26 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 
now call Hemitelia, basing his treatment mainly upon characters 
offered by supposed differences in the arrangement of the fibrovas- 
cular bundles of the stipe and upon venation. The genera recognized 
by him are: (1) Hemitelia, with two sections, Notophoria and 
Kuhemitelia; (2) Microstegnus, founded on Cyathea grandifolia 
Willd., in part; (8) Hemistegia with six “species”; (4) Actinophlebia, 
with two species; and (5) Cnemidaria, which he restricted to the 
original type species, C. speciosa, the other species having been 
removed by him to Actinophlebia and Hemistegia. The results 
of his investigation can not be regarded as satisfactory from any 
point of view, the principal objections being that, despite his 
undoubted keenness of observation, the work abounds in error 
due apparently to a willingness to accept many published observa- 
tions and citations of specimens without substantiating them, a 
marked tendency to overestimate the systematic value of trivial 
characters, a lack of sufficiently extensive and complete material, 
and a singularly inadequate conception of the requirements of nomen- 
clature, especially as regards the use of species names. Here and 
there are found in it observations and distinctions of merit; but to 
attempt, in review, to distinguish these from the many fallacious 
statements of fact or to explain in detail the various taxonomic 
errors, would be no simple task nor lead to any very useful result. 
In the writer's opinion the genus Hemitelia should be of sufficiently 
wide extent to include all these “genera” of Presl. In fact, there 
is much to be said in support of Mr. Copeland’s recent proposition ! 
to unite Hemitelia, Alsophila, and Cyathea in a single genus. The 
writer prefers for the present, however, to retain the three genera 
in their traditional sense, partly from practical considerations. 
Adhering to this view it is possible to recognize two fairly distinct 
sections of Hemitelia in the American tropics; (1) Euhemitelia and 
(2) Cnemidaria; the first, embracing large species of truly arboreous 
growth, with mainly tripinnatifid fronds and narrow, often rather 
minute segments; the latter, plants which are scarcely arborescent, 
with ample, pinnate to bipinnatifid (or rarely tripinnatifid) fronds, 
the leafy parts broad and little dissected. 
Professor Underwood, who had undertaken a study of this group, 
regarded Cnemidaria as a valid genus, its “typical members having 
a basal areole or arch formed by a union of veins rising from adjacent 
series of primary veins” of the pinne. But in about half of the 
species the veins are free, ordinarily. The indusium characters are, 
it is true, fairly constant for the species of Cnemidaria, but there is 
closer agreement in this particular between these and H. multiflora 
than there is between H. multiflora and some of its near relatives in 
Kuhemitelia. Undoubtedly there is in habit and leaf shape a close 
* Philippine Journ. Sci. C. Bot. 3: 353. 1908. 
