The following typographical or algebraic errors are known. They are given by page.
Corrections in book text, figures, or index  
page  phrase (or location):  should be: 
20  We will consider a building up  We will consider building up 
22  (in Figure 3.3)  The branch of the top
twospecies tree where species c is added to get the center threespecies tree is the wrong one. 
66  there are not or more two states that both occur (this one was different in the first edition)  there are not two or more states that occur 
107  homplasious  homoplasious 
113  that the probability becomes 1  the probability becomes 1 
119  in both bases the curves  in both cases the curves 
129  with affinities to schmoos  with affinities to shmoos 
132  (in figure caption) cytochrome  globin 
152  Gascuel (1997)  Gascuel (1997b) 
157  it will certainly rise linearly  it will certainly not rise linearly 
170  Gascuel (1997)  Gascuel (1997a) 
174  Gascuel, Desper, and Denis  Gascuel, Bryant, and Denis 
178  more then four  more than four 
178  D_{il}+D_{ij}  D_{il}+D_{jk} 
179  (in equation 12.7) max  min 
179  should not be the largest  should be the smallest 
188  (in Table 12.2)  The numbers 10 in the bottom three rows should be 5. 
226  only between amino acids can be reached  only between amino acids that can be reached 
236  more defensible that it proved  more defensible than it proved 
241  high rates of change has  high rates of change have 
242  μt = 10  μt = 5 
250  (in caption of Fig. 16.1) series of 13 independent  series of 11 independent 
269  states will be spend  states will spend 
269  as as it will with a  as they will with a 
269  and site being variable  and the site being variable 
310  (in equation 19.7)  the second θ_{0} should be θ 
310  (in equation 19.7)  C should be (C) 
310  (in equation 19.8)  C should be (C) 
310  that can takes values  that can take values 
313  have been used give a good approximation  have been used to give a good approximation 
318  (in the Figure)  quantities T should be R 
327  noted, in Chapter 16 that  noted in Chapter 16 that 
328  for each these simulated data sets  for each of these simulated data sets 
337  we would back get our original data  we would get back our original data 
340  0.68066 of the time  0.680905 of the time 
344  have randomly assigned rate  each have a randomly assigned rate 
348  (the onetailed 95% point  (the twotailed 95% point 
349  even greater when consider  even greater when we consider 
351  that lead to  that led to 
352  with more then R^{3} replicates  with more than R^{3} replicates 
355  we find ABCEDF  we find ABCDEF 
371  Data sets generated  Data sets are generated 
386  seven degrees of freedom  six degrees of freedom 
386  all seven  all six 
388  Then number of equations  The number of equations 
395  tree has the same topology as Figure 23.2  tree has the same topology as Figure 23.1 
396  The tree has the same topology as Figure 23.1.  (delete the sentence) 
402  ith character in the jth population  jth character in the ith population 
403  an p × p array  a p × p array 
402403  The values of all n characters in population 1 are on top. Below them are the values for all n characters in population 2, and so on. 
The values for all n populations of character 1 are on top. Below them are the values for all n populations of character 2, and so on. 
403  (in Figure 23.34) V^{(n)}  V^{(p)} 
403  (in equation 23.36) (mean vector 0 should be)  (a vector μ Kroneckerproduct with the identity matrix I) 
403  an (n1)p × (n1)p matrix  an (n1)p × np matrix 
410  Gegenbaur  Gegenbauer 
418  2N times the variance of change  N times the variance of change 
425  It seem that there is  It seems that there is 
430  reversion to state 1  reversion to state 0 
435  1.11666  1.1666 
436  1.11666  1.1666 
461  (in equation 26.11)  both occurrences of N_{e} should be N_{0} 
471  two occurrences of k_{i}  should both be k_{li} 
471  two occurrences of m_{jl}  should both be m_{lj} 
472  n1, n2, ..., 2 lineages  n, n1, ..., 2 lineages 
512  in its immediate ancestors  in its immediate descendants 
512  shows three unrooted trees  shows two rooted trees 
514  all possible number of loci  all possible numbers of loci 
521  placement of their root. the two trees  placement of their roots, the two trees 
559  test hypothesis about evolution  test hypotheses about evolution 
563  tips in the tree, or Nbar its mean  tips in the tree, or Nbar, its mean 
563  compared it to a the probabilities  compared it to the probabilities 
568  (in equation 33.6) ..., t_{n}  ..., t_{n1} 
576  lowermost ancestor is used  lowermost descendant is used 
588  one calls function traverse or Traverse  one calls function postorder or Traverse 
604  (Edwards 1971 reference) The likelihood for multinomial proportions under stereographic projection. Biometrics 28: 618620 
Distance between
populations on
the basis of gene frequencies. Biometrics 27: 873881. 
611  (the two listings for Gascuel, 1997)  the first should be 1997a, the second 1997b 
614  (Reference to Hein, J. 1989), title should be:  A new method that simultaneously aligns and reconstructs ancestral sequences for any number of homologous sequences, when the phylogeny is given. 
616  (Reference to Hudson, R. 1983 in Evolution)  This reference needs to start on its own (negatively indented) line. 
627  (Reference to Nielsen, R. 2002)  volume number needs to be in boldface type 
633  (Reference to Sankoff, Morel, and Cedergren, 1973) 
volume number needs to be in boldface type 
638  Swofford, D. L. and D. R. Maddison  Swofford, D. L. and W. P. Maddison 
639  (Thompson 1972 reference) Distance between populations on the basis of gene frequencies. Biometrics 27: 873881. 
The likelihood for multinomial proportions under stereographic projection. Biometrics 28: 618620 
658  indexing of Ronquist, 2000  (should have been on
page 659 among the other Ronquist page numbers) 
The following typographical errors are found in the first and second printings. They are corrected in the third printing.
Corrections in book text, figures, or index  
page  phrase (or location):  should be:  
2  Figure 1.4 shows the single reconstruction  Figure 1.4 shows the two reconstructions  
2  (in Figure 1.1)  Raise label Delta to be level with the others  
33  In Table 3.5, many of the lines for even numbers of species were computed incorrectly  The corrected lines are (showing only the ones that have changed):
 
