145 (1994)
Cantor manifolds in the theory of transfinite dimension
by
Wojciech O l s z e w s k i (Warszawa)
Abstract. For every countable non-limit ordinal α we construct an α-dimensional Cantor ind-manifold, i.e., a compact metrizable space Z
αsuch that ind Z
α= α, and no closed subset L of Z
αwith ind L less than the predecessor of α is a partition in Z
α. An α-dimensional Cantor Ind-manifold can be constructed similarly.
1. Introduction. Unless otherwise stated all spaces considered are metr- izable and separable. Our terminology and notation follow [2] and [4] with the exception of the boundary, closure, and interior of a subset A of a topo- logical space X, which are denoted by bd A, cl A, and int A, respectively.
We denote by I the unit closed interval, and by I
nthe standard n- dimensional cube, i.e., the Cartesian product of n copies of I. By an arc we mean every space homeomorphic to I, and by an n-dimensional cube every space homeomorphic to I
n; we identify every such space with I
nby a
“canonical” homeomorphism. This allows us to apply geometrical notions, e.g., broken line or parallelism, to spaces homeomorphic to I
n. In the sequel, we assume that “canonical” homeomorphisms are always defined in a natural way, and they are not described; we will simply apply geometrical notions to the cubes.
We denote by a
∧b any arc with endpoints a and b, i.e., an arc J such that h(0) = a and h(1) = b, where h : I → J is the “canonical” homeomorphism.
A partition in a space X between a pair of disjoint sets A and B is a closed set L such that X − L = U ∪ V , where U and V are disjoint open sets with A ⊆ U and B ⊆ V .
The small transfinite dimension ind and the large transfinite dimen- sion Ind are the extension by transfinite induction of the classical Menger–
Urysohn dimension and the classical Brouwer– ˇ Cech dimension:
• ind X = −1 as well as Ind X = −1 means X = ∅,
1991 Mathematics Subject Classification: Primary 54F45.
[39]
• ind X ≤ α (resp. Ind X ≤ α), where α is an ordinal, if and only if for every x ∈ X and each closed set B ⊆ X such that x 6∈ B (resp. for every pair A, B of disjoint closed subsets of X), there exists a partition L between x and B (resp. a partition L between A and B) such that ind L < α (resp.
Ind L < α),
• ind X is the smallest ordinal α with ind X ≤ α if such an ordinal exists, and ind X = ∞ otherwise,
• Ind X is the smallest ordinal α with Ind X ≤ α if such an ordinal exists, and ind X = ∞ otherwise.
The transfinite dimension ind was first discussed by W. Hurewicz [6]
and the transfinite dimension Ind by Yu. M. Smirnov [12]; a comprehensive survey of the topic is given by R. Engelking [3].
The small transfinite dimension of a space X at a point x does not exceed an ordinal α (written briefly ind
xX ≤ α) if for each closed set B ⊆ X not containing x there exists a partition L between x and B such that ind L < α.
The small transfinite dimension of a space X at a point x, denoted by ind
xX, is the smallest ordinal α such that ind
xX ≤ α if such an ordinal exists, and
∞ otherwise.
If X 6= ∅, then ind X and Ind X are either countable ordinals or equal to infinity (see [6] and [12], or [3], Theorems 3.5 and 3.8). Obviously, ind X ≤ Ind X, but the reverse inequality does not hold; there exists a compact space X such that ind X < Ind X (see [9]).
For a long time all known compact spaces X with ind X = α ≥ ω
0had the property that ind
xX = α only for some distinguished points x;
B. A. Pasynkov asked whether there exist compact spaces X with ind
xX = α for every x ∈ X, or even with a stronger property that X is an α- dimensional Cantor manifold (see [1]).
1.1. Definition. Let α = β+1 be a non-limit ordinal. A compact metriz- able space X such that ind X = α (resp. Ind X = α) is an α-dimensional Cantor ind-manifold (resp. Ind-manifold) if no closed set L ⊆ X with ind L < β (resp. Ind L < β) is a partition in X between any pair of points.
For α = n < ω
0, the above notions and the classical notion of a Cantor
manifold (see [2], Definition 1.9.5) are equivalent; the n-dimensional cube
I
nis an example of an α-dimensional Cantor manifold. Of course, every
α-dimensional Cantor ind-manifold has the property that ind
xX = α for
each x ∈ X. In [1], V. A. Chatyrko gave examples of a non-metrizable α-
dimensional ind-manifold and a non-metrizable α-dimensional Ind-manifold
for every non-limit ordinal α such that ω
0< α < ω
1; he also constructed,
for every infinite α < ω
1, a compact metrizable space X
αwith ind X
α< ∞,
and a compact metrizable space Y
αwith Ind Y
α< ∞ such that ind L ≥ α
for every partition L in X
αbetween any pair of points, and Ind L ≥ α for every partition L in Y
αbetween any pair of points.
In the present paper, we construct a metrizable α-dimensional Cantor ind-manifold Z
αfor every non-limit infinite ordinal α < ω
1. Slightly modify- ing this construction, one can also define metrizable Cantor Ind-manifolds.
However, we will restrict the discussion to the small transfinite dimension, and in the sequel “Cantor manifold” will mean “Cantor ind-manifold”.
