gto] JEFFREY—PTEROPSIDA 407 
tioning in the same way as is commonly found to be the case with 
the medullary strands of the stem,-namely as a reparatory strand 
to help fill up the leaf gap above the point of departure of the 
traces. Itis clear from what has been stated in the case of Marattia 
alata, just described, that the leaf may develop a vascular system 
resembling in a detailed manner that found in the case of sipho- 
nostelic stems. It seems eminently a sound conclusion to infer 
as a consequence, since the fundamental tissue of the cortex may 
readily be included in the fibrovascular system of the leaf trace, 
that-a similar condition may also occur in the stem. 
Only two examples have been chosen for illustration of the 
occurrence of siphonostelic leaf traces, but the genus Danaea 
among the Marattiaceae presents parallel conditions to Angiop- 
teris and Marattia. Among the other families of. ferns the phe- 
homenon is frequently represented. A notable and common case 
is that of Pieris aguilina. Here the sides of the original channel- 
shaped leaf strand become strongly pushed in, and the fusion 
takes place so that the joining bundles constituting the margins 
of the fibrovascular complex of the leaf lie above and outside the 
siphonostelic lower part of the leaf system. The branches given 
off to the main ramifications of the leaf are derived from the 
lateral margins of the siphonostelic region, but in passing off receive 
a subtending strand from the large upper fused bundle, which lies 
on the outside of the stelar tube of the leaf axis. The siphonostelic 
condition in the foliar supply is not confined to the ferns, but is 
likewise of very frequent occurrence in the angiosperms, where 
it is susceptible of a similar morphological interpretation. In the 
_ Case of certain dicotyledons, moreover, diverticula of the epidermis, 
_ lined with a hairy covering, make their way into the cavity of the 
pith of the leaf stalk, showing that here too there is no question 
of the possibility of extravascular tissues being included within 
the complex of bundles. It may be further stated that there 
appears to be no clear distinction between the condition where the 
cortical fundamental tissues become deeply embayed in the fibro- 
vascular system of the leaf, and where an actual siphonostelic 
condition results from the fusion of the marginal bundles of the 
foliar complex. Accordingly it appears strictly logical to include 
