— 577 — 
the Jaguaras in Rio Grande do Sul. None of the Specimens so far 
examined in our collection distinctly present the peculiar characters 
that specially distinguish the type, though it appears to be represented 
by poorly preserved material, that may have come from the outer 
zones of the secondary wood. 
The remarkable features of the species are (1°) the tricarinate form 
of the very large, chamberless pith which contains gvoups of secretory 
cells ; (2°) the sometimes distant arrangement of the areolale pils of 
the tracheides, which are notin regular radiate rows in the xvlem 
wedge; (8°) the exactly round form of the pore, which is often some- 
what eccentric in the areole; and (4°) the origin of the leaves and 
probably of the branches also; from the crests of the three very pro- 
minent longitudinal keels. 
The relationship of Zeiller’s Dadoxylon Pedrot and its possible 
or probable connection with some one of the foliar types already 
described in the Permian flora of Brazil remains undetermined. The 
large size of the pith, the presence of secretory vessels in the pith, 
the polygonal forms of the medullary cells in the basal portions of 
the rays and the somewhat irregular form of the tracheides in the 
slightly indefinite radial rows in the xylem wedge, agree with acyca- 
dofilicate or cycadaceous reference of the type. Bul the absence of a 
centripetal primary wood, and the strictly contrifugal development 
of the wood wedges seems to exclude it from that group. 
The .centrifugal sequence of spiral, scalaviform and areolale tra- 
cheides conforms to the Cordaitean type, though, as Zeiller points out, 
the absence of chambers distinguishes it from the wood of Cordattes. 
Zeiller is inclined nevertheless to regard the species as belonging 
to the'Cordaitean group, though probably to Noeggerathiopsis or Eu- 
ryphylum instead of Cordaites. 
In this connection it may be noted that Professor Arber is 
disposed to regard his Dado.rylon australe (1) or some of the other 
more distinctly Araucarinian types from New South Wales. as pro- 
bably the wood of Noeqggerathiopsis. Though Zeiller’s views appear 
probable it seems to me not at all impossible that Dadoxylon 
Pedroi may represent the stems or branches of Gangamopéeris or 
some other type, anatomically unknown, whose foliar characters 
or inflorescent organs would lead to its reference to the Cycado- 
filices. 
(1) The Glossopteris Vlora pay. 19L lig>. d0-4o. 
5560 G2 
