Let Φ be a set of subsets of a locally compact space X, satisfying (PCI) and (PC_{II}), and let α be a real-valued function, defined on Φ, satisfying the conditions (PM_I), (PM_{II}) and (PM_{III}). In order that there exist a positive measure μ on X such that the sets of Φ are μ-integrable and μ(M)=α(M) for all M∈Φ, it is necessary and sufficient that α satisfy in addition the following condition:*
(PM_{IV}) For every ε>0 and every M∈Φ, there exist a compact set K⊂M and an open set U⊃M such that, for every N∈Φ satisfying the relation K⊂N⊂U, one has ∣α(N)−α(M)∣≤ε.
Moreover, if the condition (PM{IV}) is satisfied, then the measure μ is unique; for every compact set K, μ(K)=infM∈Φ,M⊃Kα(M); for every open set U, μ∗(U)=supM∈Φ,M⊂Uα(M).*
Note that the condition (PM_{IV}) is equivalent to the conjunction of the following two:
(PM'_{IV}) For every ε>0 and every M∈Φ, there exists an open set U⊃M such that, for every N∈Φ contained in U, α(N)≤α(M)+ε. (PM''_{IV}) For every ε>0 and every M∈Φ, there exists a compact set K⊂M such that, for every N∈Φ containing K, α(N)≥α(M)−ε.
For, it is obvious that (PM'_{IV}) and (PM''_{IV}) imply (PM_{IV}). Conversely, let us show for example that (PM_{IV}) implies (PM'_{IV}): let K be a compact set and U an open set such that K⊂M⊂U and ∣α(P)−α(M)∣≤ε for every P∈Φ satisfying K⊂P⊂U. Then, if N∈Φ and N⊂U, M∪N belongs to Φ and K⊂M∪N⊂U, whence α(M∪N)≤α(M)+ε and a fortiori α(N)≤α(M)+ε.
When the set Φ, satisfying (PC_I) and (PC_{II}), consists of compact sets, then the condition (PM''_{IV}) is trivially verified, and (PM_{IV}) is then equivalent to (PM'_{IV}).
The condition (PM_{IV}) is necessary: this follows at once from Th. 4 of No. 6 on the 'approximation' of an integrable set by a compact set and an open set. To prove the other assertions of the theorem, we proceed in several steps.
1° Definition of a topology on P(X).
For every pair (K,U) consisting of a compact set K and an open set U in X, we denote by I(K,U) the set of subsets M⊂X such that K⊂M⊂U; in order that I(K,U) be nonempty, it is necessary and sufficient that K⊂U. If (K′,U′) is a second pair, formed by a compact set K′ and an open set U′, we have
I(K,U)∩I(K′,U′)=I(K∪K′,U∩U′). Let
T be the topology on
P(X) generated by the set of subsets
I(K,U) as
K runs over the set of compact subsets of
X, and
U over the set of open subsets of
X; by the foregoing, the
I(K,U) form a base for the topology
T (GT, I, §1, No. 3).
We observe that the definition of T implies that, in P(X), the set of compact subsets of X is dense. The condition (PC_{II}) expresses that Φ is dense in P(X), and condition (PM_{IV}) expresses that the function α is continuous on Φ for the topology induced by T. Finally, Th. 4 of No. 6 expresses that the function M↦μ(M) is continuous on the clan of μ-integrable sets, for the topology induced by T.
2° Uniqueness of μ.
We denote by Φ the set of subsets M⊂X such that α(N) tends to a finite limit as N tends to M (for the topology T) while remaining in Φ; we may then extend α in only one way to a continuous mapping α of Φ into R (GT, I, §8, No. 5, Th. 1). If there exists a measure μ meeting the requirements, the above remarks prove that the clan Ψ of μ-integrable sets is contained in Φ and that μ(M)=α(M) for every M∈Ψ; this relation holds in particular for every compact subset M of X, which proves the uniqueness of μ (No. 10, Cor. 3 of Prop. 19).
3° Extension of α to the compact sets.
Without assuming the existence of μ, we are now going to study the set Φ and the extension α of α to Φ. We first show that every compact set K belongs to Φ and that α(K)=infP∈Φ,P⊃Kα(P). Set a=infP∈Φ,P⊃Kα(P); for every ε>0, there exists an M∈Φ such that K⊂M and α(M)≤a+ε. By (PM'_{IV}), there exists an open set U⊃M such that, for every N∈Φ contained in U, we have α(N)≤α(M)+ε≤a+2ε; for every N∈Φ such that K⊂N⊂U, we therefore have a≤α(N)≤a+2ε, which, by the definitions, shows that K∈Φ and α(K)=a.
