1 6 Bower . — Studies in the Phylogeny of the Filicales. VI. 
supposing that the later helicoid branchings are omitted. The three-lobed 
condition seen in Plate I, Fig. 3 of the base of a fertile leaf would result from 
a first dichotomy, of which the left-hand shank becomes the left-hand lobe. 
The right-hand shank branching again would give the right-hand lobe and 
the apparently central or terminal lobe. Here the latter is more distinctively 
marked from the other two than it is in N eocheiropteris or Matonia . But 
this may be held as an indication, or as a consequence of its further diver- 
gence from the phyletic source. The case is, in point of branching, parallel 
to that of the three-lobed juvenile leaves of Marsilia , noted by Goebel. 1 
As bearing on this interpretation of the three-lobed leaf of Leptochilus 
tricuspis, a very interesting example was pointed out to me by Dr. R. C. 
Davie, of which a photograph supplied by the kindness of the Director of 
the Royal Botanic Garden in Edinburgh is shown in Plate I, Fig. 2. Here 
a fourth lobe is present on the left-hand side, and a very rudimentary fifth 
lobe on the right. This leaf is thus two steps nearer the condition seen in 
N eocheiropteris or Matonia than the normal for the species. Such cases 
show that the dichotomous branching of the lamina, even where there is 
helicoid development, may grade into the pinnatifid, or pinnate with 
a definitely terminal lobe. It will be seen below that similar examples 
are not uncommon in this affinity of Ferns with superficial non-indusiate 
sori. 
Christ’s description of the sori of N eocheiropteris has already been cited. 
These are of very unequal size, being sometimes relatively small and 
circular, sometimes half an inch long or more, and the largest are usually 
near the base. They are disposed in two rows, one on either side of the 
midrib of the lobe. But where the lobe merges in the helicoid curve at the 
base, the sori are only continued on the acroscopic side (Plate I, Fig. 5). The 
relations of the sori are closest near to the base of the lamina, where, being 
of elongated form, they constitute in some specimens an almost continuous 
soral tract. With a limited supply of dried material it was impossible 
to make a detailed examination; but the relation of the sori to the venation 
has been observed. Christ states that the sorus is seated on the enlarged 
end of a vein. Sori of various sizes have been examined, and this has never 
been found to be the case. They are always seated on a continuous vein, 
or on a plexus of veins ; and there seems to be no definite relation of each 
sorus to the venation. For instance, in the two large sori shown in 
Text-fig. 8, a , the upper one is seated on a single large vein, with branches ; 
the lower is inserted on a large vascular loop, from which numerous strands 
radiate. Similarly in Text-fig, 8, b, the largest is again seated above a vascu- 
lar loop : the smaller sori may be on a plexus of radiating veins, or upon 
a single vein, which then continues its course. This indefiniteness of 
vascular supply is exceptional among Ferns. It probably points to a 
1 See Trans. Roy. Soc. Edin., vol. li, pt. 3, No. 21, 1916, ‘On leaf-architecture p. 675. 
