Mesoxylou Lomaxii and M. poroxyloides. 1023 
It thus appears very probable that Cordaites Felicis was the leaf of a 
Mesoxylou of the poroxyloides type. Direct evidence from continuity may 
perhaps be hoped for. At any rate this probability may serve to call closer 
attention to the British specimens of 1 Cordaitean ’ leaves, which had been 
much neglected until the appearance of Prof. Benson’s paper. Many 
excellent specimens are now available. 
Diagnosis. 
The following is an amended diagnosis of the species : 
Mesoxylou poroxyloides, Scott and Maslen, 1910. 
Leaf-bases moderately crowded, not quite covering the surface of the 
stem. 
Pith not very large, discoid, with a persistent outer zone. 
Diaphragms sometimes splitting into finer plates towards the middle 
of the pith. 
Twin-bundles of the leaf-trace converging and uniting soon after 
reaching the pith, subdividing in the cortex to form eight bundles. 
Centripetal xylem very distinct, persisting below the fusion of the 
twin strands. 
Xylem-strands without a sheath, or with only a rudimentary one, 
limited to the region where they first reach the pith. 
Tracheides of the inner part of the intermediate secondary wood, as 
well as those of the leaf-traces, spiral, reticulate, or scalariform. 
Bordered pits in one or two rows. 
Medullary rays (with rare exceptions) uniseriate, 1-10 cells in height. 
Dictyoxylon zone of cortex very broad, fibres not much thickened. 
Seam-nodules ; Shore, Littleborough. Lower Coal Measures. 
Comparative Considerations. 
Important forms of Mesoxylou still remain to be described, but the 
three species which have now been dealt with by Mr. Maslen and myself 
are sufficient to make the position of the genus clear. The existing classical 
accounts of the structure of the stem in Cordaites (Renault, ’79 and ’ 96 ) are 
not so full as we could wish, and a reinvestigation is desirable ; so far as 
the stem is concerned we are now better acquainted, in some respects, with 
the anatomy of Mesoxylou than with that of the type-genus of the family. 
On the available data, however, it is evident that the affinity between 
Mesoxylou and Cordaites is a close one, and I agree with Mr. Maslen that 
our genus may best be included in the family Cordaiteae (Maslen, ’ll, 
p. 41 1). This is, of course, a departure from the position which I took up 
in the second edition of my £ Studies in Fossil Botany ’ (Scott, ’ 09 , p. 51 1), 
when I provisionally included what is now Mesoxylou Sutcliffii in the genus 
Poroxylou , though at the same time I referred to other forms, since added 