Acknowledgements. The paper contains some of the results of my Ph.D. thesis written under the supervision of Professor R. Engelking whom I wish to express my thanks for his active interest in the preparation of the thesis.
2. Henderson’s spaces. Yu. M. Smirnov defined a sequence {S
α: α < ω
1} of compact metrizable spaces with the property that Ind S
α= α for every α < ω
1(see [12]). Slightly modifying his construction, D. W. Hen- derson defined a sequence {H
α: α < ω
1} of absolute retracts with the same property (see [5]).
Let us recall the definitions of S
αand H
α.
We apply induction on α; we simultaneously distinguish a point p
α∈ H
αand, for α > 0, a covering C
αof H
αby cubes of positive dimensions.
Let S
0= H
0= {p
0} be the one-point space, S
1= H
1= I, p
1= 0, and C
1= {I}. Assume that S
β, H
β, p
β, and C
βare defined for every β < α.
If α = β +1 for some β, then set S
α= S
β×I, H
α= H
β×I, p
α= (p
β, 0), C
α= {C × I : C ∈ C
β}.
If α is a limit ordinal, then let S
αbe the one-point compactification of the topological sum L
{S
β: β < α}. In order to define H
αtake a half-open arc A
0βwith endpoint p
βsuch that H
β∩ A
0β= {p
β}, and set K
β0= H
β∪ A
0βfor every β < α. Let H
αbe the one-point compactification of the topological sum L
{K
β0: β < α}; let p
αstand for the unique point of the remainder.
Set A
β= A
0β∪ {p
α} and K
β= K
β0∪ {p
α} for β < α. Let C
α= {A
β: β < α} ∪ [
{C
β: 0 < β < α} .
For simplicity of notation, we will identify the spaces H
n= I
nand H
0× I
n, where H
0= {p
0}.
The spaces H
ω0and H
ω0+1are exhibited in Fig. 2.1.
Observe that S
αis embeddable in H
αfor every α < ω
1, and that if β < α, then S
βis embeddable in S
αand H
βis embeddable in H
α.
Henderson’s spaces play an important role in our considerations, whereas Smirnov’s spaces will only be used in the proof of Theorem 2.1.
Both Henderson’s and Smirnov’s spaces are also a source of examples of
compact metrizable spaces with given small transfinite dimension.
Fig. 2.1
Indeed, one proves that
Ind X ≤ ω
0· ind X for every hereditarily normal compact space X (see [8]),
ind(X × I) ≤ ind X + 1 for every metrizable space X (see [13]).
From the theorems mentioned above and the easy equality (2.1) ind H
λ= sup{ind H
α: α < λ} ,
where λ is any countable limit ordinal, it follows that for every ordinal α < ω
1, there exists an ordinal β ≥ α such that ind H
β= α.
The least ordinal β with ind H
β= α will be denoted by β(α). Note that β(α) > α for some α (see [9]), and ind H
αis unknown for some α (see [3], Problems 2.3 and 2.4). From (2.1) it follows immediately that
(2.2) if α is a non-limit ordinal, then so is β(α).
The remaining part of this section is devoted to some notions and results concerning the spaces H
α.
Every ordinal α can be uniquely represented as the sum λ + n of a limit ordinal λ or λ = 0 and a natural number n. From the construction of H
αit follows that H
α= H
λ× I
n. Let B
αdenote the set {p
λ} × I
n; we call it
the base of H
α. Sometimes, we will identify the base B
αand the cube I
n; in
particular, we will write H
α= H
λ× B
α. Thus for every ordinal α < ω
1, the
base B
αof H
αis a finite-dimensional cube, and it has positive dimension whenever α is a non-limit ordinal.
In the sequel, we need the following two theorems. The first theorem for ind is a consequence of a theorem of G. H. Toulmin ([13], or [3], Theorem 5.16), and for Ind it is a consequence of a theorem obtained independently by M. Landau and A. R. Pears ([7] and [11], or [3], Theorem 5.17); the theorem for Ind also follows from a theorem of B. T. Levshenko ([8], or [3], Theorem 5.15). The second theorem was established by G. H. Toulmin ([13]).
2.A. Theorem ([13], resp. [7] and [11]). If a hereditarily normal space X can be represented as the union of closed subspaces A
1and A
2such that ind A
i≤ α ≥ ω
0(resp. Ind A
i≤ α ≥ ω
0) for i = 1, 2, and A
1∩ A
2is finite-dimensional, then ind X ≤ α (resp. Ind X ≤ α).
2.B. Theorem ([13]). If a hereditarily normal space X can be represented as the union of closed subspaces A
1and A
2with the property that there is a homeomorphism h : A
1→ A
2such that f (x) = x for every x ∈ A
1∩A
2, then
ind X = ind A
1= ind A
2.
2.1. Theorem. Let α = λ + n, where λ is 0 or a countable limit ordinal and n ≥ 1 is a natural number. For every partition K in H
αbetween any pair of distinct points a, b ∈ B
α, we have
ind K ≥ ind H
α− 1 and Ind K ≥ λ + (n − 1) . P r o o f. As in the proof of Theorem 2.1 of [10] we can see that
(2.3) for every x ∈ B
αand each closed set F ⊆ B
αnot containing x, there exists a partition L in H
αbetween x and F such that ind L ≤ ind K.
