A short proof of nonhomogeneity of the pseudo-circle

The pseudo-circle is known to be nonhomogeneous. The original proofs of this fact were discovered independently by L. Fearnley and J.T. Rogers, Jr. The purpose of this paper is to provide an alternative, very short proof based on a result of D. Bella…

Authors: Krystyna Kuperberg (Auburn University), Kevin Gammon (Auburn University)

A SHOR T PR OOF OF NONHOMO GENEITY OF THE PSEUDO-CIR CLE KR YSTYNA KUPERBERG AND KEVIN GAMMON De dic ate d to James T. R o gers, Jr. on the o c c asion of his 65th birthday Abstra ct. The p seudo-circle is known to be nonhomogeneous. The original proofs of th is fact w ere disco vered ind ep en dently b y L. F earnley [6] and J.T. Rogers, J r. [17]. The purp ose of this paper is to provide an alternative, very short proof based on a result of D. Bellamy and W. Lewis [4]. 1. Introduction A pseudo-ar c is a hereditarily indecomp osable, c hainable con tinuum. In 194 8, E.E. Moise [16] constructed a ps eu do-arc as an indecomp osable con tin uum homeomorphic to eac h of its sub- con tinua. Moise correctly conjectured that th e hereditarily indecomp osable con tinuum giv en b y B. Knaster [11] in 1922 is a pseudo-arc. Also in 1948, R.H. Bing [1] prov ed th at Moise’s exam- ple is homogeneous. In 1951, Bing [2] p ro v ed that ev ery hereditarily indecomp osable chainable con tinuum is a pseudo-arc and that all pseudo-arcs are homeomorphic. In 1959, Bing [3] ga ve another charact erization of the p seudo-arc: a homogeneous c hainable contin uum . The history of many other asp ects of the pseudo-arc can b e found in su rv ey p ap ers b y W. Lewis [14] and [15 ]. In 195 1, Bing [2] describ ed a p seudo-cir cle , a planar hereditarily indecomp osable circularly c h ainable con tinuum whic h separates the plane and wh ose ev ery prop er su b cont inuum is a pseud o- arc. It has b een sho wn b y L. F earnley in [6] and J. T. Rogers, Jr. in [17] that the pseudo-circle is not h omogeneous. F earnley also pr o ved that the pseudo-circle is uniqu e [5 ] and [7]. T he fact that the pseud o-circle is not homogeneous also follo ws from more general theorems prov ed in [8], [10], [13], and [18]. This p ap er offers y et another, v er y sh ort p ro of, a consequence of a r esult of D. Bellam y and W. Lewis [4]. Similarly as in [18], an infinite co ve ring sp ace of a plane s eparating cont inuum is used. 2. Preliminaries Throughout the pap er, a c ontinuum will r efer to a nondegenerate compact and connected m et- ric space. A con tinuum is i nde c omp osable if it is n ot th e union of tw o prop er su b con tinua. A con tinuum is her e ditarily inde c omp osable if ev ery sub con tin uum is also in d ecomp osable. F or a p oint a in X , the c omp osant K ( a ) of a in X is the union of all prop er sub cont inua of X cont ain- ing a . An indecomp osable contin uum con tains uncounta bly many pairwise disjoint comp osan ts, see [12] Theorem 7, page 21 2. 2000 Mathematics Subje ct Classific ation. 54F15; 54F50. Key wor ds and phr ases. p seudo-circle, pseudo-arc, homogeneous, comp osan t, indecomp osable contin uum. 1 2 KR YSTYNA KU PERBERG AND K EVIN GAMMON A top ologica l space X is homo gene ous if for any tw o points in X there is a homeomorphism of X on to itself mapp ing one p oin t on to the other. Let C denote the pseudo-circle. W e may assume that C is co nta ined in a planar annulus A in suc h a wa y that the winding n umber of eac h circular chain in the sequ en ce of cro oke d circular c h ains defining C is one. Any h omeomorphism h : C → C extends to a con tin uous map f : A → A of d egree ± 1. (First extend h to a map U → A for some closed annular neighborh o o d U of C , then comp ose a retraction of A on to U w ith this extension.) Let e A b e the un iv ersal co vering sp ace of A with pr o jection p . F or an y e x ∈ e A and e y ∈ p − 1 ( f ( p ( e x ))) there is a map e f suc h that the diagram e f e A − → e A p ↓ ↓ p A − → A f comm u tes and e f ( e x ) = e y ; see for example [9], Theorem 16.3. Let b A b e the disc that is a tw o-p oint compactificatio n of e A . Denote the t w o added p oin ts of the compactificatio n by a and b . The map e f extends un iquely to a m ap F : b A → b A . Let e C = p − 1 ( C ), and let P = e C ∪ { a, b } , a t wo -p oint compactificat ion of p − 1 ( C ). D. Bellam y and W. Lewis considered this set in [4] and prov ed that P is a pseudo-arc. Denote by H th e restriction of F to P and n ote that (1) either H ( a ) = a and H ( b ) = b , or H ( a ) = b and H ( b ) = a , (2) e f ( e C ) = e C and hence H ( P ) = P , (3) e f | e C is one-to-one. Th us Lemma. H is a home om orphism fr om P to P . 