960 BROADHURST: STRUTHIOPTERIS IN NORTH AMERICA 
Other possible divisions of this large genus, based upon cha- 
racters offered by the sterile fronds, the indusium, and the spo- 
rangia and spores, have been considered by the writer and found 
unsatisfactory for the following reasons: (1) The variations in the 
sterile leaves in proportion, cutting, and texture depend greatly 
upon the age of the leaf and of the plant; (2) in the fertile leaves 
the indusium remains entire in the pinnatifid species, and though 
it usually becomes lacerate or fimbriate in the pinnate species, these 
show all degrees of laceration; (3) a careful microscopic study 
of the spores and sporangia, which Fée apparently considered 
important, reveals no constant differences in the North American 
species. 
The greatest variation is undoubtedly in the habit, for the 
North American species include: (1) Tree-climbing forms with 
pendent leaves, (2) terrestrial, cespitose forms, (3) terrestrial, 
inclined or erect forms with stems 5-45 cm. long, and (4) sub- 
arboreous forms with stems at least 20 cm. high and 12 cm. thick. 
These differences we are at present unable to correlate with any 
other character. The habits of some species are not definitely 
known, because collectors have so often brought back incomplete 
specimens without field notes; mention of the habit is rarely 
included in the original description; and further, even American 
writers have confused our own species with the descriptions of 
African and other foreign ones and attributed to them characters 
they never possess. In making the key for this genus, habit has 
therefore been almost ignored; the key is strictly artificial and 
based mainly upon the frond characters rather than upo? those 
of the whole plant. : 
The following key and descriptions have been based chiefly 
upon the specimens in the herbaria of the New York Botant 
Garden and the United States National Museum. Begun under 
the direction of Professor Lucien M. Underwood, this paper as 
been completed with assistance from Mr. William R. Maxon 
and Dr. Ralph C. Benedict. Dr. H. Christ and Dr. Be Ps 
Hochreutiner also kindly sent tracings or material from 
herbaria. Access was freely given to the herbaria at Gentva 
Kew. 
aut 
