Since geophysical methods are non-invasive, they can be of great help in so
il studies because they disturb neither the structure nor the dynamics of t
he soil. Moreover, data are acquired with reliable spatial sampling. The us
ual ways of investigation, like augering and excavation, disturb the soil a
nd are totally incompatible with a spatially dense sampling strategy, which
would destroy the object of the study. Both approaches are complementary w
hen excavations have a limited extent and are distributed according to the
information conveyed by the geophysical investigation. A basic principle of
applied geophysics is to measure different physical parameters without dir
ect access to the studied volume. Horizontal and/or vertical variations of
the parameter(s) can be recorded. Possible soil parameters should be restri
cted to measurements which do not alter the medium (reversible effect). To
be significant, the variations of the parameter(s) should exhibit a wide dy
namic range over different soil types and should be correlated in some way
to soil parameters such as particle size or hydraulic conductivity. After s
ummarising the soil properties, two examples are shown whereby electrical r
esistivity was used. The first example is a specific soil so-called hardpan
(sandy soil in arid area) in Lagadge, North Cameroon. Using resistivity su
rveys the three dimensional extension of a very coherent horizon was mapped
. This horizon is delineated by low resistivities <100 Omega m (conductivit
ies > 10 mS/m) because of the disposition of clay particles around the quar
tz grains. In a second example, a "homogeneous" area ought to be found deli
mit the extent of a surface where a pesticide transfer experiment is to tak
e place. Accurate mapping of soil horizons was not feasible by augering. Re
sistivity data have clearly shown the three-dimensional extension of clayey
horizons in the complex delta context. (C) 2000 Elsevier Science B.V. All
rights reserved.