THEOREMS, ETC., FOR MATH 516 When a statement in this list corresponds to a result in the book, I have listed the statement from the book in parentheses. Note that, for example, Proposition 1.2 means Proposition 2 from Chapter 1. Theorem 1 (=Corollary 7.41, Arzela-Ascoli Theorem). Let (X, ρ) be a separable metric space, and let F be an equicontinuous, uniformly bounded family of continuous, real-valued functions defined on X. Then every sequence in F has a subsequence that converges at each point in X. The limit function is continuous, and the subsequence converges uniformly (to the limit function) on any compact subset of X. Proposition 2 (=Proposition 8.2). Let E, A, and B be subsets of a topological space (X, T ). Then E ⊂ Ē, Ē¯ = Ē, A ∪ B = Ā ∪ B̄. Also, E is closed if and only if E = Ē, and Ē is the set of all points of closure of E. Further E o ⊂ E, (E o )o = E o , (A ∩ B)o = Ao ∩ B o , E is open if and only if E = E o , and (E o )o = E 0 . Finally, ({E)o = {(Ē). Proposition 3 (=Proposition 8.3). Let (X, T ) be a topological space and let A1 and A2 be subsets of X with X = A1 ∪ A2 . Finally, let (Y, S ) be a topological space and let f : X → Y be a function such that f|A1 and f|A2 are continuous. If A1 and A2 are both open or both closed, then f is continuous. Proposition 4 (=Proposition 8.4). If C is a collection of subsets of a non-empty set X, then there is a weakest topology T such that C ⊂ T . Proposition 5 (=Proposition 8.5). Let B be a collection of subsets of a non-empty set X. Then B is a base for a topology T on X if and only if the following two conditions are satisfied. (i) For every x ∈ X, there is a B ∈ B such that x ∈ B. (ii) For every B1 and B2 in B and every x ∈ B1 ∩ B2 , there is a B ∈ B such that x ∈ B ⊂ B1 ∩ B2 . Date: May 1, 2009. 1 2 THEOREMS, ETC., FOR MATH 516 In addition, T is the weakest topology such that B ⊂ T . Proposition 6 (=Proposition 8.6). A topological space is T1 if and only if every one-point set is closed. Lemma 7 (=Lemma 8.7, Urysohn’s Lemma). Let (X, T ) be a normal topological space and let A and B be disjoint, closed, non-empty subsets of X. Then there is a continuous function f : X → [0, 1] such that f (A) = {0} and f (B) = {1}. Theorem 8 (=Theorem 8.8, Tietze Extension Theorem). Let (X, T ) be a normal topological space, let A be a closed subset of X, and let f be a continuous real-valued function defined on A. Then there is a continuous real-valued function g defined on X such that g|A = f . Proposition 9 (=Proposition 8.10). Let (X, T ) and (Y, S ) be topological spaces, and let f : X → Y be continuous. If X is connected, then f (X) is connected. Proposition 10 (=Proposition 8.11). Let (X, T ) be a connected topological space and let f be a continuous real-valued function on X. If x and y are points of X such that f (x) < f (y) and if c ∈ (f (x), f (y)), then there is a point z ∈ X such that f (z) = c. Proposition 11 (=Proposition 8.12). A non-empty subset of R is connected if and only if it is either an interval or a one-point set. Proposition 12 (=Proposition 9.1). A topological space is compact if and only if every collection of closed sets with the finite intersection property has non-empty intersection. Proposition 13 (=Proposition 9.2). A closed subset of a compact set is compact. A compact subset of a Hausdorff space is closed. Proposition 14 (=Proposition 9.4). The continuous image of a compact set is compact. Proposition 15 (=Proposition 9.5). A one-to-one onto, continuous map from a compact space to a Hausdorff space is a homeomorphism. Proposition 16 (=Proposition 