THE PHANEROGAMS OK THE JUAN FERNANDEZ ISLANDS 163 



Boraginaceae. 

 Selkirkia Hemsl. 



94. S. Bertcrii (Colla) Hemsl. — JOHOW, Estud. 85. 



Masatierra: On the steep rocky slopes above the forests in the centre 

 of the island, rare. Co Damajuana, 500 — 550 m, very rare; the gap between 

 Damajuana and Co Ynnque, 580 m, few specimens (past fl. 18 /i2 16, no. 132); 

 northwest face of Co Piramide, narrow rock ledge, c. 600 m, few shrubs; Porte- 

 zuelo de Villagra, near the pass (also observed by JOHOW), a couple of speci- 

 mens, 570 m; O. del Monte Maderugo, rock wall 390 m, rare (past fl. 24 /i 17, 

 no. 341). 



Area of distribution: Endemic in Masatierra; monotypic. 



Verbenaceae. 

 Rhaphithamnus Miers. 



95. Rh. venustus (Phil.) Skottsb. — Syn. Citharexylon venustum Phil. 

 Anal. Univ. 1856; Rh. longiflorus Miers, Trans. Linn. Soc. XXVII (1870) 98; 

 JOHOW, Estud. 78. 



Masatierra: in the wooded region, extending from Pto Frances to the 

 south slope of Co Chumacera, hardly ever seen below 200 m and much more 

 common higher up; also in the shrubberies on the ridges. Fl. 3 /ia 16, no. 11; 

 7 /i 2 16, no. 40; 28 /i2 16, no. 198; fr. 21 /4 17, no. 625, 2i /± 17, no. lib. — A 

 form with red flowers was found near Plazoleta del Yunque, c. 250 m. 



Masafuera: Guajardo without locality. — In the forests, much more 

 scarce than in Masatierra; on the Sanchez plain, 515 m; Q. del Mono, 475 m; 

 Q. del Blindado, 440 m (fl.-fr. 19 /2 17, no. 516). 



Area of distribution: Endemic. 



Labiatae. 

 Cuminia Colla. 



While HEMSLEY, Challenger Report 51 lists three species, eriantha, fer- 

 nandezia and brevidens, JOHOW reduces brevidens to fernandezia. I have seen 

 the material on which C. brevidens was based by BENTHAM, and I think that 

 JOHOW is right. The length of the calyx lobes is submitted to some variation; 

 I admit that they are very short in C. brevidens, but otherwise no perceptible 

 difference is found. 



HEMSLEY ascribed unisexual flowers to Cuminia, JOHOW denied the cor- 

 rectness of this statement and suggested that HEMSLEY was misled by the 

 pronounced dichogamy of the flowers. My examination of numerous living 



