THE FERNS OF NICARAGUA. 



137 



the sinuses of the lobes, i. e. all the veins free, as in pi. iv, 



Fig- 3- 



In these specimens the basal veins of each lobe usually spring 

 from the midrib of the lobe, but occasionally a vein arises from 

 the mid-rib of the pinna, as shown in PI. iv, fig. 2, thus 

 approaching the following form. Figs. 2 and 3 represent 

 parts of the same pinna. 



Not rare on Ometepe, with the preceding form. 



3. A form intermediate in venation between 1 and 2, cor- 

 responding to P. nemoralis. In this form most of the pinna? 

 have free veins springing from the mid-rib between the mid- 

 ribs of the lobes, but occasionally these veins unite with the 

 basal vein from the adjoining lobe, as shown in PI. iv, fig. 4, 

 thus approaching typical P. biaurita. 



This form was very common in deep woods near Castillo, 

 growing in scattered clusters and reaching a height of 20 

 inches. Most of the fronds collected were pentagonal in form, 

 with but two lateral pinnae on each side, the lower one with a 

 prominent division near the base on the lower side. One speci- 

 men has four pinnae on a side and the basal pinna on one side 

 is simple. In form this latter frond is like 1 and 2, which have 

 the fronds elongated. It differs however, as do all the others 

 of this group, in having the pinnae tapering toward the base, as 

 represented in PI. iv, fig. 6, those of groups 1 and 2 being 

 truncate, as in tig. 5. though sometimes unequal at base. 



The venation in this group resembles that of P. anamallay- 

 ensis Beddome. 1 which, unless the termination of the veins in 

 dots remote from the margin proves a constant character, 

 should probably be added to this series. 



An examination of the entire series shows that venation is 

 not to be relied upon. — a fact already established by various 

 authors. 



Hooker. 2 referring to the fact that Linne in his herbarium 

 united P. (Compter ia \ biaurita and P '. nemoralis, says: '"The 

 Cam-pteria Agardh has no doubt correctly referred to Pt. 



1 Beddome, Ferns of S. Ind., pp. 14, 69, pi. xlv. 

 *Species Filicum, vol. 11, p. 1S1. 



