THE HEPATICAE OF THE PRINCE EDWARD ISLANDS. II. ON GYMNOCOLEOPSIS (SCHUST.) SCHUST., LOPHOZIA CYLINDRIFORMIS (MITT.) STEPH. AND THE SUBGENERIC CLASSIFICATION OF THE GENUS LOPHOZIA DUMORT.

Abstract

  Gymnocoleopsis (Schust.) Schust., described for a sole species (Lophozia multiflora Steph.) of the high mountains of the Neotropics and of Africa, is shown to include three species and two subgenera, Gymnocoleopsis s. str. and Xenolophozia (Schust.,) Schust., comb. n. G. cylindriformis (Mitt.) Schust. [based on Jungermannia (subg. Lophozia) cylindriformis Mitt.] and G. capensis (S. Arn.) Schust. (based on Lophozia capensis S. Arn.) are discussed in detail. All three taxa agree in producing gemmae; in the exclusively lateral-intercalary branching; in the autoecious inflorescences; the usually monandrous ♂ bracts without a dorsal basal tooth; the uniseriate antheridial stalks; and the 8＋4-seriate seta of the sporophyte. They all, also, appear to have bistratose capsule walls with, at best, very local differentiation of a thin, third, interior stratum. Careful study of these taxa, which had all been assigned to Lophozia (or Jungermannia subg. Lophozia), shows that an intimate affinity of Gymnocoleopsis to Lophozia exists. Their study, again, illuminates the continuing problem of how to delimit the last genus. The solution of proposing genera for every isolated species or cluster of species is rejected.

  The evolution of inflated perianths, similar to those of Gymnocolea, is regarded as an apomorphic feature, acquired independently in the holarctic genus Gymnocolea. The latter differs from Gymnocoleopsis in: (a) the exclusively Frullania-type lateral branching; (b) lack of a bracteole in the gynoecium; (c) lack of gemmae.