GA Dn a i RRS RL 5 aie eA Se aoe uy pee ae ‘gehins ens ee was Pi ih aia a Nee Ta 
5 
previously defined. It was necessary, therefore, to create two new subgenera, Para- 
polystichum and Polystichopsis, which are both well characterized and in several 
features very different from the other subgenera. I shall here briefly mention a 
single one of the characters, by which they are distinguished from other decomposite 
species, viz. the architecture of the leaf as it appears from the arrangement of the 
divisions of the 2nd, 3rd and 4th order. MerTrenivs has clearly seen the impor- 
tance of that character, and in the introduction to his monograph on Aspidium und 
Phegopteris he has given a detailed description of the different kinds, in which the 
ribs and veins are arranged in the two genera adopted by him. His division into 
Polystichoidee, Phegopteroidee, and Cyatheoidew, is based upon that character. The 
classification of METTENIUs has, however, only slightly influenced later pteri- 
dologists in their grouping of the species and only few (f. inst. KeyseRLInG and 
HILLEBRAND) have used Mertrenius’ character. The neglect of a _ character, 
which was pointed out by a pteridologist of considerable perspicacity, is probably 
due to the circumstance that other characters (presence or abscence of indusium, 
venation and others) have been considered the most important as characters of 
groups, compelling the authors to group together species that as to architecture 
are utterly different; also it is probable that the character has been underesti- 
mated, because it is scarcely available when bipinnatifid species are concerned. 
I have found, however, that METTENIUS in this case, as in several others, 
understood more clearly than several later pteridologists how to estimate, which 
characters are to be considered the most important ones. Having worked during 
a long time, as I have, with a very large material of composite species, one is 
struck by the fact that the architecture of the lamina is the very character of the 
utmost importance for the right understanding of how the species may naturally 
be grouped together. It was rather difficult to Merrentus to bring all species 
under his three groups, and he was compelled to create subdivisions connecting 
them. He wanted to prove that the arrangement of the ribs of all divisions of a leaf 
is ruled by immutable laws, which it certainly is not; still his main-divisions are 
right. With special regard to the American species the most important difference in 
the architecture of the decomposite leaves may be characterized as follows, 
t must always be remembered that a stable arrangement of the ribs can be 
found in leaves only that are divided to a certain degree, at least twice. The se- 
quence of the primary pinne is unavailable, and the most often lengthened 
pinnule of the basal pinne are equally to be excepted. Of special importance is 
the arrangement of the secondary pinnules on the medial primary pinne, 
not only because it is constant within the ‘same species, but also because it is 
always the same in all species which by other characters appear to be closely related. 
This arrangement is of two different types: 
1. The first secondary pinnula of a medial primary pinna is the eiiisc doe 2. 
that directed toward the base of the leaf; the ——— is catadromous, 
the lamina cyatheoid (fig. 1). 
