66 Palmer — Ferm of tlic Dismal Swamp, Virgbiia. 



it is stouter, more fleshy and grooved for only a short distance al)ove the 

 lower pinna, or faintly further. 



In D. crislala, Jiorkhina and in- cUnlon'uum the basal pinnules of all 

 the pinnte are largest and longest; in ceUa and goldieana, some pinnfe, 

 especially the apical ones, are similar, but the lower pinnte, especially the 

 lowermost, have the j)innules, even for several pairs, very much reduced. 

 In good fertile fronds o.f w/srt and f/o/ri/Vroia the lower basal pinnule of the 

 lowest pair of pinna; is always absent but sometimes present or appar- 

 ently present in some undersized fronds. This is often the case in gol- 

 dieana, but only occurs rarely in ccl.s(t. The basal pinnules of the upper 

 pinn;e of both these ferns are always Ofiposite and very exactly so, but 

 they begin to diverge at the centers of the pinna;. On the lower pinnje 

 this p dring is rare and it is not easy to determine whether the opposite 

 of the reduced upper basal pinnule has never been developed or whether 

 it is represented by the one occupying the adjoining position. This latter 

 view would seem to be correct, the lower pinnules having l)een gradually 

 moved along the rachis toward the tip during the evolution of the form. 

 In very young fronds ( Figs. 5, 8) there is a wide space of the lower pinna 

 beneath, the jtinnule seems forced away from the rachis and the base 

 of the midvein inclines toward the rachis of the pinna for some distance. 

 The same result is shown in numerous young fronds of both forms. Fig. 

 1-t represents the common type of goldieana, while Figs, (i and 9-12 are from 

 specimens of celsa. 



Bryopteris goldieana is extremely herbaceous and robust, its pinnules and 

 pinn;« being large and often overlapping. In ceha they are always widely 

 separated ; both are much narrower, and there is no sudden change from 

 the long, wide pinnte to the shorter, narrower one of a crowded apex as 

 in goldiraiin. Tlie reduction or absence of the lower pinnules results in 

 producing a stalk for the pinn;e, short in goldieana, longer in ceha. 

 The pinna' of reha incline upwards very decidedly, whereas in goldieana 

 they stand at a right angle to the rachis or are only slightly inclined up- 

 wards. These diflerences between the very erect narrow celm and tlie 

 broad, <lroopingand \\i^\-]v,\v%m\ii goldieana result from differences in hab- 

 itat, the dryer and lighter situation of cc/.sa contrasting in its results with 

 the gloomy, damp habitat of goldieana. 



On July 30, 18IJS), I found two clumps of (/oWtmmt on the Virginia bluffs 

 of the Potomac river o])posite Cabin John Bridge. The first contained 

 over fifty plants, all with well drooping fronds and nearly all the lower pin- 

 nules of the lower pinnae normal. These plants were growing at the 

 foot of the talus among the rocks, and the trees formed a dense canopy 

 overhead. In the second clump a mile further down, in a precisely sim- 

 ilar situation, were several dozen i)lants. But here the thinness of the 

 foliage overhead permitted the sun to shine on the plants for several 

 hours daily. The early fronds were drooping as in the first clinnp, but 

 the later and mostly fertile fronds were more erect, and the divisions were 

 less herbaceous and consequently less crowded, but in no case to the same 

 extent as in relaa. 



