FLOOD-PLAIN FOREST COMMUNITIES 197 
e) Fate of the formation—Most of our tamarack swamps are in the 
regions which are commonly dominated principally by beech and maple. 
In the higher portions of the tamarack swamps are found several species 
characteristic of beech woods and other mesophytic woods. These are 
the wood-frog, the large slug, the snail (Polygyra albolabris) and the 
red-backed salamander (Plethodon cinereus) and the spider (Castianeira 
cingulata). ‘These indicate that beech and maple are to follow. 
4. FLOOD-PLAIN AND RAVINE FOREST COMMUNITIES 
As we have noted on pp. 87-93 and 108-113, streams often develop 
by head-on erosion into uplands of rock or clay. 
a) Streams developing in rock.—In case the upland is of rock, the 
beginning of the stream is a lower place in the slope of the rock through 
which water flows when it is raining. Vegetation is usually absent. If 
there are broken pieces of rock at the sides or in the course of the inter- 
mittent stream, some of the forms mentioned on p. 218 may be present. 
Until it becomes permanent or has cut itself a deep, straight-sided 
channel, it is inhabited by the animals which inhabit the bare rock of 
hills or hillsides. After the stream has cut such a channel, there are 
always small piles of fine soils which support nettles and other mesophytic 
plants similar to those of the old mesophytic flood-plain. Flood-plain 
animals appear early in the development of the stream. 
b) Streams developing in clay.—Along the north shore we have an 
opportunity to study the vegetation of ravines of all ages corresponding 
to the aquatic stages described on pp. 87-93. The slightly lower places on 
the bluff side in which water runs only when it is raining are usually too 
steep to support plants and animals as regular residents, and have the 
same incidental forms as the steep bluff (p. 210). Later, when the sides 
of the gully become less steep, it is similar to if not identical with the 
second bluff stage (pp. 212-214); later, like the third (p. 215), and still 
later, like the young forest stage. There appears to be little or no 
difference between the bluff and the sides of young ravines. The outer 
ends of ravines as much as a mile and a half long are usually in the shrub 
stage and possess the shrub community. In favored situations the sapling 
forest, apparently identical in its animal associations with that of the bluff 
(p. 215), grows up. Up the stream, well back from the lake, a distance 
of a fourth of a mile, conditions become very different. A very meso- 
phytic forest grows up. In this we have possibly some special features 
under primeval conditions, but in the ravines along the north shore 
where the forest is so much disturbed, ravines do not differ particularly 
