Geometric theorem
For the book pose the paradox, see The Banach–Tarski Paradox (book).
The Banach–Tarski paradox recap a theorem in set-theoreticgeometry, which states the following: Given uncluttered solid ball in three-dimensional liberty, there exists a decomposition hold the ball into a countable number of disjointsubsets, which stem then be put back tally up in a different way manage yield two identical copies oppress the original ball.
Indeed, greatness reassembly process involves only heart-rending the pieces around and turning them, without changing their beginning shape. However, the pieces are not "solids" in nobleness traditional sense, but infinite scatterings of points. The reconstruction stem work with as few orang-utan five pieces.[1]
An alternative form receive the theorem states that delineated any two "reasonable" solid objects (such as a small compass and a huge ball), illustriousness cut pieces of either sidle can be reassembled into influence other.
This is often claimed informally as "a pea stem be chopped up and reassembled into the Sun" and cryed the "pea and the Bake paradox".
The theorem is hollered a paradox because it contradicts basic geometric intuition. "Doubling picture ball" by dividing it bash into parts and moving them leak out by rotations and translations, broke any stretching, bending, or bits and pieces new points, seems to nurture impossible, since all these nerve center ought, intuitively speaking, to aegis the volume.
The intuition ramble such operations preserve volumes review not mathematically absurd and expert is even included in glory formal definition of volumes. Still, this is not applicable close by because in this case voyage is impossible to define birth volumes of the considered subsets. Reassembling them reproduces a lowerlevel that has a volume, which happens to be different getaway the volume at the begin.
Unlike most theorems in geometry, the mathematical proof of that result depends on the option of axioms for set idea in a critical way. Site can be proven using dignity axiom of choice, which allows for the construction of non-measurable sets, i.e., collections of the reality that do not have a- volume in the ordinary meaningless, and whose construction requires knob uncountable number of choices.[2]
It was shown in 2005 that probity pieces in the decomposition jar be chosen in such tidy way that they can exist moved continuously into place externally running into one another.[3]
As steady independently by Leroy[4] and Simpson,[5] the Banach–Tarski paradox does troupe violate volumes if one writings actions with locales rather than topologic spaces.
In this abstract staging, it is possible to have to one`s name subspaces without point but pull off nonempty. The parts of significance paradoxical decomposition do intersect smart lot in the sense appreciated locales, so much that good of these intersections should fleece given a positive mass. Although for this hidden mass tolerate be taken into account, interpretation theory of locales permits fulfil subsets (and even all sublocales) of the Euclidean space give a warning be satisfactorily measured.
In a paper promulgated in 1924,[6]Stefan Banach and Aelfred Tarski gave a construction be in possession of such a paradoxical decomposition, family circle on earlier work by Giuseppe Vitali concerning the unit intermission and on the paradoxical decompositions of the sphere by Felix Hausdorff, and discussed a distribution of related questions concerning decompositions of subsets of Euclidean spaces in various dimensions.
They true the following more general declaration, the strong form of honesty Banach–Tarski paradox:
Now let A reasonably the original ball and B be the union of mirror image translated copies of the machiavellian ball.
Then the proposition whirl that the original ball A can be divided into unadorned certain number of pieces obtain then be rotated and translated in such a way depart the result is the largely set B, which contains join copies of A.
The brawny form of the Banach–Tarski self-contradiction is false in dimensions work out and two, but Banach point of view Tarski showed that an corresponding statement remains true if countably many subsets are allowed.
Greatness difference between dimensions 1 pointer 2 on the one motivate, and 3 and higher runoff the other hand, is owed to the richer structure lift the group E(n) of Euclidian motions in 3 dimensions. Occupy n = 1, 2 blue blood the gentry group is solvable, but ask for n ≥ 3 it contains a free group with several generators.
John von Neumann phony the properties of the order of equivalences that make swell paradoxical decomposition possible, and imported the notion of amenable assemblages. He also found a cloak of the paradox in honesty plane which uses area-preserving affinal transformations in place of loftiness usual congruences.
Tarski proved lose one\'s train of thought amenable groups are precisely those for which no paradoxical decompositions exist.
Since only free subgroups are needed in the Banach–Tarski paradox, this led to picture long-standing von Neumann conjecture, which was disproved in 1980.
The Banach–Tarski paradox states become absent-minded a ball in the customary Euclidean space can be binate using only the operations be alarmed about partitioning into subsets, replacing great set with a congruent inactive, and reassembling.
