Bower. — Studies in the Phytogeny of the Filicales. VI. 15 
from the surface. They consist for the most part of a single layer of cells, 
with dark-brown walls, except near the point of attachment, where the walls 
remain thin. Christ mentions their ciliate margin. This appearance is due 
to the thickening of the walls being continued to the septa dividing the 
marginal cells, but not to their outer walls (Text-fig. 4, b). These remaining 
very thin and becoming concave, the thicker septa appear under a low power 
as ciliar projections from the margin. The scales correspond nearly in 
structure to those of Leptochilus tricuspis , but the marginal glands were not 
found on the old scales available, though occasional appearances could 
readily be interpreted as collapsed glands. In Leptochilus they were not 
constantly present. 
I have already described the vascular arrangements in the stock of 
Neocheiropteris} It corresponds in all essentials to that of L. tricuspis : 
moreover the same gritty nests of sclerotic tissue are present, especially in 
the cortex. The stele is highly perforated in both cases, and opens on the 
obliquely upper side to receive the highly divided leaf-trace. The mature 
petiole of N eocheiropteris shows in transverse section a horseshoe of about 
six strands, of which the two adaxial are the largest. In fact, the plan 
of vascular construction of the two Ferns is practically identical. 
The lamina of N eocheiropteris helps to an understanding of that of 
L. tricuspis. Christ compared it with Dipteris ; but the branching which it 
shows is more nearly of the type of Matonia pectinata. At the end of the 
long and slender petiole the lamina widens out broadly, with pedate 
lobation right and left (Plate I, Fig. 4). The midribs of the successive lobes are 
related as a katadromic helicoid system. It may be interpreted as a sympo- 
dial development of dichotomy, that shank of each dichotomy which is not 
then branched again being directed to the anadromic side : while the kata- 
dromic shank undergoes further branching. The whole lamina may be repre- 
sented as composed of the two branches resulting from an initial dichotomy, 
as in Dipteris conjugata , where the two are very exactly equal, though 
the further branching in that Fern shows anadromic helicoid development. 
But in Neocheiropteris the two shanks of the first dichotomy do not develop 
equally. In Plate I, Fig. 4, the two are laid out slightly apart, each having 
five lobes ; and the sinus between them is very slightly deeper than any of 
the others. The sinus to the left of it is almost as deep, and the result 
of this is that the first lobe of the left-hand shank takes an almost exactly 
median position. It is also slightly the longest, and so appears as a distal 
and median lobe, while it is really the equivalent by branching of the left- 
hand lobe of the right-hand shank. This is precisely the condition already 
recognized for Matonia pectinata , and the remaining lobes of Neocheiropteris 
follow according to the same plan. 
The leaf of Leptochilus tricuspis is open to the same interpretation, 
1 Ann. of Bot., vol. xxvii, 1913, p. 473, Text-fig. B. 