35  now [(n1)1][(n2)1]/2 =  now [n1][(n1)1]/2 =  
102  they arive at  they arrive at  
113  as long as q < 1/2, p > 0, and q > 0  as long as q < 1/2, and either p > 0 or q > 0  
123  carried our by a computer  carried out by a computer  
134  prefer not to assume either reversibility  prefer not to assume either irreversibility  
155  (in equation 11.14) D  d  
155  (in equation 11.15) D  d  
156  of length t  of time t  
156  of length dt each  of time dt each  
161  is then the average of the distances  is then half the average of the distances  
161  is the average of the distances  is half the average of the distances  
172  the neighborJoining methods  the neighborjoining methods  
177  see the papers by Gascuel  see the paper by Gascuel  
200  (in equation 13.4) Q^{n2}  (^{1}/_{2} Q)^{n2}  
203  (In numerator of second line of equation 13.11)  the epsilon should be the same style as the one in the denominator  
213  all the bases frequencies  all the base frequencies  
221  Rzhetsky, 1995, Tillier, and Collins, 1998  Rzhetsky, 1995; Tillier and Collins, 1998  
229  Pagel (1995)  Pagel (1994)  
245  (in equation 15.28) max  min  
261  amounts of diveregence  amounts of divergence  
279  each entry in the this vector  each entry in this vector  
294  Figure 18.2 should have its upperright circle open to the path that comes in from its lower left.  
332  the the confidence limits  the confidence limits  
355  if we could Type I  if we count Type I  
382  the triple of species a, b, and c  the triple of species c, d, and e  
410  the OrnsteinUhlenbeck process (which will be described later in this chapter)  the OrnsteinUhlenbeck process (which will be described in the next chapter)  
418  (in equation 24.6) Var[gbar]  Var[Δ gbar]  
430  (in equation 24.18) The last integral should be from infinity to 0.  
438  x_{1}x_{2} multiplied by  x'_{1}x'_{2} multiplied by  
491  they give is no way  they give no way  
500  Wang and Jiang (1994  Wang and Jiang (1993)  
532  (in Figure 30.8) {BC  ADEFG} 0.2 0.2  {BC  ADEFG} 0.2 0.3  
532  (in equation 30.1)  0  0.2  +  0.4  0.3  +  0.3  0  +  0.1  0  +  0.05  0.2  +  0.2  0.1  = 0.95   0  0.2  +  0.4  0.3  +  0.2  0.3  +  0.3  0  +  0.1  0  +  0.05  0.2  +  0.2  0.1  = 1.05  
532  (in equation 30.2) (0  0.2)^{2} + (0.4  0.3)^{2} + (0.3  0)^{2} + (0.1  0)^{2} + (0.05  0.2)^{2} + (0.2  0.1)^{2} = 0.1825  (0  0.2)^{2} + (0.4  0.3)^{2} + (0.2  0.3)^{2} + (0.3  0)^{2} + (0.1  0)^{2} + (0.05  0.2)^{2} + (0.2  0.1)^{2} = 0.1925  
563  tree in Figure 33.1 this is 7/72  tree in Figure 33.1 this is 14/72  
578  will attached be three or more subtrees  will be three or more subtrees attached  
581  the branches leading to D and B  the branches leading to D and E  
586  composed to records  composed of records  
587  in turns of itself  in terms of itself  
591  <clade length=0.6> <clade length=0.102><name>A</name> <clade length=0.23><name>B</name> and <clade length=0.4><name>C</name></clade>  <clade length="0.6"> <clade length="0.102"><name>A</name></clade> <clade length="0.23"><name>B</name></clade> and <clade length="0.4"><name>C</name></clade>  
593  The URL given for TNT is outdated  It should instead be: http://www.zmuc.dk/public/phylogeny/tnt  
615  (in Hendy and Penny 1982 reference) 60: 133142.  59: 277290. 
The first printing contains the above errors plus these:
Corrections of page numbers in Table of Contents entries  
page  in entry:  page number should be: 
v  Tree shapes, p. 28  p. 29 
vi  Genetic algorithms, p. 44  p. 45 
vi  Tree space, p. 48  p. 50 
vi  NPhardness, p. 57  p. 59 
vii  Nonsuccessive algorithms, p. 84  p. 85 
vii  Observed fractions of the patterns, p. 110  p. 111 
vii  CaminSokal and parsimony, p. 128  p. 129 
viii  Hypotheticodeductive, p. 138  p. 139 
viii  Logical parsimony?, p. 140  p. 141 
viii  The JukesCantor model  an example, p. 155  p. 156 
ix  Codonbased models, p. 225  p. 226 
xi  Bayesian methods for phylogenies, p. 289  p. 291 
xi  Flat priors and doubts about them, p. 301  p. 302 
xi  Testing assertions about parameters, p. 311  p. 312 
xii  Parsimonybased methods, p. 322  p. 323 
xii  Tests for three species with a clock, p. 329  p. 330 
xii  Independence of characters, p. 342  p. 343 
xii  Permutation tests, p. 358  p. 359 
xii  The SH test, p. 369  p. 370 
xiii  Symmetry invariants, p. 374  p. 375 
xiii  Finding all invariants empirically, p. 387  p. 388 