This result proves at once that if K1 and K2 are two compact sets such that K1⊂K2, then α(K1)≤α(K2). If K1 and K2 are any two compact sets, we have α(K1∪K2)≤α(K1)+α(K2) by (PM_{II}). We shall see, moreover, that if K1 and K2 are disjoint then α(K1∪K2)=α(K1)+α(K2). For, there then exist two disjoint open sets U1,U2 such that K1⊂U1, K2⊂U2 (GT, II, §4, Prop. 4). Therefore, by (PC_{II}), there also exist two sets M1∈Φ, M2∈Φ such that K1⊂M1⊂U1 and K2⊂M2⊂U2. Now let P be any set of Φ containing K1∪K2; the union of the two sets P∩M1 and P∩M2 belongs to Φ by (PC_{I}), and since these two sets are disjoint, application of (PM_I) and (PM_{III}) yields
α(P)≥α(P∩M1)+α(P∩M2)≥α(K1)+α(K2),
which establishes our assertion.
4° Extension of α to the open sets.
We shall now see that, for an open set U to belong to Φ, it is necessary and sufficient that, as K runs over the set of compact subsets of U, the supremum of the numbers α(K) is finite; moreover, α(U) is then equal to this supremum.
For, let U be an open set belonging to Φ; for every ε>0 there exists a compact set K⊂U such that, for every set M∈Φ satisfying K⊂M⊂U, one has ∣α(U)−α(M)∣≤ε, whence ∣α(U)−α(K)∣≤ε; on the other hand, if K′ is any compact set contained in U, then K⊂K∪K′⊂U, whence ∣α(U)−α(K∪K′)∣≤ε and so α(U)≥α(K∪K′)−ε≥α(K′)−ε; α(U) is therefore indeed equal to the supremum of the numbers α(K) as K runs over the set of compact subsets of U.
Conversely, let U be an open set such that b=supK⊂Uα(K)<+∞ (K running over the set of compact subsets of U), and let us show that U∈Φ. For every ε>0, there exists a compact set K⊂U such that b−ε≤α(K)≤b; by (PM''_{IV}), for every set M∈Φ such that K⊂M⊂U, there exists a compact set K′⊂M such that
α(M)≤α(K′)+ε≤b+ε;
therefore b−ε≤α(M)≤b+ε, which proves that U∈Φ.
From this characterization of the open sets U∈Φ, and of α(U), it follows first of all that if U1 and U2 are two open sets such that U1⊂U2 and U2∈Φ, then U1∈Φ and α(U1)≤α(U2). On the other hand, if U1 and U2 are two open sets belonging to Φ, then the same is true of U1∪U2, and α(U1∪U2)≤α(U1)+α(U2). For, let K be any compact set contained in U1∪U2; for every point x∈K, there exists a compact neighborhood of x contained in either U1 or U2; one can therefore cover K by a finite number of these neighborhoods; if K1 (resp. K2) is the union of those that are contained in U1 (resp. U2), then K⊂K1∪K2, whence
α(K)≤α(K1∪K2)≤α(K1)+α(K2)≤α(U1)+α(U2),
which establishes the asserted property.
5° Properties of Φ and α.
The definition of Φ and α can now be transformed as follows (taking into account (PC_{II})): in order that M∈Φ, it is necessary and sufficient that, for every ε>0, there exist a compact set K and an open set U∈Φ such that K⊂M⊂U and α(U)−α(K)≤ε; α(M) is, moreover, the infimum of the α(U) for the open sets U∈Φ containing M, and the supremum of the α(K) for the compact sets K⊂M.
From this, we shall first deduce that if M1,M2 and M1∪M2 belong to Φ, then α(M1∪M2)≤α(M1)+α(M2). Indeed, if U1 and U2 are two open sets of Φ containing M1 and M2, respectively, and such that α(U1)≤α(M1)+ε and α(U2)≤α(M2)+ε, then U1∪U2 belongs to Φ, contains M1∪M2, and consequently
α(M1∪M2)≤α(U1∪U2)≤α(U1)+α(U2)≤α(M1)+α(M2)+2ε,
whence our assertion.