9.6). The continuous image of a countably compact set is countably compact. The continuous image of a Lindelöf space is Lindelöf. Proposition 17 (=Proposition 9.7). A topological space is countably compact if and only if it has the Bolzano-Weierstrass property. Proposition 18 (=Proposition 9.8). A sequentially compact space is countably compact. A first countable, countably compact space is sequentially compact. THEOREMS, ETC., FOR MATH 516 3 Proposition 19 (=Proposition 9.10). Let f be a real-valued, upper semicontinuous function defined on a countably compact space. Then f is bounded above and f attains its maximum value. Proposition 20 (=Proposition 9.11, Dini). Let (X, T ) be a countably compact space and let (fn ) be a decreasing sequence of real-valued, upper semicontinuous functions defined on X. If lim fn (x) = 0 for all x ∈ X, then (fn ) converges uniformly to zero. Lemma 21 (=Lemma 9.12). Let X be a set and let A be a collection of subsets of X with the finite intersection property. Then there exists a maximal collection B of subsets of X with the finite intersection property such that A ⊂ B. Lemma 22 (=Lemma 9.13). Let X be a set and let B be a maximal collection of subsets of X with the finite intersection property. Then any intersection of finitely many elements of B is in B, and any subset of B which has non-empty intersection with any element of B is an element of B. Theorem 23 (=Theorem 9.14, Tychonoff’s Theorem). The product of a family of compact topological spaces is also compact (in the product topology). Proposition 24 (=Proposition 9.15). Let (X, T ) be a locally compact Hausdorff space, and let K be a compact subset of X. Then there is an open set O such that O is compact and K ⊂ O. For any such set O, there is a continuous function f : X → [0, 1] such that f is zero in the complement of O and f is one on K. Proposition 25 (=Proposition 9.16). Let (X, T ) be a locally compact Hausdorff space, let K be a compact subset of X, and let (Oλ )λ∈L be an open cover of K. Then there is a partition of unity subordinate to (Oλ )λ∈L consisting of finitely many continuous functions, each of which has compact support. Proposition 26 (=Proposition 9.29). Let (X, T ) be a compact topological space. If L ⊂ C(X) is a lattice such that h, defined by h(x) = inf{f (x) : f ∈ L}, is continuous, then, for any ε > 0, there is g ∈ L such that 0 ≤ g−h < ε in X. Lemma 27 (=Lemma 9.31). Let L be a family of real-valued functions defined on a set X. Suppose L separates points and that, for any f ∈ L and any c ∈ R, c + f and cf are in L. Then, for any x 6= y in X and 4 THEOREMS, ETC., FOR MATH 516 any real numbers a and b, there is f ∈ L such that f (x) = a and f (y) = b. Lemma 28 (=Lemma 9.32). Let (X, T ) be a compact topological space and let L ⊂ C(X) be a lattice. Suppose L separates points and that, for any f ∈ L and any c ∈ R, c + f and cf are in L. Then, for any real numbers a ≤ b, any closed subset F of X, and any p ∈ X \ F , there is f ∈ L such that f ≥ a on X, f (p) = a, and f > b on F . Proposition 29 (=Proposition 9.30). Let (X, T ) be a compact topological space and let L ⊂ C(X) be a lattice. Suppose L separates points and that, for any f ∈ L and any c ∈ R, c + f and cf are in L. Then, for any h ∈ C(X) and any ε > 0, there is g ∈ L such that 0 ≤ g −h < ε on X. Lemma 30 (=Lemma 9.33). For any ε > 0, there is a polynomial Pε such that |P (s) − |s|| < ε for all s ∈ [−1, 1]. Theorem 31 (=Theorem 9.34, Stone-Weierstrass Theorem). Let (X, T ) be a compact space and let A ⊂ C(X) be an algebra that separates points and