TREE FERNS OF NORTH AMERICA MAXON. 465 



very existence is sharpest, and where, except for the acquisition of 

 this trait or of some other to give them special advantage, their life 

 would indeed be short. To meet the same conditions ferns of other 

 families and of many tribes and genera have shown wonderful adapt- 

 ability of different sorts, both in structure and change of habitat, 

 two of the most common instances being the development of the 

 climbing form and epiphytic habit of growth. In the intensely wet 

 and heavily forested mountain region of Chiriqui, for example, prob- 

 ably three-fourths of the ferns are epiphytic. To make best use of 

 the arborescent habit, the growth of the tree-fern stem must be steady 

 and of a permanent character; and we find that moist tropical con- 

 ditions usually permit tins, however slow may be the rate of develop- 

 ment from season to season. 



Before proceeding, however, to a discussion of the widely different 

 forms assumed by the many species, or of the more minute technical 

 characters which serve to distinguish the genera and species, it may 

 be well to note briefly the general factors which appear to control the 

 distribution of tree ferns in North America. 



DISTRIBUTION AND HABITAT. 



As already indicated, tree ferns are characteristically inhabitants of 

 wet, forested, tropical and subtropical regions and reach their best 

 development in mountainous districts which are not subject to 

 drought or pronounced seasonal change. In the Greater Antilles 

 they are found mainly upon the northern slopes and summits of the 

 higher mountains, as, for example, the Sierra Luquillo of Porto Rico 

 and the Blue Mountains of Jamaica, where the cool, moisture-laden 

 trade-winds from the northeast bring a constant and ample supply 

 of moisture. The fern vegetation to the south of these mountains is 

 more or less strongly xerophytic, both islands mentioned even having 

 a semiarid region of cactus and scrub growth. Similar conditions 

 were noted in the Sierra Maestra of Cuba. Here on the compara- 

 tively dry southern slopes of the peak Torquino at 3,500 feet I 

 found plants of Cyathea arborea, a species which in Jamaica and else- 

 where in the West Indies rarely ascends to more than 2,000 feet. 

 Associated with it were several polypodiaceous ferns which ordinarily 

 are characteristic of the lowland forests and whose occurrence here 

 so far above their usual limit is in all probability directly correlated 

 with moisture supply. The southern coast in the lee of the Torquino 

 is intensely hot and semiarid, with a dense "chaparal" formation 

 (including cacti), wholly unsuitable not only to tree ferns but to a 

 majority of ferns of other families, as well. 



Similar conditions upon a grander scale are observed upon the 

 continent, the tree ferns being practically confined to the humid 

 38734°— sm 1911 30 