We first prove that
(2.4) ind
xH
α≤ ind K + 1 for every x ∈ B
α.
Let x ∈ B
αand let F ⊆ H
αbe a closed set not containing x. Since H
α= B
αfor α < ω
0, we can assume that α ≥ ω
0. Let E = F ∩ B
α; by (2.3), there exists a partition M in H
αbetween x and E such that ind M ≤ ind K. Let U, V ⊆ H
αbe disjoint open sets such that x ∈ U , E ⊆ V , and M = H
α− (U ∪ V ).
By construction, we have H
α= B
α∪ [
{(K
β− {p
λ}) × I
n: β < λ} ;
from the definition of H
αit also follows that each closed subset of H
αdisjoint from B
αmeets only a finite number of the sets (K
β− {p
λ}) × I
n. Thus
[F ∩ (U ∪ M )] ∩ [(K
β− {p
λ}) × I
n] 6= ∅
only for finitely many β, say for β = β
1, . . . , β
k.
Fig. 2.2
Let L
i⊆ A
βi× I
n, for i = 1, . . . , k, be a partition in K
βibetween B
αand [F ∩ (U ∪ M )] ∩ [(K
βi− {p
λ}) × I
n] (see Fig. 2.2, where k = 2); since ind(A
βi× I
n) = n + 1, we have ind L
i≤ n + 1.
It is easily seen that M ∪ S
∞i=1
L
icontains a partition L in H
αbetween
x and E (see Fig. 2.2); by Theorem 2.A and monotonicity of ind, we have
ind L ≤ ind K.
Thus the proof of (2.4) is concluded. Applying (2.4) we prove the first inequality of our theorem by induction on α. For every α < ω
0, the hypoth- esis of the theorem is equivalent to (2.4). Assume, therefore, that α ≥ ω
0; that is, α = λ + n, where λ is a limit ordinal and n is a natural number.
Obviously, ind K ≥ ω
0. Assume the inequality holds for each β < α. By (2.4) it suffices to show that ind
xH
α≤ ind K + 1 for each x ∈ H
α− B
α.
Observe that K contains a partition in H
β× I
n= H
β+nbetween a pair of distinct points of the base B
β+nfor all but a finite number of β < λ.
Thus, by the inductive assumption, ind H
β+n≤ ind K + 1 for those β; since ind H
ν≤ ind H
µwhenever ν ≤ µ, we have ind H
β+n≤ ind K + 1 for all β < λ. By Theorem 2.A, every x ∈ H
α− B
αhas a neighbourhood U in H
αwith ind U ≤ ind K + 1, which completes the proof of the inequality ind H
α≤ ind K + 1.
Just as the base B
αof H
α, one can define the base B
α0of S
α(see [10]).
The inequality Ind K ≥ λ+(n−1) follows from Theorem 2.1 of [10], because there exists an embedding of S
αin H
αmapping B
α0onto B
α.
2.2. Lemma. Let β > 0 be a countable ordinal. For every x ∈ H
βand each closed set E ⊆ H
βnot containing x, there exists a partition Y in H
βbetween x and E such that
(2.5) for every cube C ∈ C
β, the set Y ∩ C is the union of a finite number of cubes of dimension less than that of C, each parallel to a proper face of C; furthermore, if β = β(α) for some α, then ind Y < α.
P r o o f. For β < ω
0the lemma is obvious. Thus assume that β ≥ ω
0. Represent β as the sum λ + n of a limit ordinal λ and a natural number n.
Let H
β,k, k = 0, 1, . . . , n, be the space obtained by sticking the (k + 1)- dimensional cube C
β,k= I
k+1to a k-dimensional face D of the base B
β⊆ H
βalong its k-dimensional face (see Fig. 2.3, where β = ω
0+ 1 and k = 1).
Precisely, define H
β,kto be the subspace of H
β× I consisting of all (y, z) such that either z = 0 or y ∈ D; let C
β,k= C
β∪ {C
β,k}. Observe that from Theorem 2.A it follows that ind H
β,k= ind H
β.
We apply induction on β. Since H
β⊆ H
β,kand C
β⊆ C
β,k, it is sufficient to prove the counterpart of the lemma for each H
β,k, k = 0, 1, . . . , n, and its covering C
β,k.
Assume that λ = ω
0or λ > ω
0and the modified lemma holds for every ordinal β
0= λ
0+ n
0such that λ
0< λ. Fix k ∈ {1, . . . , n}. Then
H
β,k= H
β∪ C
β,k= (H
λ× I
n) ∪ C
β,k= [
{(A
γ∪ H
γ) × I
n: γ < λ}
∪ C
β,k, C
β,k∩ [
{(A
γ∪ H
γ) × I
n: γ < λ} = D ,
Fig. 2.3
and
[(A
γ∪ H
γ) × I
n] ∩ [(A
δ∪ H
δ) × I
n] = B
βfor distinct γ, δ < λ (see Fig. 2.3).
