Assume that R is reduced and let w∈W and α∈R.
(i) We have ψ(w(α))=Uw(ψ(α)). (ii) Assume that α is positive. The root w(α) is negative if and only if
η(w−1,sα)=−1, in other words if sα∈Tw−1.
(iii) We have η(w,sα)=−1 if and only if the chambers C and w(C) are on opposite sides of the hyperplane Lα. In other words, the set Tw consists of the reflections with respect to the walls separating C and w(C).
Let β∈B(C) and put s=sβ. Clearly Ts={s} and consequently
Us(ε,t)={(ε,sts−1)(−ε,s)if t=sif t=s.
On the other hand, let ρ=∑γ∈B(C)nγ(ρ)γ be a positive root. Put
s(ρ)=γ∈B(C)∑nγ(s(ρ))γ.
If ρ=β, there exists an element γ∈B(C) with γ=β, such that nγ(ρ)>0, and we have nγ(s(ρ))=nγ(ρ)>0 (no. 1, formula (5)). Hence s(ρ) is positive. We deduce immediately that
ψ(s(ε,ρ))={(ε,ssρs−1)(−ε,s)if ρ=βif ρ=β.
Comparison of (12) and (13) now shows that Us(ψ(γ))=ψ(s(γ)) for all roots γ and all s∈S. Since S generates W, (i) follows.
On the other hand, saying that w(α) is negative is equivalent to saying that
ψ(w(α))=(−1,wsαw−1),
or, by (i), that Uw(ψ(α))=(−1,wsαw−1). If in addition α is positive, then ψ(α)=(+1,sα) and Uw(ψ(α))=(η(w−1,sα),wsαw−1), hence (ii).
Finally, by (ii), η(w,sα)=−1 if and only if one of the roots α and w−1(α) is positive and the other negative. This is equivalent to saying that (α∣x).(w−1(α)∣x)=(α∣x).(w∣x)<0 for all x∈C, hence the first assertion in (iii). The second assertion in (iii) follows immediately.
Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.
Nothing.