38 CARL CHRISTENSEN AND CARL SKOTTSBERG 



thinly herbaceous, but specimens growing on rocks (nos. 318, 1 148) have a 

 nearly coriaceous leaf with thick cartilagineous edges and the sori more di- 

 stinctely immersed. 



As stated above the species varies extraordinarily in the degree of cutting. 

 The type, first described under the name intermedium (or translucens), is con- 

 fined to Masatierra. Only once, an aberrant form was found growing on rocks 

 (no. 318); this corresponds rather to var. basicompositum from Masafuera. When 

 JOHOW states that in Masatierra the species shows »lamina pinada-tripinada», 

 he must have drawn his conclusions from Hemsley's plate, for he never col- 

 lected a much divided form in Masatierra. The plants figured by HEMSLEY 

 fig. 2, 3 were collected by Mr. DOWNTON in »Juan Fernandez*. They represent 

 two Masafueran varieties and were undoubtedly collected in Masafuera, this 

 island also having been visited by DoWNTON, and not in Masatierra, where 

 they do not occur. Masafuera is inhabited by a number of forms connected 

 by transitions. As they grow side by side, sometimes two widely distant ones 

 on the same branch, it is impossible to regard them as merely results of local 

 influences. Either, the species is composed of a number of units, elementary 

 species, giving the impression of a more or less continuous series, or we have 

 a few subspecies and their hybridogenous offspring. From taxonomic point of 

 view we have found it advisable to distinguish the following phaenotypes. 



1. typicum. 



Endemic in Masatierra, not uncommon as an epiphyte on trees, generally 

 high up above the ground. Fertile during the summer and autumn. 



Masatierra: generally found on Drimus Winteri, but rarely below 400 m, 

 e. g. on the ridge between Q. Laura and Q. Piedra agujeriada, c. 650 m; 

 Rabanal, c. 450 m; Q. Damajuana, c. 250 m (rare), above 460 m fairly com- 

 mon; N E. slope of El Yunque (no. 154); C. Central, above 400 m; V. Colonial, 

 the slope below the Portezuelo pass (no. 33); Q. Seca, c. 435 m; C. Salsipuedes, 

 670 m, and in Q. Helechos, 660 m; Pto Ingles (Johow). 



Confined to Masatierra and well illustrated by Hemsley, tab. 63 fig. I. 

 Simply pinnate with pinnae connected by a narrow wing to the rhachis, their 

 margins subentire or obscurely serrulate. Veins always forming regular costal 

 areoles, often dark. In general habit resembling large forms of the Mediter- 

 ranean P. vulgave subsp. serratum. 



An aberrant rock form corresponds to var. basicompositum described 

 below: minor, coriacea, pinnis basalibus pinnatis. — Fig. 6 d. 



Masatierra: Pto Ingles, two small specimens grew on the outside of the 

 » Robinson grotto* (no. 318). We are not sure if this form is to be considered 

 as a variety of typicum, growing on rocks, but we find it more probably that 

 it ought to be referred to var. basicompositum of the Masafueran subspecies; 

 in Masafuera, also another variety was found growing on rocks. It very likely 

 offers an example of occasional dispersal of spores from Masafuera to Masa- 

 tierra, where no forms of subsp. masafueranum occur. 



2. subsp. masafueranum C. Chr. et Skottsb. nov. subsp. 

 Fig. 6. 



