Skip to main content

Part of the book series: Universitext ((UTX))

  • 7494 Accesses

Abstract

We define Morse functions and prove the basic results (existence and genericity, Morse lemma …)

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 79.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 99.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    The main advantage of this proof is that it gives a slightly more general result than the Morse lemma: near a critical point, a function can be written as the sum of a constant, a quadratic form of rank kn and a function in the remaining nk variables whose partial derivatives of orders 1 and 2 are all zero at the point in question.

  2. 2.

    The definitions “narrow passage between two mountains” or “depression forming a passage between two mountain summits” found in dictionaries are not much more precise.

  3. 3.

    If f is a function of two arguments (u,v)∈U×VE×F, then we let (D 1 f)(u,v) denote the linear map

    $$(D_1f)_{(u,v)}:E\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}$$

    that is the partial differential of f with respect to the first variable. The analogue holds for (D 2 f)(u,v).

References

  1. Laudenbach, F.: Topologie différentielle (1992). Cours à l’École Polytechnique

    Google Scholar 

  2. Milnor, J.: Morse Theory. Princeton University Press, Princeton (1963)

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Exercises

Exercises

Exercise 1

Let U be an open subset of R n and let \(f:U\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\) be a function. Let V be an open subset of R n and let \(\varphi:V\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow}U\) be a diffeomorphism. Compute (d 2(fφ)) y (for yV).

Let M be a manifold and let \(g:M\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\) be a function. Show that the bilinear form (d 2 g) x is well defined on the vector subspace

$$\operatorname{Ker}(dg)_x\subset T_xM. $$

Exercise 2

The cotangent vector bundle T V of the manifold V is endowed with its natural manifold structure (recalled in Appendix A). Verify that we can see the differential df of a function \(f:V\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\) as a section

$$df:V\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow}T^\star V $$

of the cotangent vector bundle (which is an embedding of V in T V).

What are the critical points of f in this terminology? Show that the point a is a nondegenerate critical point of f if and only if the submanifold df(V) is transverse to the zero section at the point in question.

Prove that a nondegenerate critical point of a function is isolated (without using the Morse lemma!).

Exercise 3

(Monkey Saddle)

We consider the function

$$\begin{aligned} f:\mathbf{R}^2&\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\\ (x,y)&\mathchoice{\longmapsto}{\mapsto}{\mapsto}{\mapsto}x^3-3xy^2. \end{aligned}$$

Study the critical point (0,0). Is it nondegenerate? Draw a few regular level sets of the function f as well as its graph.

Exercise 4

If \(f:V\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\) and \(g:W\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\) are Morse functions, then \(f+g:V\times W\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\) is also a Morse function whose critical points are the pairs of critical points of f and g.

Exercise 5

(On the Complex Projective Space)

By passing to the quotient, the function on C n+1−{0} defined by

$$f(z_0,\dots ,z_{n})=\frac{\sum_{j=0}^nj\left \vert z_{j}\right \vert ^2}{\sum_{j=0}^n\left \vert z_{j}\right \vert ^2} $$

induces a function on the complex projective space P n(C), which we still denote by f.

Verify that this is a Morse function whose critical points are the points [1,0,…,0] (of index 0), [0,1,0,…,0] (of index 2) and so on up to [0,…,0,1] (of index 2n).

Readers eager to know where this function may well have come from (from [54], of course, but before that?) can turn to Remark 2.1.12.

Exercise 6

(On the sphere and on the real projective plane)

With essentially the same formula, let us now consider the function

$$\begin{aligned} f:S^2&\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\\ (x,y,z)&\mathchoice{\longmapsto}{\mapsto}{\mapsto}{\mapsto}y^2+2z^2 \end{aligned}$$

and the resulting function \(g:\mathbf{P}^{2}(\mathbf{R})\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\) on the real projective plane that follows from it by passing to the quotient. Verify that f (and therefore g) is a Morse function and show that it has six critical points (and therefore that g has three):

  • Two points of index 0, the points (±1,0,0), at level 0

  • Two points of index 1, the points (0,±1,0), at level 1

  • Two points of index 2, the points (0,0,±1), at level 2.

Figure 1.7 shows a few level sets of this function on the sphere. The critical level set containing the two points of index 1 consists of the two circles defined by intersecting the sphere with the planes zx.

Fig. 1.7
figure 7

A Morse function on the real projective plane

Exercise 7

In a square, draw the level sets of the height function on the “inner tube” torus (the answer is somewhere in this book).

Exercise 8

(A projective quadric)

We consider the “quadric” Q in P 3(C) defined by the equation

$$z_0^2+z_1^2+z_2^2+z_{3}^2=0.$$

Show that Q is a compact submanifold of the manifold P 3(C) of (real) dimension 4.

Let us write z j =x j +iy j (for 0≤j≤3) and z=x+iy for zC 4, with x,yR 4. We fix two real numbers λ and μ and, for zC 4, we set

$$\widetilde {f}(z)=\lambda(x_0y_1-x_1y_0)+\mu(x_2y_{3}-x_{3}y_2).$$

Verify that for \(\left \vert u\right \vert =1\), we have

$$\widetilde {f}(uz)=\widetilde {f}(z)$$

and deduce from it that \(\widetilde {f}\) defines a function

$$f:\mathbf{P}^{3}(\mathbf{C})\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}.$$

Suppose that the real numbers λ and μ occurring in the definition of f satisfy 0<λ<μ. Let g be the restriction of f to Q. Show that g is a Morse function that has a local minimum, a local maximum and two critical points of index 2.

Exercise 9

In this exercise, we propose a different proof of the existence of numerous Morse functions on a manifold, and more precisely of Theorem 1.2.5:

  1. (1)

    Using, for example, Exercise 2, show that the set of Morse functions on the compact manifold V is open for the topology, and therefore also for the topology.

  2. (2)

    Let aV and let U be the open set of a chart centered at a. We let denote the space of differentials of the functions defined on U (or of exact 1-forms) , seen as a subset of (for every xU, (df) x is a linear functional on R n). Note that is not an open subset. Let α be a plateau function with support in U and value 1 on a compact neighborhood K of a. For , we let

    $$\begin{aligned} F:U\times(\mathbf{R}^n)^\star&\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\\ (x, A)&\mathchoice{\longmapsto}{\mapsto}{\mapsto}{\mapsto}f(x)+\alpha(x)A\cdot x. \end{aligned}$$

    ComputeFootnote 3 (D 1 F)(x,A) for xK and show that F is a submersion

    $$K\times(\mathbf{R}^n)^\star \mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow}(\mathbf{R}^n)^\star.$$

    Show that is locally transversal to \(\left \{0\right \}\) (in the sense of Subsection A.3.c).

  3. (3)

    Show that every point a of V admits a compact neighborhood K such that the set of functions \(f:V\mathchoice{\longrightarrow}{\rightarrow}{\rightarrow}{\rightarrow} \mathbf{R}\) whose critical points in K are nondegenerate is dense in .

  4. (4)

    Deduce a proof of Theorem 1.2.5 from this.

Rights and permissions

Reprints and permissions

Copyright information

© 2014 Springer-Verlag London

About this chapter

Cite this chapter

Audin, M., Damian, M. (2014). Morse Functions. In: Morse Theory and Floer Homology. Universitext. Springer, London. https://doi.org/10.1007/978-1-4471-5496-9_1

Download citation

Publish with us

Policies and ethics