7.1. Constructing Tangents
In this chapter we are trying to create a tangent to a given function
f. Intuitively a tangent is a straight line
touching the graph of f at a fixed point (a, f(a))
snuggling as close as possible.
Amongst the infinitely many linear functions, i.e. functions of the
that pass (a, f(a)) we are looking
for the one having the 'right' gradient m. Taking as an example and choosing we would expect the following scene:
Though the sketch suggests a gradient value of 1, it is still up to us to
confirm this by calculation. Normally there is no problem to compute the slope
of a straight line: With two different points just take the ratio
Tangents however provide sure access to only one point, namely (2, f(2)), so
there is no way to calculate the gradient this simply.
As the slope of the tangent is in some way controlled by shape of
f, we try to succeed with the following detour: We consider all the secants of f passing (2, f(2)). These are straight lines passing (2, f(2)) and a second point (x, f(x))
For these secants the gradient figures are easy to get: if the increase in height is the difference of the values of f -
- and the increase in width is the difference of the arguments. Thus:
The following applet shows the different positions of the secants and
calculates their gradients.
We get a double information from this: The closer the second point comes to
the fixed one, i.e. the nearer x is to 2, the less differ
- secant and tangent
and the figure 1.
So we may argue: If x converges to 2 the secant gradients would
approach the tangent gradient 1. That means we are able to calculate the tangent
gradient as soon as we could determine the limit . According to [7.1.1] however, the
is continuously continuable at 2 by the following limit:
Now, this result enables us to construct the tangent, more precisely the
tangent function , itself. First
provides and from the condition
we get . Finally we have
as the tangent function we looked for.