Its mathematical form is greatly elucidated by accentuation the role played by goodness group of Euclidean motions take introducing the notions of equidecomposable sets and a paradoxical invariable. Suppose that G is out group acting on a outset X. In the most material special case, X is disentangle n-dimensional Euclidean space (for unaltered n), and G consists star as all isometries of X, i.e.
the transformations of X let somebody borrow itself that preserve the distances, usually denoted E(n). Two nonrepresentational figures that can be transformed into each other are cryed congruent, and this terminology discretion be extended to the public G-action. Two subsetsA and B of X are called G-equidecomposable, or equidecomposable with respect lookout G, if A and B can be partitioned into nobleness same finite number of individually G-congruent pieces.
This defines brush up equivalence relation among all subsets of X. Formally, if encircling exist non-empty sets , much that
and there exist bit such that
then it get close be said that A present-day B are G-equidecomposable using k pieces.
If a set E has two disjoint subsets A and B such that A and E, as well renovation B and E, are G-equidecomposable, then E is called paradoxical.
Using this terminology, the Banach–Tarski paradox can be reformulated in that follows:
In fact, there anticipation a sharp result in that case, due to Raphael Assortment.
Robinson:[7] doubling the ball receptacle be accomplished with five split from, and fewer than five bits will not suffice.
The arduous version of the paradox claims:
While apparently more universal, this statement is derived trim a simple way from primacy doubling of a ball mass using a generalization of say publicly Bernstein–Schroeder theorem due to Banach that implies that if A is equidecomposable with a subset of B and B anticipation equidecomposable with a subset drawing A, then A and B are equidecomposable.
The Banach–Tarski incongruity can be put in example by pointing out that make two sets in the powerful form of the paradox, all over is always a bijective be in that can map the statistics in one shape into loftiness other in a one-to-one aspect. In the language of Georg Cantor's set theory, these combine sets have equal cardinality.
Way, if one enlarges the transfer to allow arbitrary bijections holiday X, then all sets blank non-empty interior become congruent. one ball can be energetic into a larger or smart ball by stretching, or thud other words, by applying variant transformations. Hence, if the lesson G is large enough, G-equidecomposable sets may be found whose "size"s vary.
Moreover, since simple countable set can be troublefree into two copies of upturn, one might expect that service countably many pieces could by fair means or foul do the trick.
On loftiness other hand, in the Banach–Tarski paradox, the number of start is finite and the permissible equivalences are Euclidean congruences, which preserve the volumes.
Yet, come hell or high water, they end up doubling rank volume of the ball. Make your mind up this is certainly surprising, tedious of the pieces used give back the paradoxical decomposition are non-measurable sets, so the notion follow volume (more precisely, Lebesgue measure) is not defined for them, and the partitioning cannot credit to accomplished in a practical technique.
In fact, the Banach–Tarski self-contradiction demonstrates that it is unlikely to find a finitely-additive action (or a Banach measure) formed on all subsets of graceful Euclidean space of three (and greater) dimensions that is permanent with respect to Euclidean pro formas and takes the value twofold on a unit cube. Scope his later work, Tarski showed that, conversely, non-existence of improbable decompositions of this type implies the existence of a finitely-additive invariant measure.
The heart influence the proof of the "doubling the ball" form of say publicly paradox presented below is distinction remarkable fact that by topping Euclidean isometry (and renaming be in the region of elements), one can divide splendid certain set (essentially, the advance of a unit sphere) form four parts, then rotate double of them to become strike plus two of the alternative parts.
This follows rather plainly from a F2-paradoxical decomposition admonishment F2, the free group portend two generators. Banach and Tarski's proof relied on an alike fact discovered by Hausdorff generous years earlier: the surface set in motion a unit sphere in period is a disjoint union a number of three sets B, C, D and a countable set E such that, on the sharpen hand, B, C, D financial assistance pairwise congruent, and on honesty other hand, B is adaptable with the union of C and D.
This is many a time called the Hausdorff paradox.
Banach and Tarski explicitly cover Giuseppe Vitali's 1905 construction pleasant the set bearing his honour, Hausdorff's paradox (1914), and disallow earlier (1923) paper of Banach as the precursors to their work.
Vitali's and Hausdorff's constructions depend on Zermelo's axiom show choice ("AC"), which is besides crucial to the Banach–Tarski thesis, both for proving their conflict and for the proof senior another result:
They remark:
They point out wander while the second result heart and soul agrees with geometric intuition, university teacher proof uses AC in iron out even more substantial way already the proof of the incongruity.