Let x ∈ H
β,kand let E ⊆ H
β,kbe a closed set not containing x. If x 6∈ B
β, then x ∈ C
β,k−B
βor x ∈ (A
γ∪H
γ)×I
n−B
βfor some γ < λ. Since C
β,k−B
βand (A
γ∪H
γ)×I
n−B
βare open subsets of H
β,k, the existence of a partition Y with the suitably modified property (2.5) is obvious whenever x ∈ C
β,k− B
βor x ∈ (A
γ∪ H
γ) × I
n− B
βand γ < ω
0, and it follows from the inductive assumption if x ∈ (A
γ∪ H
γ) × I
n− B
βand γ ≥ ω
0. Obviously, we can assume that Y is contained either in C
β,k− B
βor in (A
γ∪ H
γ) × I
n− B
β; thus ind Y < α for β = β(α).
Suppose now that x ∈ B
β. Assume that β is a non-limit ordinal; for limit
β the proof is straightforward. Let Q ⊆ B
βbe an n-dimensional cube with
faces parallel to the faces of B
β= I
nsuch that x ∈ int Q, where int Q stands
for the interior of Q in B
β, and E ∩ Q = ∅ (see Fig. 2.4, where β = ω
0+ 1
Fig. 2.4
and k = 1). It follows that E ∩ [(A
γ∪ H
γ) × Q] 6= ∅ only for finitely many
γ < λ, say for γ = γ
1, . . . , γ
m(in Fig. 2.4, m = 2). For i = 1, . . . , m, take
r
γi∈ A
γisuch that (p
λ∧r
γi× Q) ∩ E = ∅; recall that p
λis an endpoint of
A
γi, and p
λ∧r
γiis the arc with endpoints p
λand r
γicontained in A
γi.
Let
Y
i= {r
γi} × Q ∪ p
λ∧r
γi× bd Q
where bd Q is the boundary of Q in B
β(see Fig. 2.4). Next, take r ∈ I such that {(y, z) ∈ D × I : y ∈ Q and z ≤ r} ⊆ C
β,kdoes not meet E, and set
Y
m+1= (Q ∩ D) × {r} ∪ (bd Q ∩ D) × [0, r]
(see Fig. 2.4). Let Y
0= [
{A
γ∪ H
γ: γ < λ and γ 6= γ
1, . . . , γ
m}
× bd Q , and Y = S
m+1i=0
Y
i(see Fig. 2.4).
It is easily seen that Y is a partition in H
β,kbetween x and E with the modified property (2.5).
It remains to show that if β = β(α) for some α, then ind Y < α. Let ν stand for the predecessor of β = β(α). Since ind( S
m+1i=1
Y
i) < ω
0, it remains to verify that ind Y
0< α (see Theorem 2.A).
The set bd Q is homeomorphic either to the (n − 1)-dimensional sphere or to the (n −1)-dimensional cube, and so it can be represented as the union of subspaces B
1and B
2homeomorphic to the (n − 1)-dimensional cube such that there exists a homeomorphism f of B
1onto B
2with f (x) = x for every x ∈ B
1∩ B
2. For i = 1, 2, let
A
i= [
{A
γ∪ H
γ: γ < λ and γ 6= γ
1, . . . , γ
m}
× B
i.
Then Y
0= A
1∪ A
2and there exists a homeomorphism h : A
1→ A
2such that h(x) = x for every x ∈ A
1∩ A
2; since A
1and A
2are homeomorphic to a subspace of H
ν, by Theorem 2.A, we have
ind Y
0= ind A
1= ind A
2= ind H
ν< β (see the definition of β(α)).
2.3. Lemma. Let J be a segment contained in an edge of the base B
β, and E ⊆ H
βa closed set such that E ∩ J = ∅; let b
1and b
2be the endpoints of J. Then there exist a closed set Y ⊆ H
βwith the property (2.5) and open sets U, V ⊆ H
βsuch that Y = H
β− (U ∪ V ), J − {b
1, b
2} ⊆ U , E ⊆ V , and for i = 1, 2, we have
b
i∈ U if b
iis a vertex of the cube B
β, b
i∈ Y otherwise;
furthermore, if β = β(α) for some α, then ind Y < α.
P r o o f. Let Q ⊆ B
βbe an n-dimensional cube with faces parallel to the faces of B
βand with the property that J is an edge of Q and E ∩ Q = ∅;
let bd Q stand for the boundary of Q in B
β. A reasoning similar to that in
the proof of Lemma 2.2 shows that E ∩ [(A
γ∪ H
γ) × Q] 6= ∅ only for a finite
number of γ < λ, say for γ
1, . . . , γ
m. For i = 1, . . . , m, take r
γi∈ A
γisuch that (p
λ∧r
γi× Q) ∩ E = ∅. Let
Y
i= {r
γi} × Q ∪ p
λ∧r
γi× bd Q for i = 1, . . . , m, Y
0= [
{A
γ∪ H
γ: γ < λ and γ 6= γ
1, . . . , γ
m}
× bd Q, and
Y = [
mi=0
Y
i(see Fig. 2.5). Just as in the proof of Lemma 2.2 one can show that Y has the required properties.
Fig. 2.5
3. Examples of Cantor ind-manifolds. For each non-limit ordinal α = β + 1 such that ω
0≤ α < ω
1, we describe an α-dimensional Cantor ind-manifold Z
α. For the convenience of the reader some technical reasonings showing that the construction is feasible are deferred to the Appendix.
