Fer)iS of the Dismal Swainji, Virginia. (')7 



Tlie differences in the cliaracter of tlie lower Inisal pinnules in these 

 two ferns is ecologically an interesting feature. The usual character of 

 these pinnules in goldieaua is shown in Fig. 14. In Fig. 13 is shown an- 

 other, which was growing in bright sunlight, at Great Falls, Virginia; 

 the shortening of the lower pinnules is evident. The first style of frond 

 grows in damp well-shaded situations and droops in such a way that a 

 [iractically cijual amount of light is received l)y all portions of its upper 

 surface. But a difierence occurs when the light is more aljundant; then 

 the frond becomes strengthened, that is, more erect, and consequently the 

 upper and middle i)ortious shade the lower pinnie. A struggle thus en- 

 sues between the pinnae for light. The lowermost, owing to their position, 

 are seriously handicapped, but instead of remaining in the same or 

 nearly the same plane, as in the case of well-shaded fronds, these lower 

 pinnae turn more toward the light, so that their tips approach each other 

 and their upper surfaces are turned nearly 90 degrees, so as to obtain the 

 light as nearly as possil)le i)erpendicular to their plane. In pressing 

 such specimens the stalks of one or more pinnae are necessarily frac- 

 tured where they join the stipe. In thus bringing the lower pinnae almost 

 together in order to ol)tain the greatest amount of light the greater por- 

 tion of each pinna is entirely successful, but at the expense of the lower 

 pinnules; especially so on the lowest and less so toward the middle. 

 Tliese lower pinnules are shaded not only by their own overlapping 

 when the pinnae are flexed, .but also by the stout stipe and the pin- 

 nules above. Consequently they do not receive a normal amount of 

 light and therefore during the growing period fail to develop perfectly, 

 and are outstripped by the more fortunately placed middle pinnules. 

 One extreme is shown in the usual frond oi gokileana, the other in nearly 

 every frond ot cdsa. S])ecimens of goUllmna collected about Washington, 

 an intermediate locality, altitudinally and geographically, have these basal 

 pinnules in naany cases much, and often unequally, reduced, but never 

 to the extent of celsa. Similarity of general structure and the ecological 

 character of the differences between these two ferns warrant the view 

 that ci'lm is a true subspecies of gokUeaiM, and therefore a geographical 

 i-ace or jihysiological subspecies. Our swamp plant therefore is a product 

 of abundant light, limited root moisture, and the struggle for e.Kistence 

 under peculiar conditions, wliich do not, or but very slightly, affect its 

 relative. 



\n June, 1896, near the head of Washington ditch, I found a few im- 

 mature plants ofcelsa and considered them D. c. rl'nUo)iiuna. The follow- 

 ing year, at the same place, I fouml some larger but imperfect fertile 

 fronds. This year, while penetrating the swamp north of the outlet 

 canal and about eight miles east of the other locality, I found numerous 

 plants ranging, through all stages, from those with the first fronds and 

 the remains of the prothalli, to plants over thirty inches high. It is 

 possible that this fern occurs in other localities in the same general 

 region. 



The log fern grows in several situations. About the base of a large gum 

 tree, where there was an accutnulation of waste woodv matter and an 