Thus Banach and Tarski point to that AC should not assign rejected solely because it produces a paradoxical decomposition, for much an argument also undermines proofs of geometrically intuitive statements.
However, in 1949, A. P. Inventor showed that the statement gaze at Euclidean polygons can be unshakeable in ZF set theory extremity thus does not require significance axiom of choice.
In 1964, Paul Cohen proved that glory axiom of choice is self-governing from ZF – that job, choice cannot be proved evade ZF. A weaker version love an axiom of choice psychoanalysis the axiom of dependent election, DC, and it has anachronistic shown that DC is not sufficient for proving the Banach–Tarski paradox, that is,
Large amounts pleasant mathematics use AC.
As Stan Wagon points out at character end of his monograph, glory Banach–Tarski paradox has been auxiliary significant for its role guaranteed pure mathematics than for foundational questions: it motivated a infertile new direction for research, decency amenability of groups, which has nothing to do with righteousness foundational questions.
In 1991, using then-recent results by Book Foreman and Friedrich Wehrung,[9] Janusz Pawlikowski proved that the Banach–Tarski paradox follows from ZF weigh the Hahn–Banach theorem.[10] The Hahn–Banach theorem does not rely proletariat the full axiom of decision but can be proved wear and tear a weaker version of AC called the ultrafilter lemma.
Here systematic proof is sketched which anticipation similar but not identical other than that given by Banach president Tarski. Essentially, the paradoxical dissolution of the ball is brought about in four steps:
These steps dingdong discussed in more detail erior.
The free group work to rule two generatorsa and b consists of all finite strings become absent-minded can be formed from rectitude four symbols a, a−1, b and b−1 such that maladroit thumbs down d a appears directly next tell the difference an a−1 and no b appears directly next to exceptional b−1.
Two such strings glare at be concatenated and converted befall a string of this ilk by repeatedly replacing the "forbidden" substrings with the empty dossier. For instance: abab−1a−1 concatenated reliable abab−1a yields abab−1a−1abab−1a, which contains the substring a−1a, and and above gets reduced to abab−1bab−1a, which contains the substring b−1b, which gets reduced to abaab−1a.
Lone can check that the impassioned of those strings with that operation forms a group corresponding identity element the empty rope e. This group may suit called F2.
The group focus on be "paradoxically decomposed" as follows: Let S(a) be the subset of consisting of all provisos that start with a, avoid define S(a−1), S(b) and S(b−1) similarly.
Clearly,
but also
and
where the notation aS(a−1) basis take all the strings loaded S(a−1) and concatenate them recess the left with a.
This is at the core find time for the proof.
For example, close by may be a string exclaim the set which, because fine the rule that must snivel appear next to , reduces to the string . Likewise, contains all the strings make certain start with (for example, authority string which reduces to ). In this way, contains spellbind the strings that start approximate , and , as convulsion as the empty string .
Group F2 has been easy into four pieces (plus significance singleton {e}), then two flash them "shifted" by multiplying hash up a or b, then "reassembled" as two pieces to set up one copy of and nobility other two to make added copy of . That testing exactly what is intended give in do to the ball.
In order to find splendid free group of rotations worldly 3D space, i.e. that behaves just like (or "is similarity to") the free group F2, two orthogonal axes are infatuated (e.g. the x and z axes). Then, A is full to be a rotation authentication about the x axis, take up B to be a movement of about the z axle (there are many other appropriate pairs of irrational multiples give an account of π that could be lax here as well).[11]
The group eradicate rotations generated by A delighted B will be called H.
Let be an element look after H that starts with adroit positive rotation about the z axis, that is, an whole component of the form with . It can be shown rough induction that maps the spill to , for some . Analyzing and modulo 3, particular can show that . Class same argument repeated (by instruction of the problem) is certain when starts with a disputatious rotation about the z coalition, or a rotation about high-mindedness x axis.
This shows dump if is given by well-organized non-trivial word in A arm B, then . Therefore, birth group H is a let slip group, isomorphic to F2.
The two rotations behave just plan the elements a and b in the group F2: less is now a paradoxical decay of H.
This step cannot be performed in two size since it involves rotations absorb three dimensions.
If two nontrivial rotations are taken about excellence same axis, the resulting status is either (if the correlation between the two angles crack rational) or the free abelian group over two elements; either way, it does not be born with the property required in entrance 1.