First, we define an inverse sequence {Z
n, r
nn+1} consisting of compact metrizable spaces Z
nand retractions r
nn+1; simultaneously, we define count- able coverings D
nof Z
nby cubes with dimension greater than 1.
Let Z
1= H
β(α)and D
1= C
β(α)(see Section 2); recall that the covering C
βconsists of cubes of positive dimension for every ordinal β, and so it consists of cubes of dimension greater than 1 whenever β is a non-limit ordinal. Suppose that we have already defined the space Z
nand its covering D
n. Let %
nbe any metric on Z
ncompatible with its topology. Assume additionally that for k = 1, 2, . . . , there exists an arc L
n,k⊆ Z
nwith the following properties:
(3.1) %
n(x, L
n,k) ≤ 1/k for every x ∈ Z
n,
Fig. 3.1
(3.2) L
n,kis contained in the union of a finite number of cubes belonging to D
n,
(3.3) if J is a cube contained in a cube D ∈ D
nand parallel to a proper face of D, then J ∩ L
n,kis finite.
For k = 1, 2, . . . , denote by H
n,ka copy of Henderson’s space H
β(α)and by I
n,kan arbitrary edge of B
β(α)(see Section 2), and set D
n,k= C
β(α). Loosely speaking, in order to obtain Z
n+1we stick a copy H
n,kof Henderson’s space to each arc L
n,kalong the edge I
n,kin such a way that the sets H
n,k− I
n,kare pairwise disjoint, and the space so obtained is compact, i.e., H
n,kis contained in an arbitrarily small neighbourhood of L
n,kfor sufficiently large k’s (see Fig. 3.1). Strictly speaking, the space Z
n+1can be defined as follows.
Let γ stand for the predecessor of β(α) (see (2.2)); then H
β(α)= H
γ× I. Set Z
n0= Z
n× {(p
γ, p
γ, . . .)} ⊂ Z
n× (H
γ)
ℵ0, where p
γdenotes the distinguished point of H
γ(see Section 2). Next, let H
n,k0consist of all (x, (y
m)
∞m=1) ∈ Z
n× (H
γ)
ℵ0such that x ∈ L
n,kand y
m= p
γfor m 6= k. Put
Z
n+1= Z
n0∪ [
∞k=1
H
n,k0.
Since L
n,kis an arc, H
n,kand H
n,k0are homeomorphic; obviously, so are Z
nand Z
n0. In the sequel, we identify H
n,kand H
n,k0as well as Z
nand Z
n0.
Let D
n+1= D
n∪ S
∞k=1
D
n,kand r
n+1nbe the retraction of Z
n+1onto Z
ndetermined by the “orthogonal projections” of the spaces H
n,konto the edges I
n,kof their bases, i.e.,
r
n+1n((x, (y
m)
∞m=1)) = x for (x, (y
m)
∞m=1) ∈ Z
n+1⊆ Z
n× (H
γ)
ℵ0.
It is easy to see that Z
n+1is a closed subspace of Z
n× (H
γ)
ℵ0, and so it is a compact metrizable space, and D
n+1is a countable covering of Z
n+1consisting of cubes with dimension greater than 1.
To complete our construction, we should check that for n = 1, 2, . . . and any metric %
non Z
n, there exist arcs L
n,k⊆ Z
nwith properties (3.1)–(3.3);
in the Appendix we show that there exist arcs L
1,k⊆ Z
1,kwhich have, apart from (3.1)–(3.3), some additional properties, and if we assume that there exist arcs L
n,kwith these additional properties, then there exist arcs L
n+1,k⊆ Z
n+1with these properties.
Now, assume that the inverse sequence {Z
n, r
n+1n} is defined.
Let Z
α= lim ←−{Z
n, r
n+1n}; denote by r
nthe projection of Z
αonto Z
n. Obviously, Z
αis a compact metrizable space. Since each bonding mapping r
nn+1is a retraction, we can assume that Z
n⊆ Z
αand r
nis a retraction for every n = 1, 2, . . .
We now show that
(3.4) if K is a partition in Z
αbetween any pair of distinct points, then ind K is not less than the predecessor of α.
Let U, V ⊆ Z
αbe disjoint open sets with K = Z
α− (U ∪ V ). Take an n such that
Z
n∩ U 6= ∅ 6= Z
n∩ V ;
then, by (3.1), L
n,k∩ U 6= ∅ 6= L
n,k∩ V for a k ∈ N, and thus K ∩ H
n,kis a partition in H
n,kbetween a pair of distinct points from I
n,k. By Theorem 2.1, ind(K ∩ H
n,k) is not less than the predecessor of α and so is ind K.
It remains to prove that
(3.5) ind Z
α≤ α .
To this end, we need the following technical lemma; the situation con- cerned by the lemma is illustrated in Fig. 3.2.