contains the constant functions. Then A is dense in C(X). Corollary 32 (=Corollary 9.35, Weierstrass Approximation Theorem). Every continuous function on a closed bounded subset of Rn can be uniformly approximated by a polynomial. Proposition 33 (=Proposition 10.2). A bounded linear operator is uniformly continuous. If a linear functional is continuous at a point, then it is bounded. Proposition 34 (=Proposition 10.3). The set of all bounded linear operators from a normed linear space to a Banach space is a Banach space. Theorem 35 (=Theorem 10.4, Hahn-Banach Theorem). Let X be a vector space and let p be a real-valued function on X such that p(x+y) ≤ p(x) + p(y) for all x and y in X and p(αx) = αp(x) for all α ≥ 0 and all x ∈ X. Suppose f is a linear functional defined on some subspace S of X so that f (s) ≤ p(s) for all s ∈ S. Then there is a linear functional F defined on X such that F (x) ≤ p(x) for all x ∈ X and F|S = f . Proposition 36 (=Proposition 10.5). Let X be a vector space and let p be a real-valued function on X such that p(x + y) ≤ p(x) + p(y) for all x and y in X and p(αx) = αp(x) for all α ≥ 0 and all x ∈ X, and let f be a linear functional defined on some subspace S of X. Suppose G is a collection of linear operators on X such that for all A and B in G, we have AB ∈ G and AB = BA. Suppose also that, for all THEOREMS, ETC., FOR MATH 516 5 A ∈ G, all s ∈ S, and all x ∈ X, we have p(Ax) ≤ p(x), As ∈ S, and f (As) = f (s). Then there is an extension F of f to a linear functional defined on all of X with F (x) ≤ p(x) and F (Ax) = F (x) for all x ∈ X and all A ∈ G. Proposition 37 (=Proposition 10.6). Let X be a normed vector space and let x ∈ X. Then there is a bounded linear functional F on X such that F (x) = kF kkxk. Proposition 38 (=Proposition 10.7). Let T be a subspace of a normed linear space X, and suppose there are x ∈ X and δ > 0 such that kx − tk > δ for all t ∈ T . Then there is a bounded linear functional F on X such that kF k ≤ 1, F (x) = δ, and F (t) = 0 for all t ∈ T . Theorem 39 (=Lemma 10.9+Proposition 10.10, Open Mapping Theorem). Let A be a bounded linear operator from a Banach space X onto a Banach space Y . Then A is an open mapping. Corollary 40 (=Proposition 10.11). Let X be a vector space, and suppose k · k1 and k · k2 are two norms that make X into a Banach space. If there is a constant c1 such that kxk1 ≤ c1 kxk2 for all x ∈ X, then there is a constant c2 such that kxk2 ≤ c2 kxk1 for all x ∈ X. Theorem 41 (=Theorem 10.12, Closed Graph Theorem). Let A be a linear operator from a Banach space X to a Banach space Y . Suppose that whenever (xn ) is a convergent sequence in X (with limit x) such that (Axn ) is a convergent sequence in Y (with limit y), we have y = Ax. Then A is a bounded linear operator. Proposition 42 (=Proposition 10.13, Uniform Boundedness Principle). Let F be a family of linear operators from a Banach space X to a normed vector space Y . If, for every x ∈ X, there is a nonnegative number Mx such that kf (x)kY ≤ Mx for all f ∈ F , then there is a nonnegative constant M such that kf k ≤ M for all f ∈ F . Proposition 43 (=Proposition 10.14). Let X be a vector space.If B) is a family of subsets of X all containing 0 such that (i) for all U and V in B0 , there is W ∈ B0 such that W ⊂ U ∩ V , (ii) for all U ∈ B0 and all x ∈ U , there is V ∈ B0 such that x + V ⊂ U, (iii) for all U ∈ B0 , there is V ∈ B0 such that V + V ⊂ U , 6 THEOREMS, ETC., FOR MATH 516 (iv) for all U ∈ B0 and all x ∈ X, there is a real number α such that x ∈ αU , (v) for all U ∈ B0 and all α ∈ R such that 