The Mizar article:

Upper and Lower Sequence of a Cage

by
Robert Milewski

Received August 8, 2001

Copyright (c) 2001 Association of Mizar Users

MML identifier: JORDAN1E
[ MML identifier index ]


environ

 vocabulary EUCLID, COMPTS_1, PSCOMP_1, ORDINAL2, FUNCT_5, REALSET1, FINSEQ_1,
      SPRECT_2, RELAT_1, FINSEQ_5, FINSEQ_4, ARYTM_1, PRE_TOPC, TOPREAL1,
      JORDAN3, BOOLE, GROUP_2, FUNCT_1, SPPOL_1, FINSEQ_6, JORDAN9, GRAPH_2,
      TOPREAL2, CARD_1, RELAT_2, MCART_1, GOBOARD1, JORDAN8, MATRIX_1, TREES_1,
      JORDAN1E;
 notation TARSKI, XBOOLE_0, ENUMSET1, XCMPLX_0, XREAL_0, NAT_1, FUNCT_1,
      FINSEQ_1, FINSEQ_4, FINSEQ_5, MATRIX_1, FINSEQ_6, STRUCT_0, GRAPH_2,
      PRE_TOPC, TOPREAL2, CARD_1, BINARITH, CONNSP_1, COMPTS_1, EUCLID,
      PSCOMP_1, SPRECT_2, GOBOARD1, TOPREAL1, SPPOL_1, JORDAN3, JORDAN8,
      JORDAN9;
 constructors JORDAN8, REALSET1, REAL_1, CARD_4, PSCOMP_1, BINARITH, CONNSP_1,
      JORDAN9, FINSEQ_4, GOBRD13, JORDAN3, TOPREAL2, FINSEQ_5, ENUMSET1,
      FINSOP_1, GRAPH_2, PARTFUN1;
 clusters XREAL_0, SPRECT_1, TOPREAL6, JORDAN8, REVROT_1, RELSET_1, ARYTM_3,
      PSCOMP_1, SPRECT_3, FINSEQ_5, SPPOL_2, JORDAN1A, GOBOARD4, JORDAN1B,
      MEMBERED;
 requirements NUMERALS, BOOLE, SUBSET, REAL, ARITHM;
 definitions TARSKI, TOPREAL1, SPRECT_2, XBOOLE_0;
 theorems AXIOMS, BINARITH, JORDAN8, PSCOMP_1, JORDAN1A, NAT_1, REAL_1,
      GOBOARD5, REAL_2, FINSEQ_1, FINSEQ_2, SPRECT_2, FINSEQ_4, FINSEQ_5,
      FINSEQ_6, GOBOARD7, SPPOL_2, REVROT_1, TOPREAL1, SPRECT_3, JORDAN5B,
      AMI_5, JORDAN3, JORDAN9, GOBOARD1, TARSKI, FINSEQ_3, FUNCT_1, TOPREAL5,
      JORDAN10, GRAPH_2, CARD_1, CARD_2, YELLOW_6, PRE_CIRC, ENUMSET1,
      JORDAN1B, SPRECT_5, SCMFSA_7, XBOOLE_0, XBOOLE_1, XCMPLX_0, XCMPLX_1;

begin  :: Preliminaries

  reserve n for Nat;

  theorem
     for X be non empty Subset of TOP-REAL 2
   for Y be compact Subset of TOP-REAL 2 st X c= Y holds
    N-bound X <= N-bound Y
   proof
    let X be non empty Subset of TOP-REAL 2;
    let Y be compact Subset of TOP-REAL 2;
    assume X c= Y;
    then sup (proj2||X) <= sup (proj2||Y) by PSCOMP_1:63;
    then sup (proj2||X) <= N-bound Y by PSCOMP_1:def 31;
    hence N-bound X <= N-bound Y by PSCOMP_1:def 31;
   end;

  theorem
     for X be non empty Subset of TOP-REAL 2
   for Y be compact Subset of TOP-REAL 2 st X c= Y holds
    E-bound X <= E-bound Y
   proof
    let X be non empty Subset of TOP-REAL 2;
    let Y be compact Subset of TOP-REAL 2;
    assume X c= Y;
    then sup (proj1||X) <= sup (proj1||Y) by PSCOMP_1:63;
    then sup (proj1||X) <= E-bound Y by PSCOMP_1:def 32;
    hence E-bound X <= E-bound Y by PSCOMP_1:def 32;
   end;

  theorem
     for X be non empty Subset of TOP-REAL 2
   for Y be compact Subset of TOP-REAL 2 st X c= Y holds
    S-bound X >= S-bound Y
   proof
    let X be non empty Subset of TOP-REAL 2;
    let Y be compact Subset of TOP-REAL 2;
    assume X c= Y;
    then inf (proj2||X) >= inf (proj2||Y) by PSCOMP_1:62;
    then inf (proj2||X) >= S-bound Y by PSCOMP_1:def 33;
    hence S-bound X >= S-bound Y by PSCOMP_1:def 33;
   end;

  theorem
     for X be non empty Subset of TOP-REAL 2
   for Y be compact Subset of TOP-REAL 2 st X c= Y holds
    W-bound X >= W-bound Y
   proof
    let X be non empty Subset of TOP-REAL 2;
    let Y be compact Subset of TOP-REAL 2;
    assume X c= Y;
    then inf (proj1||X) >= inf (proj1||Y) by PSCOMP_1:62;
    then inf (proj1||X) >= W-bound Y by PSCOMP_1:def 30;
    hence W-bound X >= W-bound Y by PSCOMP_1:def 30;
   end;

  theorem Th5:
   for f,g be FinSequence of TOP-REAL 2 st f is_in_the_area_of g
   for p be Element of TOP-REAL 2 st p in rng f holds
    f-:p is_in_the_area_of g
   proof
    let f,g be FinSequence of TOP-REAL 2;
    assume A1: f is_in_the_area_of g;
    let p be Element of TOP-REAL 2;
    assume A2: p in rng f;
    let n be Nat;
    assume A3: n in dom(f-:p);
    A4: p..f <= len f by A2,FINSEQ_4:31;
    A5: len(f-:p) = p..f by A2,FINSEQ_5:45;
    then n in Seg(p..f) by A3,FINSEQ_1:def 3;
    then A6: (f-:p)/.n = f/.n by A2,FINSEQ_5:46;
      n in Seg len(f-:p) by A3,FINSEQ_1:def 3;
    then 1 <= n & n <= p..f by A5,FINSEQ_1:3;
    then 1 <= n & n <= len f by A4,AXIOMS:22;
    then n in dom f by FINSEQ_3:27;
    hence thesis by A1,A6,SPRECT_2:def 1;
   end;

  theorem Th6:
   for f,g be FinSequence of TOP-REAL 2 st f is_in_the_area_of g
   for p be Element of TOP-REAL 2 st p in rng f holds
    f:-p is_in_the_area_of g
   proof
    let f,g be FinSequence of TOP-REAL 2;
    assume A1: f is_in_the_area_of g;
    let p be Element of TOP-REAL 2;
    assume A2: p in rng f;
    let n be Nat;
    assume A3: n in dom(f:-p);
    A4: len(f:-p) = len f - p..f + 1 by A2,FINSEQ_5:53;
      n in Seg len(f:-p) by A3,FINSEQ_1:def 3;
    then A5: 0+1 <= n & n <= len f - p..f + 1 by A4,FINSEQ_1:3;
    A6: n-'1 >= 0 by NAT_1:18;
      1 <= p..f by A2,FINSEQ_4:31;
    then A7: 0+1 <= n-'1+p..f by A6,REAL_1:55;
    A8: n-1 >= 0 by A5,REAL_1:84;
      n - 1 <= len f - p..f by A5,REAL_1:86;
    then n - 1 + p..f <= len f by REAL_1:84;
    then n-'1+p..f <= len f by A8,BINARITH:def 3;
    then A9: n-'1+p..f in dom f by A7,FINSEQ_3:27;
      n-'1+1 = n by A5,AMI_5:4;
    then (f:-p)/.n = f/.(n-'1+p..f) by A2,A3,FINSEQ_5:55;
    hence thesis by A1,A9,SPRECT_2:def 1;
   end;

  theorem
     for f be non empty FinSequence of TOP-REAL 2
   for p be Point of TOP-REAL 2 st p in L~f holds
    L_Cut (f,p) <> {}
   proof
    let f be non empty FinSequence of TOP-REAL 2;
    let p be Point of TOP-REAL 2;
    assume that
     A1: p in L~f and
     A2: L_Cut (f,p) = {};
      <*p*>^mid(f,Index(p,f)+1,len f) is non empty &
    (p<>f.(Index(p,f)+1) or p=f.(Index(p,f)+1));
    then A3: L_Cut(f,p) = mid(f,Index(p,f)+1,len f) & p=f.(Index(p,f)+1)
                                                         by A2,JORDAN3:def 4;
      len f in dom f by SCMFSA_7:12;
    then not Index(p,f)+1 in dom f by A2,A3,SPRECT_2:11;
    then not Index(p,f)+1 in Seg len f by FINSEQ_1:def 3;
    then Index(p,f)+1 < 1 or Index(p,f)+1 > len f by FINSEQ_1:3;
    then Index(p,f) >= len f by NAT_1:29,38;
    hence contradiction by A1,JORDAN3:41;
   end;

  theorem
     for f be non empty FinSequence of TOP-REAL 2
   for p be Point of TOP-REAL 2 st p in L~f & len R_Cut(f,p) >= 2 holds
    f.1 in L~R_Cut(f,p)
   proof
    let f be non empty FinSequence of TOP-REAL 2;
    let p be Point of TOP-REAL 2;
      len f <> 0 by FINSEQ_1:25;
    then len f > 0 by NAT_1:19;
    then A1: 0+1 <= len f by NAT_1:38;
    assume p in L~f;
    then A2: R_Cut(f,p).1 = f.1 by A1,JORDAN1B:4;
    assume 2 <= len R_Cut(f,p);
    hence f.1 in L~R_Cut(f,p) by A2,JORDAN3:34;
   end;

