developed a considerable area of uncut lamina, whether by the shortening 
of the sinuses between lobes, or otherwise, there has been a tendency 
toward an anastomosis of the veins. The illustrations of this are very 
numerous, as in the evolution of Goniopteris and Aspidium from Las- 
treea; of Callipteris from Diplazium; of the Nidus group in Asplenium ; 
of Pteris biaurita from P. quadriaurita ; of Goniophlebium and Phyma- 
todes from Eupolypodium. In Prosaptia, the least divided species, with 
broadest segments, P. Toppingi, has usually some anastomosing veins. 
From the Grammitis section has been derived Loxogramme, characterized 
by the anastomosing veins of the usually ample but simple and entire 
fronds, and also by the total loss of articulate stipes. These two char- 
acters together, in a group certainly independent of Phymatodes, amply 
warrant the recognition of Loxogramme as a distinct genus. Its most 
Grammitis-like species is L. parallela; at tlje other extreme, it contains 
one species with dimorphous fronds, L. conferta. 
It is more than possible that Blume was correct in treating An- . 
trophyum and Loxogramme as relatives; I know of no' other likely origin 
for the Vittariece , and the resemblance of these two genera is certainly 
strong. If this be the origin of the tribe, Vittaria is an offshoot of 
Antrophyum and Monogramma is, in any case, derived through Pleuro- 
gramme from Vittaria. 
Goniophlebium 37 and Phymatodes are independent in origin from all 
the preceding offshoots of Polypodium and nearly or quite independent 
of each other. The former is a small, compact, uniform, therefore very 
natural group, confined to this part of the world, the union of which 
with the American so-called Goniophlehia is quite unwarranted. 
Phymatodes , 38 too, is a natural group, but very far from compact and 
uniform, being internally diverse and the parent of numerous highly 
specialized groups commonly treated as genera. In view of their un- 
stable venation and evidently broad affinities, it seems probable that 
the oldest representatives of Phymatodes are small ferns with simple 
fronds, the fertile and sterile alike, and without specialized fertile region 
( Craspedaria , in part). Such fronds are found on young plants of 
various Phymatodes species, more complicated at maturity. Grammitis 
is the probable source of this group. Within the Craspedaria group, 
specialization has taken place in the dimorphism of the fronds of most 
species, and in the nearly related species, P. accedens, the sori are 
restricted to the' specialized apex. I believe that both Hymenolepis and 
Niphobolus have been derived from some ancestor not very unlike P. 
37 J. Smith’s Schellolepis is the original Goniophlebium of Blume, P. cuspidatum 
being the first species mentioned, and the whole subgenus being Malayan. Flora 
Javae (1828), II, 132. 
38 If raised to generic rank, it must probably be called Selliguea. Pleopeltis 
would be an older name; but it (P. angustum) has certainly been distinct from 
Phymatodes much longer than Selliguea has. 
