MATH 221

Arun Ram
Department of Mathematics and Statistics
University of Melbourne
Parkville, VIC 3010 Australia
aram@unimelb.edu.au

Last update: 13 August 2014

Lecture 18

Rolle's Theorem and the Mean Value Theorem

Rolle's Theorem If f(a)=f(b), and f(x) is continuous between a and b, and
f(x) is differentiable between a and b,
then there is a point c between a and b such that dfdx|x=c=0
a c c c c b x y y=f(x) 4 possible choices forc!

Mean Value Theorem If f(x) is continuous between a and b, and
f(x) is differentiable between a and b,
then there is a point c between a and b such that dfdx|x=c =f(b)-f(a)b-a.
a c c c c c b x f ( a ) f ( b ) y This line has slope f(b)-f(a)b-a Note: If f(a)=f(b) then the line connecting (a,f(a)) and (b,f(b)) has slope 0 and so Rolle's theorem is a special case of the mean value theorem.

Verify Rolle's theorem for the function f(x)=e-xsinx on the interval [0,π]. π 4 π x y f(0) = e0sin0=0, f(π) = eπsinπ=0, dfdx = e-xcosx-e-x sinx = e-x (cosx-sinx). So, if dfdx=0 then cosx-sinx=0. So cosx=sinx. So x=π4. So c=π4 when dfdx|x=c=0.

Verify the mean value theorem for f(x)=ex in the interval [0,1]. 1 x 1 3 y y=ex line with slope f(1)-f(0)1-0. f(0) = e0=1, f(1) = e1=e, so f(1)-f(0)1-0 = e-11-0=e-1. dfdx=ex and we want c so that dfdx|x=c=e-1. If dfdx|x=c=ec=e-1. Then c=ln(e-1)ln(1.78) which is between 0 and 1.

Consider the mean value theorem for f(x)=1x in the interval [-1,1]. f(1)=11=1 andf(-1)= 1-1=-1. So f(1)-f(-1)1-(-1) =1-(-1)2=22 =1. So we want c so that dfdx|x=c=f(1)-f(-1)1-(-1)=1. dfdx|x=c= d1xdx|x=c =-1x2|x=c =-1c2. Find c so that -1c2=1. (IMPOSSIBLE with real numbers.) What went wrong? 1 -1 x 1 -1 y line with slope f(1)-f(-1)1-(-1)=1 f(x) is not continuous or differentiable at x=0!! So the mean value theorem does not apply.

Show that the equation x5+10x+3=0 has exactly one real root.

Let f(x)=x5+10x+3. We have to show that there is only one real number that can be plugged into f(x) to get 0.

Notes:

(a) As x, f(x).
(b) As x-, f(x)-.
x y y=f(x) (a) and (b) tell us that y=f(x) must cross the x axis. Suppose it crosses twice, at x=a and x=b. a b c x y Then dfdx|x=c=0 for some c between a and b. So 0=dfdx|x=c= d(x5+10x+3)dx |x=c=5x4+10 |x=c=5c4+10. But 5c4+10 is never 0, no matter what c is. So y=f(x) can't cross the x axis twice. So it must cross it only once. So there is exactly one number (real number) that can be plugged into f(x) to get 0.

Discuss Rolle's theorem for f(x)=(x-1)(2x-3) in the interval 1x3. 1 3 x 6 y f(1) = 0, f(3) = (3-1)(6-3)=6. Since f(1)f(3) we can't apply Rolle's theorem with x=1 and x=3. Are there two points in the interval [1,3] where f(a)=f(b)? Yes, f(1)=0 and f(32)=0. so we should be able to find c between 1 and 32 so that d|x=cdx=0. dfdx|x=c= (x-1)·2+1· (2x-3)|x=c= 4x-5|x=c=4c-5. so dfdx|x=c=0 when c=54, and 54 is between 1 and 32.

Notes and References

These are a typed copy of Lecture 18 from a series of handwritten lecture notes for the class MATH 221 given on October 18, 2000.

page history