28 1. Various Ways of Representing Surfaces and Examples
that it have maximal rank; the matrix is given by
and so our requirement is that the two tangent vectors to the surface,
given by the columns of Df, be linearly independent. Under this con-
dition, the Implicit Function Theorem guarantees that the parametric
representation is locally bijective and that its inverse is differentiable.
Parametric representations may of course have singularities. A
good example is the representation of the sphere given by the inverse
map to the geographic coordinates, which maps an open disc regularly
onto the sphere with a point removed, and collapses the boundary of
the disc into this single point.
a. Remarks on metric spaces and topology. Geometry in its
most immediate form deals with measuring
For this rea-
son, metric spaces are fundamental objects in the study of geometry.
In the geometric context, the distance function itself is the object of
interest; this stands in contrast to the situation in analysis, where
metric spaces are still fundamental (as spaces of functions, for exam-
ple), but where the metric is introduced primarily in order to have a
notion of convergence, and so the topology induced by the metric is
the primary object of interest, while the metric itself stands somewhat
in the background.
A metric space is a set X, together with a metric, or distance
function, d: X × X → R0
which satisfies the following axioms for all
values of the arguments:
(1) Positivity: d(x, y) ≥ 0, with equality iff x = y.
(2) Symmetry: d(x, y) = d(y, x).
(3) Triangle inequality: d(x, z) ≤ d(x, y) + d(y, z).
reader should be aware, however, that in modern mathematical terminology,
the word ‘geometry’ may appear with adjectives like ‘aﬃne’ or ‘projective’. Those
branches of geometry study structures which do not involve distances directly.