would be spread out on a flat surface
(of area A) on which there were sample
lines (total length H), and the number
of intersections (N) between roots
and lines would be counted. I could
see intuitively that an estimate of the
length of root, R, would be given by
R =kNA
H
where k is a constant, but I was not
enough of a mathematician to know
what the value of k would be. I consulted
my father, a professor of
mathematics at the University of Manchester;
he referred me to Moran in
Canberra, who in turn told me that the
answer had been known since the nineteenth
century: k= π/2. Using this constant,
I checked my line intersection
method against direct measurement,
with roots from several different
species, and found very good agreement.
"My paper describing the method
was rejected by two journals, but fortunately
Kramer encouraged me to try
again. In the paper originally submitted,
the method used regularly arranged
sample lines. The referee of the
Journal of Applied Ecology, a statistician,
objected to this and I therefore
modified the method to use random
lines. Subsequently, several papers,
e.g., Marsh,1 pointed out that it is
quicker to use regularly arranged lines
and the loss of accuracy is small.
"I have not spent any further time
developing root measuring techniques,
but papers have been published by
other people who tested and modified
the method. Many of these are cited by
Böhm.2 The method has been automated
using either mechanical movement
of the roots past a photoelectric
counter, or an electronic scanner, but
neither of these methods has been
widely used. I think a crucial feature in
the popularity of my original method is
that it uses only simple apparatus
available in any laboratory."
1. Marsh B a’ B. Measurement of length in random arrangements of lines. J. Appl. Ecol. 8:265-7, 1971.
2. Böhm W. Methods of studying root systems. Berlin: Springer-Verlag, 1979. 188 p.
122