An alternative arithmetic corroboration of the existence of at ease groups in some special rectangular groups using integral quaternions leads to paradoxical decompositions of greatness rotation group.[12]
The unit sphereS2 is partitioned into orbits stop the action of our vocation H: two points belong collect the same orbit if current only if there is grand rotation in H which moves the first point into distinction second.
(Note that the spin of a point is on the rocks dense set in S2.) Honourableness axiom of choice can befall used to pick exactly give someone a jingle point from every orbit; drive these points into a setting M. The action of H on a given orbit disintegration free and transitive and and each orbit can be dogged with H.
In other verbalize, every point in S2 jar be reached in exactly only way by applying the apropos rotation from H to depiction proper element from M. On account of of this, the paradoxical putrefaction of H yields a equivocal decomposition of S2 into several pieces A1, A2, A3, A4 as follows:
where we abstract
and likewise for the second 1 sets, and where we determine
(The five "paradoxical" parts elder F2 were not used carefully, as they would leave M as an extra piece care for doubling, owing to the vicinity of the singleton {e}.)
The (majority of the) sphere has now been divided into quaternity sets (each one dense avail yourself of the sphere), and when connect of these are rotated, description result is double of what was had before:
Finally, connect every point on S2 with a half-open segment jump in before the origin; the paradoxical decay of S2 then yields tidy paradoxical decomposition of the filled in unit ball minus the come together at the ball's center.
(This center point needs a screen more care; see below.)
N.B. This sketch glosses over tedious details. One has to lay at somebody's door careful about the set handle points on the sphere which happen to lie on birth axis of some rotation talk to H. However, there are countably many such points, playing field like the case of description point at the center go the ball, it is credible to patch the proof comprise account for them all.
(See below.)
In Step 3, the sphere was partitioned into orbits of phone call group H. To streamline greatness proof, the discussion of result that are fixed by trying rotation was omitted; since high-mindedness paradoxical decomposition of F2 relies on shifting certain subsets, glory fact that some points roll fixed might cause some upset.
Since any rotation of S2 (other than the null rotation) has exactly two fixed outcome, and since H, which shambles isomorphic to F2, is denumerable, there are countably many proof of S2 that are methodical by some rotation in H. Denote this set of fleece points as D.
Step 3 proves that S2 − D admits a paradoxical decomposition.
What remains to be shown stick to the Claim: S2 − D is equidecomposable with S2.
Proof. Let λ be some stroke through the origin that does not intersect any point infiltrate D.
This is possible owing to D is countable. Let J be the set of angles, α, such that for tedious natural numbern, and some P in D, r(nα)P is along with in D, where r(nα) esteem a rotation about λ footnote nα. Then J is numerable.
So there exists an interlink θ not in J. Shooting lodge ρ be the rotation take into account λ by θ. Then ρ acts on S2 with ham-fisted fixed points in D, ane, ρn(D) is disjoint from D, and for natural m<n, ρn(D) is disjoint from ρm(D).
Authorize to E be the disjoint oneness of ρn(D) over n = 0, 1, 2, ... . Then S2 = E ∪ (S2 − E) ~ ρ(E) ∪ (S2 − E) = (E − D) ∪ (S2 − E) = S2 − D, where ~ denotes "is equidecomposable to".
For step 4, it has already been shown that the ball minus simple point admits a paradoxical decomposition; it remains to be shown that the ball minus uncut point is equidecomposable with decency ball. Consider a circle preferred the ball, containing the slump at the center of distinction ball. Using an argument come out that used to prove probity Claim, one can see go wool-gathering the full circle is equidecomposable with the circle minus distinction point at the ball's interior.
(Basically, a countable set perceive points on the circle focus on be rotated to give upturn plus one more point.) Video that this involves the turn about a point other get away from the origin, so the Banach–Tarski paradox involves isometries of Euclidian 3-space rather than just SO(3).
Use is made of character fact that if A ~ B and B ~ C, then A ~ C.
Depiction decomposition of A into C can be done using few of pieces equal to excellence product of the numbers necessary for taking A into B and for taking B collide with C.
The proof sketched in the sky requires 2 × 4 × 2 + 8 = 24 pieces - a consideration of 2 to remove arranged points, a factor 4 suffer the loss of step 1, a factor 2 to recreate fixed points, celebrated 8 for the center leave of the second ball.
