06.07.2013 Views

The Picard-Lefschetz theory of complexified Morse functions 1 ...

The Picard-Lefschetz theory of complexified Morse functions 1 ...

The Picard-Lefschetz theory of complexified Morse functions 1 ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Complexified <strong>Morse</strong> <strong>functions</strong> 45<br />

<strong>The</strong>n we will define<br />

π = π0#π2#π4, E = E0#E2#E4.<br />

<strong>The</strong> main step is to specify exact symplectic identifications<br />

Ψ02 : π −1<br />

0 (c0 + 1) −→ π −1<br />

2 (c2 − 1), and<br />

Ψ24 : π −1<br />

2 (c2 + 1) −→ π −1<br />

4 (c4 − 1).<br />

First, Ψ24 is easy because, for 0 < s ≤ 1, there are canonical maps<br />

So we set<br />

ρ 2 s : π −1<br />

2 (c2 + s) −→ M2,<br />

ρ 4 −s : π −1<br />

4 (c4 − s) −→ M2.<br />

Ψ24 = (ρ 4 −1 )−1 ◦ ρ 2 1 .<br />

<strong>The</strong> next lemma will tell us how to define Ψ02.<br />

Lemma 6.1 For 0 < s ≤ 1, there is a canonical exact symplectic isomorphism<br />

Pro<strong>of</strong> Note that<br />

ν 2 −s : π −1<br />

2 (c2 − s) −→ M.<br />

M2 \ (∪ j=k<br />

j=1 φ (L j<br />

2 )′(D r/2(T ∗ S 3 ))<br />

is canonically isomorphic to<br />

M \ (∪ j=k<br />

j=1φL j(Dr/2(T<br />

2<br />

∗ S 3 )).<br />

From now on we will identify these. <strong>The</strong>n, by definition, π −1<br />

2 (c2 − s) is equal to the<br />

quotient manifold obtained from the disjoint union:<br />

M \ (∪ j=k<br />

where for each j we identify<br />

with<br />

using the gluing maps<br />

j=1φL j(Dr/2(T<br />

2<br />

∗ S 3 )) ⊔ [⊔ j=k<br />

j=1 [(π2 loc) j ] −1 (c2 − s)]<br />

Uj = (Φ 2 −s) −1 (D (r/2,r](T ∗ S 3 )) ⊂ [(π 2 loc) j ] −1 (c2 − s)<br />

Wj = φ j(D<br />

L2 gS3 [r/2,r] (T∗S 3 )) ⊂ M,<br />

(φ j<br />

L ◦ σπ/2 ◦ Φ<br />

2<br />

2 −s)|Uj : Uj −→ Wj.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!