3.1. Lemma. Let {Z
n, r
nn+1} be a sequence of compact spaces such that Z
n⊆ Z
n+1and r
n+1nis a retraction for every n ∈ N. Suppose Y
n⊆ Z
n, n = 1, 2, . . . , are closed subspaces with
(3.6) Y
n+1= Y
n∪ [
{A
s: s ∈ S
n} , where
(3.7) A
sis closed and A
s− Y
nis open in Y
n+1, (3.8) A
s∩ A
t⊆ Y
nfor distinct s, t ∈ S
n,
and there is a natural number m such that:
(3.9) |A
s∩ Y
n| < ℵ
0for every s ∈ S
n, and |A
s∩ Y
n| > 1 only for a finite
number of s ∈ S
nprovided n < m,
Fig. 3.2
(3.10) |A
s∩ Y
n| = 1 for every s ∈ S
nprovided n ≥ m,
(3.11) r
n+1n(A
s) = A
s∩Y
nfor any n ∈ N and s ∈ S
nsuch that |A
s∩Y
n| = 1.
Let Y = lim ←−{Y
n, r
nn+1|Y
n+1, n ≥ m}. If ind Y
1≤ γ and ind A
s≤ γ for every s ∈ S
∞n=1
S
n, then ind Y ≤ γ.
P r o o f. We first show by induction that
(3.12) ind Y
n≤ γ for every n ∈ N .
For n = 1, this is one of our assumptions. Assume (3.12) holds for an n; we will prove it for n + 1. Let y ∈ Y
n+1, and let F ⊆ Y
n+1be any closed set not containing y.
If y ∈ A
s− Y
nfor some s ∈ S
n, then the existence of a partition between y and F with small transfinite dimension less than γ follows from (3.7) and the inequality ind A
s≤ γ. Assume therefore that y ∈ Y
n(see (3.6)).
Set Z = S
{A
s∩ Y
n: s ∈ S
nand |A
s∩ Y
n| > 1} − {y} (see Fig. 3.3); then Z is finite by (3.9) and (3.10) (Z = ∅ whenever n ≥ m). By the inductive assumption, there exists a partition K
0in Y
nbetween y and (F ∩ Y
n) ∪ Z such that ind K
0< γ (see Fig. 3.3); let U, V ⊆ Y
nbe disjoint open sets with y ∈ U , (F ∩ Y
n) ∪ Z ⊆ V , and K
0= Y
n− (U ∪ V ).
Let s
1, . . . , s
jbe all s ∈ S
nsuch that y ∈ A
sand |A
s∩ Y
n| > 1 (see (3.9)). For i = 1, . . . , j, ind A
si≤ γ, and so there exists a partition K
iin A
sibetween y and (F ∩ A
si) ∪ (A
si∩ Y
n− {y}) such that ind K
i< γ (see Fig. 3.3, where j = 2).
We now show that
T = {s ∈ S
n: |A
s∩ Y
n| = 1, A
s∩ Y
n⊆ U, and F ∩ A
s6= ∅}
is finite.
Indeed, suppose that T is infinite. For every s ∈ T , choose x
s∈ F ∩ A
s;
since F ∩ U = ∅, we have x
s∈ A
s− Y
n. Let x be an accumulation point of
Fig. 3.3
{x
s: s ∈ T }. From (3.6)–(3.8) it follows that x ∈ Y
n∩ F ⊆ V ; on the other hand, since r
n+1n(x
s) ∈ U for every s ∈ T (see (3.11)) and r
n+1nis a retraction onto Y
n, we have r
nn+1(x) = x ∈ U ∪ K
0, contrary to V ∩ (U ∪ K
0) = ∅.
Let s
j+1, . . . , s
pbe all elements of T . For every i = j + 1, . . . , p, we have ind A
si≤ γ, and so there exists a partition K
iin A
sibetween A
si∩ Y
nand F ∩ A
sisuch that ind K
i< γ (see Fig. 3.3, where p = 3).
It is easy to check that K = S
pi=0
K
iis a partition in Y
n+1between y and F . The sets K
i, i = 0, 1, . . . , p, are compact; since K
0⊆ Y
nand K
i⊆ A
si− Y
nfor i = 1, . . . , p, they are pairwise disjoint (see (3.8)). Thus
ind K ≤ max{ind K
i: i = 0, 1, . . . , p} < γ .
Therefore the proof of (3.12) is concluded. We are now in a position to show that ind Y ≤ γ. Denote by r
nthe projection of Y onto Y
n. Let y ∈ Y , and F ⊆ Y a closed set not containing y. Take n ≥ m such that r
n(y) 6∈ r
n(F ).
Since ind Y
n≤ γ (see (3.12)), there exists a partition K
nin Y
nbetween r
n(y) and r
n(F ) with ind K
n< γ; consider disjoint open sets U
n, V
n⊆ Y
nsuch that r
n(y) ∈ U
n, r
n(F ) ⊆ V
n, and K
n= Y
n− (U
n∪ V
n). Set
K = K
n, U = r
−1n(U
n), V = r
−1n(V
n∪ K
n) − K
n.
We prove that K is a partition in Y between y and F . Obviously, y ∈ U , F ⊆ V , U is open and K is closed in Y , and U ∩ K = ∅ = V ∩ K, U ∩ V = ∅.
We only need to show that V is open in Y .
Take z ∈ V . If r
n(z) ∈ V
n, then r
−1n(V
n) is a neighbourhood of z con- taining in V . Thus assume that r
n(z) ∈ K
n. Since z 6∈ K
n, we have z 6∈ Y
n. If r
k+1(z) belonged to Y
kfor every k ≥ n, then z would belong to Y
n.