  theorem Th9:
   for f be non empty FinSequence of TOP-REAL 2 st f is_S-Seq
   for p be Point of TOP-REAL 2
    st p in L~f holds not f.1 in L~mid(f,Index(p,f)+1,len f)
   proof
    let f be non empty FinSequence of TOP-REAL 2 such that
     A1: f is_S-Seq;
    let p be Point of TOP-REAL 2 such that
     A2: p in L~f and
     A3: f.1 in L~mid(f,Index(p,f)+1,len f);
    A4: 1 <= Index(p,f)+1 by NAT_1:29;
      Index(p,f) < len f by A2,JORDAN3:41;
    then A5: Index(p,f)+1 <= len f by NAT_1:38;
      1 in dom f by FINSEQ_5:6;
    then f/.1 in L~mid(f,Index(p,f)+1,len f) by A3,FINSEQ_4:def 4;
    then Index(p,f)+1 = 0+1 by A1,A4,A5,JORDAN5B:16;
    then Index(p,f) = 0 by XCMPLX_1:2;
    hence contradiction by A2,JORDAN3:41;
   end;

  theorem Th10:
   for i,j,m,n be Nat holds
    i+j = m+n & i <= m & j <= n implies i = m
   proof
    let i,j,m,n be Nat;
    assume that
     A1: i+j = m+n and
     A2: i <= m and
     A3: j <= n;
    consider k be Nat such that
     A4: m = i + k by A2,NAT_1:28;
    consider l be Nat such that
     A5: n = j + l by A3,NAT_1:28;
      i+(k+n) = i+j by A1,A4,XCMPLX_1:1;
    then k+n = j by XCMPLX_1:2;
    then j + (l+k) = j+0 by A5,XCMPLX_1:1;
    then l+k = 0 by XCMPLX_1:2;
    then k = 0 by NAT_1:23;
    hence i = m by A4;
   end;

  theorem
     for f be non empty FinSequence of TOP-REAL 2 st f is_S-Seq
   for p be Point of TOP-REAL 2 st p in L~f & f.1 in L~L_Cut(f,p) holds
    f.1 = p
   proof
    let f be non empty FinSequence of TOP-REAL 2; assume
     A1: f is_S-Seq;
then A2:f is one-to-one unfolded s.n.c. special & len f >= 2 by TOPREAL1:def 10
;
    let p be Point of TOP-REAL 2 such that
     A3: p in L~f and
     A4: f.1 in L~L_Cut(f,p) and
     A5: f.1 <> p;
    set g = mid(f,Index(p,f)+1,len f);
    A6: len f in dom f by FINSEQ_5:6;
      1 in dom f by FINSEQ_5:6;
    then A7: f/.1 = f.1 by FINSEQ_4:def 4;
    A8: not f.1 in L~g by A1,A3,Th9;
    then p<>f.(Index(p,f)+1) by A4,JORDAN3:def 4;
    then A9: L_Cut(f,p) = <*p*>^g by JORDAN3:def 4;
    per cases;
     suppose g is empty;
      then L_Cut(f,p) = <*p*> by A9,FINSEQ_1:47;
      then len L_Cut(f,p) = 1 by FINSEQ_1:56;
      hence contradiction by A4,TOPREAL1:28;
     suppose g is non empty;
      then L~L_Cut(f,p) = LSeg(p,g/.1) \/ L~g by A9,SPPOL_2:20;
      then A10: f.1 in LSeg(p,g/.1) by A4,A8,XBOOLE_0:def 2;
      consider i being Nat such that
       A11: 1 <= i & i+1 <= len(<*p*>^g) and
       A12: f/.1 in LSeg(<*p*>^g,i) by A4,A7,A9,SPPOL_2:13;
      A13: 1 <= Index(p,f) by A3,JORDAN3:41;
        LSeg(<*p*>^g,i) c= LSeg(f,Index(p,f)+i-'1) by A3,A11,JORDAN3:49;
      then A14: Index(p,f)+i-'1 = 1 by A2,A12,JORDAN5B:33;
        1+1 <= Index(p,f)+i by A11,A13,REAL_1:55;
      then 1 <= Index(p,f)+i by AXIOMS:22;
      then A15: Index(p,f)+i = 1+1 by A14,AMI_5:4;
      then Index(p,f) = 1 by A11,A13,Th10;
      then A16: p in LSeg(f,1) by A3,JORDAN3:42;
      A17: i = 1 by A11,A13,A15,Th10;
      A18: 1+1 <= len f by A1,TOPREAL1:def 10;
      then A19: p in LSeg(f/.1,f/.(1+1)) by A16,TOPREAL1:def 5;
        2 in dom f by A18,FINSEQ_3:27;
      then g/.1 = f/.(1+1) by A6,A15,A17,SPRECT_2:12;
      hence contradiction by A5,A7,A10,A19,SPRECT_3:16;
   end;

begin  :: About Upper and Lower Sequence of a Cage

  definition
   let C be compact non vertical non horizontal Subset of TOP-REAL 2;
   let n be Nat;
   func Upper_Seq(C,n) -> FinSequence of TOP-REAL 2 equals :Def1:
    Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n);
   coherence;
  end;

  theorem Th12:
   for C be compact non vertical non horizontal Subset of TOP-REAL 2
   for n be Nat holds len Upper_Seq(C,n) =
    (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))
   proof
    let C be compact non vertical non horizontal non empty
        Subset of TOP-REAL 2;
    let n be Nat;
    A1: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A2: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A1,FINSEQ_6:96;
    thus len Upper_Seq(C,n) =
          len (Rotate(Cage(C,n),W-min L~Cage(C,n))-:(E-max L~Cage(C,n)))
                                                                  by Def1
       .= (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A2,FINSEQ_5:45;
   end;

  definition
   let C be compact non vertical non horizontal Subset of TOP-REAL 2;
   let n be Nat;
   func Lower_Seq(C,n) -> FinSequence of TOP-REAL 2 equals :Def2:
    Rotate(Cage(C,n),W-min L~Cage(C,n)):-E-max L~Cage(C,n);
   coherence;
  end;

  theorem Th13:
   for C be compact non vertical non horizontal Subset of TOP-REAL 2
   for n be Nat holds len Lower_Seq(C,n) =
    len Rotate(Cage(C,n),W-min L~Cage(C,n))-
    (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+1
   proof
    let C be compact non vertical non horizontal Subset of TOP-REAL 2;
    let n be Nat;
    A1: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A2: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A1,FINSEQ_6:96;
    thus len Lower_Seq(C,n) =
          len (Rotate(Cage(C,n),W-min L~Cage(C,n)):-(E-max L~Cage(C,n)))
                                                                 by Def2
       .= len Rotate(Cage(C,n),W-min L~Cage(C,n))-
          (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+1
                                                           by A2,FINSEQ_5:53;
   end;

  definition
   let C be compact non vertical non horizontal Subset of TOP-REAL 2;
   let n be Nat;
   cluster Upper_Seq(C,n) -> non empty;
   coherence
   proof
    A1: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A2: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A1,FINSEQ_6:96;
      len Upper_Seq(C,n) = (E-max L~Cage(C,n))..Rotate(Cage(C,n),
                          W-min L~Cage(C,n)) by Th12;
    then len Upper_Seq(C,n) >= 0+1 by A2,FINSEQ_4:31;
    then len Upper_Seq(C,n) > 0 by NAT_1:38;
    hence thesis by FINSEQ_1:25;
   end;
   cluster Lower_Seq(C,n) -> non empty;
   coherence
   proof
    A3: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A4: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A3,FINSEQ_6:96;
      len Lower_Seq(C,n) = len Rotate(Cage(C,n),W-min L~Cage(C,n))-
     (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+1 by Th13;
    then len Lower_Seq(C,n)-1 = len Rotate(Cage(C,n),W-min L~Cage(C,n))-
      (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)) by XCMPLX_1:26
       .= len Rotate(Cage(C,n),W-min L~Cage(C,n))+-
     (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)) by XCMPLX_0:def 8
;
    then len Lower_Seq(C,n)-1+(E-max L~Cage(C,n))..Rotate(Cage(C,n),
     W-min L~Cage(C,n)) = len Rotate(Cage(C,n),W-min L~Cage(C,n)) by XCMPLX_1:
139;
    then (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)) =
     len Rotate(Cage(C,n),W-min L~Cage(C,n))-(len Lower_Seq(C,n)-1)
                                                                 by XCMPLX_1:26
       .= len Rotate(Cage(C,n),W-min L~Cage(C,n))-len Lower_Seq(C,n)+1
                                                                by XCMPLX_1:37;
    then len Rotate(Cage(C,n),W-min L~Cage(C,n))-len Lower_Seq(C,n)+1 <=
     len Rotate(Cage(C,n),W-min L~Cage(C,n)) by A4,FINSEQ_4:31;
    then len Rotate(Cage(C,n),W-min L~Cage(C,n))-len Lower_Seq(C,n) <=
     len Rotate(Cage(C,n),W-min L~Cage(C,n))-1 by REAL_1:84;
    then len Lower_Seq(C,n) >= 0+1 by REAL_2:105;
    then len Lower_Seq(C,n) > 0 by NAT_1:38;
    hence thesis by FINSEQ_1:25;
   end;
  end;

