- Page xiv, line 1:
**I**would like to thank... should be**We**would like to thank... - Page 4, line 16: increase with
**out**bound - Page 13, line 10 (and elsewhere on this page):
The partial of
**f**with respect to**x**is really the partial of_{0}**f**with respect to the "variable"**phi**. - Page 27, last line should read: That is,
**two nonzero**vectors**V**and**W**... - Page 38, Exercise 13 should read:
**A( alpha V**._{1}+ beta V_{2}) = alpha A V_{1}+ beta A V_{2} - Page 58, very interesting way to enumerate the phase portraits in
**Figure 3.9**, don't you think? - Page 59, Exercise 9 should read complex eigenvalues
**with nonzero imaginary part.** - Page 66: Note that, in the proof of the Theorem, the "only if" part is
missing. This part
follows immediately, however, from the assertion at the top of
page 66 regarding the fact that, if the eigenvalues are negative, then the
conjugacy takes curves tending to the equilibrium point to other such curves.
So if there is only one negative eigenvalue, then we only get a single curve
of such solutions, but if there are two eigenvalues with negative real parts,
we get a whole open set of such solutions and so there is no conjugacy.

- Page 86, last line: first component of
**V**should be**x**._{1}+ iy_{1} - Page 88, line 18 Section 5.1, not 5.2.
- Page 88, last line:
**W**should be_{2j}**E**_{2j} - Page 102. Fifth line of the proof should read "Since
**V**is open..."._{2}

- Page 111, line 3: in the
**k**component of the solution replace_{1}**c**with_{k}**c**._{k1} - Page 116. In the general solution
**Y(t)**at the bottom of the page, the terms**y**should each be multiplied by_{i}(t)**omega**_{i} - Page 118: The formulas for
**theta**and_{1}**theta**are wrong, they should read_{2}**theta**, a minus sign rather than a plus sign in each of these cases. This does not affect the Poincare map, however._{1}= theta_{1}(0) - omega_{1}(t) - Page 144, Equation 3 (as well as just above) should read:
**u**instead of_{k+1}(t)**u**. Similarly,_{r+1}(t)**u**instead of_{k}(s)**u**. Most importantly, inside the integral should read_{r}(s)**F(u**, not just_{k}(s))**u**._{k}(s) - Page 147. Remove the first
**K**in the Theorem. - Page 164, last equation:
**theta'**should be**1**, not**-1**. - Page 169, line 11:
**h**should be_{s}(y)**h**.^{s}(y) - Page 177, line 13: The bifurcation should have gone "the other
way", i.e., with
**a < a**._{0} - Page 178, line 3: The second term in the numerator should be
**G**, but of course this does not change anything since this entire term is zero because_{x}G_{xa}**G**_{x}= 0 - Page 192, line 6:
**x + or - 1**should read**x = + or -1**, the equals sign is missing.

- Page 195, the example: Technically, we have not shown that linearization
implies that the origin is unstable when
**epsilon > 0**. However, this follows immediately from the fact that the**x, y**-plane is invariant and the system is actually linear on this plane for all**epsilon**. - Page 196, line 2 and line 6: L' should be L'/2.
- Page 204, line -6: the words "isolated minimum" should be
replaced by "isolated critical point that is a minimum."
- Page 242 line -10: The proof follows from Corollary 5 in 10.6, not
Corollary 6.

- Page 251 line -10: The expression
**xM**should read_{x}+ yM_{y}**xM**._{x}+ yN_{y}

- Page 251 line -3: lies over mu should read to the left of nu.
- Page 295 last line: v_theta should be u_r
- Page 297, Exploration 13.8: The potential energy
**U(X)**should be the negative of what is written, i.e.,**U(X) = -1 / |X|^v**.

- Page 326, problem 9: This problem should read "show that the system is
not chaotic in the region where
**x, y**, and**z**are all positive." - Page 332, figure 15.4: The labels of the graphs should be
**y = f(x)**and**y = f**instead of^{2}(x)**y = xf(x)**and**y = xf**^{2}(x) - Page 333, figure 15.5: The captions should read
**f**rather than_{c}(x)**f**. Also, there is a missing parenthesis in the first displayed equation._{c(x)} - Page 335, figure 15.6: The captions should read
**f**rather than_{c}(x)**f**._{c(x)} - Page 337, figure 15.7: In the caption the minus sign should be an equals
sign.

- Page 347, line 19:
**t**should be_{n+1}**s**._{n+1} - Page 348, line -4: the superscripts should be
**n+1**, not**n**. - Page 379. The third equation in the Chua circuit system should
read
**z' = - by**, not**-bz**.

Special thanks to the following people for catching the above errors:

- Jamil Abreu
- Larry Baker
- Jeff Haag and his students at Humboldt State
- Kresimir Josic and his students at the University of Houston
- John Josef Leth
- Benjamin Lotto
- Tiennyu Ma
- Luis Fernando Mello
- Bogdan Przeradzki
- Charles Pugh
- Jeffrey Rabin
- Martin Raussen
- Julian Wilson