xiii  The REML approach, p. 400  p. 401 
xiii  Using approximate Brownian motion, p. 411  p. 412 
xiv  Selection for an optimum, p. 420  p. 421 
xv  Fu's method, p. 484  p. 485 
xv  Methods of inferring the species phylogeny, p. 490  p. 491 
xv  Parsimony method, p. 497  p. 498 
xv  Reconciled trees, p. 509  p. 510 
xvi  Strict consensus, p. 521  p. 522 
xvi  The symmetric difference, p. 528  p. 529 
xvi  Stratocladistics, p. 549  p. 550 
xvii  Index, p. 644  p. 645 
Corrections in book text, figures, or index  
page  phrase (or location):  should be:  
xiii  23 Brownian motion and gene frequencies  23 Brownian motion and gene frequencies  
13  only A or G are possible  only A or C are possible  
15  In Figure 2.2, the fourth tip has: infinity, infinity, infinity, 0  should be: 0, infinity, infinity, infinity  
15  these total to 2.5  these total to 4.5  
16  Wheeler and Nixon (1990)  Wheeler and Nixon (1994)  
23  (2n3)!  2^{n1} (n1)!  (2n3)!  2^{n2} (n2)!  
24  rise to a rooted bifurcating trees  rise to rooted bifurcating trees  
29  Edwards and CavalliSforza  CavalliSforza and Edwards  
30  (in Table caption): Edwards and CavalliSforza  CavalliSforza and Edwards  
35  connected with of genes  connected with coalescent trees of genes  
36  program and want to know  program and wants to know  
38  because they sieze the first  because they seize the first  
42  (in figure caption) separates B and C  separates G and C  
44  304  288  
62  (which requires 3 changes)  (which requires 5 changes)  
63  (in Figure 5.4) bottommost node has number 3  should be 5  
66  there are not two states that both occur  there are not two or more states that occur  
69  reconstruction of states at interior nodes is  reconstructions of states at interior nodes are  
73  any one of them are present  any one of them is present  
75  (in middle row of equation 7.1) x_{l} ≤ x_{r}  x_{l} ≤ x ≤ x_{r}  
79  (in figure caption) which shows that  that shows which  
88  (in Table 8.2)  the table should be transposed so that the 01 combination is the one missing  
92  Table 1.1  Table 8.1  
92  {Alpha, Delta, Gamma}  {Alpha, Gamma, Delta}  
94  if there is missing data  if there are missing data  
99  index i in product in equation 9.3  index should be j  
102  pioneering attempts  pioneering attempt  
103  branch that applies to all characters  character that applies to all branches  
105  so do the number of parameters  so does the number of parameters  
105  this connected between  this connection between  
106  In equation 9.12, the expression is not correct. The product over branches should not be there at all. The top and middle cases also need an extra factor of 1/2. The assertions below the expression are still correct, however!  
112  and has presumable been  and has presumably been  
112  (in Equation 9.16) P_{1100}  P_{0011}  
112  1100, but there is also the pattern 0011  0011, but there is also the pattern 1100  
115  that there is probability  there is probability  
117  as only parallel changes  as parallel changes  
142  Sober chooses former  Sober chooses the former  
143  when the data is randomized  when the data are randomized  
144  regard the experiment as repeatible  regard the experiment as repeatable  
150  Equation 11.4  Before the big left parenthesis there should be a term of x_{ij,k}  
151  (in equations 11.8 and 11.9) D  d  
152  (in equations 11.11 and 11.12) D  d  
152  11.10 (or 11.7)  X^{T}X (or X^{T} W X)  
152  11.12 (or 11.9)  11.9 (or 11.12)  
161  indexclusteringclustering  clustering  
163  The two distance matrices on this page are identical, when they should not be.  The first should have no boldfacing, boxes or asterisks. It should be placed after "branch length" in line 3. The second should be placed after "borders of the table".  
169  indexclusteringclustered  clustered  
173  extent to which the data departs  extent to which the data depart  
180  it could still be fairly slow  it could still be fairly fast  
181  likelihood methods infer from this data  likelihood methods infer from these data  
183  {ABCDE  EFGHI}  {ABCD  EFGHI}  
183  is discussed is their paper  is discussed in their paper  
186  this data with a parsimony method  these data with a parsimony method  
191  for the finding it  for finding it  
191  and, if they are compatible finds  and, if they are compatible, finds  
191  the same as the any  the same as any  
197 