  definition
   let C be compact non vertical non horizontal Subset of TOP-REAL 2;
   let n be Nat;
   cluster Upper_Seq(C,n) -> one-to-one special unfolded s.n.c.;
   coherence
   proof
    A1: Upper_Seq(C,n) =
     Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n) by Def1;
    set f = Rotate(Cage(C,n),W-min L~Cage(C,n));
    A2: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A3: E-max L~Cage(C,n) in rng f by A2,FINSEQ_6:96;
    A4: len Upper_Seq(C,n) = (E-max L~Cage(C,n))..f by Th12;
      now let i1,j1 be set;
     assume that
      A5: i1 in dom Upper_Seq(C,n) and
      A6: j1 in dom Upper_Seq(C,n) and
      A7: Upper_Seq(C,n).i1 = Upper_Seq(C,n).j1 and
      A8: i1 <> j1;
     A9: L~Cage(C,n) = L~f by REVROT_1:33;
     reconsider i2 = i1, j2 = j1 as Nat by A5,A6;
     A10: i2 in Seg((E-max L~Cage(C,n))..f) & j2 in Seg((E-max L~Cage(C,n))..f)
                                              by A4,A5,A6,FINSEQ_1:def 3;
     A11: Upper_Seq(C,n).i1 = Upper_Seq(C,n)/.i2 by A5,FINSEQ_4:def 4
        .= f/.i2 by A1,A3,A10,FINSEQ_5:46;
     A12: Upper_Seq(C,n).j1 = Upper_Seq(C,n)/.j2 by A6,FINSEQ_4:def 4
        .= f/.j2 by A1,A3,A10,FINSEQ_5:46;
     A13: 1 <= i2 & 1 <= j2 by A10,FINSEQ_1:3;
     A14: j2 <= (E-max L~Cage(C,n))..f & i2 <= (E-max L~Cage(C,n))..f
                                                           by A10,FINSEQ_1:3;
       (E-max L~f)..f < len f by SPRECT_5:17;
     then A15: j2 < len f & i2 < len f by A9,A14,AXIOMS:22;
     per cases by A8,AXIOMS:21;
      suppose i2 < j2;
       hence contradiction by A7,A11,A12,A13,A15,GOBOARD7:38;
      suppose j2 < i2;
       hence contradiction by A7,A11,A12,A13,A15,GOBOARD7:38;
    end;
    hence Upper_Seq(C,n) is one-to-one by FUNCT_1:def 8;
    thus Upper_Seq(C,n) is special by A1;
    thus Upper_Seq(C,n) is unfolded by A1;
    let i,j be Nat;
    assume A16: i+1 < j;
      len f <> 0 by FINSEQ_1:25;
    then len f > 0 by NAT_1:19;
    then 0+1 <= len f by NAT_1:38;
    then A17: len f in dom f by FINSEQ_3:27;
      now per cases;
     suppose A18: j+1 <= len Upper_Seq(C,n);
      then A19: LSeg(Upper_Seq(C,n),j) = LSeg(f,j) by A1,SPPOL_2:9;
        j < len Upper_Seq(C,n) by A18,NAT_1:38;
      then i+1 <= len Upper_Seq(C,n) by A16,AXIOMS:22;
      then A20: LSeg(Upper_Seq(C,n),i) = LSeg(f,i) by A1,SPPOL_2:9;
      A21: j+1 <= (E-max L~Cage(C,n))..f by A1,A3,A18,FINSEQ_5:45;
      A22: (E-max L~Cage(C,n))..f <= len f by A3,FINSEQ_4:31;
        now per cases by A21,REAL_1:def 5;
       suppose j+1 < (E-max L~Cage(C,n))..f;
        hence j+1 < len f by A22,AXIOMS:22;
       suppose A23: j+1 = (E-max L~Cage(C,n))..f;
        assume j+1 >= len f;
        then (E-max L~Cage(C,n))..f = len f by A22,A23,AXIOMS:21;
        then E-max L~Cage(C,n) = f.(len f) by A3,FINSEQ_4:29
           .= f/.(len f) by A17,FINSEQ_4:def 4
           .= f/.1 by FINSEQ_6:def 1
           .= W-min L~Cage(C,n) by A2,FINSEQ_6:98;
        hence contradiction by TOPREAL5:25;
      end;
      hence LSeg(Upper_Seq(C,n),i) misses LSeg(Upper_Seq(C,n),j)
                                           by A16,A19,A20,GOBOARD5:def 4;
     suppose j+1 > len Upper_Seq(C,n);
      then LSeg(Upper_Seq(C,n),j) = {} by TOPREAL1:def 5;
      then LSeg(Upper_Seq(C,n),i) /\ LSeg(Upper_Seq(C,n),j) = {};
      hence LSeg(Upper_Seq(C,n),i) misses LSeg(Upper_Seq(C,n),j)
        by XBOOLE_0:def 7;
    end;
    hence thesis;
   end;
   cluster Lower_Seq(C,n) -> one-to-one special unfolded s.n.c.;
   coherence
   proof
    A24: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A25: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A24,FINSEQ_6:96;
    A26: Lower_Seq(C,n) =
     Rotate(Cage(C,n),W-min L~Cage(C,n)):-E-max L~Cage(C,n) by Def2;
    set f = Rotate(Cage(C,n),W-min L~Cage(C,n));
    A27: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A28: E-max L~Cage(C,n) in rng f by A27,FINSEQ_6:96;
    A29: len Lower_Seq(C,n) = len f-(E-max L~Cage(C,n))..f+1 by Th13;
    A30: L~Cage(C,n) = L~f by REVROT_1:33;
    then A31: f/.1 = W-min L~f by A27,FINSEQ_6:98;
    then A32: (W-max L~Cage(C,n))..f <= (N-min L~Cage(C,n))..f
                                                      by A30,SPRECT_5:24;
    A33: (N-min L~Cage(C,n))..f < (N-max L~Cage(C,n))..f
                                                      by A30,A31,SPRECT_5:25;
    A34: (N-max L~Cage(C,n))..f <= (E-max L~Cage(C,n))..f
                                                      by A30,A31,SPRECT_5:26;
      (W-max L~Cage(C,n))..f > 1 by A30,A31,SPRECT_5:23;
    then (N-min L~Cage(C,n))..f > 1 by A32,AXIOMS:22;
    then (N-max L~Cage(C,n))..f > 1 by A33,AXIOMS:22;
    then A35: (E-max L~Cage(C,n))..f > 1 by A34,AXIOMS:22;
      now let i1,j1 be set;
     assume that
      A36: i1 in dom Lower_Seq(C,n) and
      A37: j1 in dom Lower_Seq(C,n) and
      A38: Lower_Seq(C,n).i1 = Lower_Seq(C,n).j1 and
      A39: i1 <> j1;
     reconsider i2 = i1, j2 = j1 as Nat by A36,A37;
     A40: i2 in Seg len Lower_Seq(C,n) & j2 in Seg len Lower_Seq(C,n)
                                                  by A36,A37,FINSEQ_1:def 3;
     then i2 >= 1 & j2 >= 1 by FINSEQ_1:3;
     then A41: i2 = i2-'1+1 & j2 = j2-'1+1 by AMI_5:4;
     A42: Lower_Seq(C,n).i1 = Lower_Seq(C,n)/.i2 by A36,FINSEQ_4:def 4
        .= f/.(i2-'1+(E-max L~Cage(C,n))..f) by A26,A28,A36,A41,FINSEQ_5:55;
     A43: Lower_Seq(C,n).j1 = Lower_Seq(C,n)/.j2 by A37,FINSEQ_4:def 4
        .= f/.(j2-'1+(E-max L~Cage(C,n))..f) by A26,A28,A37,A41,FINSEQ_5:55;
       0+1 <= i2 & 0+1 <= j2 by A40,FINSEQ_1:3;
     then A44: i2-1 >= 0 & j2-1 >= 0 by REAL_1:84;
       j2 <= len f-(E-max L~Cage(C,n))..f+1 &
      i2 <= len f-(E-max L~Cage(C,n))..f+1 by A29,A40,FINSEQ_1:3;
     then j2-1 <= len f-(E-max L~Cage(C,n))..f &
      i2-1 <= len f-(E-max L~Cage(C,n))..f by REAL_1:86;
     then j2-1+(E-max L~Cage(C,n))..f <= len f &
      i2-1+(E-max L~Cage(C,n))..f <= len f by REAL_1:84;
     then A45: j2-'1+(E-max L~Cage(C,n))..f <= len f &
      i2-'1+(E-max L~Cage(C,n))..f <= len f by A44,BINARITH:def 3;
       i2-'1+(E-max L~Cage(C,n))..f >= (E-max L~Cage(C,n))..f &
      j2-'1+(E-max L~Cage(C,n))..f >= (E-max L~Cage(C,n))..f by NAT_1:29;
     then A46: 1 < i2-'1+(E-max L~Cage(C,n))..f &
      1 < j2-'1+(E-max L~Cage(C,n))..f by A35,AXIOMS:22;
     per cases by A39,AXIOMS:21;
      suppose i2 < j2;
       then i2-1 < j2-1 by REAL_1:54;
       then i2-1 < j2-'1 by A44,BINARITH:def 3;
       then i2-'1 < j2-'1 by A44,BINARITH:def 3;
       then i2-'1+(E-max L~Cage(C,n))..f < j2-'1+(E-max L~Cage(C,n))..f
                                                                by REAL_1:53;
       hence contradiction by A38,A42,A43,A45,A46,GOBOARD7:39;
      suppose j2 < i2;
       then j2-1 < i2-1 by REAL_1:54;
       then j2-1 < i2-'1 by A44,BINARITH:def 3;
       then j2-'1 < i2-'1 by A44,BINARITH:def 3;
       then j2-'1+(E-max L~Cage(C,n))..f < i2-'1+(E-max L~Cage(C,n))..f
                                                                by REAL_1:53;
       hence contradiction by A38,A42,A43,A45,A46,GOBOARD7:39;
    end;
    hence Lower_Seq(C,n) is one-to-one by FUNCT_1:def 8;
    thus Lower_Seq(C,n) is special by A25,A26,SPPOL_2:41;
    thus Lower_Seq(C,n) is unfolded by A25,A26,SPPOL_2:28;
    let i,j be Nat;
    assume A47: i+1 < j;
      i+1 >= 1 by NAT_1:29;
    then A48: j > 0+1 by A47,AXIOMS:22;
    then j = j-'1+1 by AMI_5:4;
    then A49: LSeg(Lower_Seq(C,n),j) = LSeg(f,j-'1+(E-max L~Cage(C,n))..f)
                                                         by A25,A26,SPPOL_2:10;
      now per cases by NAT_1:18;
     suppose i > 0;
      then A50: i >= 0+1 by NAT_1:38;
      then i = i-'1+1 by AMI_5:4;
      then A51: LSeg(Lower_Seq(C,n),i) = LSeg(f,i-'1+(E-max L~Cage(C,n))..f)
                                                         by A26,A28,SPPOL_2:10;
      A52: j-1 > 0 by A48,REAL_1:86;
      A53: i-1 >= 0 by A50,REAL_1:84;
        i+1-1 < j-1 by A47,REAL_1:54;
      then i-1+1 < j-1 by XCMPLX_1:29;
      then i-1+1 < j-'1 by A52,BINARITH:def 3;
      then i-'1+1 < j-'1 by A53,BINARITH:def 3;
      then i-'1+1+(E-max L~Cage(C,n))..f < j-'1+(E-max L~Cage(C,n))..f
                                                                by REAL_1:53;
      then A54: i-'1+(E-max L~Cage(C,n))..f+1 < j-'1+(E-max L~Cage(C,n))..f
                                                                by XCMPLX_1:1;
        i-'1 >= 0 by NAT_1:18;
      then A55: i-'1+(E-max L~Cage(C,n))..f > 0+1 by A35,REAL_1:67;
        now per cases;
       suppose j-'1+(E-max L~Cage(C,n))..f+1 <= len f;
        then j-'1+(E-max L~Cage(C,n))..f < len f by NAT_1:38;
        then LSeg(Lower_Seq(C,n),i) misses LSeg(Lower_Seq(C,n),j)
                                      by A49,A51,A54,A55,GOBOARD5:def 4;
        hence LSeg(Lower_Seq(C,n),i) /\ LSeg(Lower_Seq(C,n),j) = {}
                                      by XBOOLE_0:def 7;
       suppose j-'1+(E-max L~Cage(C,n))..f+1 > len f;
        then LSeg(Lower_Seq(C,n),j) = {} by A49,TOPREAL1:def 5;
        hence LSeg(Lower_Seq(C,n),i) /\ LSeg(Lower_Seq(C,n),j) = {};
      end;
      hence LSeg(Lower_Seq(C,n),i) /\ LSeg(Lower_Seq(C,n),j) = {};
     suppose i = 0;
      then LSeg(Lower_Seq(C,n),i) = {} by TOPREAL1:def 5;
      hence LSeg(Lower_Seq(C,n),i) /\ LSeg(Lower_Seq(C,n),j) = {};
    end;
    hence thesis by XBOOLE_0:def 7;
   end;
  end;

