GLEASON: SOME APPLICATIONS OF THE QUADRAT METHOD 23 
very long period. Minor differences in environment may lead 
to the predominance of certain species in favorable places, to the 
persistence of relics, or to the early appearance of pioneers of other 
associations. Uniformity can not be perfect, although the general 
tendency of vegetation (apart from the effect of disturbing physical 
factors) is always toward uniformity, and it is most nearly attained 
in old (climax) closed associations, where entrance of new species 
is prevented by vegetative control of the environment, or in 
associations of extreme environment, where a small group of 
adjusted species enjoy a monopoly of the area. 
If the association were absolutely homogeneous, if the plant 
individuals of the different component species were as regularly 
distributed as water and alcohol in a mixture of the two liquids, 
all quadrats of appreciable size would be similar and any quad- 
rat could be chosen to represent the vegetation. Since no 
association is perfectly uniform, any one quadrat may by its 
structure accentuate the variability instead of concealing it. 
The value of a single quadrat also depends on its size and on the 
number of species in the association. In communities of numerous 
species (75 or more), a meter quadrat seldom includes as many as 
fifteen of them and usually contains still fewer. In very open 
associations, a meter quadrat frequently includes but a single 
individual plant. In either case, the error is reduced by increasing 
the size of the quadrat, but this simultaneously increases the labor 
of charting and decreases the value of a photograph. 
The single quadrat of small size, one to two meters square, is 
therefore essential for graphic illustration, either by photograph or 
map, and is undoubtedly of value also for experimental work, 
but it fails to give a fair representation of the whole association, 
and is of less value as a basis for written description or as a unit 
for intensive study in field work. To be sure, a suitable quadrat 
may be chosen, but the quadrat method itself, as ordinarily used, 
offers no aid in the selection of this typical area, so the actual 
choice invariably represents the observer’s idea rather than the 
impersonal facts. 
The chief value of the quadrat method lies in the results ob- 
tained from the study of many quadrats, chosen at random to 
avoid the personal element, scattered over an area wide enough 
