300 BOTANICAL GAZETTE [OCTOBER 
accordance with definite factors. This grouping, and its causes, 
so far as known, we have now to consider for the marshland. 
The vegetation of any region develops, in response to the 
great primary ecological factors of temperature and precipita- 
tion, a characteristic climatic type, which prevails wherever no 
limiting factor becomes prepotent. For the region in which the 
marshland lies, this type is a mixed mesophytic forest, as shown 
upon the neighboring uplands. Further, according to Cowles’ 
theory, which seems to me in general well-grounded, all the 
vegetation of a region is tending to approach this type, because 
the vegetation is very closely correlated with physiographic fac- 
tors, and physiographically any region is in general tending to 
approach a base-level uniformity. This tendency involves the 
elimination of prepotent s2condary factors, and hence _the 
approach of the vegetation to the climatic type. The marshes 
appear to offer an exception to this rule, an exception more 
seeming than real, as will later be noticed. 
On the marshland, however, certain limiting factors do 
become prepotent, determining marked deviations from the 
climatic type and hence divisions of the vegetation as a whole, 
namely, the presence of abundant salt in the soil, determining a 
HALOPHYTIC DIVISION, an accumulation of fresh water determining 
a HYDROPHYTIC DIVISION, and ¢he influence of man, which removes 
large sections from the halophytic to the MESOPHYTIC DIVISION, 
or rather to a special section of it which may be called the 
culture section. The fourth of the principal vegetation divisions, 
the xerophytic, is not represented in the marshland. These 
divisions, however, are by no means of homogeneous appear- 
ance, but their vegetation falls into distinct groups of plants, 
the prominent forms of which have the same general aspect and 
adaptations, and occupy and are correlated with distinct physio- 
graphic positions. Such groups are called formations. They 
usually form distinct features of the landscape, and are known 
by distinctive local names expressing the physical habitat, such 
as bog, marsh, etc. But the formations, more closely observed, 
do not show an even mixture of plants, for these are grouped or 
segregated into definite assemblages, dependent in some part 