 
204  pi_{i} in line 11  the pi should be the Greek letter π instead  
204  first n_{i} in equation 13.12  m_{i}  
204  many C's at one end a branch  many Cs at one end of a branch  
204  and many T's at the other  and many Ts at the other  
204  the C's were in the ancestor  the Cs were in the ancestor  
204  the T's in the descendant  the Ts in the descendant  
206  eigenvectors lambda_{i} and eigenvalues  eigenvalues lambda_{i} and eigenvectors  
206  (in lowerright corner of first matrix in equation 13.18) π_{A}  π_{T}  
207  to give Table 13.5  to give Table 13.4  
208  as in Table 13.5  as in Table 13.4  
210  such as Table 13.5  such as Table 13.4  
213  approximate,du  approximate,  
215 

 
215  remarkable level  remarkably level  
215  power of 1.0  power of 2.0  
215  w_{ij} = 1/D_{ij}  w_{ij} = 1/D_{ij}^{2}  
217  x^{α1}e^{x/β}  r^{α1}e^{r/β}  
220  A's, B's and C's  As, Bs and Cs  
220  such as exp(brt)  such as exp(bt)  
220  E[e^{brt}]  E_{r}[e^{brt}]  
220  difference between sequences (horizontal axis)  difference between sequences (vertical axis)  
225  is contributed by assignments  contributed by the assignments  
226  (CGT, CGC)  (AGC, AGT)  
232  (in equation 15.3) max[0,lk]  max[0,kl]  
232  The Kimura 2parameter does this  The Kimura 2parameter model does this  
233  factor of 64  factor of 16  
233  2 x 10^{8}  3 x 10^{5}  
238  (equation 15.22)

