated what I believed to be the relationship of these 
Zygopterid petioles to one another. ‘The table is inserted below (text-fig. 5), and in it 
M. duplex has been placed in close relation to Dinewron ; in fact, it has been considered 
one of the forms directly derived from a Dinewron ancestry. Zygopteris and Diplolabis 
have been grouped together as the second derived form, and Ktapteris as the third. It 
is exceedingly interesting to find that the discovery of the stem and early stages of the 
petiole-trace of M. duplex has entirely confirmed the view set forth in that table. 
* Gorpon, “On the Structure and Affinities of Diplolabis rémeri (Solms),” Trans, Roy. Soc. Edin., vol. xlvil. 
pt. iv. 
