238 C. SKOTTSBKRG 



south, one species ^oin<^ nortli as tar as Mexico and the West Indies, i extending 

 east from I^^ie^ia to the Falkhmd Is. and Tristan da Cunha. On the opposite side 

 of Antarctica are 18 species. 14 of tliese indit^enous in New Zealand, where 8 are 

 endemic, the remainder scattered alon^ tiie route Macquarie I.-Auckland and Camp- 

 bell Is.-Tasmania-Australia-New Zealand-Lord Howe and Norfolk Is.; one New 

 Zealand species reappears in Hawaii, another on Kerguelen and New Amster- 

 dam I., and Marion 1. has an endemic species. Finally, 2 species are found in New 

 (iuinea, one of them also reported from Borneo, and the Philippines have I en- 

 demic species, l^oth sections of sub^. liuriicijiia are represented in America, only 

 Stenandrae in the opj)osite sector. The monotypical subg. Pseudocarcx is Magel- 

 lanian. 



It is not easy to find one's way through the labyrinth of the enormous and 

 still (growing ^cnus Carcx, world-wide but unbalanced as the tropics are poor in 

 species in comparison with the temperate and cold zones. Our two island species, 

 the endemic C. hfrttrojiiivia and the south Andean Baiiksii, belong to different 

 sections and different geographic groups, the former to sect. Echinochlaenae: of 

 20 species 16 are endemic in New Zealand (one with a variety on Norfolk I.), 

 I in Australia, 1 in Tasmania; the little known Chilean C. laniprocarpa Phil, and 

 L . bfrti-yojiiaiia are far-flung outposts, but it lies near at hand to assume that they 

 or their ancestors migratetl across Antarctica. C. Bavksii belongs to PVigidae- 

 I'"uli<4inosae, a boreal group centering in luu'asia with one species in Pacific North 

 America, but I cannot tell if luviksii comes near this species. 



The s\stematic position of Juiuiia austraiis was briefly discussed in 22g. 

 109 and above p. 202, but whether we bring it to Morenieae or Iriarteae or let it 

 form a separate subtribe it remains a member of the neotropical element. HUTCIIIN- 

 S(>N followed Hi:n'1 HAM and HooKKR in placing it next to Ceroxylon, but this 

 genus is |)ol\gam()-monoecious and the stigma becomes basal in fruit. To me 

 Ckni/A 1 s opinion lacks foundation; he solves what he calls "a hopeless conflict 

 among taxonomists" (7/. 85) by deriving Juajiia from "a massive center of origin 

 of angiospermy at the Mascarenes " (p. 103). 



Ochai:;ai'ia and I It'sf^crooyeiiria belong to an Andean-neotropical assemblage 

 of genera, with close relati\es in Chile. 



To judge from Hr( IIK\ \i s monogra[)h of the Juncaceae in Pflanzenreich 

 f.uzula uiasafucnvia nuist be referred to a grouj) of Andean species, L. racentosa 

 Desv. (.Mex. S. Chile), cxcelsa Huch. (Boliv.), ///^vyw]'/;// Buch. (Argent.), /.^'^^^//V 

 Buch. (C^hile), and cliiloisis .\ees et Mey. (Chile, south to l^'uegia), but in the same 

 grou}) we find A. spicata (L.) DC. (Arct. circumj). and Alpine) and ahyssinica Pari. 

 (ICthiop., Brit. \\. Afr.), and the |)()ssibilit}' that the Andine species are of boreal 

 origin should be considered. On the other hand much speaks in favour of a south- 

 ern origin of Juncaceae: the subantarctic genera Marsippospennuui dind RostkovJay 

 the isolated rnoinuiu in South Africa, endemic Andine genera like Oxychloc and 

 Patosia, well marked endennc species of I.uzula in New Zealand and the Magel- 

 lanian region, and the subantarctic-bicentric Jinicns scluuchzcrioides group. It is 

 true that 4 of the 5 J// ;/r//.s- species reported from Juan Fernandez — some of them 

 perhaps not native— inhabit the Andes, 3 going east to Brazil, Argentina and Uru- 