 
238  we can use numerical methods to iterate Q in equation (15.22) until the expected fraction of restriction fragments equals the observed fraction F,  we can note that equation (15.22) is a quadratic equation in Q and solve it for Q,  
241  concerned with of genes  concerned with coalescent trees of genes  
242  (in equation 15.26, twice) k=infinity  k=0  
244  it does has some robustness  it does have some robustness  
245  (in equation 15.28) μ t √(2 π)  √(μ t) √(2 π)  
249 

 
250  Thus is does not show  Thus it does not show  
253  in equation 16.11  in equation 16.10  
253  Equation 16.11 can be  Equation 16.10 can be  
256  (at end of equation 16.17)  add these factors: L_{6}^{(i)}(y) L_{8}^{(i)}(z)  
256  (at end of equation 16.19)  add these factors: L_{6}^{(i)}(y) L_{8}^{(i)}(z)  
257  [figure caption] around nodes 6 and 8  around branch 7  
257258  all occurences of node 6 or branch length t_{6}  ... should be node 7 or branch length t_{7}  
259  In equation 16.11 that would  In equation 16.10 that would  
263  (in equation 16.28) r_{ij}, r_{ij+1}, ..., r_{im}  r_{ij}  
264  in equation 16.29, before the summation sign add a factor Prob(D^{(j)}  T, r_{ij})  
265  sometimes too more flexibility  sometimes more flexibility  
268  C0 G0  G0 C0  
269  Using Gu's model,  Using a similar model,  
271  number of character patters  number of character patterns  
273  showed a particular interesting  showed a particularly interesting  
276  H_{3} = H_{2}⊗H_{2}  H_{3} = H_{1}⊗H_{2}  
277  branch in common.  branch in common, their path lengths add to these sums.  
279  r_{1}+r_{2}+r_{3}+r_{4} Note  r_{1}+r_{2}+r_{3}+r_{4}. Note  
280  (equation 17.12) q  g  
280  (equation 17.13) q  g  
283  are denoted by q_{i}  g_{i}  
283  (equation 17.15) q_{i}  g_{i}  
284  those that have nonzero q_{i}  those that have nonzero g_{i}  
284  minimizing C in equation 17.15  minimizing Γ^{2} in equation 17.15  
289  Now some data is examined.  Now some data are examined.  
289  less influence with 22 tosses  less influence with 44 tosses  
289  data is half as probable as it is  data are half as probable as they are  
291  Gomberg (1966)  Gomberg (1968)  
293  drawn will be less that R  drawn will be less than R  
300  followed by (Yang and Rannala, 1997) by  also be followed (Yang and Rannala, 1997) by  
309310  the symbol for the vector theta should be boldface throughout  
310  we look this up on  we look this up in  
312  (in legend of Figure 19.1) hightest likelihood  highest likelihood  
314  Thus we can find  Thus, if we can find  
315  branch lengths of evolutionary rates  branch lengths or evolutionary rates  
316  penalize it by subtracting twice  penalize it by adding twice  
319  one tree on that we see  one tree on which we see  
328  whether a two data sets  whether two data sets  
330  the closest fit to this data  the closest fit to these data  
332  equal numbers changes in interior  equal numbers of changes in interior  
333  (in equation 19.19)

 
333  p_{1} = q_{1} = 1/3  p_{2} = q_{2} = 1/3  
333  alternative values of p_{2}  alternative values of p_{1}  
340  between the delete1/e jackknife and the delete1/e jackknife  between the delete1/e jackknife and the bootstrap  
347  In Figure 20.4, the central curve is too dark, and the rightmost curve has its left part too dark  
350  triple bootsrap  triple bootstrap  
356  rather that distance between  rather than distance between  
364  their parsimony or likelihood or likelihood scores  their parsimony or likelihood scores  
364  PHYLIP  (Should be in regular Palatino font)  
365  Heading "An example"  Should be a section head: Roman font, bold face  
368  13514  13,514  
375  2415  2,415  
375  62,741  (shift it half a digit to the right)  
376  equal numbers of A's, C's, G's, and T's  equal numbers of As, Cs, Gs, and Ts  
381  (in equation 22.17, first line) P_{xyyx}  P_{xyxz} (terms in equation 22.17 have also been arranged in a more logical order)  
393  Gomberg (1966)  Gomberg (1968)  
393  (in equation 23.4) σ^{2} v_{12}+σ^{2} v_{11}+σ^{2} v_{7}  σ^{2} v_{12}+σ^{2} v_{11} +σ^{2}v_{10}+σ^{2} v_{7}  
394  v_{4} in Figure 23.1  the v should be italic font  
395  species 1 by x_{i1}  species 1 by x_{1i}  
395  species 2 by x_{i2}  species 2 by x_{2i}  
397  (in equation 23.15)