0 < |α| ≤ 1, we have αU ⊂ U and αU ∈ B0 , then B0 is a base at 0 for a topology that makes X into a topological vector space. Conversely, if T is a topology that makes X into a topological vector space, then there is a base at 0, B0 , satisfying conditions (i)–(v). In addition (X, T ) is Hausdorff if and only if (vi) the only element in U for all U ∈ B0 is 0. Proposition 44 (=Proposition 10.15, Tychonoff). Let X be a Hausdorff topological space of finite dimension n. Then X is topologically isomorphic to Rn . Proposition 45 (=Proposition 10.16). A subspace of a Banach space is weakly closed if and only if it is strongly closed. Theorem 46 (=Theorem 10.17, Alagoulu’s Theorem). The closed unit ball in a Banach space is weakly compact. Lemma 47 (=Lemma 10.18). If K1 and K2 are convex subsets of a vector space X and if λ ∈ R, then K1 ∩ K2 , λK1 , and K1 + K2 are also convex. Lemma 48 (=Lemma 10.19). Let K be a convex subset of a vector space X and suppose that 0 is an internal point of K. Then (1) p(λx) = λp(x) for all λ ≥ 0 and all x ∈ X. (2) p(x + y) ≤ p(x) + p(y) for all x and y in X. (3) If p(x) < 1, then x ∈ K. If x ∈ K, then p(x) ≤ 1. Theorem 49 (=Theorem 10.20). Let K1 and K2 be two disjoint convex subsets of a vector space X. If one has an internal point, then the two sets are separated by a non-zero linear functional. Proposition 50 (=Proposition 10.22). Let X be a locally convex topological vector space, and let F be a closed convex subset of X. Then, for any x0 ∈ X \ F , there is a continuous linear functional f on X such that f (x0 ) < inf{f (x) : x ∈ F }. Lemma 51 (=Lemma 10.25). Let f be a continuous linear functional on a closed convex subset K of a topological vector space. Then the set S = {x : f (x) = max f } is a supporting set for K. THEOREMS, ETC., FOR MATH 516 7 Theorem 52 (=Theorem 10.26, Krein-Milman). Let K be a nonempty compact convex subset of a locally convex topological vector space. Then K is the closed convex hull of its extreme points. Proposition 53 (=Proposition 10.27). Every Hilbert space H has a complete orthonormal system. If H is separable, then every orthonormal system is countable. Moreover, if H is separable and (ϕn ) is a complete orthonormal system for K, then, for every x ∈ H, we have x= ∞ X (x, ϕn )ϕn n=1 and 2 kxk = ∞ X (x, ϕn )2 . n=1 Theorem 54 (Projection Theorem, not in Royden). Let M be a closed subspace of a Hilbert space X. Then for every x ∈ X, there is a unique p(x) ∈ M such that x − p(x) is orthogonal to every element of M . Theorem 55 (=Proposition 10.28, Riesz Representation Theorem). If f is a bounded linear functional on a Hilbert space X, then there is a unique z ∈ X such that f (x) = (x, z) for all x ∈ X. Moreover kzk = kf k. Proposition 56 (=Proposition 11.1). Let (X, B, µ) be a measure space. If A and B are measurable with A ⊂ B, then µA ≤ µB. Proposition 57 (=Proposition 11.2). Let (X, B, µ) be a measure space and let (Ei ) be a sequence of measurable sets with Ei+1 ⊂ Ei for all i and µE1 < ∞. Then Ã∞ ! \ µ Ei = lim µEn . i=1 Proposition 58 (=Proposition 11.3). Let (X, B, µ) be a measure space and let (Ei ) be a sequence of measurable sets. Then Ã∞ ! ∞ X [ µEi . µ Ei ≤ i=1 i=1 Proposition 59 (=Proposition 11.4). Let (X, B, µ) be a measure space. Then there is a unique complete measure space (X, B0 , µ0 ) such that (1) B ⊂ B0 , (2) µE = µ0 E for all E ∈ B, 8 THEOREMS, ETC., FOR MATH 516 (3) E ∈ B0 if and only if there are sets B and C in B and a subset A of C such that µC = 0 and E = A ∪ B. Proposition 60 (=Proposition 11.5). Let (X, B) be a measurable space, and