200 CARL SKOTTSBERG 



Johow based his opinion on BERTERO himself, who was inclined to regard it 

 as a noticeable form of pinnata. While I regard pruinata as a species, I follow 

 Johow with respect to Berteriana. Finally, Johow added one new species, 

 D. gigantea, which has remained very little known until now. 



Returning to D. macrophylla, it is evident that Rea macrantha is different 

 from this. Both have incised leaves, but their shape is not the same and there 

 are many differences in the flower and fruit. D. macrophylla is endemic in 

 Masafuera, macrantha in Masatierra. Rea marginata is certainly distinct from 

 both, not only in the leaves, which are entire, but also in flower and fruit, 

 described here for the first time. Finally, var. marginata of JOHOW, at home 

 on the coast rocks, is distinct from the true marginata and will receive the 

 new name D. litoralis. Thus, D. macrophylla in Johow's sense includes four 

 species, one from Masafuera, two confined to Masatierra and one to a rock off 

 Masatierra and to Santa Clara. 



These four species differ from the rest in many respects. The stem is 

 branched, hollow with numerous diaphragms, the receptacle is also hollow, its 

 surface plane, with numerous fibrils round the points of insertion of the achenes. 

 The head is very large, the flowers bright orange, the achenes distinctly 

 winged, the pappus formed by short, irregular, ± rufous bristles. I am pretty 

 sure that the ssplittingup school* would reserve the name Dendroseris for these 

 four species. My reasons for not doing so were explained under Robinsonia. 

 The remaining species have creamy white flowers and unvinged, obovoid, 

 slightly compressed achenes; here the accordance between them ends, and they 

 belong to two rather different types. In D. pinnata and D. regia (nov. spec.) 

 the stem is simple, but otherwise as in D. macrophylla; the receptacle (unknown 

 in D. regia) is hollow, but concave and only slightly fibrillose. The remaining 

 species (D. micrantha, pruinata, neriifolia and gigantea) have a branched and 

 solid stem, and a solid, concave and naked receptacle. D. pinnata and regia 

 are intermediate between the macrophylla-type and the micrantha-type. Either, 

 we should retain Dendroseris in the old sense, or three genera must be 

 recognized. Further, D. gigantea differs from all its congeners in the shape of 

 the corolla. In view of these circumstances I propose to subdivide the genus 

 in the following manner. 



I. Stem branched, hollow. Head 3 — 5 cm wide; receptacle fibrillose; flowers 

 orange yellow; achenes winged Subgen. I. Eudendroseris Skottsb. 



A. Leaves chartaceous, grossly dentate-lobate. 



1. Leaves oblong-ovate, ± regularly lobate. Anthers dark. Achenes 

 mostly broader than long. Cotyledons entire. 



D. macrantha (Bert, et Dene) Skottsb. 



2. Leaves often as broad as long, irregularly lobate-serrate. Anthers 

 yellow. Achenes twice as long as broad. Cotyledons trilobate. 



D. macrophylla D. Don. 



B. Leaves coriaceous, entire. 



1. Anthers dark. Achenes thick. 



D. marginata (Bert, et Dene) Hook, et Arn. 



2. Anthers yellow. Achenes very flat. D. litoralis Skottsb. 



