[PENHALLOw] DEVELOPMENT OF CERTAIN MARSH LANDS 29 
the minority and not recognizable without much care in assorting the 
material. 
At the extreme northern end of the trench, as shown in figure 5, 
just where the drainage water falls into the channel of the marsh, a 
vast number of fibrous structures are to be met with. These were at 
first regarded as representing the fibrous roots of the associated trees 
and shrubs. Mr. Lewis Goodwin, in charge of the excavation, informed 
me that this material occurred in vast quantities all through the trench 
at the lowest level reached, and that it could be handled with a fork 
_ precisely like hay. I subsequently found this to be a perfectly correct 
description, and I also found that the material was in a condition of 
remarkably perfect preservation, due to some extent, possibly, to the 
later action of salt water. 
A recognition of the various features thus described, made it impor- 
tant to secure a complete vertical section of the marsh, and this I was 
fortunate in securing owing to the fact that the trench was in places 
deep enough to penetrate to the basic formation. By simply slicing 
down the side of the trench on the line of section, a complete exposure 
of all the successive phases in development was made. The results 
are represented diagramatically in figure 8, and the details are as 
follows :— 
Proceeding in the inverse order, (Fig. 8) the uppermost layer 
consists of turf formed by the growth and decay of the characteristic 
salt marsh vegetation, consolidated by its own weight; by the action of 
water and decay, and by the cementing action of a large proportion of 
silt brought in by the tide and held in place by the vegetation itself. 
This layer is fairly uniform in thickness and may be said to have a 
maximum thickness of eighteen inches. Its surface is a little above 
mean high water, and this relation must have obtained from the very 
first period of development. It completely encloses the stumps of the 
original forest as already shown, and its base rests upon the stratum 
in which the trees had grown and found anchorage. 
The second layer, seven inches in thickness, consists of peat derived 
from the decay of herbaceous vegetation and fallen leaves. It is pene- 
trated in all directions by the roots of herbaceous plants and by struc- 
tures probably representing the rhizomes of Cyperaceæ. There are also 
more or less numerous fragmented roots belonging to the trees of the 
same zone. This is the stratum to which the white pine belongs, though 
it is altogether probable from the age of the latter, that the seeds must 
have sprouted on the surface of the layer below. From this we draw the 
inference that the trees and the peat developed concurrently, and that the 
latter was derived largely from the decaying foliage of the former. 
The stumps are often of large size, and they afford evidence that the 
