80 BOTANICAL RESULTS OF THE SCOTTISH NATIONAL ANTARCTIC EXPEDITION. 



Habitat. Scotia Bay, South Orkneys, July 1903. 



This species is most nearly related to Delesseria sanguinea, Lam. , from which it 

 differs in having the cystocarps not smooth, but more or less clothed with simple 

 tapering appendages, chiefly disposed around the sides, and leaving the top bare. But 

 for this wreath of appendages we should regard the plant as no more than an old narrow- 

 leaved and very proliferous form of D. sanguinea. 



The question whether or not H. stephanocarpum belongs rightly to the genus 

 Delesseria depends on the view held of the systematic position of D. sanguinea 

 ( Hijdrolapatliwn sanguineum, J. Ag.), with which species our plant must stand or 

 fall. Agardh, attaching primary importance to the structure of the mature fruit, 

 regarded Hydrolapathum as forming a distinct genus in Rhodymeniacese near Rltodo- 

 phyllis, on account of its composite cystocarp with " nucleoli " separated by radiating 

 columns of sterile threads, and on account of the carpostomium-structure. Schmitz, on 

 the other hand, reunited Hydrolapathum with Delesseria on account of the similarity 

 of procarpial development, which is a more primitive character than the mature fruit 

 on which Agardh founded his conclusions. If we follow Schmitz and De Toni, our 

 plant would be called Delesseria stephanocarpa. Our own inclination is, however, 

 to follow a middle course. Instead of sinking Hydrolapathum into Delesseria, from 

 which so many less well-marked genera have been quarried, we would maintain 

 Hydrolapathum as an independent genus on the score of the structure of its cystocarps 

 and sporophylls ; but we would place it in the Delesseriese, and not in Rhodymeniaceae. 



25. PTERIDIUM PROLIFERUM, A. and E. S. Gcpp in Journ. ofBot., xliii., 1905, p. 107, 

 tab. 470, figs. 7-9. 



Frons fruticulosa, circa 12 cm. alta, alterne dichotoma (sed ramificatio ob prolifi- 

 catioues copiosas obscura) ; rarni complanati, costati, alati, costa inferne conspicua, 

 superne attenuata, omniuo sine venis lateralibus ; rami ramulique laciuiati, a marginibus 

 costaque prolificantes, alterne et irregulariter dichotomi. Ramuli ultimi membranacei, 

 ligulati vel cuneato-ligulati, usque ad apices obsolete et simpliciter costati, irregulariter 

 lacerati vel grosse dentati, prolificantes. Cellulse pagiualcs homceocystideae omnes 

 rotundato-angulatse. Tetrasporangia sine online utroque latere costee phyllorum par- 

 vorum disposita, soros nee in unum conflueiites, nee ad apicem attinentes formantia. 

 (Figs. 15-17.) 



Habitat. Scotia Bay, South Orkneys, 9-10 fathoms, May 1903. 



We should have preferred to style our plant simply Delesseria prolifera, using 

 Delesseria in the old wide sense. But that genus, as emended by J. G. Agardh, is now 

 so limited in its scope that we are compelled to refer the plant to Pteridium, although 

 we regard it and certain other genera latterly split off Delesseria as too nearly allied to 

 be worthy of generic rank. In our species the mode of branching is very much masked 

 by the abundant proliferations. It is in habit most like P. alata and P. pleurosporum, 

 but differs from the former in being much more irregularly branched, and in having no 



