26 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 



sufficient material for study and comparison, the difficulties in so 

 far as the plants are concerned largely disappear, and the range of 

 variation, great in some cases, rather slight in others, may be made 

 out with tolerable accuracy. 



This, at least, is true of the West Indian species, as observed by 

 the writer in Cuba and Jamaica, and as represented by a large suite 

 of specimens in the U. S. National Herbarium and the herbarium of 

 the New York Botanical Garden. The difficulties are mostly in the 

 matter of applying the early names, many of which were given origi- 

 nally to plates rather than to plants. The historic P. triangulum 

 itself offers greater difficulties than any other, and even as now dealt 

 with is probably not closely enough delimited, further material from 

 Santo Domingo being necessary to a clearer understanding of the 

 species. 



The characters of the genus as a whole are too well known to 

 require elaboration here, but several secondary features not com- 

 monly regarded as diagnostic may be pointed out, as these have 

 seemed to the writer to be quite constant. These are: (1) The pres- 

 ence or absence of a proliferous bud upon the rachis; and (2) the 

 position of this, whether (a) terminal at the truncate or retuse apex, 

 (b) borne some distance below the foliose apex, or (c), as is often 

 the case, at the end of a cirrhate or flagelliform prolongation of the 

 rachis. The extension of the rachis into a whip-like tail serves to 

 increase the chances of successful proliferation; and in general this 

 tendency seems to be associable with a moist shady habitat. Such 

 plants have an arching or even a pendent habit, are less fertile than 

 those growing in drier or more exposed situations, and obviously 

 depend more upon vegetative reproduction and less upon the dis- 

 persal of spores." Whatever may have been the cause of this corre- 

 lation between habitat and form, the characters appear to be fixed 

 and constant and the species amply distinct and recognizable. 

 Thus no. form of true P. triangvlum, so far as the writer has observed, 

 is proliferous, though this is a marked feature of several of its closest 

 relatives. 



The status of P. heterolepis, discussed below, is interesting from 

 another point of view; that of a species apparently not re-collected 

 in its typical form from the time of its discovery up to 1907. The 

 P. viviparum of Fee is now found to be merely a young state of this 

 strictly Cuban species and quite different from the Jamaican plant 

 described as viviparum by Jenman and others. These and several 

 others are here illustrated from photographs of the most typical 



o- In the ease of P. rhizophorum this is manifest in a strong tendency toward dimorph- 

 ism, the fertile fronds being rigidly erect and usually very much smaller than the 

 elongate and radicant sterile ones. 