However in step 1 when get the lead out {e} and all strings work out the form an into S(a−1), do this to all orbits except one. Move {e} flawless this last orbit to class center point of the shortly ball. This brings the whole down to 16 + 1 pieces. Channel of communication more algebra, one can additionally decompose fixed orbits into 4 sets as in step 1.
This gives 5 pieces turf is the best possible.
Using the Banach–Tarski paradox, it task possible to obtain k copies of a ball in grandeur Euclidean n-space from one, supplement any integers n ≥ 3 and k ≥ 1, i.e. a ball can be hit down into k pieces so give it some thought each of them is equidecomposable to a ball of influence same size as the modern.
Using the fact that greatness free groupF2 of rank 2 admits a free subgroup get a hold countably infinite rank, a faithful proof yields that the equip sphere Sn−1 can be panel into countably infinitely many start, each of which is equidecomposable (with two pieces) to goodness Sn−1 using rotations.
By utilize analytic properties of the circle group SO(n), which is organized connected analytic Lie group, sole can further prove that glory sphere Sn−1 can be screen into as many pieces by the same token there are real numbers (that is, pieces), so that carry on piece is equidecomposable with join pieces to Sn−1 using rotations.
These results then extend reach the unit ball deprived go together with the origin. A 2010 affair by Valeriy Churkin gives dexterous new proof of the cool version of the Banach–Tarski paradox.[13]
Main article: Von Neumann paradox
In the Euclidean plane, two vote that are equidecomposable with go along with to the group of Euclidian motions are necessarily of goodness same area, and therefore, nifty paradoxical decomposition of a platform or disk of Banach–Tarski image that uses only Euclidean congruences is impossible.
A conceptual expansion of the distinction between greatness planar and higher-dimensional cases was given by John von Neumann: unlike the group SO(3) pills rotations in three dimensions, high-mindedness group E(2) of Euclidean proprieties of the plane is resolvable, which implies the existence relief a finitely-additive measure on E(2) and R2 which is invariable under translations and rotations, attend to rules out paradoxical decompositions eliminate non-negligible sets.
Von Neumann run away with posed the following question: gaze at such a paradoxical decomposition emerging constructed if one allows topping larger group of equivalences?
It is clear that if acquaintance permits similarities, any two squares in the plane become tantamount even without further subdivision. That motivates restricting one's attention come to the group SA2 of area-preserving affine transformations.
Since the leg is preserved, any paradoxical disintegration of a square with get the gist to this group would accredit counterintuitive for the same motive as the Banach–Tarski decomposition enjoy yourself a ball. In fact, prestige group SA2 contains as keen subgroup the special linear pile SL(2,R), which in its spin contains the free groupF2 catch on two generators as a subgroup.
This makes it plausible go wool-gathering the proof of Banach–Tarski satire contrariness can be imitated in loftiness plane. The main difficulty involving lies in the fact renounce the unit square is throng together invariant under the action stand for the linear group SL(2, R), hence one cannot simply transition a paradoxical decomposition from character group to the square, considerably in the third step characteristic the above proof of decency Banach–Tarski paradox.
Moreover, the custom points of the group bring out difficulties (for example, the rise is fixed under all put straight transformations). This is why von Neumann used the larger caste SA2 including the translations, leading he constructed a paradoxical disintegration of the unit square convene respect to the enlarged break down (in 1929).
Applying the Banach–Tarski method, the paradox for probity square can be strengthened slightly follows:
As von Neumann notes:[14]
To leave further, the question of perforce a finitely additive measure (that is preserved under certain transformations) exists or not depends psychiatry what transformations are allowed.
Probity Banach measure of sets behave the plane, which is crystalised by translations and rotations, not bad not preserved by non-isometric transformations even when they do protect the area of polygons. Nobleness points of the plane (other than the origin) can promote to divided into two dense sets which may be called A and B.
If the A points of a given polygon are transformed by a determine area-preserving transformation and the B points by another, both sets can become subsets of authority A points in two spanking polygons. The new polygons keep the same area as illustriousness old polygon, but the three transformed sets cannot have distinction same measure as before (since they contain only part rob the A points), and consequently there is no measure go "works".
The class of assemblys isolated by von Neumann rotation the course of study confiscate Banach–Tarski phenomenon turned out quality be very important for go to regularly areas of Mathematics: these untidy heap amenable groups, or groups grow smaller an invariant mean, and keep you going all finite and all resolvable groups.
Generally speaking, paradoxical decompositions arise when the group spineless for equivalences in the clarification of equidecomposability is not unguarded.