  theorem Th14:
   for C be compact non vertical non horizontal Subset of TOP-REAL 2
   for n be Nat holds
    len Upper_Seq(C,n) + len Lower_Seq(C,n) = len Cage(C,n)+1
   proof
    let C be compact non vertical non horizontal Subset of TOP-REAL 2;
    let n be Nat;
    thus len Upper_Seq(C,n) + len Lower_Seq(C,n) =
          (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+
          len Lower_Seq(C,n) by Th12
       .= (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+
          (len Rotate(Cage(C,n),W-min L~Cage(C,n))-
          (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+1)
                                                                  by Th13
       .= (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+
          (len Rotate(Cage(C,n),W-min L~Cage(C,n))-
          (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)))+1
                                                                 by XCMPLX_1:1
       .= (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+
          len Rotate(Cage(C,n),W-min L~Cage(C,n))-
          (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+1
                                                                 by XCMPLX_1:29
       .= len Rotate(Cage(C,n),W-min L~Cage(C,n))+1 by XCMPLX_1:26
       .= len Cage(C,n)+1 by REVROT_1:14;
   end;

  theorem Th15:
   for C be compact non vertical non horizontal Subset of TOP-REAL 2
   for n be Nat holds
    Rotate(Cage(C,n),W-min L~Cage(C,n)) = Upper_Seq(C,n) ^' Lower_Seq(C,n)
   proof
    let C be compact non vertical non horizontal Subset of TOP-REAL 2;
    let n be Nat;
    A1: len (Upper_Seq(C,n) ^' Lower_Seq(C,n)) +1 =
          len Upper_Seq(C,n) + len Lower_Seq(C,n) by GRAPH_2:13
       .= len Cage(C,n)+1 by Th14
       .= len Rotate(Cage(C,n),W-min L~Cage(C,n))+1 by REVROT_1:14;
    then A2: len (Upper_Seq(C,n) ^' Lower_Seq(C,n)) =
     len Rotate(Cage(C,n),W-min L~Cage(C,n)) by XCMPLX_1:2;
      now let i be Nat;
     assume A3: i in Seg len Rotate(Cage(C,n),W-min L~Cage(C,n));
     then A4: 1 <= i & i <= len Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                               by FINSEQ_1:3;
     A5: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
       E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
     then A6: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A5,FINSEQ_6:96;
     A7: i in dom Rotate(Cage(C,n),W-min L~Cage(C,n)) by A3,FINSEQ_1:def 3;
     per cases;
      suppose A8: i <= len Upper_Seq(C,n);
       then i <= (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                                 by Th12;
       then A9: i in Seg ((E-max L~Cage(C,n))..Rotate(Cage(C,n),
                                       W-min L~Cage(C,n))) by A4,FINSEQ_1:3;
         len (Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n)) =
        (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A6,FINSEQ_5:45;
       then A10: i in dom (Rotate(Cage(C,n),W-min L~Cage(C,n))-:
                         E-max L~Cage(C,n)) by A9,FINSEQ_1:def 3;
       thus (Upper_Seq(C,n) ^' Lower_Seq(C,n)).i = Upper_Seq(C,n).i
                                                         by A4,A8,GRAPH_2:14
          .= (Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n)).i
                                                                  by Def1
          .= (Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n))/.i
                                                        by A10,FINSEQ_4:def 4
          .= Rotate(Cage(C,n),W-min L~Cage(C,n))/.i by A6,A9,FINSEQ_5:46
          .= Rotate(Cage(C,n),W-min L~Cage(C,n)).i by A7,FINSEQ_4:def 4;
      suppose i > len Upper_Seq(C,n);
       then i >= len Upper_Seq(C,n)+1 by NAT_1:38;
       then consider j be Nat such that
        A11: i = len Upper_Seq(C,n)+1+j by NAT_1:28;
       A12: i = len Upper_Seq(C,n)+(j+1) by A11,XCMPLX_1:1;
       A13: j+1 >= 1 by NAT_1:29;
         i < len (Upper_Seq(C,n) ^' Lower_Seq(C,n))+1 by A1,A4,NAT_1:38;
       then i < len Lower_Seq(C,n) + len Upper_Seq(C,n) by GRAPH_2:13;
       then i-len Upper_Seq(C,n) < len Lower_Seq(C,n) by REAL_1:84;
       then A14: j+1 < len Lower_Seq(C,n) by A12,XCMPLX_1:26;
       A15: i = (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+(j+1)
                                                             by A12,Th12;
       A16: j+1+1 >= 1 by NAT_1:29;
       A17: len (Rotate(Cage(C,n),W-min L~Cage(C,n)):-E-max L~Cage(C,n)) =
        len Rotate(Cage(C,n),W-min L~Cage(C,n))-
        (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n))+1
                                                           by A6,FINSEQ_5:53;
         j+1+(E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)) <=
        len Rotate(Cage(C,n),W-min L~Cage(C,n)) by A4,A12,Th12;
       then j+1 <= len Rotate(Cage(C,n),W-min L~Cage(C,n))-
        (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)) by REAL_1:84;
       then j+1+1 <= len (Rotate(Cage(C,n),W-min L~Cage(C,n)):-
        (E-max L~Cage(C,n))) by A17,REAL_1:55;
       then j+1+1 in Seg len (Rotate(Cage(C,n),W-min L~Cage(C,n)):-
        (E-max L~Cage(C,n))) by A16,FINSEQ_1:3;
       then A18: j+1+1 in dom (Rotate(Cage(C,n),W-min L~Cage(C,n)):-
        (E-max L~Cage(C,n))) by FINSEQ_1:def 3;
       thus (Upper_Seq(C,n) ^' Lower_Seq(C,n)).i = Lower_Seq(C,n).(j+1+1)
                                                    by A12,A13,A14,GRAPH_2:15
          .= (Rotate(Cage(C,n),W-min L~Cage(C,n)):-(E-max L~Cage(C,n))).
             (j+1+1) by Def2
          .= (Rotate(Cage(C,n),W-min L~Cage(C,n)):-(E-max L~Cage(C,n)))/.
             (j+1+1) by A18,FINSEQ_4:def 4
          .= Rotate(Cage(C,n),W-min L~Cage(C,n))/.
             (j+1+(E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)))
                                          by A6,A18,FINSEQ_5:55
          .= Rotate(Cage(C,n),W-min L~Cage(C,n)).i by A7,A15,FINSEQ_4:def 4;
    end;
    hence Rotate(Cage(C,n),W-min L~Cage(C,n)) =
     Upper_Seq(C,n) ^' Lower_Seq(C,n) by A2,FINSEQ_2:10;
   end;

  theorem
     for C be compact non vertical non horizontal Subset of TOP-REAL 2
   for n be Nat holds
    L~Cage(C,n) = L~(Upper_Seq(C,n) ^' Lower_Seq(C,n))
   proof
    let C be compact non vertical non horizontal Subset of TOP-REAL 2;
    let n be Nat;
      Rotate(Cage(C,n),W-min L~Cage(C,n)) = Upper_Seq(C,n) ^' Lower_Seq(C,n)
                                                                  by Th15;
    hence L~Cage(C,n) = L~(Upper_Seq(C,n) ^' Lower_Seq(C,n)) by REVROT_1:33;
   end;

