60 MISSOURI BOTANICAL GARDEN. 
The comparatively smooth bark of the cypress does not 
furnish so good a foothold for the Porella, Polypodium 
and other epiphytes as does that of the tupelo, hence there 
is little epiphytic vegetation upon the living trees. 
The changes one notes in ascending the Saint Francis 
River from Bertig to Kennett, some thirty-five miles, are 
not so much in the forms present as in the distribution of 
those forms. There is a gradual narrowing of the area 
occupied by the Polygonum-Zizaniopsis association and 
an encroachment upon it of the Taxodium-Nyssa associa- 
tion. While there is a great deal of the tupelo, in pure 
or mixed groves, distributed all the way up the river to 
Kennett, it is noticeable that the cypress increases decid- 
edly. The nature of the woods marginalto the Taxodium- 
Nyssa association is described below. 
Wherever the land appears above the water, in small or in 
large islands, the forms mentioned usually appear upon it. 
From experiments made with the cypress as illustrated in 
Plate 16, it is not necessary for it to be surrounded with 
water in order to grow vigorously. It seems reasonable to 
suppose that the conditions upon these islands are suitable 
for the growth of the cypress, and that it would be found 
there, were it not for the encroachment upon it of the 
more vigorous broad-leaved forms. As the latter are evi- 
dently stronger in the struggle for occupancy of an area 
possessing the conditions they require, the cypresses are 
driven off the land where the broad-leaved trees can live, 
into the water where they cannot follow. This is well 
illustrated on all the marginal uplands and islands, for 
example, at Bear Island where the land is occupied with 
oaks, sweet gum, etc., while the cypress crowds up to 
them on all sides, the amount of water marking the ten- 
sion line between the two types. 
In open areas in which the soil of the cypress forest is 
never covered with a great depth of water, but is dry a 