let f be an extended real-valued function on X. Then the following conditions are equivalent. (1) {x : f (x) < α} ∈ B for all αinR. (2) {x : f (x) > α} ∈ B for all αinR. (3) {x : f (x) ≤ α} ∈ B for all αinR. (4) {x : f (x) ≥ α} ∈ B for all αinR. Theorem 61 (=Theorem 11.6). Let (X, B) be a measurable space. If f and g are measurable functions and if c ∈ R, then f + c, cf, f + g, f g, min{f, g}, max{f, g} are measurable. If (fn ) is a sequence of measurable functions, then sup fn , inf fn , lim fn , lim fn are measurable. Proposition 62 (=Proposition 11.7). Let (X, B, µ) be a measure space, and let f be a nonnegative measurable function. Then there is an increasing sequence (ϕn ) of simple functions such that ϕn → f . If X is σ-finite, then each ϕn can be chosen to vanish outside a set of finite measure. Proposition 63 (=Proposition 11.8). Let (X, B, µ) be a complete measure space, and let f be a measurable function. If g is an extended real-valued function with f = g a.e., then g is measurable. Lemma 64 (=Lemma 11.9). Let (X, B) be a measurable space, let D be a dense subset of R, and suppose that, for each α ∈ D, there is a measurable set Bα such that α < β implies that Bα ⊂ Bβ . Then there is a unique measurable function f such that f (x) ≤ α for all x ∈ Bα and f (x) ≥ α for all x ∈ X \ Bα . Proposition 65 (=Proposition 11.10). Let (X, B, µ) be a measure space, let D be a dense subset of R, and suppose that, for each α ∈ D, there is a measurable set Bα such that α < β implies that µ(Bα \ Bβ ) = 0. Then there is a measurable function f such that f ≥ α a.e. on Bα and f ≤ α a.e. on X \ Bα . If g is any other function with this property, then f = g a.e. Lemma 66 (=Theorem 11.11 , Fatou’s Lemma). Let (fn ) be a sequence of nonnegative measurable functions and suppose that f = lim fn exists THEOREMS, ETC., FOR MATH 516 a.e. Then Z 9 Z f dµ ≤ lim E fn dµ E for any measurable set E. Theorem 67 (=Theorem 11.12, Monotone Convergence Theorem). Let (fn ) be an increasing sequence of nonnegative measurable functions, and set f = fn . Then Z Z f dµ = lim fn dµ E E for any measurable set E. Proposition 68 (=Proposition 11.13). Let f and g be nonnegative measurable functions, and let a and b be nonnegative constants. Then Z Z Z (af + bg) dµ = a f dµ + b g dµ E E E R R for any measurable E. In addition, E f ≥ 0 and E f = 0 if and only if f = 0 a.e. on E. Corollary 69 (=Corollary 11.14). Let (fn ) be a sequence of nonnegative measurable functions. Then Z X ∞ ∞ Z X fn dµ = fn dµ E n=1 n=1 E for any measurable E. Proposition 70 (=Proposition 11.15). Suppose f and g are integrable functions and E is a measurable set. (1) If a and b are real numbers, then Z Z Z (af + bg) dµ = a f dµ + b g dµ. E E E (2) If h is a measurable function with |h| ≤ |f | a.e., then h is integrable. R R (3) If f ≥ g a.e., then E f dµ ≥ E g dµ. Theorem 71 (=Theorem 11.16, Lebesgue Dominated Convergence Theorem). Let (fn ) be a sequence of measurable functions with f = lim fn a.e., and suppose that there is a nonnegative integrable function g such that |fn | ≤ g a.e. for all n. Then f is integrable and Z Z f dµ = lim fn dµ. E E 10 THEOREMS, ETC., FOR MATH 516 Proposition 72 (=Proposition 11.17). Let (µn ) be a sequence of measures on a measurable space (X, B) which converges setwise to a measure µ. Then µ is a measure on (X, B). In addition, if (fn ) is a sequence of nonnegative measurable functions such that f = lim fn exists, then Z Z f dµ ≤ lim fn dµn . E E Proposition 73 (=Proposition 11.18). Let (µn ) be a sequence of measures on a measurable space (X, B) which