  theorem
     for C be compact non vertical non horizontal non empty Subset of TOP-REAL
2
   for n be Nat holds
    L~Cage(C,n) = L~Upper_Seq(C,n) \/ L~Lower_Seq(C,n)
   proof
    let C be compact non vertical non horizontal non empty
     Subset of TOP-REAL 2;
    let n be Nat;
    A1: Upper_Seq(C,n) = Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n)
                                                                 by Def1;
    A2: Lower_Seq(C,n) = Rotate(Cage(C,n),W-min L~Cage(C,n)):-E-max L~Cage(C,n)
                                                                by Def2;
    A3: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A3,FINSEQ_6:96;
    then L~Rotate(Cage(C,n),W-min L~Cage(C,n)) =
     L~Upper_Seq(C,n) \/ L~Lower_Seq(C,n) by A1,A2,SPPOL_2:25;
    hence thesis by REVROT_1:33;
   end;

  theorem Th18:
   for P be Simple_closed_curve holds W-min(P) <> E-min(P)
   proof
    let P be Simple_closed_curve;
      now assume A1: W-min(P) = E-min(P);
     A2:|[W-bound P, inf (proj2||W-most P)]|=W-min(P) by PSCOMP_1:def 42;
       |[E-bound P, inf (proj2||E-most P)]|=E-min(P) by PSCOMP_1:def 47;
     then W-bound P= E-bound P by A1,A2,SPPOL_2:1;
     hence contradiction by TOPREAL5:23;
    end;
    hence W-min(P) <> E-min(P);
   end;

  theorem Th19:
   for C be compact non vertical non horizontal Subset of TOP-REAL 2
   for n be Nat holds
    len Upper_Seq(C,n) >= 3 & len Lower_Seq(C,n) >= 3
   proof
    let C be compact non vertical non horizontal Subset of TOP-REAL 2;
    let n be Nat;
    set pWi = W-min L~Cage(C,n);
    set pWa = W-max L~Cage(C,n);
    set pEi = E-min L~Cage(C,n);
    set pEa = E-max L~Cage(C,n);
    A1: pWi <> pWa by SPRECT_2:62;
    A2: pWa <> pEa by JORDAN1B:6;
      pWi <> pEa by TOPREAL5:25;
    then A3: Card {pWi,pWa,pEa} = 3 by A1,A2,CARD_2:77;
    A4: Upper_Seq(C,n) = Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n)
                                                                 by Def1;
    A5: Lower_Seq(C,n) = Rotate(Cage(C,n),W-min L~Cage(C,n)):-E-max L~Cage(C,n)
                                                                by Def2;
    A6: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A7: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A6,FINSEQ_6:96;
      (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)) <=
     (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n));
    then A8: pEa in rng Upper_Seq(C,n) by A4,A7,FINSEQ_5:49;
    A9: Upper_Seq(C,n)/.1 = Rotate(Cage(C,n),W-min L~Cage(C,n))/.1
                                                       by A4,A7,FINSEQ_5:47;
    then A10: Upper_Seq(C,n)/.1 = W-min L~Cage(C,n) by A6,FINSEQ_6:98;
    then A11: pWi in rng Upper_Seq(C,n) by FINSEQ_6:46;
      pWa in rng Cage(C,n) by SPRECT_2:48;
    then A12: pWa in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                          by A6,FINSEQ_6:96;
    set f=Rotate(Cage(C,n),W-min L~Cage(C,n));
    A13: L~Cage(C,n) = L~f by REVROT_1:33;
    then A14: (W-max L~f)..f <= (N-min L~f)..f by A9,A10,SPRECT_5:24;
      (N-min L~f)..f < (N-max L~f)..f by A9,A10,A13,SPRECT_5:25;
    then A15: (W-max L~f)..f < (N-max L~f)..f by A14,AXIOMS:22;
      (N-max L~f)..f <= (E-max L~f)..f by A9,A10,A13,SPRECT_5:26;
    then pWa..Rotate(Cage(C,n),W-min L~Cage(C,n)) <
     pEa..Rotate(Cage(C,n),W-min L~Cage(C,n)) by A13,A15,AXIOMS:22;
    then A16: pWa in rng Upper_Seq(C,n) by A4,A7,A12,FINSEQ_5:49;
      {pWi,pWa,pEa} c= rng Upper_Seq(C,n)
    proof
     let x be set;
     assume x in {pWi,pWa,pEa};
     hence x in rng Upper_Seq(C,n) by A8,A11,A16,ENUMSET1:def 1;
    end;
    then A17: Card {pWi,pWa,pEa} c= Card rng Upper_Seq(C,n) by CARD_1:27;
      Card rng Upper_Seq(C,n) c= Card dom Upper_Seq(C,n) by YELLOW_6:3;
    then Card rng Upper_Seq(C,n) c= len Upper_Seq(C,n) by PRE_CIRC:21;
    then 3 c= len Upper_Seq(C,n) by A3,A17,XBOOLE_1:1;
    hence len Upper_Seq(C,n) >= 3 by CARD_1:56;
    A18: pWi <> pEi by Th18;
    A19: pEi <> pEa by SPRECT_2:58;
      pWi <> pEa by TOPREAL5:25;
    then A20: Card {pWi,pEi,pEa} = 3 by A18,A19,CARD_2:77;
    A21: pEa in rng Lower_Seq(C,n) by A5,FINSEQ_6:66;
      Lower_Seq(C,n)/.(len Lower_Seq(C,n)) =
          Rotate(Cage(C,n),W-min L~Cage(C,n))/.
              (len Rotate(Cage(C,n),W-min L~Cage(C,n))) by A5,A7,FINSEQ_5:57
       .= Rotate(Cage(C,n),W-min L~Cage(C,n))/.1 by FINSEQ_6:def 1
       .= W-min L~Cage(C,n) by A6,FINSEQ_6:98;
    then A22: pWi in rng Lower_Seq(C,n) by REVROT_1:3;
      pEi in rng Cage(C,n) by SPRECT_2:49;
    then A23: pEi in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                          by A6,FINSEQ_6:96;
      pEi..Rotate(Cage(C,n),W-min L~Cage(C,n)) >
     pEa..Rotate(Cage(C,n),W-min L~Cage(C,n)) by A9,A10,A13,SPRECT_5:27;
    then A24: pEi in rng Lower_Seq(C,n) by A5,A7,A23,FINSEQ_6:67;
      {pWi,pEi,pEa} c= rng Lower_Seq(C,n)
    proof
     let x be set;
     assume x in {pWi,pEi,pEa};
     hence x in rng Lower_Seq(C,n) by A21,A22,A24,ENUMSET1:def 1;
    end;
    then A25: Card {pWi,pEi,pEa} c= Card rng Lower_Seq(C,n) by CARD_1:27;
      Card rng Lower_Seq(C,n) c= Card dom Lower_Seq(C,n) by YELLOW_6:3;
    then Card rng Lower_Seq(C,n) c= len Lower_Seq(C,n) by PRE_CIRC:21;
    then 3 c= len Lower_Seq(C,n) by A20,A25,XBOOLE_1:1;
    hence len Lower_Seq(C,n) >= 3 by CARD_1:56;
   end;

  definition
   let C be compact non vertical non horizontal Subset of TOP-REAL 2;
   let n be Nat;
   cluster Upper_Seq(C,n) -> being_S-Seq;
   coherence
   proof
      len Upper_Seq(C,n) >= 3 by Th19;
    then len Upper_Seq(C,n) >= 2 by AXIOMS:22;
    hence thesis by TOPREAL1:def 10;
   end;
   cluster Lower_Seq(C,n) -> being_S-Seq;
   coherence
   proof
      len Lower_Seq(C,n) >= 3 by Th19;
    then len Lower_Seq(C,n) >= 2 by AXIOMS:22;
    hence thesis by TOPREAL1:def 10;
   end;
  end;

