CHILEAN SPECIES OF METZGEFJA. 277 



pubescens by Taylor and Hooker, 27, p. 445); Staten Island, 1882, 

 Spegazzini 65 in part (Massal., Y., 11, p. 257). 



The following additional stations may be cited from the literature: 

 Hanover and Atalaya Islands, Skottsberg (24, p. 10); Clarence and 

 Hoste Islands, Hariot (2, p. 247); Tuesday Bay and Punta Arenas, 

 Straits of Magellan, Naurnann (16, p. 43); Brecknock Pass, Spegaz- 

 zini (11, p. 257); Tekenika Bay, Almirantazgo, and Lake Fagnano, 

 Tierra del Fuego, Skottsberg (23, p. 9, and 24, p. 10) ; Mount Sarmiento, 

 Tierra del Fuego, Spegazzini (11, p. 257). 



The presence of hairs on the dorsal surface of the thallus is the most 

 remarkable feature of M. frontipilis and will at orice distinguish it 

 from all the other known species of the genus except M. pubescens. 

 In the northern species, however, the thallus is scarcely convex and 

 the hairs are equally abundant on the ventral surface, being situated 

 on both costa and wings ; whereas in M . frontipilis the thallus is usu- 

 ally distinctly convex and the ventral hairs are largely restricted to 

 the costa, the wings being almost or entirely free from them. To a 

 certain extent the female branches offer an exception to this descrip- 

 tion, so far as the arrangement of the hairs is concerned, since in these 

 both surfaces are equally hairy throughout. The male branches are 

 still unknown. 



Other noteworthy characters are derived from the costa and espe- 

 cially from its cortical cells, which show marked variations in number. 

 According to Lindberg's original account these cortical cells are in 

 eight to twelve rows both dorsally and ventrally. Stephani (19, p. 

 932), with a larger series of specimens at his disposal, places the 

 extremes at six and eighteen and associates the lower numbers with the 

 branches and the higher with the main axis. As a matter of fact only 

 four rows are present in some of the slender branches studied by the 

 writer. Stephani brings out in addition the interesting fact that the 

 wings are sometimes two or three cells thick at their junction with the 

 costa, a feature overlooked by Lindberg. 



Although the published descriptions of M. frontipilis state emphati- 

 cally that the ventral surface of the wings is wholly naked, this condi- 

 tion is by no means always realized. The hairs show a tendency to 

 encroach, as it were, upon the ventral surface, not only from the 

 margins but also (less frequently) from the costa. This tendency 

 sometimes expresses itself by a very slight displacement of the mar- 

 ginal hairs, so that they are not truly marginal (Fig. 1, A); but the 

 displacement may be much more marked than this, so that the hairs 

 appear on the ventral surface one, two, three or even four cells away 



