ORNITHOLOGICAL LITERATURE 
653 
Chapter 12 sets forth a scenario to explain the 
evolution of W-c G. This involves a wave of 
distribution, first from west to east, starting with 
the origin of the complex, perhaps before the first 
Pleistocene glaciations, and from south to north as 
glaciers retreated. Having reached the east, the 
population wave rebounded back westward. As 
the evolving populations moved they would find 
already occupied habitat or collide with a 
population moving another direction, and might 
Dot find suitable unoccupied breeding habitat for 
some distance, thus being separated from their 
closest relatives by intervening unrelated popula¬ 
tions. These waves would have continued through 
the Pleistocene with new patches of habitat newly 
opening for eolonization with the retreat of 
glaciers. This pattern explains why geographically 
adjacent populations are not phenotypically sim¬ 
ilar, and why continental variation in the W-c G is 
not clinal. as noted in earlier chapters. A major 
problem with this interesting hypothesis is that 
breeding areas for most of these “taxa" are 
unknown, according to Hanson’s original descrip¬ 
tions. Anderson gives latitude and longitude tor 
each of them (Tables 10.2 and 11.2), but these can 
be based only on Hanson's conjectures of 
breeding range of taxa that he named solely on 
the basis of migrant or wintering birds. Without 
knowledge of the actual breeding areas of all 
subspecies, this patchwork scheme unravels. 
This analysis sets the stage for a revision of the 
taxonomy that Hanson had proposed, A synony¬ 
my of Hanson’s subspecies with AOU races is 
given again in Chapter 13; this one differs in 
substantial ways from that in Table 6.1 with 
different lists of Hanson’s taxa under each AOU 
subspecies. Anderson’s own recommended clas¬ 
sification. presented in Tables 13.4 and 13.6. 
elevates canadensis and hutchinsii to superspecies 
representing large and small W-c G. respectively. 
Each major stem from the dendrograms of 
Chapters 10 and 11 is considered a species—nine 
in canadensis and six in hutchinsii, for a total of 
15 species in the W-c C« complex. He uses some 
of Hanson's subspecific names for these species 
names, continuing to recognize others, and 
currently used AOU names, as subspecies. There 
is no characterization of these species level taxa 
or indication of geographic range. This cannot be 
considered a validly proposed classification be¬ 
cause the author ignores all rules (c.g.. priority) of 
the International Code of Zoological Nomencla¬ 
ture. 
The final chapter in the book is a guide to 
identification by region, as not all taxa will be 
found throughout North America. This is accom¬ 
panied by 23 color plates each showing dorsal and 
ventral views of specimens of 4-5 taxa. These 
plates give a good idea of the range of variation in 
color in the complex. 
Finally, a word about the Literature Cited and 
the Index. Some entries in the Literature Cited 
lack dates of publication, names of all authors, or 
complete titles, or give erroneous pagination. 
While drafting this review 1 had occasion to try' to 
find mention of Wilks Lambda in the text. It is 
not in the index under either word, but the name 
of an author caught my eye. Three page references 
were to citations of papers by that author, and two 
were to that author’s name in the Literature Cited. 
Not recalling that I have ever seen the Literature 
Cited indexed, I checked further, for other authors 
and for some keywords. The indexing of authors’ 
names is haphazard, but among other points there 
arc five pages under ‘’management’’ in the index 
that point to use of that word in citations in the 
Literature Cited, such as Journal of Wildlife 
Management. Under “wildlife management, man¬ 
agers" six indexed pages are all in the Literature 
Cited where a journal is cited, none to the text of 
the book. Software run amok. 
Anderson's primary statistical tool (page 55), 
used throughout the study, is discriminant func¬ 
tion analysis. One statistic used in this analysis, to 
test differences between means of two identified 
groups, is Wilks’ lambda. This is misspelled 
"Wilke's lamda" in five table headings in 
Chapter 2; “lambda" is correct in the text and 
in the tables of the next chapter, but the term is 
not mentioned in the rest of the hook although the 
same analysis is used throughout. In addition to 
errors earlier noted in Chapters 1 and 2,1 found a 
dozen errors there of grammar, poor or incomplete 
sentences, incorrect homophones, etc., not to 
mention questionable and confusing punctuation; 
there are dozens on dozens of such errors in the 
rest of the book. Most errors of this sort may be 
insignificant as far as understanding the geese is 
concerned, but 1 firmly believe that the quality of 
presentation of information, or production of a 
book, is somehow proportional to the quality of 
the gathering and assessment of the information 
presented. 
It is obvious that a great deal of effort went into 
this attempt to provide an understanding of and 
taxonomic basis for the extreme variation in the 
