28 GLEASON: SOME APPLICATIONS OF THE QUADRAT METHOD 
distributed absolutely at random over the association, that is, if 
the association were absolutely uniform throughout, separated 
quadrats would never be necessary. But plants are not distri- 
buted entirely at random. Each plant-parent gives rise to dis- 
seminules which occupy, roughly speaking, a circle about it, with 
a diameter depending upon the mobility of the disseminules. 
As early as 1903 the writer (1907, p. 159) was able to demonstrate 
these in his field work. As a check, the following data from 240 
one-meter quadrats are offered. The 27 species have frequency 
indices ranging from 0 (1 quadrat only) to 99, based upon the 
whole series. If any fourth part of the series is taken as a basis, 
composed of 60 contiguous quadrats, as shown in the last four 
columns of TaBLE I, the average discrepancy between the two 
series of indices is 6.8, or for the 15 commonest species 10.7, 
and the difference may be as high as 44 for a species which normally 
grows in patches. But if every fourth quadrat is taken as a basis, 
that is, 60 separated quadrats, as shown in column 4 of Taste I, 
the greatest discrepancy is only 9 and the average only 1.9, Or 2.7 
for the commonest species. In other words, 60 quadrats well 
scattered give results practically as reliable as 240 contiguous ones. 
From a list of the frequency indices, made from quadrats of 
suitable size and number, the species are arranged in order of 
their general distribution in the association, and any number of 
them may be segregated for further study as the most typical or 
commonest. Since the frequency index increases with the size 
of the quadrat, a quadrat of proper size may now be chosen which 
will normally include all the more important species as well as a 
number of the less important ones also. The size of this major 
quadrat may be determined in advance and, in field work with 
students, one of them assigned to each student for detailed study 
and description. Although every such major quadrat may not 
actually contain all the important species, the student may be 
assured that it is nevertheless a fair sample of the association as @ 
whole. The value of setting a definite area before each student, 
in which he may do his more intensive work, will be at once appar- 
ent to every teacher who has tried to present this phase of ecology 
in the field. The method of determining the major quadrat wil 
be discussed below. , 