converges setwise to a measure µ. Suppose that (fn ) and (gn ) are sequences of measurable functions such that f = lim fn and g = lim gn exist everywhere. Suppose also that |fn | ≤ gn for all n and that Z Z g dµ = lim gn dµn < ∞. Then Z Z f dµ = lim fn dµn . Lemma 74 (=Lemma 11.19). Every measurable subset of a positive set is positive. The union of countably many positive sets is positive. Lemma 75 (=Lemma 11.20). Let (X, B, ν) be a signed measure space and let E be a measurable set with 0 < νE < ∞. Then there is a positive set A with A ⊂ E and νA > 0. Theorem 76 (=Proposition 11.21, Hahn Decomposition Theorem). Let (X, B, ν) be a signed measure space. Then there are disjoint measurable sets A and B with A ∪ B = X such that A is a positive set and B is a negative set. Proposition 77 (=Proposition 11.22). Let (X, B, ν) be a signed measure space. Then there are two mutually singular measures ν + and ν − such that ν = ν + − ν − . Moreover, there is only pair of such measures. Theorem 78 (=Theorem 11.23, Radon-Nikodym Theorem). Let (X, B) be a measurable space and let µ and ν be two measures on B. If µ is σ-finite and if ν is absolutely continuous with respect to µ, then there is a nonnegative measurable function f such that Z νE = f dµ E for all measurable sets E. If g is any other such function, then f = g a.e.[µ]. THEOREMS, ETC., FOR MATH 516 11 Theorem 79 (=Theorem 11.24, Lebesgue Decomposition). Let (X, B, µ) be a σ-finite measure space, and let ν be a σ-finite measure on (X, B). Then there are unique measures ν0 and ν1 such that ν0 ⊥ µ, ν1 ¿ µ and ν = ν0 + ν1 . Theorem 80 (=Theorem 11.25). For 1 ≤ p ≤ ∞, Lp (µ) is a Banach space. For p1 + 1q = 1, if f ∈ Lp (µ) and g ∈ Lq (µ), then f g ∈ L1 (µ) and ¯ Z ¯Z ¯ ¯ ¯ f g dµ¯ ≤ |f g| dµ ≤ kf kp kgkq . ¯ ¯ Proposition 81 (=Proposition 11.26). Let 1 ≤ p < ∞. For every f ∈ Lp (µ) and every ε > 0, there is a simple function ϕ which vanishes outside a set of finite measure such that kf − ϕkp < ε. Lemma 82 (=Lemma 11.27). Let (X, B, µ) be a finite measure space and let 1 ≤ p < ∞. If g ∈ L1 (µ) and if there is a constant M such that ¯Z ¯ ¯ ¯ ¯ gϕ dµ¯ ≤ M kϕkp ¯ ¯ for all simple functions ϕ, then g ∈ Lq (µ) for p1 + 1q = 1, and kgkq ≤ M . Lemma 83 (=Lemma 11.28). Let (X, B, µ) be a measure space, let (En ) be a sequence of disjoint measurable sets, let (fn ) be a sequence in LpP (µ) for some p ∈ [1, ∞) such thatP fn = 0 a.e outside En , and set f = Pfn . Then f ∈ Lp if and only if kfn kp P < ∞. In this case, the sum fn converges in Lp (µ) to f and kf kpp = kfn kpp . Theorem 84 (=Theorem 11.29, Riesz Representation). Let 1 ≤ p < ∞ and let (X, B, µ) be a σ-finite measure space. Then, for every bounded linear function F on Lp (µ), there is a unique g ∈ Lq (µ) (with 1 + 1q = 1) such that p Z F (f ) = f g dµ for all f ∈ Lp (µ). In addition, kgkq = kF k. Theorem 85 (=Theorem 11.30, Riesz Representation). Let 1 < p < ∞ and let (X, B, µ) be a measure space. Then, for every bounded linear function F on Lp (µ), there is a unique g ∈ Lq (µ) (with p1 + 1q = 1) such that Z F (f ) = f g dµ for all f ∈ Lp (µ). In addition, kgkq = kF k. 12 THEOREMS, ETC., FOR MATH 516 Theorem 86 (=Theorem 12.1). Let µ∗ be an outer measure for a set X. Then B, the collection of all measurable sets, is a σ-algebra, and µ̄, the restriction of µ∗ to B, is a complete measure on B. Lemma 87 (= Lemma 12.2). Let µ be a measure on A , an algebra of subsets of a setSX, and let (Ai ) P be a sequence of elements of A . If A ∈ A and A ⊂ Ai , then µA ≤ µAi . Corollary 88 (=Corollary 12.3). The restriction of µ∗ to A is just