128 CARL SKOTTSBERG 



tion. The colour of the lower Surface shows the same considerable variation 

 as in other forms. The flowers are not dioecious, but bisexual and quite typical, 

 and do not differ in size or shape from Chilean specimens. The only character 

 remaining is the clustered leaves, but even in this case there is a considerable 

 variation between different trees and also between different branches of the 

 same tree. Generally, the short internodes give to the island Drimys a somewhat 

 different habit, but cultivation alone will decide upon the value of this character. 

 A distinct tendency to the development of a form with shorter internodes was 

 observed in the introduced Aristotelia maqui, both by JOHOW (1. c. 108) and by 

 the writer. 



Area of distribution: Not counting the tropical forms, D. Winteri in- 

 habits Chile from Frai Jorge (an isolated locality) to Cape Horn, and Juan 

 Fernandez. 



Lactoridaceae. 

 Lactoris Phil. 



59. L. fernandeziana Phil. — Johow, Estud. 114. — Plate 12, fig. 1. 



Masatierra: in the damp montane region, confined to the dense and humid 

 forests above 500 m, rare. — C. Chifladores, steep slope above Pto Frances, 

 c. 500 m, about a dozen plants (past fl. 17 /4 17, no. 622); forests of El Yunque 

 (Johow); V. Colonial, C. Central, one plant 510 m, several 520 m (fl. 18 /i 17, 

 no. 303); Portezuelo de Villagra (also observed by Joi-iow), one small specimen, 

 570 m; mountains at the back of Pto Ingles (Johow); mountain ridge west of 

 El Yunque, c. 500 m, very rare (fr. 24 /4 17, no. 629); highest part of the Villagra 

 Valley below Portezuelo, some fine shrubs at the roadside, 510 m (fl. 8 /i 17, no. 230). 



Area of distribution: Endemic in Masatierra. 



Cruciferae. 

 Cardamine L. 



60. C. chenopodiifolia Pers. — Syn. C. fernandeziana (Phil.) Joh.; Johow, 

 Estud. 1 10. 



Masatierra: Germain; Reed! 



Unfortunately, no trace was seen of this interesting plant. Germain col- 

 lected it in October, 1854, Reed in September, 1872; it is an annual, developed 

 in the spring and probably of short duration, and this explains that we did not 

 find it. Perhaps it did not appear at all in 1916 on account of the very dry 

 spring. It is also possible that it has become very rare or that it has disap- 

 peared altogether. As it is unknown in Peru and Chile, we have no good 

 reason to regard it as accidentally introduced. 



I am convinced that SCHULZ (ENGLER's Jahrb. XXXII.443) was right in bringing 

 Heterocarpus fernandezianus Phil, to Cardamine chenopodiifolia; when comparing 

 the former with material of the latter from Uruguay, Brazil and Bolivia, I came 



