30 BOTANICAL GAZETTE [JAN UARY 
there is no specially differentiated epithelium, and in this respect 
we meet with a feature which serves to sharply separate these 
structures from those of the Cycadacee. On these grounds I 
should feel no hesitation in deciding as to whether a given plant 
were cycadaceous or filicoid in its affinities. From this point of 
view, then, it would seem that the Topeka specimen is more 
nearly allied to ferns, and the same would hold true of Myelox- 
ylon, if we are to base an opinion upon the excellent figures of 
Mr. Seward. 
SuUB-CORTICAL SCLERENCHYMA.—The distribution of the scler- 
enchyma can hardly be taken as an argument one way or the 
other, since in both ferns and cycads there is such wide variation. 
I should consider this a specific rather than a generic character. 
In all of the myeloxylons so far studied, the sclerenchyma is 
distributed in separate strands. In the cycads studied by me 
this tissue forms a continuous band in all cases where strongly 
developed. In Angiopteris it forms a continuous zone of con- 
siderable thickness, with separate strands lying along the inner 
face. 
A résumé of the results above detailed shows that in the 
Topeka specimen there are characters which directly connect it 
with Rachiopteris Williamsoni, and also with other European spe- 
cies of Myeloxylon, and the evidence would seem to indicate 
that few of these can be separated generically. Admitting the 
force of some of _ objections raised by Mr. Seward respecting 
the filicoid character of Myeloxylon, there are, nevertheless, 
nee pacerusiel in | favor of this view, which seem to me to | pes 
le scan while 
7 : | 7 y to any modern 
type would : seem n to | raise z a 1 question as to the possible correctness 
of the view oy peas by Goeppert that these plants 
in reality represe : lized bie Slatin a oa 
between the 5. 
