The correct name of the type species of Lophosiphonia has a problematic history. When establishing the genus, Falkenberg (in Schmitz & Falkenberg, 1897: 460) specified that the type was ``L. obscura (C. Agardh) Fkbg. (= Polysiphonia obscura C. Ag.), in den wärmeren Teilen des atlantischen Oceans''. In addition, he mentioned ``etwa 3 andere (ungenügend bekannte) Arten in verschiedenen wärmeren Meeren''. In his monograph, however, Falkenberg (1901: 495) stated that the type was L. subadunca (Kützing) Falkenberg, based on Polysiphonia subadunca Kützing (1843b: 418; type locality: Corsica). While this subsequent typification may be ignored, a fundamental problem remains: to which species does C. Agardh's name apply?
It should be pointed out immediately that C. Agardh did not publish the name Polysiphonia obscura. That author (C. Agardh, 1828: 108) described the species as Hutchinsia obscura on the basis of a collection from Cádiz, Atlantic Spain, the transfer to Polysiphonia being made by J. Agardh (1842: 123). Although C. Agardh did not mention the number of pericentral cells, J. Agardh specified 5--7. In the protologue of the generic name, Falkenberg did not specify a definite number of pericentral cells, stating merely ``(meist zahlreiche)'', but in his monograph he specified 6 for L. subadunca and 12--18 for L. obscura (Falkenberg, 1901: 496, 500).
Howe (1920: 574) first pointed out Falkenberg's confusion, having found that an authentic specimen of Hutchinsia obscura in the Agardh Herbarium (LD) is conspecific with L. subadunca. The confusion seems to have been initiated either by J. Agardh in his selection of a representative specimen that he sent to Harvey, or by Harvey (1846 [1846--1851]: pl. CII.A) in his interpretation of that specimen. In any event, Polysiphonia obscura as illustrated and described by Harvey has 12 or 13 rather than 5--7 pericentral cells. Kützing (1849: 808; 1863a: 13, pl. 40: figs. a, b) perpetuated the confusion by specifying 11--14 pericentral cells for P. obscura while J. Agardh (1863 [1851--1863]: 943) contradicted his previous statement by attributing 12 pericentral cells to this species.
Although Howe had shown that the name Lophosiphonia subadunca should be replaced by L. obscura, he did not make the change, probably because he thought that it would lead to further confusion. The species to which Falkenberg had misapplied C. Agardh's name was listed as L. obscura auct. Howe's solution is not nomenclaturally acceptable, however, and the type species of the genus must be called L. obscura (C. Agardh) Falkenberg regardless of Falkenberg's misapplication of the name.
For the species with the larger number of pericentral cells, Kylin (1956: 539) adopted the name L. reptabunda (Suhr) Kylin, based on ``Polysiphonia reptabunda Suhr in Kützing 1849 S. 806''. The identity of this species is uncertain. Nomenclaturally, the first step was the description of Hutchinsia reptabunda Suhr (1831: 684) from Biarritz, Atlantic France. This species was transferred to Polysiphonia by Kützing (1843b: 417), who said it had 10 pericentral cells but gave no provenance. In a later treatment, Kützing (1849: 806) repeated his earlier description and gave as provenance ``In mari atlantico subtropico ad Algas parasitica'', while still later (Kützing, 1863a: 12) he wrote ``Ad oras africanas''. There seems to be no published information on the type of Hutchinsia reptabunda Suhr. Maggs & Hommersand (1993: 382) erred in treating Polysiphonia reptabunda ``Suhr ex Kützing'' (1843b: 417) as a new species with a holotype in Kützing's herbarium (in L). Polysiphonia adunca Kützing (1849: 808; 1863a: 13, pl. 40: figs. c--e), based on a collection by Solier from Marseille and said to have 10--15 pericentral cells, is a likely synonym of L. reptabunda.
Wynne (1986b: 2248, 2261) attributed the name Lophosiphonia obscura to (J. Agardh) Falkenberg, indicating by referring to Art. 63.1 (cited in error as Art. 73.1) of the Sydney Code that Hutchinsia obscura C. Agardh was superfluous when published and hence illegitimate. Polysiphonia obscura J. Agardh (1842: 123) was treated by Wynne as a priorable nomen novum in accordance with Art. 72.2, Note 1 (Art. 58.3, Tokyo Code). The element in the protologue of H. obscura that would seem to render that name superfluous is the citation ``Conferva intertexta, Roth Cat. II. pag. 188. tab. 3. fig. 5''. This citation entails a bibliographic error, since Roth's name was published in the first rather than the second volume of his Catalecta botanica. Previously, C. Agardh (1824: 164) had placed Conferva intertexta Roth in the synonymy of Sphacelaria cirrosa (Roth) C. Agardh (q.v.). However, Conferva intertexta Roth (1797: 188--189, pl. III: fig. 6; type locality: unspecified but probably the Mediterranean Sea) is a later homonym of C. intertexta Withering (1796: 132), so that C. Agardh was not obligated to adopt Roth's name. (The identity of C. intertexta Withering, based on collections by Velley from Weymouth (Dorset) and by Stackhouse from Cornwall, is not known.) Two interpretations of the name Hutchinsia obscura are thus possible: it could be treated either as a nomen novum for Conferva intricata Roth or as the name of a new species. Since the type of C. intricata Roth probably is not extant while the specimen from Cádiz upon which C. Agardh based his description of H. obscura is in the Agardh Herbarium (in LD), nomenclature would be best served by treating H. obscura as a new species.