Note . 
535 
In various respects Lyginodendron and Heterangiam have points in 
common with Gleicheniaceae, Osmundaceae, Marattiaceae, Ophio- 
glosseae, and Cycadeae. The view of their affinities, which we suggest, 
is that they are derivatives of an ancient generalized race of ferns, 
from which they have already diverged considerably in the cycadean 
direction. Of the two genera, Heterangium appears to be geologi- 
cally the more ancient, and certainly stands nearer to the filicinean 
stock. Lyginodendron , while retaining conspicuous fern-like charac- 
ters, has advanced much further on cycadean lines. This view by 
no means involves the improbable assumption that these plants were 
the actual ancestors of existing Cycadeae. How far their divergence 
from the fern stock had proceeded cannot be determined until we 
are acquainted with their organs of reproduction. 
The existence of a fossil group on the border land of ferns and 
Cycads seems now to be well established. Count Solms-Laubach 
places his Protopitys in this position, which is probably shared by 
Myeloxylon and Poroxylon. Messrs. Bertrand and Renault have indeed 
endeavoured to derive the last-named genus from Lycopodiaceae, and 
have extended the same view to Lyginodendron and Heterangium. In 
the latter cases their theory is completely negatived by the organization 
of the leaves, and by many structural details. 
The relation of the genera which we have described to those 
ancient gymnosperms, the Cordaiteae , will form one of the most 
interesting palaeobotanical problems of the future. 
The paper is illustrated by micro-photographs and by camera-lucida 
drawings. 