  theorem
     for C be compact non vertical non horizontal Subset of TOP-REAL 2
   for n be Nat holds
    L~Upper_Seq(C,n) /\ L~Lower_Seq(C,n) =
     {W-min L~Cage(C,n),E-max L~Cage(C,n)}
   proof
    let C be compact non vertical non horizontal Subset of TOP-REAL 2;
    let n be Nat;
    A1: Upper_Seq(C,n) = Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n)
                                                                 by Def1;
    A2: Lower_Seq(C,n) = Rotate(Cage(C,n),W-min L~Cage(C,n)):-E-max L~Cage(C,n)
                                                                by Def2;
    A3: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A4: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A3,FINSEQ_6:96;
    A5: (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n)) <=
        (E-max L~Cage(C,n))..Rotate(Cage(C,n),W-min L~Cage(C,n));
      len Lower_Seq(C,n) >= 3 by Th19;
    then len Lower_Seq(C,n) >= 2 by AXIOMS:22;
    then A6: rng Lower_Seq(C,n) c= L~Lower_Seq(C,n) by SPPOL_2:18;
    A7: E-max L~Cage(C,n) in rng Lower_Seq(C,n) by A2,FINSEQ_6:66;
    A8: len Upper_Seq(C,n) >= 3 by Th19;
    then len Upper_Seq(C,n) >= 2 by AXIOMS:22;
    then A9: rng Upper_Seq(C,n) c= L~Upper_Seq(C,n) by SPPOL_2:18;
    A10: E-max L~Cage(C,n) in rng Upper_Seq(C,n) by A1,A4,A5,FINSEQ_5:49;
      Upper_Seq(C,n)/.1 = Rotate(Cage(C,n),W-min L~Cage(C,n))/.1
                                                         by A1,A4,FINSEQ_5:47
       .= W-min L~Cage(C,n) by A3,FINSEQ_6:98;
    then A11: W-min L~Cage(C,n) in rng Upper_Seq(C,n) by FINSEQ_6:46;
      Lower_Seq(C,n)/.(len Lower_Seq(C,n)) =
          Rotate(Cage(C,n),W-min L~Cage(C,n))/.
           (len Rotate(Cage(C,n),W-min L~Cage(C,n))) by A2,A4,FINSEQ_5:57
       .= Rotate(Cage(C,n),W-min L~Cage(C,n))/.1 by FINSEQ_6:def 1
       .= W-min L~Cage(C,n) by A3,FINSEQ_6:98;
    then A12: W-min L~Cage(C,n) in rng Lower_Seq(C,n) by REVROT_1:3;
    thus L~Upper_Seq(C,n) /\ L~Lower_Seq(C,n) c=
     {W-min L~Cage(C,n),E-max L~Cage(C,n)}
    proof
     let x be set;
     assume A13: x in L~Upper_Seq(C,n) /\ L~Lower_Seq(C,n);
     then A14: x in L~Upper_Seq(C,n) & x in L~Lower_Seq(C,n) by XBOOLE_0:def 3;
     reconsider x1=x as Point of TOP-REAL 2 by A13;
     consider i1 be Nat such that
      A15: 1 <= i1 & i1+1 <= len Upper_Seq(C,n) and
      A16: x1 in LSeg(Upper_Seq(C,n),i1) by A14,SPPOL_2:13;
     consider i2 be Nat such that
      A17: 1 <= i2 & i2+1 <= len Lower_Seq(C,n) and
      A18: x1 in LSeg(Lower_Seq(C,n),i2) by A14,SPPOL_2:13;
     set f=Rotate(Cage(C,n),W-min L~Cage(C,n));
     set pE=E-max L~Cage(C,n);
     set pW=W-min L~Cage(C,n);
     A19: LSeg(Upper_Seq(C,n),i1) = LSeg(f,i1) by A1,A15,SPPOL_2:9;
     set i3=i2-'1;
     A20: i3+1 = i2 by A17,AMI_5:4;
     then A21: LSeg(Lower_Seq(C,n),i2) = LSeg(f,i3+pE..f) by A2,A4,SPPOL_2:10;
     A22: len Upper_Seq(C,n) = pE..f by Th12;
     A23: len Lower_Seq(C,n) = len f-pE..f+1 by Th13;
     A24: i2-1 >= 1-1 by A17,REAL_1:49;
       i2 < len f-pE..f+1 by A17,A23,NAT_1:38;
     then i2-1 < len f-pE..f by REAL_1:84;
     then i2-1+pE..f < len f by REAL_1:86;
     then A25: i3+pE..f < len f by A24,BINARITH:def 3;
     A26: i3 >= 0 by NAT_1:18;
     A27: pE..f >= 1 by A4,FINSEQ_4:31;
       i3+1 < len f-pE..f+1 by A17,A20,A23,NAT_1:38;
     then i3 < len f-pE..f by REAL_1:55;
     then A28: i3+pE..f < len f by REAL_1:86;
     A29: i1+1 < pE..f+1 by A15,A22,NAT_1:38;
       1+pE..f <= i3+1+pE..f by A17,A20,REAL_1:55;
     then i1+1 < i3+1+pE..f by A29,AXIOMS:22;
     then i1+1 < i3+pE..f+1 by XCMPLX_1:1;
     then A30: i1+1 <= i3+pE..f by NAT_1:38;
     A31: i3+pE..f+1 <= len f by A28,NAT_1:38;
     A32: len f > 4 by GOBOARD7:36;
     A33: i3+pE..f >= 0 & i1 >= 0 by NAT_1:18;
     A34: pE..f-'1+1 = pE..f by A27,AMI_5:4;
     assume A35: not x in {W-min L~Cage(C,n),E-max L~Cage(C,n)};
       now per cases by A15,A30,REAL_1:def 5;
      suppose i1+1 < i3+pE..f & i1 > 1;
       then LSeg(f,i1) misses LSeg(f,i3+pE..f) by A28,GOBOARD5:def 4;
       then LSeg(f,i1) /\ LSeg(f,i3+pE..f) = {} by XBOOLE_0:def 7;
       hence contradiction by A16,A18,A19,A21,XBOOLE_0:def 3;
      suppose A36: i1 = 1;
         i3+pE..f >= 0+3 by A8,A22,A26,REAL_1:55;
       then A37: i1+1 < i3+pE..f by A36,AXIOMS:22;
         now per cases by A31,REAL_1:def 5;
        suppose i3+pE..f+1 < len f;
         then LSeg(f,i1) misses LSeg(f,i3+pE..f) by A37,GOBOARD5:def 4;
         then LSeg(f,i1) /\ LSeg(f,i3+pE..f) = {} by XBOOLE_0:def 7;
         hence contradiction by A16,A18,A19,A21,XBOOLE_0:def 3;
        suppose i3+pE..f+1 = len f;
         then i3+pE..f = len f-1 by XCMPLX_1:26;
         then i3+pE..f = len f-'1 by A33,BINARITH:def 3;
         then LSeg(f,i1) /\ LSeg(f,i3+pE..f) = {f/.1} by A32,A36,REVROT_1:30;
         then x1 in {f/.1} by A16,A18,A19,A21,XBOOLE_0:def 3;
         then x1 = f/.1 by TARSKI:def 1
            .= pW by A3,FINSEQ_6:98;
         hence contradiction by A35,TARSKI:def 2;
       end;
       hence contradiction;
      suppose A38: i1+1 = i3+pE..f;
         0+pE..f <= i3+pE..f by A26,REAL_1:55;
       then pE..f = i1+1 by A15,A22,A38,AXIOMS:21;
       then i1 = pE..f-1 by XCMPLX_1:26;
       then A39: i1 = pE..f-'1 by A33,BINARITH:def 3;
         i3+pE..f >= pE..f by NAT_1:29;
       then pE..f < len f by A25,AXIOMS:22;
       then pE..f+1 <= len f by NAT_1:38;
       then pE..f-'1 + (1+1) <= len f by A34,XCMPLX_1:1;
       then LSeg(f,i1) /\ LSeg(f,i3+(pE..f)) = {f/.(pE..f)}
                                          by A15,A34,A38,A39,TOPREAL1:def 8;
       then x1 in {f/.(pE..f)} by A16,A18,A19,A21,XBOOLE_0:def 3;
       then x1 = f/.(pE..f) by TARSKI:def 1
          .= pE by A4,FINSEQ_5:41;
       hence contradiction by A35,TARSKI:def 2;
     end;
     hence contradiction;
    end;
    thus {W-min L~Cage(C,n),E-max L~Cage(C,n)} c=
     L~Upper_Seq(C,n) /\ L~Lower_Seq(C,n)
    proof
     let x be set;
     assume A40: x in {W-min L~Cage(C,n),E-max L~Cage(C,n)};
     per cases by A40,TARSKI:def 2;
      suppose x = W-min L~Cage(C,n);
       hence x in L~Upper_Seq(C,n) /\ L~Lower_Seq(C,n)
                                               by A6,A9,A11,A12,XBOOLE_0:def 3;
      suppose x = E-max L~Cage(C,n);
       hence x in L~Upper_Seq(C,n) /\ L~Lower_Seq(C,n)
                                                by A6,A7,A9,A10,XBOOLE_0:def 3;
    end;
   end;

  theorem
     for C be compact non vertical non horizontal Subset of TOP-REAL 2 holds
    Upper_Seq(C,n) is_in_the_area_of Cage(C,n)
   proof
    let C be compact non vertical non horizontal Subset of TOP-REAL 2;
    A1: Upper_Seq(C,n) = Rotate(Cage(C,n),W-min L~Cage(C,n))-:E-max L~Cage(C,n)
                                                                 by Def1;
    A2: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A3: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A2,FINSEQ_6:96;
      Rotate(Cage(C,n),W-min L~Cage(C,n)) is_in_the_area_of Cage(C,n)
                                                              by JORDAN1B:11;
    hence Upper_Seq(C,n) is_in_the_area_of Cage(C,n) by A1,A3,Th5;
   end;

  theorem
     for C be compact non vertical non horizontal Subset of TOP-REAL 2 holds
    Lower_Seq(C,n) is_in_the_area_of Cage(C,n)
   proof
    let C be compact non vertical non horizontal Subset of TOP-REAL 2;
    A1: Lower_Seq(C,n) = Rotate(Cage(C,n),W-min L~Cage(C,n)):-E-max L~Cage(C,n)
                                                                 by Def2;
    A2: W-min L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:47;
      E-max L~Cage(C,n) in rng Cage(C,n) by SPRECT_2:50;
    then A3: E-max L~Cage(C,n) in rng Rotate(Cage(C,n),W-min L~Cage(C,n))
                                                           by A2,FINSEQ_6:96;
      Rotate(Cage(C,n),W-min L~Cage(C,n)) is_in_the_area_of Cage(C,n)
                                                              by JORDAN1B:11;
    hence thesis by A1,A3,Th6;
   end;

  theorem
     for C be compact connected non vertical non horizontal Subset of TOP-REAL
2
     holds (Cage(C,n)/.2)`2 = N-bound L~Cage(C,n)
   proof
    let C be compact connected non vertical non horizontal
     Subset of TOP-REAL 2;
      Cage(C,n)/.1 = N-min L~Cage(C,n) by JORDAN9:34;
    then Cage(C,n)/.2 in N-most L~Cage(C,n) by SPRECT_2:34;
    then (Cage(C,n)/.2)`2 = (N-min L~Cage(C,n))`2 by PSCOMP_1:98;
    hence thesis by PSCOMP_1:94;
   end;

