a)
Suppose
x→alimf(x)=l.By definition,
∀ϵ>0, ∃δ>0, 0<∣x−a∣<δ⟹∣f(x)−l∣<ϵ.By the reverse triangle inequality, we have
∣∣f(x)∣−∣l∣∣≤∣f(x)−l∣<ϵ.Hence,
x→alim∣f(x)∣=∣l∣.b)
Suppose limx→af(x)=l and limx→ag(x)=m.
Without loss of generality, suppose also that max(l,m)=l. For contradiction, assume that
x→alimmax(f,g)(x)=n=l.Let ϵ=2l−n. By definition,
∃δ1,0<∣x−a∣<δ1∃δ2,0<∣x−a∣<δ2⟹∣f(x)−l∣<ϵ,⟹∣max(f,g)(x)−n∣<ϵ.Now, let δ=min(δ1,δ2) and consider x such that 0<∣x−a∣<δ. We have
−2l−n+l<f(x)<2l−n+l,−2l−n+n<max(f,g)(x)<2l−n+n.Consider two subcases: l>n and l<n. If l>n, then f(x)>max(f,g)(x). If l<n, then either limx→ag(x)=n or f is not a function.
Hence,
x→alimmax(f,g)(x)=max(x→alimf(x),x→alimg(x)).