DERIVATION OF THE FLORA AND FAUNA 209 



See also 112. F. B. H. Brown (j',-. III. 262) suggested an affinity between cordifolia 

 2iV\dfatii/ih'e7is7s F. B. H. Brown (Marquesas), but they belong to different subgenera 

 and the latter is referred to X. fragrans Hook, from N. Caled. as a variety {112). 



Scrophulariaceae. 



Mimulus L. About 130; widely dispersed, [)redominantly western N. Amer. (80 

 Calif.), south to S. Chile; 2 Afr., Madag.. 4 E. Asia, 5 Austral., Tasm., N. Zeal. 



glahratus H. B. K. N. Amer. to Boliv., Argent, and Chile; polymorphous, the 

 island form very close to var. parvijiorus (Lindl.) Grant [114). 

 EupJirasia L. About 1 00; bor.-circump. with isolated populations in the S. Andes, 

 south to P'ueg. and Falkl., and in Austral.-X. Zeal., the gap between E. Asia 

 and Australia bridged over by trop. mountain stations; see map 7<5'. 224. 



formosissima Skottsb. The fact that this species is very unlike the Trifidae 

 of Chile makes it particularly interesting. In 22g. 169 I emphasized the differ- 

 ence between forjuosissiuta and the Australes of X. Zealand and placed it nearer 

 to the boreal Semicalcaratac. W'ETTSTEIX, in his contribution to my paper (I.e. 209), 

 expressed the opinion that it could be attached to a Japanese group of species, 

 intermediate between Australes and Semicalcaratac. The question was taken up 

 by Dr RlETZ who thinks, with good reason I believe, that I overrated its rela- 

 tions to boreal species (77. 533) and that, in most respects, it is more nearly related 

 to X. Zealand forms. 



Plantaginaceae. 



Plavtago L. About 270; world-wide, essentially temp. 



fernandezia Bert. The nearest relative appears to be P. princeps Cham, et 

 Schld. of Hawaii. Both belong to sect. Palaeopsyllium, scattered over the S. hemisph. 

 and extending north to X. Amer., Hawaii and S. Eur.: X. Amer. 5, S. Amer. i, 

 S. Eur. 2, Afr. 5, St. Helena i, Madag. i, Lord Howe I. i, Auckl. Is. i, Rapa 2, 

 and Hawaii 9. 



tnincata Cham. Centr. and S. Chile. PiLGER distinguished the island form as 

 a separate subspecies close to ssp. firnia Pilger, but I doubt that it deserves the 

 rank assigned to it, and it is even possible that it is a very late arrival in the 

 islands. 



Rubiaceae. 



Hedyotis L. A large pantrop. genus. 



thesiifolia St. Hil. Trop.-subtrop. S. Amer., in Chile south to Chiloe. Very 

 likely not truly indigenous in J. Fern. 



Nertera Banks et Sol. 10-12; centering in X. Zeal. (4 endem. sp.), north to Malaya 

 and Haw^aii; Tristan da C; S. Amer. from Falkl. and Fueg. to Colomb. and Mex. 



granadejisis (L. fil.) Druce. A polymorphous species, reported from S. Amer., 

 Tristan da C, Austral., Tasm., X. Zeal., Java and Hawaii — see 244, where I pointed 

 out that it is heterogeneous but that the plant from ]\Iasafuera seems to be iden- 

 tical with the form common in Magell. and Falkl. 

 Coprosma Forst. 90; a western Pacific genus centering in X. Zeal. (39) and distri- 



14 ~ 557857 The \at. Hist, of Juan Fernandez and Easier Isl. Vol. I 



