26 CARL SKOTTSBERG, MARINE ALG^E 2. RHODOPHYCEiE. 



Only a part of a medium-sized leaf. The cystocarps are situated as described by 

 Cotton, Crypt. Fälld. p. 184, on a multitude of very small proliferations from the veins. 

 One of these cystocarpophylls is figured, fig. 8 d, but there are others much larger, up 

 to 1 cm. Many of them are sterile. Even in D. sanguinea, there is no fundamental 

 difference between fertile leaflets and vegetative branches from the costa, as a cysto- 

 carpophyll, if fertilization does not take place, may develop into a leaf (Phillips, 1. c. p. 

 179). 



According to Agardh 1. c. III p. 496 tetraspores occur both on leaflets and on the 

 main frond. 



The apical growth is the same as in D. sanguinea (fig. 7 c), as is also the anato- 

 mical structure. I have examined Hooker's type material in this respect. It consists 

 of several fine specimens, one of enormous size, and I think his sanguinea and var. lan- 

 cifolia are the same thing. His declaration that the former could not be separated from 

 the northern sanguinea depends on insufficient knowledge of this. D. lancifolia is charac- 

 terized by the generally quite simple, large (to 40 cm or more) and delicate frond, the 

 coarse undulating lateral veins, and the numerous fertile leaflets borne on the veins of the 

 still intact lamina. The colour varies from a fine crimson to a dilute pink in the larger 

 fronds. Agardh, who quotes a Cape Horn specimen as type of his lancifolia telling, 

 however, that he had not seen Harvey's type (1 c), did not feel sure of the identity. 

 The specimen in Herb. Agardh consists of three leaves, but it is hardly possibly to tell 

 how these are united (Kylin in letter to the writer); it was collected in 1842 and undoubt- 

 edly belongs to Hooker's material, although it may have reached Agardh through 

 Herb. Gray. There cannot be much doubt that it is the same as the rest. 



Distribution: Fuegia, Falkl. 



D. lancifolia and epiglossum form Agardh' s genus Paraglossum. It should be 

 borne in mind that Agardh excluded D. sanguinea from the family, and that he chose 

 another type for the genus, viz. D. sinuosa: »et huic generi Deless. sinuosam quasi ty- 

 picam speciem hodiernis Delesseriis considerandam esse» (Sp. Alg. III: 3 p. 160). He 

 foimd that he could not bring his former section Paraglossum to Delesseria but had to 

 create a new genus for it, referred to the section Pteridieae. Certainly there are con- 

 spicuous specif ic differences between D. sanguinea and lancifolia, in the seasonal develop- 

 ment, the latter probably of one year's duration only, and in the shape of the fertile 

 leaflets, which are much less reduced in the latter, indicating a more primitive type, but 

 I am afraid there will be too much splitting if we attribute generic value to characters 

 of this kind. In D. fuegiensis we find transitions between special fertile leaflets and or- 

 dinary leaves, bearing cystocarps or tetraspores, and in D. epiglossum there are no such 

 leaflets at all, so that I judge it far better to keep them all under one genus. Future 

 investigations comprising all the species involved will show whether we shall be able 

 to establish natural sections. 



D. Larsenii Skottsb. in Kylin & Skottsb. p. 41, f. 20, 21 a. — Fig. 7 d — e, 8. 



Unfortunately, I did not obtain any fresh material of this. It seems to come close 

 to D. lancifolia. The mode of growth and the structure of the costa are the same, as 