 
398  (in equation 23.16)



402  (in equation 23.33) x  x (boldfaced)  
413 

 
417  frac1N  ^{1}/_{N}  
418  V_{A}/(2N)  V_{A}/N  
418  (in equation 24.6) (2NV_{M})/(2N) = V_{M}  (2NV_{M})/N = 2V_{M}  
418  N times the variance of change  2N times the variance of change  
424  1973b  1973a  
435  (in equation 25.1) Y_{2} = y_{2}/square root(0.75)  Y_{2} = y_{2}/square root(0.975)  
439  (in equation 25.2) A⊗T + E⊗I  T⊗A + I⊗E  
439  matrix, E⊗I  matrix, I⊗E  
440  (in equation 25.3) I⊗μ, A⊗T + E⊗I  1⊗μ, T⊗A + I⊗E  
457  have any number lineages  have any number of lineages  
461  of equation 26.9  of equation 26.10  
466  3 x 2 / 10^{6} = 0.000006  3 x 2 /(4 x 10^{6}) = 0.0000015  
466  4 x 3 / 10^{6} = 0.000012  4 x 3 / (4 x 10^{6}) = 0.000003  
466  0.000012+0.00000508  0.000003+0.00000608  
467  ancestry of below  ancestry below  
471  were one  were once  
472 

 
475  Prob(G  Θ_{0}) at start of line in middle of page  Prob(G  Θ)  
481 

 
482  Heading "MCMC for a variety of coalescent models"  Should be a section heading: Roman font, bold face  
486  the data was represented  the data were represented  
488  the copies from B and C are  the copies from A and C are  
488  than either is to the copy from A  than either is to the copy from B  
494  the N  i would actually  the N_{i} would actually  
498  (in Figure caption): are shown by by the hash marks  are shown by the hash marks  
499  an alignment of of position  an alignment of position  
501  all the G's on the same side  all the Gs on the same side  
501  the sequence in the two parts of the tree  the sequences in the two parts of the tree  
504  Two T's are inserted to its right.  Two Ts are inserted to its right  
504  Two T's are deleted.  Two Ts are deleted.  
504  T's in the middle of the sequence  Ts in the middle of the sequence.  
507  adjacent A's are deleted  adjacent As are deleted  
507  turn out to be A's,  turn out to be As,  
525  Heading: "A dismaying result"  Should be a subsection heading: Italic font and not boldfaced  
531  based on NNI's  based on NNIs  
532  ranch lengths  branch lengths  
544  parasite species of a species  parasite species or a species  
548  in Figure 32.1, for each large ellipse the leftmost small ellipse within it is too dark  
557  (in Figure 32.6)

