1 6 Seward. — On the genus Myeloxylon ( Brong .). 
weaker subdivisions of the petioles of many genera. Grant- 
ing, however, that this arrangement of xylem and phloem 
may have been common in palaeozoic fern-petioles, we have 
still the important fact to remember that in recent Ferns the 
protoxyiem is not on the side of the xylem next to the phloem : 
in Myeloxylon and recent Cycads it is so situated . 
2. Nature of the tracheids. 
I have referred to the tracheids of Myeloxylon as reticulate, 
scalariform, and spiral. In Fig. 1 8 is represented part of an 
oblique section of the xylem of a specimen of Myeloxylon from 
Robert Brown’s collection in the botanical department of the 
British Museum. This shows fairly well the character of some 
of the tracheids : at p we see what may possibly be the remains 
of spiral thickenings of protoxyiem. In no case have I made 
out true bordered-pits, such as we find in recent Cycads, 
but it would be rash to attach any great weight to negative 
evidence in dealing with the minute structure of fossil tracheids. 
In spite of the absence of bordered-pits in Myeloxylon , there 
are various points of resemblance between its xylem-elements 
and those of Cycads. In some respects, for example in their 
reticulate structure, the tracheids of the fossil specimens do not 
correspond very closely either with Ferns or Cycads. 
3. Arrangement and form of the vascular bundles. 
The disposition of the bundles, as shown in transverse 
section, is in many recent Cycadean petioles very similar to 
that in Myeloxylon : in Ferns, on the contrary, the bundles 
are usually arranged in more distinctly concentric lines, such 
as we find in Angiopteris . As we have seen, the common 
form of the bundle in Myeloxylon is that of Fig. 13 ; this 
corresponds very closely with the shape of recent Cycadean 
bundles. In some specimens of Myeloxylon there are other 
bundles than the usual oval type, and these are also re- 
peated in some recent species. If we compare Fig. 15, drawn 
from Binney’s specimen, with the bundle of Cycas Thouarsii 
sketched in outline in Fig. 19, we see at once a close corre- 