µ. Lemma 89 (= Lemma 12.4). The function µ∗ is an outer measure. Lemma 90 (= Lemma 12.5). If A ∈ A , then A is measurable with respect to µ∗ . Proposition 91 (=Proposition 12.6). Let µ be a measure on an algebra A of subsets of a set X and let µ∗ be the outer measure induced by µ. Then, for any E ⊂ X and any ε > 0, there is A ∈ Aσ with E ⊂ A and µ∗ A ≤ µ∗ E + ε. In addition, for any E ⊂ X, there is B ∈ Aσδ such that E ⊂ B and µ∗ B = µ∗ E. Proposition 92 (=Proposition 12.7). Let µ be a σ-finite measure on an algebra A of subsets of a set X and let µ∗ be the outer measure induced by µ. Then E ⊂ X is measurable if and only if there are sets A ∈ Aσδ and B ⊂ A with µ∗ B = 0 such that E = A \ B. In addition, if B is any subset of X with µ∗ B = 0, then there is C ∈ Aσδ with µ∗ C = 0 and B ⊂ C. Theorem 93 (=Theorem 12.9, Carathéodory). Let µ be a measure on an algebra A of subsets of a set X and let µ∗ be the induced outer measure. Then µ̄, the restriction of µ∗ to the σ-algebra B of all µ∗ measurable sets, is an extension of µ to the σ-algebra B and A ⊂ B. If µ is finite, so is µ̄. If µ is σ-finite, so is µ̄, and µ̄ is the unique measure on the σ-algebra generated by A which extends µ. Proposition 94 (=Proposition 12.10). Let C be a semialgebra of subsets of a set X, and let µ be a set function on C such that µ∅ = 0 if ∅ ∈ C . Suppose also that (1) If C ∈ C is a union Pof finitely many disjoint elements C1 , . . . , Cn of C , then µC = µCi . (2) If C ∈ C is a union of P countably many disjoint elements C1 , . . . of C , then µC ≤ µCi . Then there is a unique extension of µ to a measure on the algebra generated by C . THEOREMS, ETC., FOR MATH 516 13 Lemma 95 (=Lemma 12.14). Let {Ri } be a countable collection of disjoint measurable P rectangles such that R = ∪Ri is a measurable rectangle. Then λR = λRi . Lemma 96 (=Lemma 12.15). Let x ∈ X and E ∈ Rσδ . Then Ex is measurable. Lemma 97 (=Lemma 12.16). Let E ∈ Rσδ and suppose µ × νE < ∞. Then g, defined by g(x) = ν(Ex ) is measurable and Z g dµ = µ × ν(E). Lemma 98 (=Lemma 12.17). If µ × ν(E) = 0, then ν(Ex ) = 0 for almost all x. Proposition 99 (=Proposition 12.18). Let E be a measurable subset of X × Y with µ × ν(E) < ∞. Then for almost all x, Ex is measurable, the function g, defined by g(x) = ν(Ex ), is measurable, and Z g dµ = µ × νE. In the next two theorems, if f is a function on X × Y , we define fx and f y by fx (y) = f (x, y) and f y (x) = f (x, y). Theorem 100 (=Theorem 12.19, Fubini). Let (X, A , µ) and (Y, B, ν) be complete measure spaces and let f be integrable on X × Y . Then (i) (i)0 (ii) (ii)0 (iii) For almost all x, fx is integrable. For almost all x, fx is integrable. For almost all x, fx is integrable. For ¢ R ¡Ralmost ¢all x, fxR is integrable. R ¡R f dν dµ = f d(µ × ν) = f dµ dν. X Y X×Y Y X Theorem 101 (=Theorem 12.20, Tonelli). Let (X, A , µ) and (Y, B, ν) be σ-finite complete measure spaces and let f be nonnegative and measurable on X × Y . Then (i) (i)0 (ii) (ii)0 (iii) For almost all x, fx is measurable. For almost all x, fx is measurable. For almost all x, fx is measurable. For ¢ R ¡Ralmost ¢all x, fxR is measurable. R ¡R f dν dµ = X×Y f d(µ × ν) = Y X f dµ dν. Y X Proposition 102 (=Proposition 12.40). If µ∗ is a Carathéodory outer measure with respect to Γ, then every function in Γ is µ∗ -measurable. 14 THEOREMS, ETC., FOR MATH 516 Proposition 103 (=Proposition 12.41). Let µ∗ be an outer measure on a metric space (X, ρ) and suppose that, for any two subsets A and B of X with ρ(A, B) > 0, we have µ∗ A + µ∗ B = µ∗ (A + B). Then every closed set is measurable.