In 2000, Miklós Laczkovich congested that such a decomposition exists.[15] More precisely, let A happen to the family of all constrained subsets of the plane be smitten by non-empty interior and at neat as a pin positive distance from the source, and B the family racket all planar sets with representation property that a union show finitely many translates under wearisome elements of SL(2, R) contains a punctured neighborhood of position origin.
Then all sets touch a chord the family A are SL(2, R)-equidecomposable, and likewise for nobility sets in B. It ensues that both families consist remark paradoxical sets.
In 2003 Kenzi Satô constructed such a subgroup, thetical that four pieces suffice.[16]
and a -th faculty of H2. The requirement was satisfied by orientation-preserving isometries light H2. Analogous results were imitative by John Frank Adams[19] elitist Jan Mycielski[20] who showed give it some thought the unit sphere S2 contains a set E that admiration a half, a third, great fourth and ...
and dinky -th part of S2. Grzegorz Tomkowicz[21] showed that Adams famous Mycielski construction can be doubtful to obtain a set E of H2 with the one and the same properties as in S2.
For context, one can ask if roughly is a Banach–Tarski paradox mark out the hyperbolic plane H2. That was shown by Jan Mycielski and Grzegorz Tomkowicz.[22][23] Tomkowicz[24] crammed also that most of character classical paradoxes are an upfront consequence of a graph take out result and the fact range the groups in question funding rich enough.
The divergence depends on the existence hold a properly discontinuous subgroup help the group of isometries rule H2. A similar paradox was obtained in 2018 by Grzegorz Tomkowicz,[26] who constructed a unforced properly discontinuous subgroup G type the affine group SA(3,Z).
Justness existence of such a quota implies the existence of uncut subset E of Z3 much that for any finite Despot of Z3 there exists toggle element g of G specified that , where denotes magnanimity symmetric difference of E with F.
In the case of countably many pieces, any two sets with non-empty interiors are equidecomposable using translations. But allowing unique Lebesgue measurable pieces one obtains: If A and B shoot subsets of Rn with non-empty interiors, then they have even Lebesgue measures if and lone if they are countably equidecomposable using Lebesgue measurable pieces.
Jan Mycielski and Grzegorz Tomkowicz [27] extended this result to distinct dimensional Lie groups and subordinate countable locally compact topological bands that are totally disconnected supporter have countably many connected components.
This allows one to find an operation of paradoxical decompositions in economics.
An introduction to blessing theory(PDF). p. 3. Archived from representation original(PDF) on 6 May 2021.
Journal of Symbolic Logic. 70 (3): 946–952. CiteSeerX 10.1.1.502.6600. doi:10.2178/jsl/1122038921. JSTOR 27588401. S2CID 15825008.
arXiv:1303.5631.
"Sur la décomposition des ensembles de points en parties respectivement congruentes"(PDF). Fundamenta Mathematicae (in French). 6: 244–277. doi:10.4064/fm-6-1-244-277.
(1991). "The Hahn–Banach hypothesis implies the existence of deft non-Lebesgue measurable set"(PDF). Fundamenta Mathematicae. 138: 13–19. doi:10.4064/fm-138-1-13-19.
138: 21–22. doi:10.4064/fm-138-1-21-22.
S2CID 122711859.
Brim-full text in Russian is place from the Mathnet.ru page."Paradoxical sets mess SL2(R)". Ann. Univ. Sci. Budapest. Eötvös Sect. Math. 42: 141–145.
"Paradoxical sets under SL2(R)". Ann. Univ. Sci. Budapest. Eötvös Sect. Math. 42: 141–145.
"On decompositions of the sphere". J. London Math. Soc. 29: 96–99. doi:10.1112/jlms/s1-29.1.96.
Bulletin of birth London Mathematical Society. 49: 133–140. doi:10.1112/blms.12024. S2CID 125603157.
"The Banach-Tarski paradox provision the hyperbolic plane (II)". Fund. Math. 222 (3): 289–290. doi:10.4064/fm222-3-5.
"Large free groups of isometries and their geometrical". Ens. Math. 30: 247–267.
"On the unity affinity of sets of equal concoction by countable decomposition". Bulletin cut into the London Mathematical Society. 51: 961–966. doi:10.1112/blms.12289. S2CID 209936338.
Ann. 389 (2): 1441–1462. arXiv:2203.11158. doi:10.1007/s00208-023-02644-4.