Next, let us show that if K is a compact set and U is an open set of Φ such that K⊂U, then α(U−K)=α(U)−α(K). By the foregoing, we have α(U)≤α(K)+α(U−K). On the other hand, for every compact set K′⊂U−K,
α(K∪K′)=α(K)+α(K′)≤α(U);
since U−K is open and belongs to Φ, α(U−K) is the supremum of the α(K′), which shows that α(K)+α(U−K)≤α(U).
The definition of Φ may therefore now be expressed in the following way: in order that M∈Φ, it is necessary and sufficient that, for every ε>0, there exist a compact set K and an open set U∈Φ such that K⊂M⊂U and α(U−K)≤ε.
We are now in a position to prove that Φ is a clan and α an additive set function on Φ. We first show that if M and N belong to Φ then so do M∩CN and M∪N. By hypothesis, for every ε>0 there exist two compact sets K,K′ and two open sets U,U′ of Φ such that
K⊂M⊂U,K′⊂N⊂U′,α(U−K)≤ε,α(U′−K′)≤ε.
The set K′′=K∩CU′ is compact, the set U′′=U∩CK′ is open and belongs to Φ, and K′′⊂M∩CN⊂U′′; on the other hand, U′′−K′′ is contained in the union of U∩CK and U′∩CK′, whence α(U′′−K′′)≤2ε, which proves that M∩CN∈Φ. Similarly, U1=U∪U′ is open and belongs to Φ, K1=K∪K′ is compact, and K1⊂M∪N⊂U1; on the other hand, U1−K1 is contained in the union of U−K and U′−K′, whence again α(U1−K1)≤2ε, and M∪N belongs to Φ. Finally, if M and N are disjoint, then
α(K1)=α(K)+α(K′)≥α(M)+α(N)−2ε,
consequently α(M∪N)≥α(M)+α(N)−2ε; since ε is arbitrary, we have α(M∪N)=α(M)+α(N).
6° Existence of the measure μ.
By Prop. 18 of No. 9, there exists one and only one positive linear form β on the vector space E(Φ) of Φ-step functions, such that β(φM)=α(M) for all M∈Φ. For every compact subset K of X, let us denote by G(K) the space of uniform limits of functions of E(Φ) whose support is contained in K. Since β is positive, ∣β(f)∣≤α(K)⋅∥f∥ for every function f∈E(Φ) whose support is contained in K; the restriction of β to the space of these functions is a continuous linear form for the topology of uniform convergence; it may therefore be extended to a positive continuous linear form βK on G(K). Moreover, if K⊂K1, then the restriction of βK1 to G(K) is identical to βK, therefore there exists a positive linear form β on the union G of the G(K), that extends each of the forms βK.
Now, since every compact set belongs to Φ, the space H of continuous real-valued functions with compact support is a subspace of G (No. 10, Prop. 19); the restriction to H of the positive linear form β is therefore a positive measure μ. Let us show that for every compact set K, μ(K)=α(K). For every ε>0, there exists an open set U∈Φ such that K⊂U, μ(U)≤μ(K)+ε and α(U)≤α(K)+ε. Let f be a continuous mapping of X into [0,1] whose support is contained in U and such that f(x)=1 on K (Ch. III, §1, No. 2, Lemma 1). Then μ(K)≤μ(f)≤μ(U)≤μ(K)+ε, and, on the other hand,
α(K)=β(φK)≤β(f)≤β(φU)=α(U)≤α(K)+ε;
since μ(f)=β(f), we see that ∣μ(K)−α(K)∣≤ε, and since ε is arbitrary, μ(K)=α(K).
The characterization of the open sets belonging to Φ, combined with Cor. 4 of Th. 4 of No. 6, then shows that the open sets belonging to Φ are none other than the μ-integrable open sets, and that, for such a set U, we have μ(U)=α(U). Th. 4 of No. 6 and the characterization of the sets of Φ given in 5° then show that the μ-integrable sets are the sets of Φ and that, for such a set M, μ(M)=α(M). Finally, the fact that μ∗(U)=supM∈Φ,M⊂Uα(M) for every open set U follows at once from (PCII) and Cor. 4 of Th. 4 of No. 6.
Theorem 5 is thus completely proved.
Bourbaki puts the proof directly after the statement, so what is above is the statement and its proof together.