  theorem
     for C be compact connected non vertical non horizontal Subset of TOP-REAL
2
   for k be Nat st 1 <= k & k+1 <= len Cage(C,n) &
    Cage(C,n)/.k = E-max L~Cage(C,n) holds
     (Cage(C,n)/.(k+1))`1 = E-bound L~Cage(C,n)
   proof
    let C be compact connected non vertical non horizontal
     Subset of TOP-REAL 2;
    A1: Cage(C,n) is_sequence_on Gauge(C,n) by JORDAN9:def 1;
    let k be Nat;
    assume that
     A2: 1 <= k and
     A3: k+1 <= len Cage(C,n) and
     A4: Cage(C,n)/.k = E-max L~Cage(C,n);
    A5: Cage(C,n)/.1 = N-min L~Cage(C,n) by JORDAN9:34;
    then A6: 1 < (N-max L~Cage(C,n))..Cage(C,n) by SPRECT_2:73;
      (N-max L~Cage(C,n))..Cage(C,n) <= (E-max L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:74;
    then A7: (E-max L~Cage(C,n))..Cage(C,n) > 1 by A6,AXIOMS:22;
    A8: k < len Cage(C,n) by A3,NAT_1:38;
    then A9: k in dom Cage(C,n) by A2,FINSEQ_3:27;
    then (E-max L~Cage(C,n))..Cage(C,n) <= k by A4,FINSEQ_5:42;
    then A10: k > 1 by A7,AXIOMS:22;
    then consider i1,j1,i2,j2 be Nat such that
     A11: [i1,j1] in Indices Gauge(C,n) and
     A12: Cage(C,n)/.k = Gauge(C,n)*(i1,j1) and
     A13: [i2,j2] in Indices Gauge(C,n) and
     A14: Cage(C,n)/.(k+1) = Gauge(C,n)*(i2,j2) and
     A15: i1 = i2 & j1+1 = j2 or i1+1 = i2 & j1 = j2 or
         i1 = i2+1 & j1 = j2 or i1 = i2 & j1 = j2+1 by A1,A3,JORDAN8:6;
    A16: 1 <= i1 & i1 <= len Gauge(C,n) & 1 <= j1 & j1 <= width Gauge(C,n)
                                                            by A11,GOBOARD5:1;
    A17: 1 <= i2 & i2 <= len Gauge(C,n) & 1 <= j2 & j2 <= width Gauge(C,n)
                                                            by A13,GOBOARD5:1;
    A18: len Gauge(C,n) = width Gauge(C,n) by JORDAN8:def 1;
      Gauge(C,n)*(i1,j1)`1 = E-bound L~Cage(C,n) by A4,A12,PSCOMP_1:104
       .= Gauge(C,n)*(len Gauge(C,n),j1)`1 by A16,A18,JORDAN1A:92;
    then A19: i1 <= len Gauge(C,n) & i1 >= len Gauge(C,n) by A16,GOBOARD5:4;
    then A20: i1 = len Gauge(C,n) by AXIOMS:21;
    reconsider k'=k-1 as Nat by A9,FINSEQ_3:28;
      k >= 1+1 by A10,NAT_1:38;
    then A21: k' >= 1+1-1 by REAL_1:49;
    A22: k'+1 < len Cage(C,n) by A8,XCMPLX_1:27;
    then consider i3,j3,i4,j4 be Nat such that
     A23: [i3,j3] in Indices Gauge(C,n) and
     A24: Cage(C,n)/.k' = Gauge(C,n)*(i3,j3) and
     A25: [i4,j4] in Indices Gauge(C,n) and
     A26: Cage(C,n)/.(k'+1) = Gauge(C,n)*(i4,j4) and
     A27: i3 = i4 & j3+1 = j4 or i3+1 = i4 & j3 = j4 or
         i3 = i4+1 & j3 = j4 or i3 = i4 & j3 = j4+1 by A1,A21,JORDAN8:6;
    A28: k'+1 = k by XCMPLX_1:27;
    then A29: i1 = i4 & j1 = j4 by A11,A12,A25,A26,GOBOARD1:21;
    A30: 1 <= i3 & i3 <= len Gauge(C,n) & 1 <= j3 & j3 <= width Gauge(C,n)
                                                            by A23,GOBOARD5:1;
    A31: j3 = j4
    proof
     assume A32: j3 <> j4;
     per cases by A27,A32;
      suppose i3 = i4 & j3+1 = j4;
       hence contradiction by A19,A21,A22,A23,A24,A25,A26,A29,JORDAN10:1;
      suppose A33: i3 = i4 & j3 = j4+1;
         k' < len Cage(C,n) by A22,NAT_1:38;
       then Gauge(C,n)*(i3,j3) in E-most L~Cage(C,n)
                                        by A20,A21,A24,A29,A30,A33,JORDAN1A:82;
       then A34: (Gauge(C,n)*(i4,j4+1))`2 <= (Gauge(C,n)*(i4,j4))`2
                                                by A4,A26,A28,A33,PSCOMP_1:108;
         j4 < j4+1 by NAT_1:38;
       hence contradiction by A16,A29,A30,A33,A34,GOBOARD5:5;
    end;
      i1 = i2
    proof
     assume A35: i1 <> i2;
     per cases by A15,A35;
      suppose i1+1 = i2 & j1 = j2;
       hence contradiction by A17,A19,NAT_1:38;
      suppose i1 = i2+1 & j1 = j2;
       then A36: Cage(C,n)/.(k+1) = Cage(C,n)/.k'
                             by A14,A19,A24,A27,A29,A30,A31,NAT_1:38,XCMPLX_1:2
;
         k'+1+1 <= len Cage(C,n) by A22,NAT_1:38;
       then k'+(1+1) <= len Cage(C,n) by XCMPLX_1:1;
       then A37: LSeg(Cage(C,n),k') /\ LSeg(Cage(C,n),k) = {Cage(C,n)/.k}
                                                    by A21,A28,TOPREAL1:def 8;
         Cage(C,n)/.k' in LSeg(Cage(C,n),k') &
        Cage(C,n)/.(k+1) in LSeg(Cage(C,n),k) by A2,A3,A8,A21,A28,TOPREAL1:27;
       then Cage(C,n)/.(k+1) in {Cage(C,n)/.k} by A36,A37,XBOOLE_0:def 3;
       then Cage(C,n)/.(k+1) = Cage(C,n)/.k by TARSKI:def 1;
       hence contradiction by A11,A12,A13,A14,A35,GOBOARD1:21;
    end;
    then Gauge(C,n)*(i1,j1)`1 = Gauge(C,n)*(i2,1)`1 by A16,GOBOARD5:3
       .= Gauge(C,n)*(i2,j2)`1 by A17,GOBOARD5:3;
    hence thesis by A4,A12,A14,PSCOMP_1:104;
   end;

  theorem
     for C be compact connected non vertical non horizontal Subset of TOP-REAL
2
   for k be Nat st 1 <= k & k+1 <= len Cage(C,n) &
    Cage(C,n)/.k = S-max L~Cage(C,n) holds
     (Cage(C,n)/.(k+1))`2 = S-bound L~Cage(C,n)
   proof
    let C be compact connected non vertical non horizontal
     Subset of TOP-REAL 2;
    A1: Cage(C,n) is_sequence_on Gauge(C,n) by JORDAN9:def 1;
    let k be Nat;
    assume that
     A2: 1 <= k and
     A3: k+1 <= len Cage(C,n) and
     A4: Cage(C,n)/.k = S-max L~Cage(C,n);
    A5: Cage(C,n)/.1 = N-min L~Cage(C,n) by JORDAN9:34;
    then A6: 1 < (N-max L~Cage(C,n))..Cage(C,n) by SPRECT_2:73;
      (N-max L~Cage(C,n))..Cage(C,n) <= (E-max L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:74;
    then A7: 1 < (E-max L~Cage(C,n))..Cage(C,n) by A6,AXIOMS:22;
      (E-max L~Cage(C,n))..Cage(C,n) < (E-min L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:75;
    then A8: 1 < (E-min L~Cage(C,n))..Cage(C,n) by A7,AXIOMS:22;
      (E-min L~Cage(C,n))..Cage(C,n) <= (S-max L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:76;
    then A9: (S-max L~Cage(C,n))..Cage(C,n) > 1 by A8,AXIOMS:22;
    A10: k < len Cage(C,n) by A3,NAT_1:38;
    then A11: k in dom Cage(C,n) by A2,FINSEQ_3:27;
    then (S-max L~Cage(C,n))..Cage(C,n) <= k by A4,FINSEQ_5:42;
    then A12: k > 1 by A9,AXIOMS:22;
    then consider i1,j1,i2,j2 be Nat such that
     A13: [i1,j1] in Indices Gauge(C,n) and
     A14: Cage(C,n)/.k = Gauge(C,n)*(i1,j1) and
     A15: [i2,j2] in Indices Gauge(C,n) and
     A16: Cage(C,n)/.(k+1) = Gauge(C,n)*(i2,j2) and
     A17: i1 = i2 & j1+1 = j2 or i1+1 = i2 & j1 = j2 or
         i1 = i2+1 & j1 = j2 or i1 = i2 & j1 = j2+1 by A1,A3,JORDAN8:6;
    A18: 1 <= i1 & i1 <= len Gauge(C,n) & 1 <= j1 & j1 <= width Gauge(C,n)
                                                            by A13,GOBOARD5:1;
    A19: 1 <= i2 & i2 <= len Gauge(C,n) & 1 <= j2 & j2 <= width Gauge(C,n)
                                                            by A15,GOBOARD5:1;
      Gauge(C,n)*(i1,j1)`2 = S-bound L~Cage(C,n) by A4,A14,PSCOMP_1:114
       .= Gauge(C,n)*(i1,1)`2 by A18,JORDAN1A:93;
    then A20: j1 <= 1 & j1 >= 1 by A18,GOBOARD5:5;
    then A21: j1 = 1 by AXIOMS:21;
    reconsider k'=k-1 as Nat by A11,FINSEQ_3:28;
      k >= 1+1 by A12,NAT_1:38;
    then A22: k' >= 1+1-1 by REAL_1:49;
    A23: k'+1 < len Cage(C,n) by A10,XCMPLX_1:27;
    then consider i3,j3,i4,j4 be Nat such that
     A24: [i3,j3] in Indices Gauge(C,n) and
     A25: Cage(C,n)/.k' = Gauge(C,n)*(i3,j3) and
     A26: [i4,j4] in Indices Gauge(C,n) and
     A27: Cage(C,n)/.(k'+1) = Gauge(C,n)*(i4,j4) and
     A28: i3 = i4 & j3+1 = j4 or i3+1 = i4 & j3 = j4 or
         i3 = i4+1 & j3 = j4 or i3 = i4 & j3 = j4+1 by A1,A22,JORDAN8:6;
    A29: k'+1 = k by XCMPLX_1:27;
    then A30: i1 = i4 & j1 = j4 by A13,A14,A26,A27,GOBOARD1:21;
    A31: 1 <= i3 & i3 <= len Gauge(C,n) & 1 <= j3 & j3 <= width Gauge(C,n)
                                                            by A24,GOBOARD5:1;
    A32: i3 = i4
    proof
     assume A33: i3 <> i4;
     per cases by A28,A33;
      suppose j3 = j4 & i3+1 = i4;
       hence contradiction by A20,A22,A23,A24,A25,A26,A27,A30,JORDAN10:3;
      suppose A34: j3 = j4 & i3 = i4+1;
         k' < len Cage(C,n) by A23,NAT_1:38;
       then Gauge(C,n)*(i3,j3) in S-most L~Cage(C,n)
                                        by A21,A22,A25,A30,A31,A34,JORDAN1A:81;
       then A35: (Gauge(C,n)*(i4+1,j4))`1 <= (Gauge(C,n)*(i4,j4))`1
                                               by A4,A27,A29,A34,PSCOMP_1:118;
         i4 < i4+1 by NAT_1:38;
       hence contradiction by A18,A30,A31,A34,A35,GOBOARD5:4;
    end;
      j1 = j2
    proof
     assume A36: j1 <> j2;
     per cases by A17,A36;
      suppose j1 = j2+1 & i1 = i2;
       hence contradiction by A19,A20,NAT_1:38;
      suppose A37: j1+1 = j2 & i1 = i2;
         k'+1+1 <= len Cage(C,n) by A23,NAT_1:38;
       then k'+(1+1) <= len Cage(C,n) by XCMPLX_1:1;
       then A38: LSeg(Cage(C,n),k') /\ LSeg(Cage(C,n),k) = {Cage(C,n)/.k}
                                                    by A22,A29,TOPREAL1:def 8;
         Cage(C,n)/.k' in LSeg(Cage(C,n),k') &
        Cage(C,n)/.(k+1) in LSeg(Cage(C,n),k) by A2,A3,A10,A22,A29,TOPREAL1:27;
       then Cage(C,n)/.(k+1) in {Cage(C,n)/.k}
                  by A16,A20,A25,A28,A30,A31,A32,A37,A38,NAT_1:38,XBOOLE_0:def
3;
       then Cage(C,n)/.(k+1) = Cage(C,n)/.k by TARSKI:def 1;
       hence contradiction by A13,A14,A15,A16,A36,GOBOARD1:21;
    end;
    then Gauge(C,n)*(i1,j1)`2 = Gauge(C,n)*(1,j2)`2 by A18,GOBOARD5:2
       .= Gauge(C,n)*(i2,j2)`2 by A19,GOBOARD5:2;
    hence thesis by A4,A14,A16,PSCOMP_1:114;
   end;

