274 EVANS. 



accredited it to Corral. He likewise cited M. Dusenii from the same 

 locality and described M. brevialata from San Pedro Island as a new 

 species. In the second paper he listed M. jmbescens from Tierra del 

 Fuego, although in his monograph he had restricted it to the Northern 

 Hemisphere. Other data regarding stations will be noted later in 

 connection with individual species. Strange to say Stephani made no 

 mention in either paper of M. chilcnsis, M. decrescens, or M. terricola, 

 in spite of the fact that he had used Dusen's specimens in drawing up 

 the original descriptions of these species. It might at first appear that 

 he had repudiated them, but this is clearly not the case, so far as M. 

 chilcnsis and M. terricola are concerned, because he has referred to 

 both of them in his later writings. In 1901, for example, he reported 

 M. chilevsis from Clarence Island in the Straits of Magellan (22, p. 4) ; 

 in 1911 he reported the same species from Chiloe and Juan Fernandez 

 (23, p. 10); and in 1916 he reported M. terricola from Bolivia (25, p. 

 180). These reports were based on specimens collected by Racovitza, 

 Skottsberg and Herzog, respectively. In his report on Skottsberg's 

 ample Chilean collections Stephani cited new stations for several other 

 species of Metzgeria and also listed, as an addition to the flora, M. 

 albinca Spruce, previously known from a single locality in Brazil. His 

 record was based on a specimen from Huafo Island. In 1917 (26, p. 

 47) he made his last contribution to our knowledge of the Chilean 

 Metzgeriae by proposing as a new species M. antarctica from Punta 

 Arenas, basing his description on a specimen collected by Von Schrenk. 

 According to his published writings Stephani recognizes twenty species 

 of Metzgeria as members of the flora. 



As in other large and natural genera of plants the species of Metz- 

 geria are often difficult to distinguish. This is partly because most 

 of the differential characters are drawn from variable structures and 

 partly because the plants sometimes remain for a long time in an 

 embryonic or juvenile stage of development, during which certain of 

 the specific features fail to reveal themselves. Var. ulmda Nees of 

 M. /areata, as shown by Goebel (5), is an excellent example of this 

 condition, and equally good examples occur among the Chilean species. 

 One unfortunate result of the difficulties involved in the determination 

 of specimens has been an accumulation of incorrect records, not only 

 in herbaria but also in the literature. This has been strikingly shown 

 by Schiffner (17) in the case of M. dichotoma (Sw.) Nees, a species 

 first described in 1788 and therefore one of the earliest, to be recognized. 

 In the Lindenberg Herbarium at Vienna he found eleven specimens 

 bearing this name. Three of these were from Jamaica and represented 