3. Proof of no nhomogeneity of the pseudo-circle Theorem 1. The pseudo-cir cle is not homo gene ous. Pr o of. Let K ( a ) and K ( b ) b e the comp osan ts of a and b , resp ectiv ely , in the pseudo-arc P . Let e x and e y b e tw o p oints in P su c h that e x ∈ ( K ( a ) ∪ K ( b )) − { a, b } and e y ∈ P − ( K ( a ) ∪ K ( b )). If C w ere homogeneous, then there would b e a homeomorph ism h : C → C taking x = p ( e x ) on to y = p ( e y ). Th en there w ould b e a homeomorphism H : P → P as d escrib ed in section 2 taking e x on to e y . This is not p ossible since under eve ry such h omeomorphism, the s et K ( a ) ∪ K ( b ) is in v arian t; the image of a comp osan t is a comp osant.  Remark. It is not imp ortant for th is pro of that K ( a ) and K ( b ) are n ot the same set, but th e authors are grateful to D. Bellam y and W. Lewis for sho wing that K ( a ) and K ( b ) were indeed differen t co mp osan ts. Theorem 2. If for some x , the c omp osant K ( a ) interse cts the fib er p − 1 ( x ) , then it c ontains p − 1 ( x ) . A S HOR T PROOF OF NONHOMOGENEITY OF THE PSEUDO-CIRCLE 3 Pr o of. If y ∈ p − 1 ( x ) ∩ K ( a ), then by the definition of a comp osan t, there is a prop er sub con tin uum W of P that con tains b oth a and y . Let g : e C → e C b e a deck transf ormation suc h that p − 1 ( x ) = { g n ( y ) } n ∈ Z , Z b eing the set of intege rs. Denote b y G the exte nsion of g to P . The set W n = G n ( W ) is a con tinuum conta ining a and g n ( y ). Thus p − 1 ( x ) ⊂ K ( a ).  Note that T h eorem 2 and Remark ab o v e imply that p ( K ( a ) − { a } ) ∩ p ( K ( b ) − { b } ) = ∅ . Question. Ca n the sets p ( K ( a ) − { a } ) and p ( K ( b ) − { b } ) b e use d to classify the c omp osants of the pseudo-cir cle C ? The authors w ould lik e to thank Jim Rogers, Da vid Bellam y , and W a yn e Lewis for th eir com- men ts. Referen ces [1] R.H. Bing, A homo gene ous inde c omp osable plane c ontinuum , Duke Math. J. 15 (1948), 729-742. [2] R.H. Bing, Conc erning her e di tarily inde c omp osable c ontinua , P acific J. Math. 1 (1951), 43-51. [3] R.H. Bing, Each homo gene ous nonde gener ate chainable c ontinuum i s a pseudo-ar c , Proc. Amer. Math. S oc. 10 (1959), 345-346 [4] D.P . Bellam y and W. Lewis, An orientation r eversing home omorphism of the plane wi th i nvariant pseudo-ar c , Proc. A mer. Math. So c. 114 (1992), 1145-1149. [5] L. F earnley , The pseudo-cir cle i s unique , Bull. A mer. Math. So c. 75 (1969), 398-401. [6] L. F earnley , The pseudo-cir cle i s not homo gene ous , Bull. Amer. Math. So c. 75 (1969), 554-558. [7] L. F earnley , The pseudo-cir cle i s unique , T rans. Amer. Math. Soc. 149 (1970), 45-64. [8] C.L. Hagopian, The fixe d-p oint pr op erty for almost chainable homo gene ous c ontinua , Illinois J. Math. 20 (1976), 650-652. [9] S.T. Hu, H omotop y Theory , Elsevier Science and T ec hnology Bo oks, 1959. [10] J. K ennedy , and J.T. Rogers, Jr., Orbits of the pseudo cir cle , T rans. A mer. Math. So c. 296 (1986), 327-340. [11] B. K naster, Un c ontinu dont tout so´ us-c ontinu est ind´ ec omp osable , F und. Math. 3 (1922), 247-286. [12] K. Kuratowski, T opology , V ol. I I, Academic Press, 1968. [13] W. Lewis, Almost chainable homo gene ous c ontinua ar e chainable , Houston J. Math. 7 ( 1981), 373-377. [14] W. Lewis, T he pseudo-ar c , Con temp. Math. 117 (1991), 103-123. [15] W. Lewis, T he pseudo-ar c , Bol. So c. Mat. Mexicana 5 (1999), 25-77. [16] E.E. Moise, An i nde c omp osable plane c ontinuum which is home omorphic to e ach of its nonde gener ate sub c on- tinua , T rans. Amer. Math. S oc. 63 (1948), 581-594. [17] J.T. Rogers, Jr., The pseudo-cir cle is not homo gene ous , T rans. Amer. Math. Soc. 148 (1970), 417-42 8. [18] J.T. Rogers, Jr., Homo gene ous, sep ar ating pl ane c ontinua ar e de c omp osable , Michigan Math. J. 28 (1981), 317-322. Dep ar tmen t of Ma thema tics and St a ti stics, Auburn University, Auburn, AL 36849, USA E-mail addr ess : kuperkm@auburn .edu Dep ar tmen t of Ma thema tics and St a ti stics, Auburn University, Auburn, AL 36849, USA E-mail addr ess : gammokb@auburn .edu

Original Paper

Loading high-quality paper...

Comments & Academic Discussion

Loading comments...

Leave a Comment