 
557  exp[  exp[  
561  probability that then urn  probability that the urn  
578  subtrees of size 5, 7, and 1  subtrees of size 8, 4, and 1  
578  these are species AE, FL, and M  these are species AH, IL, and M  
578  angles 2.4166, 3.33833, and 0.24166  angles 3.8666, 1.9333, and 0.48332  
578  Thus the fivespecies subtree has a branch length 10  Thus the eightspecies subtree has a branch length 13  
578  one with four species and one with one  one with five species and one with three  
578  for the fivespecies subtree  for the eightspecies subtree  
578  branch of length 10  branch of length 13  
578  Thus the fivespecies subtree  Thus the eightspecies subtree  
578  starts with an angle of 2.4166 radians  starts with an angle of 3.8666 radians  
578  moves out a distance 10  moves out a distance 13  
578  with four and one species  with five and three species  
578  2.4166 radian sector  3.8666 radian sector  
578  branch of length 10  branch of length 13  
578  one of 1.9333 and the other of 0.48332  one of 2.4166 and the other of 1.4500  
578  lengths 3 and 21  lengths 10 and 13  
582  dashed curves in Figure 34.6 need to connect more accurately with dashed lines  
612  Gomberg, D. 1966.  Gomberg, D. 1968.  
619  Kishino, H., T. Miyata and  Kishino, H., T. Miyata, and  
621  (in Lande 1976 reference) 31: 442446.  30: 314334.  
623  (in reference for Lundy, 1985) 198 .  198.  
627  (in Nielsen et al. 1998 reference): maximumlikelihood  maximumlikelihood  
630  Purvis and Garland reference  It is out of alphabetical order. It should follow Purdom et al.  
641  (add reference after Wakeley 1998:)  Wald, A. 1949. Note on the consistency of the maximum likelihood estimate. Annals of Mathematical Statistics 20: 595601.  
644  posledobatel'nostyam. (in Russian)  posledovatel'nostyam (in Russian).  
645  (after Angielczyk entry)  Antonov, 631  
646  Bousquet, 325  Bousquet, 325, 622  
648  Cucumel, 191, 527  Cucumel, 191, 527, 621  
651  Hannenhalli, 516, 614  Hannenhalli, 516, 597, 614  
654  Landrum, 95  Landrum, 95, 605  
655  Margoliash, 131, 132, 134, 148, 222  Margoliash, 131, 132, 134, 148, 222, 609  
655  Markowitz, 217, 268  Markowitz, 217, 268, 609  
655  Meacham, 92, 95, 530, 578, 590, 624  Meacham, 92, 95, 530, 578, 590, 591, 605, 624  
660  Schoenberg, xx, 51  Schoenberg, xx, 40, 51  
660  (after Solovyov entry)  Soules, 597  
663  Wald, 269272  Wald, 269272, 641  
664  Zharkikh, 65, 66, 120, 213, 346, 349, 351, 631, 644  Zharkikh, 65, 66, 120, 213, 346, 349, 351, 622, 631, 644 
These typos are corrected in the second printing.
I'd appreciate being told of any other typos that are discovered.
There will be a chance to correct them at the next reprinting.
Thanks for pointing out some of these to:  
 
I am particularly grateful to John Trueman, Simon Ho, Peter Morcos, and Nik Schueler, who reported many of these typos. 
Of course the book omits all work after early August, 2003. But some earlier works were not discussed that should have been. Particularly egregious omissions include:
As for incorrect arguments, there will be those who say most of the book is incorrect. Obviously I don't agree. The arguments I made that may be wrong are:
We could check whether all quartets of species satisfy the FPM condition, but this does not guarantee that there is one tree that implies all of those individual FPM conditions.is blatantly incorrect. If all possible FPM conditions are satisfied, it has been known at least since Buneman's work that there must be a tree that implies all of them. A better statement would have been:
We could check whether all quartets of species satisfy the FPM condition; unless all possible quartets satisfy it, there is no guarantee that there is one tree that implies all of the individual FPM conditions.
In their trees the long branches are in adjacent pairs. This results in parsimony doing better than likelihood. As they lengthen the long branches, both do worse, as expected. At a certain length, parsimony falls prey to a long branch attraction effect, causing likelihood to outperform it beyond that length.There are really two effects of long branch attraction in their simulation, one which brings adjacent branches together, and one, which occurs more when branches are longer, which attracts more distant long branches. This is left fairly unclear by the current wording.
Murtagh, F. 1984. Complexities of hierarchic clustering algorithms: state of the art. Computational Statistics Quarterly 1: 101113.for which see scans of the paper at this URL.
There has been thre further printing, and there will be additional reprintings as necessary, with correction of typos. I don't think there is ever likely to be a full second edition. Work on phylogenies is growing rapidly, and the second edition would have to be twice the size. Aside from whether people would buy or read such a massive tome, there is the issue of who has the energy to write it. I don't. Followup books probably either will have to be more narrowly focussed on particular subfields, or will skim more lightly through the literature, citing much less of it.
It is possible that a Revised Edition might be produced, correcting the above errors, adding a smallish amount of material to bring the coverage more up to date, while compensating for the extra length by dropping coverage of some of the more obscure methods. We're thinking about this.
Those interested in downloading the data sets and trees that were used for the examples in the book should look at the web page here.
A list of reviews of the book that have appeared, and some reactions to them, will be found here.
Joe Felsenstein
