The function γn defined on Rn by γn(x)= exp(−∥x∥2) belongs to S(Rn). In fact, one proves by induction on k that, for every integer k∈N, there exists a polynomial Pk∈R[X] such that ∂kγ1=Pkγ1.
For all α=(αi)∈Nn and β=(βi)∈Nn, and every x=(xi)∈Rn, one then has
∣(Xβ∂αγn)(x)∣=i=1∏n∣xi∣βi∣Pαi(xi)∣γ1(xi)
which is a bounded quantity when x varies in Rn.
Let α∈Nn and β∈Nn. If φ∈S(Rn), then Xβ∂α(φ) is again a Schwartz function; the resulting mapping φ↦Xβ∂αφ of S(Rn) into itself is continuous.
The space S(Rn) is a topological algebra. More precisely, if φ1 and φ2 belong to S(Rn), then φ1φ2 is a Schwartz function such that
qα,β(φ1φ2)⩽∑qγ,β(φ1)qα−γ,0(φ2)(2)
0⩽γ⩽α
for all α and β∈Nn (prop. 1 of IV, p. 196).
The canonical inclusion of S(Rn) in the space C∞(Rn), endowed with the topology described in No.o3 of IV, p. 200, is continuous, since
supx∈K∣∂αφ(x)∣⩽qα,0(φ)
for every compact subset K of Rn, every α∈Nn and every Schwartz function φ.
Bourbaki sets the proof directly after the statement, and the corpus does not separate them, so what is above is the statement and its proof together.
Nothing.
Nothing.