  theorem
     for C be compact connected non vertical non horizontal Subset of TOP-REAL
2
   for k be Nat st 1 <= k & k+1 <= len Cage(C,n) &
    Cage(C,n)/.k = W-min L~Cage(C,n) holds
     (Cage(C,n)/.(k+1))`1 = W-bound L~Cage(C,n)
   proof
    let C be compact connected non vertical non horizontal
     Subset of TOP-REAL 2;
    A1: Cage(C,n) is_sequence_on Gauge(C,n) by JORDAN9:def 1;
    let k be Nat;
    assume that
     A2: 1 <= k and
     A3: k+1 <= len Cage(C,n) and
     A4: Cage(C,n)/.k = W-min L~Cage(C,n);
    A5: Cage(C,n)/.1 = N-min L~Cage(C,n) by JORDAN9:34;
    then A6: 1 < (N-max L~Cage(C,n))..Cage(C,n) by SPRECT_2:73;
      (N-max L~Cage(C,n))..Cage(C,n) <= (E-max L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:74;
    then A7: 1 < (E-max L~Cage(C,n))..Cage(C,n) by A6,AXIOMS:22;
      (E-max L~Cage(C,n))..Cage(C,n) < (E-min L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:75;
    then A8: 1 < (E-min L~Cage(C,n))..Cage(C,n) by A7,AXIOMS:22;
      (E-min L~Cage(C,n))..Cage(C,n) <= (S-max L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:76;
    then A9: 1 < (S-max L~Cage(C,n))..Cage(C,n) by A8,AXIOMS:22;
      (S-max L~Cage(C,n))..Cage(C,n) < (S-min L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:77;
    then A10: 1 < (S-min L~Cage(C,n))..Cage(C,n) by A9,AXIOMS:22;
      (S-min L~Cage(C,n))..Cage(C,n) <= (W-min L~Cage(C,n))..Cage(C,n)
                                                           by A5,SPRECT_2:78;
    then A11: (W-min L~Cage(C,n))..Cage(C,n) > 1 by A10,AXIOMS:22;
    A12: k < len Cage(C,n) by A3,NAT_1:38;
    then A13: k in dom Cage(C,n) by A2,FINSEQ_3:27;
    then (W-min L~Cage(C,n))..Cage(C,n) <= k by A4,FINSEQ_5:42;
    then A14: k > 1 by A11,AXIOMS:22;
    then consider i1,j1,i2,j2 be Nat such that
     A15: [i1,j1] in Indices Gauge(C,n) and
     A16: Cage(C,n)/.k = Gauge(C,n)*(i1,j1) and
     A17: [i2,j2] in Indices Gauge(C,n) and
     A18: Cage(C,n)/.(k+1) = Gauge(C,n)*(i2,j2) and
     A19: i1 = i2 & j1+1 = j2 or i1+1 = i2 & j1 = j2 or
         i1 = i2+1 & j1 = j2 or i1 = i2 & j1 = j2+1 by A1,A3,JORDAN8:6;
    A20: 1 <= i1 & i1 <= len Gauge(C,n) & 1 <= j1 & j1 <= width Gauge(C,n)
                                                            by A15,GOBOARD5:1;
    A21: 1 <= i2 & i2 <= len Gauge(C,n) & 1 <= j2 & j2 <= width Gauge(C,n)
                                                            by A17,GOBOARD5:1;
    A22: len Gauge(C,n) = width Gauge(C,n) by JORDAN8:def 1;
      Gauge(C,n)*(i1,j1)`1 = W-bound L~Cage(C,n) by A4,A16,PSCOMP_1:84
       .= Gauge(C,n)*(1,j1)`1 by A20,A22,JORDAN1A:94;
    then A23: i1 <= 1 & i1 >= 1 by A20,GOBOARD5:4;
    then A24: i1 = 1 by AXIOMS:21;
    reconsider k'=k-1 as Nat by A13,FINSEQ_3:28;
      k >= 1+1 by A14,NAT_1:38;
    then A25: k' >= 1+1-1 by REAL_1:49;
    A26: k'+1 < len Cage(C,n) by A12,XCMPLX_1:27;
    then consider i3,j3,i4,j4 be Nat such that
     A27: [i3,j3] in Indices Gauge(C,n) and
     A28: Cage(C,n)/.k' = Gauge(C,n)*(i3,j3) and
     A29: [i4,j4] in Indices Gauge(C,n) and
     A30: Cage(C,n)/.(k'+1) = Gauge(C,n)*(i4,j4) and
     A31: i3 = i4 & j3+1 = j4 or i3+1 = i4 & j3 = j4 or
         i3 = i4+1 & j3 = j4 or i3 = i4 & j3 = j4+1 by A1,A25,JORDAN8:6;
    A32: k'+1 = k by XCMPLX_1:27;
    then A33: i1 = i4 & j1 = j4 by A15,A16,A29,A30,GOBOARD1:21;
    A34: 1 <= i3 & i3 <= len Gauge(C,n) & 1 <= j3 & j3 <= width Gauge(C,n)
                                                            by A27,GOBOARD5:1;
    A35: j3 = j4
    proof
     assume A36: j3 <> j4;
     per cases by A31,A36;
      suppose i3 = i4 & j3 = j4+1;
       hence contradiction by A23,A25,A26,A27,A28,A29,A30,A33,JORDAN10:2;
      suppose A37: i3 = i4 & j3+1 = j4;
         k' < len Cage(C,n) by A26,NAT_1:38;
       then Gauge(C,n)*(i3,j3) in W-most L~Cage(C,n)
                                        by A24,A25,A28,A33,A34,A37,JORDAN1A:80;
       then A38: (Gauge(C,n)*(i3,j3+1))`2 <= (Gauge(C,n)*(i3,j3))`2
                                                by A4,A30,A32,A37,PSCOMP_1:88;
         j3 < j3+1 by NAT_1:38;
       hence contradiction by A20,A33,A34,A37,A38,GOBOARD5:5;
    end;
      i1 = i2
    proof
     assume A39: i1 <> i2;
     per cases by A19,A39;
      suppose i1 = i2+1 & j1 = j2;
       hence contradiction by A21,A23,NAT_1:38;
      suppose A40: i1+1 = i2 & j1 = j2;
         k'+1+1 <= len Cage(C,n) by A26,NAT_1:38;
       then k'+(1+1) <= len Cage(C,n) by XCMPLX_1:1;
       then A41: LSeg(Cage(C,n),k') /\ LSeg(Cage(C,n),k) = {Cage(C,n)/.k}
                                                    by A25,A32,TOPREAL1:def 8;
         Cage(C,n)/.k' in LSeg(Cage(C,n),k') &
        Cage(C,n)/.(k+1) in LSeg(Cage(C,n),k) by A2,A3,A12,A25,A32,TOPREAL1:27;
       then Cage(C,n)/.(k+1) in {Cage(C,n)/.k}
                  by A18,A23,A28,A31,A33,A34,A35,A40,A41,NAT_1:38,XBOOLE_0:def
3;
       then Cage(C,n)/.(k+1) = Cage(C,n)/.k by TARSKI:def 1;
       hence contradiction by A15,A16,A17,A18,A39,GOBOARD1:21;
    end;
    then Gauge(C,n)*(i1,j1)`1 = Gauge(C,n)*(i2,1)`1 by A20,GOBOARD5:3
       .= Gauge(C,n)*(i2,j2)`1 by A21,GOBOARD5:3;
    hence thesis by A4,A16,A18,PSCOMP_1:84;
   end;

Back to top