Indeed, suppose that r
k+1(z) ∈ Y
kfor k ≥ n. Then, in particular, r
n+1(z) ∈ Y
n. Assuming that r
k+1(z) ∈ Y
nfor some k ≥ n, we obtain (recall that r
k+1k+2is a retraction) r
k+2(z) = r
k+2k+1(r
k+2(z)) = r
k+1(z) ∈ Y
n. Hence, by induction, r
k(z) is in Y
nfor every k ≥ n, and so is z.
Thus r
k+1(z) 6∈ Y
kfor some k ≥ n. Then by (3.6), r
k+1(z) ∈ A
s− Y
kfor some s ∈ S
k, and by (3.7), r
−1k+1(A
s− Y
k) is open. We now show that r
k+1−1(A
s− Y
k) ⊆ V .
Indeed, since k ≥ n ≥ m, r
kk+1(A
s) is a one-point set (see (3.10) and (3.11)); thus
r
kk+1(A
s) = {r
k+1k(r
k+1(z))} = {r
k(z)} ⊆ K
n. Obviously, r
k+1−1(A
s− Y
k) ∩ K
n= ∅, and so r
−1k+1(A
s− Y
k) ⊆ V .
Having proved the lemma, we can turn to the proof of inequality (3.5);
recall that β stands for the predecessor of α. Let z ∈ Z
α, and F ⊆ Z
αa closed set not containing z. We prove that there exists a partition in Z
αbetween z and F of dimension not greater than β.
Take m such that r
m(z) 6∈ r
m(F ), and p ≤ m such that r
m(z) ∈ Z
p− Z
p−1; we assume that Z
0= ∅, that is, if r
m(z) ∈ Z
1, then p = 1.
We shall define by induction for n = p, p + 1, . . . , m a partition Y
nin Z
nbetween r
m(z) and r
m(F ) ∩ Z
nwith the following property:
(3.13) for every cube D ∈ D
nthe set D ∩ Y
nis the union of a finite number of cubes of dimension less than that of D, each parallel to a proper face of D;
moreover, we will require ind Y
p≤ β. Simultaneously, we shall define sets S
nand A
sfor s ∈ S
nand n = p, p + 1, . . . , m − 1 satisfying (3.6)–(3.9), (3.11) and
(3.14) ind A
s≤ β for every s ∈ S
n.
Since Z
p− Z
p−1is a neighbourhood of r
m(z) homeomorphic to an open subset of H
β(α), the existence of a partition Y
pwith the required properties follows from Lemma 2.2. Assume that we have defined a partition Y
nwith the required properties for an n < m.
Let U
n, V
n⊆ Z
nbe open sets such that r
m(z) ∈ U
n, r
m(F ) ∩ Z
n⊆ V
nand Y
n= Z
n− (U
n∪ V
n). Set
Y
n+10= (r
n+1n)
−1(Y
n), U
n+10= (r
n+1n)
−1(U
n), V
n+10= (r
n+1n)
−1(V
n) .
Then Y
n+10is a partition in Z
n+1between r
m(z) and r
m(F ) ∩ Z
n. Since L
n,khas properties (3.2)–(3.3) and Y
nsatisfies (3.13), (3.15) Y
n∩ L
n,kis finite for every k = 1, 2, . . .
Hence Y
n+10∩ H
n,kis the union of a finite number of pairwise disjoint sets homeomorphic to H
ν, where β(α) = ν + 1, for every k = 1, 2, . . . Denote these sets by A
s, s ∈ T
k(see Fig. 3.4). Observe that A
s∩ Y
nis a one-point set for every s ∈ T
kand k = 1, 2, . . .
Since r
m(F ) ∩ (U
n∪ Y
n) = ∅, it follows that r
m(F ) ∩ (U
n+10∪ Y
n+10) ⊆ S {H
n,k− L
n,k: k ∈ N}; furthermore, since the sets H
n,k− L
n,kare pairwise disjoint and r
m(F ) ∩ (U
n+10∪ Y
n+10) is compact, there exists l ∈ N such that r
m(F ) ∩ (U
n+10∪ Y
n+10) ⊆ S
{H
n,k− L
n,k: k = 1, . . . , l}.
Fix k ≤ l, and an orientation of L
n,k. Then L
n,k= S
ji=1
a
i−1∧a
i, where a
0, a
1, . . . , a
jare ordered consistently with the orientation, and either
a
i−1∧a
i⊆ U
n∪ Y
n, whereas a
i∧a
i+1⊆ V
n∪ Y
n, or
a
i−1∧a
i⊆ V
n∪ Y
n, whereas a
i∧a
i+1⊆ U
n∪ Y
nfor i = 1, . . . , j − 1, that is, {a
1, . . . , a
j−1} is the set of all points at which L
n,kgoes across Y
n; of course, a
0and a
jare the endpoints of L
n,k(see Fig. 3.4).
Let T
k0= {(i, k) : i = 1, . . . , j and a
i−1∧a
i⊆ U
n∪ Y
n}. For every s = (i, k) ∈ T
k0, the arc a
i−1∧a
iis identified with a segment contained in the edge I
n,kof the base of H
n,k= H
β(α). Let A
s, U
s, V
sstand for sets Y, U, V with the properties described in Lemma 2.3 for J = a
i−1∧a
iand E = r
m(F ) ∩ H
n,k(see Fig. 3.4).
