MAXON STUDIES OF TROPICAL AMERICAN FERNS. 29 



furnished the basis for Gmelin's peculiarly descriptive specific name. Figure 5, 

 indeed, shows a plant with crenate pinnae, which can not well be an example of this 

 species; the only material seen which approaches it is a single sheet of poor specimens 

 from the vicinity of Troy, Jamaica, altitude 450 to 660 meters, Underwood 2837. 

 These are an abnormal state apparently and can not be referred definitely to any 

 described species; they are perhaps nearest triangulum, but differ not only in form 

 but in the soft ferruginous vestiture of the rhizome. Further collections may estab- 

 lish their claim to rank as a distinct species; in any event Gmelin's name can not 

 possibly apply to them . In this connection may be noted ( 'hristensen's error in apply- 

 ing the name echinatum to the plant wrongly called mucronatum by Hooker. See 

 under P. struthionis. 



3. Aspidium trapezoides Sw., 1801. A name given by Swartz to West Indian plants, 

 the diagnosis being short, unsatisfactory, and without reference to any published fig- 

 ure. The description is repeated at page 44 of the Synopsis Filicum, and is elaborated 

 in the larger work of the same year° with citation of Sloane's plate 36, figure 1 . This 

 figure represents a dwarf relatively broad-leaved Jamaican plant of a type certainly 

 not very common in Jamaica but probably referable to P. triangulum. It is assuredly 

 no form of "P. viriparum," as Christensen suggests with doubt; that is to say, no 

 form of P. hetcrolepis nor of P. dissimulaus of the present paper. 



4. Polystichum falcatum Fee, 1850-1852. Described from Santo Domingo and best 

 regarded as an extreme form of P. triangulum. 



5. Polystichum cyphochlamys Fee, 1850-1852. Founded upon Cuban specimens col- 

 lected by Linden (no. 2175), the precise locality not stated. It is presumably one of 

 the complex of variable Cuban forms discussed below and retained for the present 

 under P. triangulum. 



Polystichum triangulum, even when restricted as in the present treatment, is an 

 exceedingly variable species. The typical form of Plunder's plate 72 is possibly not 

 very common in Santo Domingo, the few specimens seen from Haiti always showing 

 broader fronds. Matching the plate best are certain Jamaican specimens ( Maxon 1 201, 

 1489, 1873, 1883, 1884, 1887, 2207, 2555, 2591, 2827, 2968; Underwood 1167, 1822, 2838). 

 Among these some are narrower and some broader than indicated in plate 72. The 

 narrower form is illustrated at plate 33 of Hooker's Exotic Ferns (1859), but the pinnae 

 are almost invariably serrate-spinescent . Between these extremes of size in Jamaica 

 there seems to be every intergradation , and there is a similar intergradation in various 

 other characters directly associable with habitat; hence, far apart as are the extremes, 

 there appears to be no sufficient reason for separating any of the forms specifically. 

 The Cuban plants average much broader than the Jamaican and show far greater vari- 

 ation in cutting; some few of them agree closely with Haitian specimens collected by 

 Nash and Taylor (nos. 1341 and 1352) which probably represent the P. falcatum of 

 Fee. These last may, indeed, prove to be a distinct species, yet they seem to connect 

 with the narrow form through the following Cuban series, all from the province of 

 Oriente: Maxon 4243, 4260, 4267, 4459, 4461. P. triangulum as thus defined is not 

 unlikely an aggregate, requiring to be restudied later with the aid of more abundant 

 material from Santo Domingo. To attempt at present to distinguish by name the 

 large number of supposed or probable intermediates even varietally or subspecifically 

 would serve no useful purpose. 



Jenman's description b is excellent for the ordinary form, though the measurements 

 are less than for many of the Cuban and Haitian specimens. The species, as here 

 understood, is never proliferous. 



Known from the islands of Santo Domingo, Cuba, and Jamaica. In Jamaica it is 

 common in rocky situations at from 600 to 1,800 meters. 



a Fl. Ind. Occ. 3: 1647. 1806. 



6 Hull. Dot. Dept. Jamaica II. 2: 296. 1895. 



