THE FERN GENUS CHEILANTHES IN CONTINENTAL UNITED STATES John T. Mickel The New York Botanical Garden, Bronx, NY 10458 Ever since Cheilanthes was described by Swartz in 1806, botanists have had great difficulty, in delimiting the genus. It is clearly part of a laxge complex in which the species are fairly distinct but the generic limits are unclear. Customari-ly three larger genera have been recognized — Cheilanthes , Noth-olaena, and Pellaea — and occasionally several small splinter genera have been recognized as well. However, there is a great problem in distinguishing Cheilanthes from Notholaena and to a lesser extent from Pellaea, It is the purpose of this paper to excimine the distinction between Cheilanthes and Notholaena. Classically, Cheilanthes has been distinguished from Noth-olaena by the former having a recurved, differentiated margin to protect the sori, whereas Notholaena possesses a plane or slightly recurved, undifferentiated margin (many authors, in-cluding Cronquist et al. , 1972; Dittmer et al. , 1954; Knobloch cuid Correll, 1962; Munz, 1965). In preparing keys to the ferns of the united States, it has become quite clear to me that" the two genera cannot in fact be distinguished using this character. Careful examination of the species of Cheilanthes and Notholaena shows that many species of Cheilanthes in fact do not have a well differentiated margin. Frequently, the margin curves back slightly without developing a false indusium and does not cover the sori. In most species of Notholaena, the margin curves back in precisely the same manner. In the northwestern United States, for exan^Jle, Notholaena parryi is distinguished on the meurgin character from Cheilanthes feei (Cronquist et al., 1972), but the margins are so covered with hairs that only careful observa-tion will show that the margins of both species are the same. Another cuiomally lies in that Cheilanthes cooperae has never been placed in Notholaena even though its margin is almost per-fectly plcine and entirely undifferentiated. Efforts have been made to find additional characters, such as stipe and blade suiatomy, hairs, spore morphology, euid chromo-some numbers (Knobloch, 1969; Knobloch et aL, 1975; Knobloch and Vblz, 1964, 1968; Lellinger, 1965; Tryon and Tryon, 1973), but none have been found. A basic problem seems to lie in the assumption that Notholaena and Cheilanthes are indeed sepaurate genera and that therefore there are characters to be found to distinguish them. I think this is a misconception. I have examined the plants thoroughly and ccin find no character that will support the maintencuice of Notholaena as ccxranonly circum-scribed. Rather, I see distinct g'roups of species that can be
BioStor