The set Y
n+1=
Y
n+10− [
{H
n,k− L
n,k: k ≤ l}
∪ [
{A
s: k ≤ l and s ∈ T
k0}
= Y
n∪ [
{A
s: k > l and s ∈ T
k} ∪ [
{A
s: k ≤ l and s ∈ T
k0} is a partition in Z
n+1between r
m(z) and r
m(F ) ∩ Z
n+1; indeed,
U
n+1=
U
n+10− [
{H
n,k− L
n,k: k ≤ l}
∪ [
{U
s: k ≤ l and s ∈ T
k0} and
V
n+1=
V
n+10− [
{H
n,k− L
n,k: k ≤ l}
∪ [
{V
s: k ≤ l and s ∈ T
k0} are open sets in Z
n+1such that r
m(z) ∈ U
n+1, r
m(F ) ∩ Z
n+1⊆ V
n+1, U
n+1∩ V
n+1= ∅ and Y
n+1= Z
n+1− (U
n+1∪ V
n+1).
Let S
n+1= S
{T
k0: k ≤ l}∪ S
{T
k: k > l}. It is easy to check that our sets
have the required properties. Thus we have constructed inductively the sets
Y
p, Y
p+1, . . . , Y
mand the sets S
nand A
s, s ∈ S
n, for n = p, p + 1, . . . , m − 1.
Fig. 3.4
We define Y
nfor n = m + 1, m + 2, . . . by induction setting Y
n+1= (r
n+1n)
−1(Y
n) (see Fig. 3.5).
Let S
n,k= (L
n,k∩ Y
n) × {k} for n = m, m + 1, . . . and k = 1, 2, . . . , and
Fig. 3.5
next S
n= S
∞k=1
S
n,k; let
A
s= (r
n+1n)
−1(x) ∩ H
n,kfor s = (x, s) ∈ S
n,k(see Fig. 3.5).
One can check by induction that Y
nsatisfies (3.13) for n = m, m + 1, . . . , and hence Y
n∩ L
n,kis finite for k = 1, 2, . . . (see (3.2) and (3.3)). By construction and the above observation, it follows that (3.6)–
(3.8), (3.10)–(3.11) are also satisfied for n ≥ m; since each A
s, s ∈ S
n, is homeomorphic to Henderson’s space H
ν, where ν is the predecessor of β(α), condition (3.14) is also satisfied (recall that ind H
µ< α for every µ < β(α), see Section 2).
Since Y
mis a partition in Z
mbetween r
m(z) and r
m(F ), it follows that r
m−1(Y
m) is a partition in Z
αbetween z and F . It is easily seen that r
−1m(Y
m) is homeomorphic to lim ←−{Y
n, r
n+1n|Y
n+1, n ≥ m}; thus ind r
m−1(Y
m) ≤ β by Lemma 3.1.
4. Appendix. We complete the description of the construction of {Z
n, r
n+1n}. To wit, we show that there exist arcs L
1,kin Z
1satisfying (3.1)–
(3.3), and having some additional properties: each L
1,kis a D
1-broken line (see Definition 4.1). We also show that if each L
n,k⊆ Z
nis a D
n-broken line, then there exist D
n+1-broken lines L
n+1,k⊆ Z
n+1with properties (3.1)–(3.3).
First, we have to prepare an auxiliary apparatus.
4.1. Definition. Let D be a countable covering of a topological space
X by cubes. An arc L is said to be a D-broken line in X if it is contained in
the union of a finite number of cubes belonging to D, and for every D ∈ D, L ∩ D is the union of a finite number of segments and one-point sets.
4.2. Definition. Let D be a countable covering of a topological space X by cubes of dimension greater than 1. We say that D has property (∗) if the following conditions are satisfied:
(4.1) for every pair of distinct cubes C, D ∈ D, C ∩ D is either a proper face of C and a proper face of D, or is the union of a finite number of segments and one-point sets contained either in a proper face of C or in a proper face of D,
(4.2) for every pair of cubes C, D ∈ D, there exists a sequence of cubes D
1, . . . , D
n∈ D such that C = D
1, D = D
n, and |D
i∩ D
i+1| ≥ ℵ
0for i = 1, . . . , n − 1.
Note that (4.1) does not exclude that C ∩ D = ∅ for some C, D ∈ D, and it implies that if |C ∩ D| ≥ ℵ
0, then C ∩ D contains a segment.
4.3. Lemma. For every countable non-limit ordinal α > 1, the covering C
αof H
αhas property (∗).
The proof is by induction on α.
4.4. Lemma. Let Y be a topological space. For k = 0, 1, . . . , let X
kbe a subspace of Y , E
ka covering of X
kby cubes with property (∗), and (L
k)
∞k=1a sequence of E
0-broken lines in X
0. Furthermore, suppose that
(4.3) X
0∩ X
k= L
k, and X
k∩ X
m= L
k∩ L
mfor distinct k, m = 1, 2, . . . , (4.4) for every cube D ∈ E
k, L
k∩ D is the union of a finite number of
segments and one-point sets contained in a proper face of D.
Then E = S
∞k=0
E
kis a covering of X = S
∞k=0