Since f is monotonic increasing, f is Riemann ingegrable and F is well-defined. We assume F(x) is differentiable at some q∈Q∩[0,1], then if q=0,q=1, F is differentiable at q iff we have x→q;x∈[0,q)limx−qF(x)−F(q)=x→q;x∈(q,1]limx−qF(x)−F(q) From Exercise 9.8.5 we know that f is not continuous at q and if we locate q=q(n) for some n∈N as in Exercise 9.8.5, we shall have f(x)≥f(q)+2−n,x>q, so x→q;x∈[0,q)limx−qF(x)−F(q)=x→q;x∈[0,q)limx−q∫xqf(y)dy≤x→q;x∈[0,q)limx−q∫xqf(q)dy=f(q)x→q;x∈(q,1]limx−qF(x)−F(q)=x→q;x∈(q,1]limx−q∫qxf(y)dy≥x→q;x∈(q,1]limx−q∫qx(f(q)+2−n)dy=f(q)+2−n Thus x→q;x∈(q,1]limx−qF(x)−F(q)−x→q;x∈[0,q)limx−qF(x)−F(q)≥2−n contradicts the result that the two limits are equal. If q=0 or q=1, F may be differentiable at q by definition, since f is continuous at 0 or 1 when restricted on [0,1].
Exercise 11.9.2
When I is empty or a single point then the statement is trivial. Now suppose I is a nonempty interval, choose some x0∈I, then for any x∈I,x=x0, the function F−G is differentiable on [x0,x] or [x,x0], thus we have some ξ between x and x0, s.t. x−x0(F−G)(x)−(F−G)(x0)=(F−G)′(ξ)=f(ξ)−f(ξ)=0 So we have (F−G)(x)=F(x)−G(x)=F(x0)−G(x0) This means if we let C=F(x0)−G(x0), we will have F(x)=G(x)+C,∀x∈I,x=x0, since we also have F(x0)=G(x0)+C valid, the conclusion is true. Alternatively, we can use the second Fundamental theorem of calculus, since (F−G)′(x)=0,∀x∈I the function F−G:I→R is an antiderivative of the function g(x)=0,x∈I. Since g is constant, it is Riemann integrable, so choose some x0∈I, then for any x∈I,x=x0, we can use the second Fundamental theorem of calculus on [x0,x] or [x,x0] and have ∫[x0,x]g=0=(F−G)(x)−(F−G)(x0) or ∫[x,x0]g=0=(F−G)(x0)−(F−G)(x) In either case we have (F−G)(x0)=(F−G)(x), This means if we let C=F(x0)−G(x0), we will have F(x)=G(x)+C,∀x∈I,x=x0. Since we also have F(x0)=G(x0)+C valid, the conclusion is true.
Exercise 11.9.3
Since f is monotone increasing, f is Riemann integrable, thus if f is continuous at x0, then F is differentiable at x0 by the first Fundamental theorem of calculus. Conversely we let F is differentiable at x0 and assume f is not continuous at x0, then we shall have f(x0−)<f(x0+), let η=(f(x0−)+f(x0+))/2, then we can find ϵ>0, s.t. f(x0−)<η−ϵ<η<η+ϵ<f(x0+) As f is monotone increasing, we know f(x)<η−ϵ if x<x0 and f(x)>η+ϵ if x>x0. Then for y>x0, we have F(y)−F(x0)=∫[x0,y]f>∫[x0,y](η+ϵ)=(η+ϵ)(y−x0) And for y<x0, we have F(x0)−F(y)=∫[y,x0]f<∫[y,x0](η−ϵ)=(η−ϵ)(x0−y) So we have y→x0+limy−x0F(y)−F(x0)≥η+ϵ>ηy→x0−limy−x0F(y)−F(x0)=y→x0−limx0−yF(x0)−F(y)≤η−ϵ<η Thus F′(x0+)>F′(x0−), a contradiction to F being differentiable at x0.