... inequality,1
The triangle inequality, which is one of the defining properties of a norm, states that

\begin{displaymath}\Vert x+y\Vert\le\Vert x\Vert+\Vert y\Vert\ \mbox{for all}\x,y\in{\bf {\rm R}}^n.
\end{displaymath}

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
... inequality,2
The reverse triangle inequality, which can be proved from the triangle inequality, states that

\begin{displaymath}\Vert x-y\Vert\ge\left\vert\,\Vert x\Vert-\Vert y\Vert\,\right\vert\ \mbox{for all}\x,y\in{\bf {\rm R}}^n.
\end{displaymath}

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
... integrand.3
A digression: Equation (4) is commonly used when one wants to use a Mean Value Theorem (MVT). For a function $f:{\bf {\rm R}}\rightarrow{\bf {\rm R}}$ of one variable, the MVT (a special case of Taylor's theorem) states that if f is sufficiently smooth, then there exists $c\in (a,b)$ such that

f(b)=f(a)+f'(c)(b-a).

However, the MVT does not hold for vector-valued functions, because the number c is typically different for each component Fi. Equation (4) is usually an adequate substitute.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Mark S. Gockenbach
2003-01-23